ANIMAL MODELS OF HUMAN DISEASE Vet Pathol 45:551–562 (2008) Pathology of Experimental SARS Coronavirus Infection in Cats and Ferrets J. M. A. VAN DEN BRAND, B. L. HAAGMANS, L. LEIJTEN, D. VAN RIEL, B. E. E. MARTINA, A. D. M. E. OSTERHAUS, AND T. KUIKEN Department of Virology, Erasmus Medical Center, Rotterdam, the Netherlands Abstract. The pathology of severe acute respiratory syndrome-coronavirus (SARS-CoV) infection in cats and ferrets is poorly described, and the distribution of angiotensin-converting enzyme 2 (ACE2), a receptor for SARS-CoV, in the respiratory tracts of these species is unknown. We observed SARS-CoV antigen expression and lesions in the respiratory tracts of 4 cats and 4 ferrets at 4 days postinoculation and ACE2 expression in the respiratory tracts of 3 cats and 3 ferrets without infection. All infected cats and ferrets had diffuse alveolar damage associated with SARS-CoV antigen expression. A novel SARS-CoV– associated lesion was tracheo-bronchoadenitis in cats. SARS-CoV antigen expression occurred mainly in type I and II pneumocytes and serous cells of tracheo-bronchial submucosal glands of cats and in type II pneumocytes of ferrets. ACE2 expression occurred mainly in type I and II pneumocytes, tracheo-bronchial goblet cells, serous epithelial cells of tracheo-bronchial submucosal glands in cats, and type II pneumocytes and serous epithelial cells of tracheo-bronchial submucosal glands in ferrets. In conclusion, the pathology of SARS-CoV infection in cats and ferrets resembles that in humans except that syncytia and hyaline membranes were not observed. The identification of tracheo-bronchoadenitis in cats has potential implications for SARS pathogenesis and SARS-CoV excretion. Finally, these results show the importance of ACE2 expression for SARS-CoV infection in vivo: whereas ACE2 expression in type I and II pneumocytes in cats corresponded to SARS-CoV antigen expression in both cell types, expression of both ACE2 and SARS-CoV antigen in ferrets was limited mainly to type II pneumocytes. Key words: Angiotensin-converting enzyme (ACE2); cats; coronavirus; ferrets; histology; immuno- histochemistry; respiratory system; severe acute respiratory syndrome (SARS). Severe acute respiratory syndrome (SARS) emerged in the human population in November 2002 and spread rapidly across Asia, Europe, and North America in subsequent months.29 Although the causative virus, SARS-coronavirus (SARS- CoV),5,6,19,28 was eradicated from the human population by July 2003, its progenitors are likely still present in animal reservoirs21 and could again cross the species barrier into humans. In total, SARS caused 774 deaths out of more than 8,000 people with confirmed infection. Because the pathogenesis of SARS in humans is poorly understood, it is difficult to develop preventive and therapeutic strategies against this disease.29 The primary lesion of SARS in humans is diffuse alveolar damage (DAD), indicating injury to the alveolar septa.4,7,12,25,27 This lesion corresponds in part to the cell types in the respiratory tract in which SARS-CoV antigen has been detected: alveolar epithelial cells (primarily type II pneumocytes), bronchial epithelial cells, and alveolar macrophag- es.3,29,33,38,40 Cases of longer duration (more than 10 days) demonstrated features of organizing-phase or late-stage DAD.7,27 Angiotensin-converting enzyme 2 (ACE2) has been identified as a receptor for the attachment to and uptake of SARS-CoV in host cells.20 The distribution of ACE2 in human tissues corresponds largely to the cell types in which SARS-CoV replication has been observed: in addition to type II pneumocytes and bronchial epithelial cells, ACE2 expression has been observed in type I pneumocytes and endothelial cells as well as smooth muscle cells of blood vessels, but not alveolar macrophages.15,37 Since the appearance of SARS, several animal models have been developed for SARS-CoV infection in humans: SARS-CoV infection in 551 at UNIV OF BIRMINGHAM on March 20, 2015 vet.sagepub.com Downloaded from macaques,6 marmosets,10 mice,8,31 golden Syrian hamsters,32 rats,24 cats, and ferrets.22 Subbarao and Roberts35 reviewed the advantages and disadvan- tages of the above animal models. The limitations of rodents such as hamsters and young inbred mice are that they do not show illness35 and their lung structure differs from that of the human lung;26 the limitations of nonhuman primates are availability, cost, and housing.35 SARS-CoV infections in cats and ferrets do not show these limitations and may therefore be valuable as animal models. Previous experiments showed that domestic cats and ferrets were susceptible to SARS-CoV infec- tion and that they were able to transmit the virus efficiently to previously uninfected sentinel animals that were housed with them.22 The acute pulmo- nary lesions seen in those species were reported to be similar to those in humans and macaques, but were not described.22,36 Chronic lesions in ferrets euthanized at 23 days after infection were bronchi- al and bronchiolar hyperplasia as well as peribron- chiolar and perivascular lymphocytic cuffing.1,2 However, the pathology and viral distribution in cats and ferrets with acute SARS-CoV infection are poorly understood. Our goal was to describe the pathology of acute SARS-CoV infection in cats and ferrets and to determine the expression of ACE2 in the respira- tory tracts of these species. Therefore, we experi- mentally infected cats and ferrets with SARS-CoV and examined respiratory tract tissues at 4 days postinoculation (d.p.i.) by histopathology, immu- nohistochemistry, and immunofluorescence. Materials and Methods Virus preparation Strain HKU 39849 of SARS-CoV was passaged 4 times on Vero 118 cells cultured in Iscove’s Modified Dulbeco’s Medium (Bio Whitaker, Walkersville, MD). After centrifugation at 270 3 g for 5 minutes, 1 ml samples were made from the supernatant and the pelleted cells, and were resuspended in 5 ml medium. The titer of this virus stock was 1 3 106 median tissue culture infectious dose (TCID50) per ml. All cell cultures were done under biosafety level 3 conditions. Experimental protocol The experiments were performed under biosafety level 3 conditions at the Erasmus Medical Center in Rotterdam under an animal study protocol approved by the Institutional Animal Welfare Committee. Specified- pathogen-free cats and ferrets were purchased from commercial breeders (Harlan, Indianapolis, IN and Schimmel, Uddel, The Netherlands, respectively), main- tained in standard housing, and provided with commer- cial food pellets and water ad libitum until the start of the experiment. Before infection, they were examined clinically and determined to be healthy by a registered veterinarian and were placed in negative-pressurized glove boxes. Four cats and 4 ferrets were inoculated intratracheally in 1–5 seconds under ketamine anesthe- sia with 106 TCID50 of SARS-CoV. The animals were checked daily for clinical signs. At 4 days d.p.i., they were euthanized by exsanguination under ketamine anesthesia. Three cats and 3 ferrets were not inoculated and were used as negative controls. Pathologic examination The animals were necropsied according to a standard protocol by opening the thoracic and abdominal cavities and the skull and examining all major organs, including the brain. Samples were collected for histologic exam- ination and were stored in 10% neutral-buffered formalin (lungs after inflation with formalin), embedded in paraffin, sectioned at 4 mm, and stained with HE for examination by light microscopy. Selected lung sections also were stained with periodic acid Schiff (PAS) for detection of mucoid substances. The following tissues were examined by light microscopy: lung (cranial, middle, and caudal lobes of one lung), liver, spleen, kidney, trachea, urinary bladder, mesenteric lymph node, stomach (ferrets only), pancreas, duodenum, jejunum, ileum, and tracheo-bronchial lymph node. Semiquantitative assessment of SARS-CoV infection– associated inflammation in the lung was performed as reported earlier:14 each slide was examined for inflam- matory foci at 103 objective, and each focus was scored for size (1] smaller or equal than area of 103 objective, 2] larger than area of 103 objective and smaller than area of 23 objective, and 3] larger than area of 23 objective) and severity of inflammation (1] mild, 2] moderate, and 3] marked). Slides were examined without knowledge of the identity of the animals. The cumulative scores for the inflammatory foci provided the total score per animal. Immunohistochemistry Immunohistochemical examination on all tissues as examined by using light microscopy was performed as reported previously,19 except that mouse anti-SARS- nucleocapsid IgG2a (clone Ncap4, Imgenex, San Diego, CA, USA) (0.3 mg/ml) and goat anti-mouse IgG2a (SouthernBiotech, Birmingham, AL, USA) were used as primary and secondary antibodies, respectively. Omission of the primary antibody or replacement of the primary antibody by an irrelevant mouse IgG2a antibody (clone 20102, R&D, Abingdon, UK) (0.3 mg/ ml) were included as negative controls in each staining. The same tissues as mentioned above from noninfected cats and ferrets were used as negative controls. Lung sections from an experimentally inoculated ferret euthanized on 1 d.p.i. were used as positive controls. Semiquantitative assessment of SARS-CoV antigen expression was performed as reported earlier:14 25 arbitrarily chosen 203 objective fields of lung paren- chyma in each lung section were examined by using light 552 van den Brand, Haagmans, Leijten, van Riel, Martina, Osterhaus, and Kuiken Vet Pathol 45:4, 2008 at UNIV OF BIRMINGHAM on March 20, 2015 vet.sagepub.com Downloaded from microscopy for the presence of SARS-CoV antigen expression, without the knowledge of the identity of the animals. The cumulative scores for each animal were expressed as the number of positive fields per 100 fields. For double staining of SARS-CoV antigen and PAS- positive material, staining for SARS-CoV antigen was performed as described above with the following modifications. Antigen retrieval was performed at 100uC for 15 minutes with citric acid buffer, pH 6.0. Sections were cooled for 20 minutes on ice in the same buffer. To allow double staining, mouse anti-SARS- nucleocapsid IgG1 (Ncap17, Imgenex) (1.25 mg/ml) and biotin-labeled goat anti-mouse IgG1 (SouthernBiotech) were used as primary and secondary antibodies, respectively. After incubation with the secondary antibody, the labeled biotin was coupled to avidin- biotin-HRP complexes (DAKO, Heverlee, Belgium) for 1 hour at room temperature. Peroxidase activity was revealed by incubating slides in 3,39–diaminobenzidine- tetrachlorhydrate (Sigma, St Louis, MO) for 3 to 5 minutes, resulting in a brown precipitate. After washing in phosphate buffered saline (PBS), slides were stained with PAS according to standard methods. For detection of ACE2, tissue sections were pretreat- ed as described above and incubated with a rabbit polyclonal antibody to ACE2 (Abcam, Cambridge, UK) (5 mg/ml) or rabbit IgG control (R&D) (5 mg/ml) in PBS/0.1% body surface area (BSA) for 1 hour at room temperature. After washing with PBS, sections were incubated with HRP-labeled goat-anti-rabbit IgG (DAKO) in PBS/0.1% BSA for 1 hour at room temperature. Peroxidase activity was revealed by incu- bating slides in 3-amino-9-ethylcarbazole (Sigma) for 10 minutes and counterstaining with hematoxylin. To identify the ACE2 expression at the time of infection, we used the 3 noninfected cats and 3 noninfected ferrets, since spike protein binding during SARS-CoV infection results in down-regulation of ACE2 expression.18 For comparison of ACE2 expres- sion and histologic architecture of the tracheal submu- cosal glands among cats, ferrets, and humans, archival human tracheal tissues of 3 individuals, without histologic lesions or evidence of respiratory tract infection at the time of death, also were stained with PAS and tested for ACE2 expression. Immunofluorescence For double staining of SARS-CoV antigen and keratin, 3-mm-thick, formalin-fixed, paraffin-embedded tissue sections were pretreated as described above. Autofluorescence in the sections was blocked by incubation with 0.3 M glycine for 10 minutes. Sections were washed with PBS/0.05% Tween 20 and incubated with mouse IgG1 against human pankeratin AE1/AE3 (Neomarkers, Fremont, CA) (8 mg/ml) for 1 hour at room temperature in the dark. After washing with PBS, sections were incubated with mouse IgG1 against SARS-CoV (clone Ncap17, Imgenex) (1.25 mg/ml) for 1 hour at room temperature in the dark. The pankeratin antibody was labeled by using the Zenon Alexa Fluor 488 Mouse IgG1 labeling kit (cat. no. Z25090, Molec- ular Probes, Invitrogen, Breda, The Netherlands) according to the manufacturer’s protocol. The SARS- CoV antibody was labeled by using a Zenon Alexa Fluor 594 Mouse IgG1 labeling kit (cat. no. Z25007, Molecular Probes). The slides were mounted by using the Vectashield, hard set with 4,6-diamidino-2-phenyl- indole (Vector, Burlingame, CA). Statistical analysis One-way analysis of variance was used for comparing the scores of pulmonary lesions and viral antigen expression. Differences were considered significant at P , .05. Results Clinical findings The cats showed no clinical signs such as lethargy or dyspnea. The ferrets, however, became lethargic from 2 d.p.i. onward, and 1 of these ferrets (No. 3) died at 4 d.p.i. Gross examination All 4 cats had multifocal pulmonary consolida- tion characterized by red, firm, level areas of approximately 1 cm in diameter in the cranial and medial lobes. Cat No. 4 had multifocal consolida- tion in all lung lobes. All 4 ferrets also had multifocal pulmonary consolidation similar to that in the cats, except that the foci were smaller, ranging from 0.1 to 1 cm in diameter, and were localized mainly in the caudal lobes. Ferret No. 2 also had dark red and enlarged mesenteric lymph nodes, and ferret No. 3 had a dark red, friable liver and a mottled, red-and-pink spleen. Other organs of the cats and ferrets showed no significant lesions. Histology The primary damage from SARS-CoV infection was seen in the respiratory tract in both cats and ferrets and was comparable in character and severity. In the cats, histologic lesions were seen in the lung, tracheo-bronchial submucosal glands, spleen, tracheo-bronchial and mesenteric lymph nodes, and Peyer’s patches (Figs. 1–4). In the lung, all cats had multifocal, mild-to-moderate, exuda- tive DAD characterized by cellular debris in the alveolar lumen and epithelial cells with karyor- rhexis, karyopyknosis, and loss of cellular detail (multifocal necrosis), sparse type II pneumocyte hyperplasia, and infiltration with few neutrophils in the alveolar septa (Fig. 4). In the lumina of terminal bronchioles and adjacent alveoli, there were few alveolar macrophages and neutrophils (Figs. 3, 4). In the tracheo-bronchial submucosal Vet Pathol 45:4, 2008 Pathology of SARS in Cats and Ferrets 553 at UNIV OF BIRMINGHAM on March 20, 2015 vet.sagepub.com Downloaded from Figs. 1–4. Respiratory tract. Histologic and immunohistochemical staining for SARS-CoV antigen and ACE2 in SARS-CoV–infected cats and control cats. Detailed descriptions can be found in the text. Fig. 1a. Trachea; cat No. 2. Moderate lymphoplasmacytic tracheoadenitis. HE. Fig. 1b. Trachea; cat No. 2. SARS-CoV antigen expression in epithelial cells of submucosal glands. Immunoperoxidase with hematoxylin counterstain. Fig. 1c. Trachea; control cat. ACE2 expression in epithelial cells of mucosa and submucosal glands. Immunoperoxidase with hematoxylin counterstain. Fig. 2a. Bronchus; cat No. 1. Moderate lymphoplasmacytic bronchoadenitis. HE. Fig. 2b. Bronchus; cat No. 1. SARS-CoV antigen expression in epithelial cells of submucosal glands. Immunoperoxidase with hematoxylin counterstain. Fig. 2c. Bronchus; control cat. ACE2 expression in epithelial cells of mucosa and submucosal glands. Immunoperoxidase with hematoxylin counterstain. Fig. 3a. Bronchiole; cat No. 1. Mild lymphoplasmacytic and neutrophilic bronchiolitis. HE. Fig. 3b. Bronchiole; cat No. 1. Lack of SARS- CoV antigen expression. Immunoperoxidase with hematoxylin counterstain. Fig. 3c. Bronchiole; control cat. ACE2 expression in epithelial cells of mucosa. Immunoperoxidase with hematoxylin counterstain. Fig. 4a. Alveoli; cat No. 4. Mild DAD. HE. Figs. 4b, c. Alveoli; cat No. 3. SARS-CoV antigen expression in type I (k) and II (l) pneumocytes. Immunoperoxidase with hematoxylin counterstain. Figs. 4d, e. Alveoli; control cat. ACE2 expression in type I (m) and II (n) pneumocytes. Immunoperoxidase with hematoxylin counterstain. 554 van den Brand, Haagmans, Leijten, van Riel, Martina, Osterhaus, and Kuiken Vet Pathol 45:4, 2008 at UNIV OF BIRMINGHAM on March 20, 2015 vet.sagepub.com Downloaded from glands, all cats had multifocal, mild-to-moderate tracheo-bronchoadenitis (Figs. 1, 2, 5, 6). This was characterized by multifocal epithelial necrosis in submucosal glands and by infiltration with moder- ate numbers of plasma cells and lymphocytes and with few macrophages and neutrophils in sur- rounding connective tissue. With PAS staining, these submucosal glands displayed narrow, empty lumina, a high proportion of serous cells compared with mucous cells, and mild granular PAS staining in mucous cells (Figs. 6, 7), which was also seen in noninfected cats. In cats Nos. 1 and 3, the spleen had moderate lymphoid hyperplasia, while tracheo- bronchial and mesenteric lymph nodes and Peyer’s patches had both mild-to-moderate lymphoid hyperplasia and moderate-to-marked sinusoidal histiocytosis. No significant lesions were seen in other tissues examined or in the tissues of the negative control cats. In the ferrets, histologic lesions were seen in the lung, liver, spleen, and tracheo-bronchial lymph nodes (Figs. 8–11). In the lung, all ferrets had multifocal, mild-to-severe, subacute, exudative DAD characterized by variable numbers of alveo- lar macrophages and neutrophils, mixed with proteinaceous exudate in alveolar and bronchiolar lumina (Fig. 11). The alveolar septa were moder- ately thickened with infiltrating neutrophils, mac- rophages, and multifocal moderate epithelial ne- crosis (Fig. 11). There was multifocal hyperplasia and hypertrophy of type II pneumocytes and bronchiolar epithelial cells (Fig. 10). There was hyperplasia of bronchus-associated lymphoid tis- sue, and additionally, ferret No. 3 had lymphoid aggregates around pulmonary blood vessels. Over- all, the severity of pulmonary lesions in the ferrets and cats showed no significant differences (F 5 0.48, P 5 .51) (Fig. 12). In the liver, all ferrets had mild diffuse hepatic lipidosis, while ferret No. 3 also had mild diffuse hepatic congestion and hemorrhage. The spleen and tracheo-bronchial lymph nodes of all ferrets showed mild lymphoid hyperplasia. In contrast to the cats, no significant lesions were seen in the tracheo-bronchial submu- cosal glands of the ferrets (Fig. 8). PAS staining of these glands (in both infected and noninfected animals) showed wide lumina filled with PAS- positive material, a low proportion of serous cells compared with mucous cells, and intense, diffuse PAS-expression in mucous cells (Figs. 6, 7). No significant lesions were seen in other tissues examined or in the tissues of negative control ferrets. In human tissues, the tracheal and bronchial submucosal glands had an almost equal mixture of serous and mucous cells. The lumina varied in size, and some were filled with large amounts of PAS- positive material. SARS-CoV antigen expression The cell types expressing SARS-CoV antigen, visible as diffuse red staining of the cytoplasm, differed between cats and ferrets (Table 1, Figs. 1– 4, 8–11). In cats, SARS-CoV antigen expression was limited to the respiratory tract and the intestine. In the respiratory tract of all cats, SARS-CoV antigen expression was seen mainly in type I pneumocytes, type II pneumocytes (Fig. 4), and serous cells of the tracheo-bronchial submu- cosal glands (Figs. 1, 2); occasionally in alveolar macrophages; and rarely in tracheal, bronchial, and (Cat No. 3 only) bronchiolar ciliated epithelial cells. The identity of SARS-CoV-antigen-positive type I and type II pneumocytes was based on their morphology (flat and cuboidal, respectively) and confirmed by double staining with pankeratin to Fig. 5. Bronchus; cat No. 1. Histologic staining of a bronchus in a cat infected with SARS-CoV. Moderate lymphoplasmacytic bronchoadenitis with mild necrosis of glandular epithelial cells. HE. Inset: Higher magni- fication of a submucosal gland with loss of epithelial cells. Vet Pathol 45:4, 2008 Pathology of SARS in Cats and Ferrets 555 at UNIV OF BIRMINGHAM on March 20, 2015 vet.sagepub.com Downloaded from exclude macrophages (Figs. 13–15). The identity of SARS-CoV-antigen-positive serous cells in the tracheo-bronchial submucosal glands was based on the central-to-paracentral round nucleus and finely granular cytoplasm. It was confirmed by double staining with PAS: serous cells showed no or only weak staining with PAS, in contrast to intense staining of mucous cells also present in these glands (Figs. 6, 7). In the other tissues, a few enterocytes in the ileum had expression of SARS- CoV antigen, but only in the intestine of cat No. 2. In all ferrets, SARS-CoV antigen expression was limited to the respiratory tract, where it was seen mainly in type II pneumocytes (Figs. 8–11), and rarely in type I pneumocytes and alveolar macro- phages. Again, the identity of SARS-CoV-antigen- positive type I and type II pneumocytes was determined by their morphology and double staining with pankeratin (Figs. 13–15). Neither ferrets nor cats had viral antigen expression in the other tissues. Virus antigen expression was present in positive tissue controls and absent in negative tissue as well as isotype and omission controls. The expression of SARS-CoV antigen in the respiratory tract of both cats and ferrets was usually associated with the presence of histologic lesions. The level of SARS-CoV antigen expression in the lungs of ferrets and cats showed no significant differences (F 5 1.34, P 5 .29) (Fig. 12). ACE2 expression Expression of ACE2, visible as diffuse red staining of the cytoplasm and cell membrane, differed between the respiratory tract of cats and ferrets (Table 1, Figs. 1–4, 8–11). In cats, strong ACE2 expression was seen mainly in tracheal and bronchial goblet cells (Figs. 1, 2), serous epithelial cells of tracheo-bronchial submucosal glands (Figs. 1, 2), and type I and type II pneumocytes (Fig. 4); moderate expression was seen in tracheal, bronchial, and bronchiolar ciliated cells, bronchi- olar nonciliated cells, and mucous cells of the tracheo-bronchial submucosal glands (Figs. 1–3); and weak expression was seen in alveolar macro- phages. In addition, strong ACE2 expression was seen in the endothelium and smooth muscle of pulmonary blood vessels. In ferrets, the pattern of ACE2 expression resembled that in cats, except that it was absent in type I pneumocytes and in Figs. 6,7. Bronchus. Histologic, histochemical, and immunohistochemical staining of bronchi comparing cats and ferrets infected with SARS-CoV. Fig. 6a. Cat No. 1. Moderate lymphoplasmacytic bronchoadenitis with mild necrosis of glandular epithelial cells. HE. Fig. 6b. Cat No. 1. Few PAS-positive cells in bronchial submucosal glands and scant mucus in acinar lumina. Fig. 6c. Cat No. 3. SARS-CoV antigen expression in serous glandular epithelial cells. Double staining with PAS and immunoperoxidase. Fig. 7a. Ferret No. 2. No lesions in bronchial submucosal glands. HE. Fig. 7b. Ferret No. 3. Many PAS-positive cells in bronchial submucosal glands and abundant mucus in acinar lumina. PAS. Fig. 7c. Ferret No. 2. Lack of SARS-CoV antigen expression. Double staining with PAS and immunoperoxidase. 556 van den Brand, Haagmans, Leijten, van Riel, Martina, Osterhaus, and Kuiken Vet Pathol 45:4, 2008 at UNIV OF BIRMINGHAM on March 20, 2015 vet.sagepub.com Downloaded from Figs. 8–11. Respiratory tract. Histologic and immunohistochemical staining for SARS-CoV antigen and ACE2 in SARS-CoV–infected ferrets and control ferrets. Detailed descriptions can be found in the text. Fig. 8a. Trachea; ferret No. 2. No lesions. HE. Fig. 8b. Trachea; ferret No. 2. Lack of SARS-CoV antigen expression. Immunoperoxidase with hematoxylin counterstain. Fig. 8c. Trachea; control ferret. ACE2 expression in epithelial cells of mucosa and submucosal glands. Immunoperoxidase with hematoxylin counterstain. Fig. 9a. Bronchus; ferret No. 3. No lesions. HE. Fig. 9b. Bronchus; ferret No. 3. Lack of SARS-CoV antigen expression. Immunoperoxidase with hematoxylin counterstain. Fig. 9c. Bronchus; control ferret. ACE2 expression in epithelial cells of mucosa and submucosal glands. Immunoperoxidase with hematoxylin counterstain. Fig. 10a. Bronchiole; ferret No. 2. Mild lymphoplasmacytic and neutrophilic bronchiolitis. HE. Fig. 10b. Bronchiole; ferret No. 2. Lack of SARS-CoV antigen expression. Immunoperoxidase with hematoxylin counterstain. Fig. 10c. Bronchiole; control ferret. ACE2 expression in epithelial cells of mucosa. Immunoperoxidase with hematoxylin counterstain. Fig. 11a. Alveoli; ferret No. 3. DAD with intraluminal protein. HE. Fig. 11b. Alveoli; ferret No. 3. SARS-CoV antigen expression in type II pneumocytes. Immunoperoxidase with hematoxylin counterstain. Fig. 11c. Alveoli; ferret. ACE2 expression in type II pneumocytes. Immunoperoxidase with hematoxylin counterstain. Vet Pathol 45:4, 2008 Pathology of SARS in Cats and Ferrets 557 at UNIV OF BIRMINGHAM on March 20, 2015 vet.sagepub.com Downloaded from tracheal and bronchial goblet cells (Figs. 8–11). In human trachea, the ACE2 expression was stronger in serous cells than in mucous cells of mucosa and submucosal glands. Discussion The localization, character, and severity of respiratory tract lesions from acute SARS-CoV infection in cats and ferrets show both similarities to and differences from those in acute human cases.4,7,11,27,33 First, the localization of the lesions in cats and ferrets is similar to that in humans: in all three species, the lesions affect mainly the alveoli and bronchioles. A clear difference is SARS-CoV– associated tracheo-bronchoadenitis, which we ob- served in cats, but which has not been reported in humans. Second, the character of the lesions in alveoli and bronchioles of cats and ferrets shows some similarities to that of humans. In all three species, these lesions are characterized by necrosis and subsequent hyperplasia of alveolar and bron- chiolar epithelium, infiltration by neutrophils, and increased numbers of alveolar macrophages. In humans and ferrets, proteinaceous exudate was seen in alveolar and bronchiolar lumina. A notable difference is the lack of syncytia in the pulmonary lesions of cats and ferrets, whereas this feature is considered characteristic for the disease in humans. Third, the severity of the lesions in cats and ferrets shows the most marked difference from that in humans. The pulmonary lesions in cats and ferrets are multifocal, in contrast to the more diffuse lesions in humans. Also, cats and ferrets lack fibrin exudation and hyaline membrane formation, both of which are features of severe damage to the alveolar septum, while these features are commonly reported for human SARS. However, these differ- ences in severity are largely a reflection of the fact that pathologic reports of human SARS are generally limited to fatal cases, which only com- prise about 10% of the total number of confirmed cases.29 It is likely that many people who recovered from SARS had less severe respiratory tract lesions. Overall, this comparison suggests that SARS-CoV infections in cats and ferrets are suitable animal models to study the pathogenesis and pathology of human SARS in general. However, in their present form they do not replicate the lesions of fatal cases of human SARS. Adapting this virus to cats and ferrets might increase its pathogenicity, as has been done for mice.30 Tracheo-bronchoadenitis caused by SARS-CoV infection is a novel finding, with potentially important implications for SARS pathogenesis and SARS-CoV excretion. Pathogenesis of SARS could be affected because the mucus produced by tracheo-bronchial submucosal glands is an impor- tant part of the mucociliary escalator that removes foreign material from the airways. Inflammation of these glands could reduce the efficiency of this Fig. 12. Semiquantitative scoring of histologic le- sions (white bars) and SARS-CoV antigen expression (black bars) in the lung. These scores did not differ significantly between cats and ferrets. Table 1. Expression of ACE2 and SARS-CoV antigen in the respiratory tract of cats, ferrets, and humans with SARS. Species Antigens Tissues and Cell Types Alveolus Bronchiole Type I Pneumocyte Type II Pneumocyte Alveolar Macrophage Ciliated Epithelial Cell Nonciliated Epithelial Cell Cat ACE2 +++ +++ + ++ ++ SARS-CoV +++ +++ + + 2 Ferret ACE2 2 +++ + ++ ++ SARS-CoV 2 +++ + 2 2 Human ACE215 ++ +++ + ++ ++ SARS-CoV3,29,33 2 +++ + 2 2 NOTE: Abundance of ACE2 or SARS-CoV antigen expression: +++ 5 abundant; ++ 5 occasional; + 5 rare; 2 5 absent. 558 van den Brand, Haagmans, Leijten, van Riel, Martina, Osterhaus, and Kuiken Vet Pathol 45:4, 2008 at UNIV OF BIRMINGHAM on March 20, 2015 vet.sagepub.com Downloaded from Tissues and Cell Types Bronchus Trachea Pulmonary Blood Vessel Ciliated Epithelial Cell Goblet Cell Glandular Epithelial Cell Ciliated Epithelial Cell Goblet Cell Glandular Epithelial Cell Endothelial cell Smooth Muscle Cell ++ +++ +++ ++ +++ +++ ++ ++ + 2 +++ + 2 +++ 2 2 ++ 2 +++ ++ 2 +++ ++ ++ 2 2 2 2 2 2 2 2 ++ 2 +++ ++ 2 +++ ++ ++ 2 2 2 2 2 2 2 2 Table 1. Extended. Figs. 13–15. Alveoli. Immunofluorescence double staining for SARS-CoV antigen and pankeratin in cats and ferrets infected with SARS-CoV antigen. Fig. 13a. Cat No. 3. Type I pneumocyte with SARS-CoV antigen expression (red). Immunofluorescence. Fig. 13b. Cat No. 3. Type I pneumocyte with pankeratin expression (green). Immunofluorescence. Fig. 13c. Cat No. 3. Type I pneumocyte with colocalization of SARS-CoV antigen and pankeratin (orange). Immunofluorescence. Fig. 14a. Cat No. 3. Type II pneumocyte with SARS-CoV antigen expression (red). Immunoflu- orescence. Fig. 14b. Cat No. 3. Type II pneumocyte with pankeratin expression (green). Immunofluorescence. Fig. 14c. Cat No. 3. Type II pneumocyte with colocalization of SARS-CoV antigen and pankeratin (orange). Immunofluorescence. Fig. 15a. Ferret No. 2. Type II pneumocyte with SARS-CoV antigen expression (red). Immunofluorescence. Fig. 15b. Cat No. 2. Type II pneumocyte with pankeratin expression (green). Immunofluorescence. Fig. 15c. Cat No. 2. Type II pneumocyte with colocalization of SARS-CoV antigen and pankeratin (orange). Immunofluorescence. Vet Pathol 45:4, 2008 Pathology of SARS in Cats and Ferrets 559 at UNIV OF BIRMINGHAM on March 20, 2015 vet.sagepub.com Downloaded from defense system, thereby increasing the risk of lower respiratory tract infection by SARS-CoV and concurrent pathogens. Excretion of SARS-CoV might increase as a result of SARS-CoV infection of the tracheo-bronchial submucosal glands: virus secreted by these glands into the trachea and bronchi is more likely to be expectorated than virus produced lower in the respiratory tract. The susceptibility of tracheo-bronchial submu- cosal glands to SARS-CoV infection shows clear host species differences, which may be relevant for human infection. Although the glands of both cats and ferrets expressed ACE2, only cat glands became infected and inflamed (Figs. 1, 5, 6). One explanation may be that ferret glands lack a cofactor for SARS-CoV replication, as discussed above. Another explanation may be species differ- ences in the histologic architecture of the tracheo- bronchial submucosal glands. The glands of ferrets had a higher proportion of mucous cells and more mucus in the acinar lumina than those of cats (Figs. 6, 7), and this may have inhibited the attachment of SARS-CoV to target cells. There- fore, we hypothesize that SARS-CoV infects mainly serous cells. To our knowledge, SARS- associated tracheo-bronchoadenitis has not been reported in humans. However, the expression of SARS-CoV antigen in the tracheal and bronchial submucosal glands of fatal human cases3 suggests that inflammation of these glands may be an as- yet-unreported feature of human SARS that requires further attention. This study shows the importance of ACE2 expression for SARS-CoV infection in vivo. This is based on our observation that SARS-CoV antigen expression occurred only in ACE2-positive cells (Table 1, Figs. 1–4, 8–11). SARS-CoV S1 protein is reported to bind efficiently to both ferret and feline ACE2.13,41 The importance of ACE2 expression is well illustrated by differences that we observed between cats and ferrets at the level of the alveoli: whereas cat type I pneumocytes abundantly expressed both ACE2 and SARS-CoV antigen, the absence of ACE2 expression in ferret type I pneumocytes corresponded with the absence of expression of SARS-CoV antigen (Figs. 1–4, 8–11). However, it must be realized that colocalization of ACE2 and SARS-CoV antigen expression is consistent with, but not proof of, a functional relationship. Also, several ACE2-positive cell types did not express SARS-CoV antigen (Table 1). Possible reasons for this could be that other receptors besides ACE2 are required for SARS- CoV attachment to the host cell, as is also described in humans.9,17,23,39 For example, recent in vitro studies have shown that the presence of cathepsin L is required for productive SARS-CoV infection.16,34 Alternatively, SARS-CoV may infect other cell types at an earlier or later stage of infection. Alveolar macrophages in both cats and ferrets demonstrated weak expression of ACE2 and sparse expression of SARS-CoV antigen. This presence of SARS-CoV antigen indicates either virus infection of alveolar macrophages or, as is suggested for humans,38,39 phagocytosis of viral antigen produced in other cells. In conclusion, our study shows that SARS-CoV infections in cats and ferrets are suitable models for studying the pathology and pathogenesis of human SARS. We also have identified tracheo-broncho- adenitis in cats as a novel SARS-CoV–induced lesion, which could potentially occur in human SARS and affect the pathogenesis and virus excretion of this disease. Finally, we have shown in vivo that ACE2 expression is important for SARS-CoV infection and subsequent development of lesions in the respiratory tract. Acknowledgements We are grateful for the help of F. van der Panne with microphotographs and G. van Amerongen and R. Dias D’Ullois with animal experiments. This work was supported by European grant SP22-CT-2004-511060 (DISSECT). References 1 Czub M, Weingartl H, Czub S, He R, Cao J: Evaluation of modified vaccinia virus Ankara based recombinant SARS vaccine in ferrets. Vaccine 23: 2273–2279, 2005 2 Darnell ME, Plant EP, Watanabe H, Byrum R, St Claire M, Ward JM, Taylor DR: Severe acute respiratory syndrome coronavirus infection in vac- cinated ferrets. J Infect Dis 196:1329–1338, 2007 3 Ding Y, He L, Zhang Q, Huang Z, Che X, Hou J, Wang H, Shen H, Qiu L, Li Z, Geng J, Cai J, Han H, Li X, Kang W, Weng D, Liang P, Jiang S: Organ distribution of severe acute respiratory syndrome (SARS) associated coronavirus (SARS-CoV) in SARS patients: implications for pathogenesis and virus transmission pathways. J Pathol 203:622–630, 2004 4 Ding Y, Wang H, Shen H, Li Z, Geng J, Han H, Cai J, Li X, Kang W, Weng D, Lu Y, Wu D, He L, Yao K: The clinical pathology of severe acute respiratory syndrome (SARS): a report from China. J Pathol 200:282–289, 2003 5 Drosten C, Gunther S, Preiser W, van der Werf S, Brodt HR, Becker S, Rabenau H, Panning M, Kolesnikova L, Fouchier RA, Berger A, Burguiere AM, Cinatl J, Eickmann M, Escriou N, Grywna K, 560 van den Brand, Haagmans, Leijten, van Riel, Martina, Osterhaus, and Kuiken Vet Pathol 45:4, 2008 at UNIV OF BIRMINGHAM on March 20, 2015 vet.sagepub.com Downloaded from