Infection of ciliated cells by human parainfluenza virus type 3 in an in vitro model of human airway epithelium

Abstract

We constructed a human recombinant parainfluenza virus type 3 (rPIV3) that expresses enhanced green fluorescent protein (GFP) and used this virus, rgPIV3, to characterize PIV3 infection of an established in vitro model of human pseudostratified mucociliary airway epithelium (HAE). The apical surface of HAE was highly susceptible to rgPIV3 infection, whereas only occasional cells were infected when virus was applied to the basolateral surface. Infection involved exclusively ciliated epithelial cells. There was little evidence of virus-mediated cytopathology and no spread of the virus beyond the ciliated cell types. Infection of ciliated cells by rgPIV3 was sensitive to a neuraminidase specific for alpha2-6-linked sialic acid residues, but not to a neuraminidase that cleaves alpha2-3- and alpha2-8-linked sialic acid residues. This provided evidence that rgPIV3 utilizes alpha2-6-linked sialic acid residues for initiating infection, a specificity also described for human influenza viruses. The PIV3 fusion (F) glycoprotein was trafficked exclusively to the apical surface of ciliated cells, which also was the site of release of progeny virus. F glycoprotein localized predominately to the membranes of the cilial shafts, suggesting that progeny viruses may bud from cilia per se. The polarized trafficking of F glycoprotein to the apical surface also likely restricts its interaction with neighboring cells and could account for the observed lack of cell-cell fusion. HAE derived from cystic fibrosis patients was not more susceptible to rgPIV3 infection but did exhibit limited spread of virus due to impaired movement of lumenal secretions due to compromised function of the cilia.

FIG. 1.

Recombinant PIV3 expressing enhanced GFP from an added reporter gene (rgPIV3). (A) The coding sequence for GFP was inserted into the downstream noncoding region of the P/C/D/V gene. Nucleotides 3693 to 3698 of the PIV3 genome were modified into an AflIII site, which was used to accept an insert that contained PIV3 gene-end (GE), intergenic (IG), and gene-start (GS) signals, followed by NcoI and HindIII sites. These latter sites were used to accept an NcoI-HindIII fragment containing the open reading frame for enhanced GFP (shaded rectangle, with the ATG and TAA initiation and termination codons indicated). (B) Fluorescent photomicrographs showing the intracellular expression of GFP in cell culture at 36 h following rgPIV3 infection of HEp-2 cells (i), A549 cells (ii), or nonpolarized, PD primary airway epithelial cells (iii). Original magnification, ×10.

FIG. 2.

Expression of GFP in HAE infected with rgPIV3 (10⁶ PFU; MOI, ∼3). (A and B) GFP expression in HAE viewed en face 48 h after apical (A) or basolateral (B) inoculation of HAE by rgPIV3. Original magnification, ×10. (C) Representative confocal optical section of rgPIV3-mediated GFP expression (green) in HAE 48 h following apical infection. HAE was also probed with β-tubulin IV antibody (red), revealing that GFP was present only in ciliated columnar epithelial cells. (D) Representative histological cross-section of HAE infected with rgPIV3. At 48 h p.i., no obvious cytopathic effects or cell-cell fusion was apparent. Counterstain is hematoxylin and eosin. Bar, 20 μm.

FIG. 3.

Susceptibility of human airway cells to infection with rgPIV3 at various times postseeding (A) or following treatment of the lumenal surfaces of cultures with NA from V. cholerae (B). (A) Quantitation of GFP-positive human airway cells 24 h after inoculation of the cells with rgPIV3 (10⁶ PFU) (filled triangles) or AdVGFP (10¹⁰ p/ml) (open triangles) at different days postseeding revealed that airway epithelial cells are initially susceptible to rgPIV3 and AdVGFP infection at early time points in culture (1 to 6 days), followed by a refractory period to both viruses that coincided with the establishment of an ALI (arrow). By 16 days, HAE regained susceptibility to infection by rgPIV3 but not AdVGFP. (B) Quantitation of GFP-positive cells present 24 h after rgPIV3 inoculation of PD airway cells or HAE either without (closed bars) or with (open bars) pretreatment of the lumenal surfaces with NA III derived from V. cholerae (160 mU/ml for 3 h).

FIG. 4.

Localization of cell surface HS and sialic acid linkages on PD airway cells, viewed en face (A, C, and E) and HAE, viewed in histological cross-section (B, D, and F). (A) Representative fluorescent photomicrographs of HS localization, detected with antibody F58-10E4, on the cell surface of a subpopulation of PD cells (red). (B) In HAE, HS localization was restricted to basal epithelial cells (red, arrowheads). (C and D) Sialic acid residues with α2-3 linkages, detected with MAA lectin, were present on the cell surface of a subpopulation of PD cells (C) (red), whereas in HAE (D) they were localized on the apical surface (green), viewed against a counterstain of antibody to β-tubulin IV (red). Staining with MAA was predominantly on ciliated cells (arrows in panel D) but was also detected on some nonciliated cells (arrowhead in panel D). (E and F) Sialic acid residues with α2-6 linkages, detected with SNA lectin, were present on the cell surface of PD cells (E) (red), whereas in HAE (F) localization was predominately detected on the apical surface (green), viewed against a counterstain of antibody to β-tubulin IV (red). In panel F, SNA lectin was detected on ciliated (arrows) and nonciliated (arrowheads) cells. Bar, 10 μm.

FIG. 5.

Inhibition of rgPIV3 infection of HAE following lumenal treatment with NAs from different organisms. (A) Representative en face photomicrographs of HAE exposed to vehicle control alone (i) or treated with NA from V. cholerae, which cleaves α2-3, -6, and -8 linkages (ii); Newcastle disease virus NA, specific for α2-3 and α2-8 linkages (iii); or A. ureafaciens NA, which is specific for α2-6 linkages. NAs were applied to the apical surface for 3 h immediately prior to washing and inoculation with rgPIV3. GFP-positive cells were photographed 24 h later. (B) Quantitation of GFP expression from experiments such as shown in panel A, representing cultures from three different patients, each in duplicate. Cultures of HAE were treated with the NAs described above (shown as sialic acid linkages cleaved, with the control being vehicle alone) and infected with rgPIV3 (filled bars) or rgRSV (open bars). The expression of GFP was quantified 24 h p.i. The data shown represent means ± standard errors (SE) of the GFP-positive surface area. (C) Recovery of susceptibility to infection with rgPIV3 following treatment with V. cholerae NA III that cleaves α2-3, -6, and -8 linkages. Replicate cultures of HAE were treated on the apical surface with NA III for 3 h, and the apical surface was inoculated with rPIV3 immediately (0 h) or following 6, 12, 24, 48, or 72 h of recovery. A control culture (Ctl) was infected without prior NA III treatment. The expression of GFP was quantified 24 h p.i. The data shown represent means ± SE of the GFP-positive surface area from duplicate cultures derived from two different patient samples.

FIG. 6.

Localization of the PIV3 F glycoprotein in rgPIV3-infected PD airway cells and HAE. (A) Photomicrographs of PD airway cells viewed en face (i) and HAE in histological cross-sections (ii, iii, and iv). The cells were infected with rgPIV3 (i, ii, and iii) or rgRSV (iv), incubated for 24 h, and stained with a monoclonal antibody specific to the F glycoprotein (red). The F glycoprotein was localized in the surface membrane of PD airway cells (i) and in the membranes of the cilial shafts of HAE (ii and iii). No immunoreactivity for PIV3 F glycoprotein was detected with rgRSV-inoculated cultures probed under identical conditions to those described above (iv). Bars, 10 μm (ii) or 3 μm (iii and iv). (B) Electron transmission photomicrographs of the apical surface of rgPIV3-infected HAE cells, probed with a monoclonal antibody specific to the F glycoprotein and detected with an immunogold secondary antibody (12 nm). Immunoreactivity to F glycoprotein was localized predominantly to membranes of the cilial shafts (arrows in panels i and ii) and occasionally in ciliated cell microvilli projections (arrowheads in panels i and ii). No immunogold was observed in rgRSV-infected HAE or in HAE inoculated with vehicle control alone and probed with antibodies under identical conditions to those described above (data not shown). Bars, 0.2 μm.

FIG. 7.

Spread of rgPIV3 infection is compromised in HAE derived from CF airway epithelium. rgPIV3-mediated expression of GFP viewed en face in HAE derived from a non-CF patient (A, C, E, and G) and from a CF patient (B, D, F, and H). Inoculation of the apical surfaces of non-CF (A) or CF (B) HAE with a high titer of rgPIV3 (10⁶ PFU) resulted in similar levels of infection at 24 h p.i. Similarly, with a low-titer rgPIV3 inoculum (10³ PFU), lower but comparable numbers of GFP-positive cells were detected 24 h after inoculation in non-CF (C) and CF (D) HAE. Following the pattern of viral spread with time revealed that viral spread was facilitated by coordinated cilial beat in non-CF HAE by 48 h p.i. (E), leading to homogenous distribution of infection by 72 h (G). In contrast, for CF HAE, viral spread was restricted to cells in close proximity to the primary infected cells and spread remained limited at both 48 (F) and 72 (H) h p.i. Original magnification, ×10.

FIG. 8.

rgPIV3-induced ciliated cell shedding leads to mucin-containing cell metaplasia in HAE. (A) Quantitation of the percentage of GFP-positive cells over time after apical inoculation of HAE with rgPIV3 (10⁶ PFU; closed triangles) or AdVGFP (10¹⁰ p/ml; inoculated after sodium caprate treatment to open epithelial tight junctions; closed circles). (B) Transepithelial resistance measurements over time after inoculation of HAE with rgPIV3 (open bars) or vehicle control only (filled bars). (C) Representative histological sections of perfluorocarbon-osmium tetroxide-fixed HAE inoculated with either vehicle control (i and iii) or rgPIV3 (ii and iv) and fixed 2 days p.i. (i and ii) or 13 days p.i. (iii and iv). Note the increased presence of mucin-containing cells (indicated by arrows) in HAE inoculated with rgPIV3 and fixed 13 days p.i. Bar, 20 μm.