Severe acute respiratory syndrome coronavirus infection of human ciliated airway epithelia: role of ciliated cells in viral spread in the conducting airways of the lungs

Abstract

Severe acute respiratory syndrome coronavirus (SARS-CoV) emerged in 2002 as an important cause of severe lower respiratory tract infection in humans, and in vitro models of the lung are needed to elucidate cellular targets and the consequences of viral infection. The SARS-CoV receptor, human angiotensin 1-converting enzyme 2 (hACE2), was detected in ciliated airway epithelial cells of human airway tissues derived from nasal or tracheobronchial regions, suggesting that SARS-CoV may infect the proximal airways. To assess infectivity in an in vitro model of human ciliated airway epithelia (HAE) derived from nasal and tracheobronchial airway regions, we generated recombinant SARS-CoV by deletion of open reading frame 7a/7b (ORF7a/7b) and insertion of the green fluorescent protein (GFP), resulting in SARS-CoV GFP. SARS-CoV GFP replicated to titers similar to those of wild-type viruses in cell lines. SARS-CoV specifically infected HAE via the apical surface and replicated to titers of 10(7) PFU/ml by 48 h postinfection. Polyclonal antisera directed against hACE2 blocked virus infection and replication, suggesting that hACE2 is the primary receptor for SARS-CoV infection of HAE. SARS-CoV structural proteins and virions localized to ciliated epithelial cells. Infection was highly cytolytic, as infected ciliated cells were necrotic and shed over time onto the luminal surface of the epithelium. SARS-CoV GFP also replicated to a lesser extent in ciliated cell cultures derived from hamster or rhesus monkey airways. Efficient SARS-CoV infection of ciliated cells in HAE provides a useful in vitro model of human lung origin to study characteristics of SARS-CoV replication and pathogenesis.

FIG. 1.

Localization of hACE2 on the apical surface of human airway epithelium ex vivo and in vitro. Representative histological frozen sections of freshly excised human nasal (A and B) or human tracheobronchial (C and D) airway tissues or HAE (E and F) were probed with either goat polyclonal anti-hACE (A, C, and E) or goat polyclonal anti-biotin (B and F) as a species-specific negative-control antibody. Bound primary antibody was visualized using donkey anti-goat secondary antibody conjugated to Texas Red. Immunofluorescence indicative of hACE (red) was detected in nasal and tracheobronchial tissue as well as HAE and was localized specifically to the apical membrane of ciliated cells (arrows). Nonciliated cell types present in the airway tissue or HAE were negative for hACE immunolocalization (arrowheads). Panel D shows Alcian blue (pH 2.5)- periodic acid-Schiff staining of human tracheobronchial airway tissue to highlight nonciliated cells (mucin-containing cells) present in the epithelium. Representative images are from specimens obtained from three different patients. Bar, 30 μm.

FIG. 2.

Schematics of SARS-CoV GFP construct and SARS-CoV infectious cloning strategy and subgenomic RNA transcription levels. (A) Schematic representation of the SARS-CoV genome, with all defined ORFs indicated by rectangles. Viral leader sequences are indicated by dark gray squares at the 5′ end of the genome (arrowhead) and between each ORF. The arrow indicates the 5′ viral leader transcription regulatory sequence (TRS). (B) Schematic representation of the SARS-CoV infectious cloning strategy and the mutations to engineer the GFP into ORFs7a/7b. The fragments of the genome are indicated by rectangles with the ORFs in each, as shown in panel A. The ORFs of the F clone have been expanded to indicate the location of GFP within the SARS-CoV GFP infectious-clone construct. (C) To determine levels of viral RNA synthesis from wild-type and SARS-CoV GFP viruses, total RNA was isolated from infected cells at 12 h postinfection and probed with a nucleocapsid-specific leader-containing probe. mRNAs 2 through 9 were detected for all samples, demonstrating that deletion of ORF7a/7b allowed the synthesis of all subgenomic RNA species. The filled arrowhead indicates wild-type mRNA 7, and the open arrowhead indicates the increased size of mRNA 7 associated with ORF7a/b excision and replacement with GFP. Samples are indicated across the top of the gel, and the individual RNA species are specified to the left and right of the image.

FIG. 3.

SARS-CoV GFP infects human airway epithelial cells derived from nasal and tracheobronchial epithelia but not alveolar epithelium. The apical surfaces of HAE derived from nasal (A) or tracheobronchial (B) airway tissues or alveolar regions of the human lung (C and D) were inoculated with SARS-CoV GFP (A, B, and C) or human PIV3 expressing GFP (D), and 48 h later, GFP-positive cells were assessed with fluorescent microscopy. Although nasal and tracheobronchial HAE were efficiently infected by SARS-CoV GFP, alveolar-derived cells were poorly infected by SARS-CoV (C) but efficiently infected by PIV3 (D). Similar data were obtained for alveolar cultures derived from A549 cells.

FIG.4.

Infection and spread of SARS-CoV GFP infection in HAE over time after apical or basolateral inoculation. HAE were inoculated via the apical (A, C, E, G, and I) or basolateral (B, D, F, and H) compartments with SARS-CoV GFP and GFP-positive cells and assessed over time (1 to 5 days postinfection). HAE inoculated with vehicle alone showed no GFP-positive cells (J). Apical inoculation resulted in significant numbers of GFP-positive cells at 40 h postinfection (C), with efficient spread of infection by 90 h postinfection (G). In contrast, basolateral inoculation resulted in a low proportion of cells positive for GFP only at 68 h postinfection (F). These images are representative of duplicate cultures from at least three different patient sets. Original magnification, ×10. (K) Apical inoculation of HAE with Urbani or icSARS-CoV was performed, and apical washes and basolateral media were harvested at the indicated time points postinfection. Collected samples were serially diluted, and titers were determined by plaque assay with Vero E6 cells. Titers are expressed as PFU/ml. Both Urbani and icSARS-CoV replicated to high titers in the apical compartment of HAE within 24 h, whereas progeny virions were detected in the basolateral compartment at later time points and to lower levels. All infections were performed in duplicate. Filled circles, Urbani apical; open circles, Urbani basolateral; filled squares, icSARS-CoV apical; open squares, icSARS-CoV basolateral.

FIG. 5.

SARS-CoV infects ciliated cells after apical inoculation of HAE. Representative histological sections of HAE 48 h postinfection with icSARS-CoV (A, B, and C) probed with mouse anti-S (A and B), mouse anti-N (C), or a mouse irrelevant anti-hemagglutinin (D) and visualized with anti-mouse secondary antibodies conjugated to Texas Red (red). Detection of S or N immunoreactivity was localized specifically to the ciliated cells of HAE (arrows), indicating that SARS-CoV infects this cell type after apical inoculation. No immunoreactivity was observed for nonciliated cell types (arrowheads). Images were obtained with a tricolor fluorescent filter to define the morphology of the tissue (gray/blue), with original magnifications of ×40 (A) and ×100 (B, C, and D).

FIG. 6.

Ultrastructural localization of SARS-CoV in HAE. Representative transmission electron microscopic photomicrographs of HAE infected with Urbani SARS-CoV. (A) HAE inoculated with vehicle alone, demonstrating the typical morphological features of the apical surfaces of ciliated cells with prominent cilia and microvilli. (B to E) HAE inoculated with Urbani SARS-CoV 48 h before fixation and showing the presence of large numbers of virus particles in vesicles inside ciliated cells (B and E) and on the surface of ciliated cells (B, D, and E) or shed into pericilial regions (C). Large quantities of virions were noted on the surface of ciliated cells, where ciliated cells were identified by cilial basal bodies (E). (F to H) To confirm that the observed virions were SARS-CoV, immuno-EM was performed using polyclonal mouse antisera against S with secondary antibodies conjugated to 12-nm colloidal gold (F). SARS-CoV infection resulted in extrusion and shedding of infected ciliated cells into the airway surface microenvironment (G and H). Similar observations were seen with HAE infected with icSARS-CoV and SARS-CoV GFP. Scale bars are shown for each panel. Filled arrowheads, cilia; filled arrows, microvilli; open arrowheads, virions; thin-tailed arrow in panel F, immuno-EM colloidal gold.

FIG. 7.

hACE2 is the primary receptor for SARS-CoV entry into HAE. (A) HAE were pretreated with polyclonal or monoclonal antisera directed against hACE2 (pACE2 and mACE2, respectively) or an anti-MUC1 negative-control antibody prior to inoculation with SARS-CoV GFP (black bar) or PIV3 expressing GFP (gray bar). Thirty-six hours postinfection, the numbers of GFP-positive cells for each antiserum treatment were assessed, demonstrating that pACE2 could ablate SARS-CoV infection of HAE but that mACE2 or MUC1 antisera had no effect on SARS-CoV infection. None of the antiserum treatments significantly affected PIV3 infection of HAE. (B) To assess effects of pretreatment of HAE with receptor-specific antisera on the growth kinetics of SARS-CoV infection, apical washes at the indicated time points were collected and results were determined by plaque assay with VeroE6 cells. Results represent data obtained from HAE derived from two different patients. Titers are expressed as PFU/ml. Filled squares, polyclonal hACE2 only; open squares, polyclonal and monoclonal hACE2; filled circles, monoclonal hACE2 only; filled triangles, MUC1; filled diamonds, no antiserum.

FIG. 8.

Morphological consequences of SARS-CoV infection of ciliated cells. Representative histological sections of HAE inoculated with SARS-CoV GFP and probed with mouse anti-S conjugated to Texas Red at 48, 96, 120, and 144 h postinfection (A to D). S localization (red) was restricted to the apical surface of ciliated cells by 48 h (A), but by 96 h postinfection, in some cases S immunoreactivity was detected around the periphery of the ciliated cells (B) (arrow), suggesting basolateral shedding of SARS-CoV. By 120 h postinfection, infected ciliated cells began to extrude from the epithelium (C) (arrow), and by 144 h postinfection, infected ciliated cells or components thereof were shed into the luminal compartment of HAE (D) (arrow). Note that the shed ciliated cell was positive for S immunoreactivity around the periphery of the cell, suggesting that virus was being shed from these regions. Bar, 5 μm. Images were obtained with a tricolor fluorescent filter to define the morphology of the tissue (gray/blue), with an original magnification at ×100.

FIG. 9.

SARS-CoV GFP infection of mouse, hamster, and rhesus monkey airway epithelial cell cultures. Ciliated airway epithelial cell cultures were derived from mice (A and D), hamsters (B and E), and rhesus monkeys (C and F) and inoculated via the apical surface with SARS-CoV GFP (10⁶ PFU). Hematoxylin and eosin-stained histological sections are shown for each species (A, B, and C), and GFP fluorescent images were recorded 48 h postinfection (D, E, and F). Original magnifications were ×40 (A, B, and C) and ×10 (D, E, and F). Bar, 5 μm (A, B, and C).