Reduced Replication of Highly Pathogenic Avian Influenza Virus in Duck Endothelial Cells Compared to Chicken Endothelial Cells Is Associated with Stronger Antiviral Responses

Abstract

Highly pathogenic avian influenza viruses (HPAIVs) cause fatal systemic infections in chickens, which are associated with endotheliotropism. HPAIV infections in wild birds are generally milder and not endotheliotropic. Here, we aimed to elucidate the species-specific endotheliotropism of HPAIVs using primary chicken and duck aortic endothelial cells (chAEC and dAEC respectively). Viral replication kinetics and host responses were assessed in chAEC and dAEC upon inoculation with HPAIV H5N1 and compared to embryonic fibroblasts. Although dAEC were susceptible to HPAIV upon inoculation at high multiplicity of infection, HPAIV replicated to lower levels in dAEC than chAEC during multi-cycle replication. The susceptibility of duck embryonic endothelial cells to HPAIV was confirmed in embryos. Innate immune responses upon HPAIV inoculation differed between chAEC, dAEC, and embryonic fibroblasts. Expression of the pro-inflammatory cytokine IL8 increased in chicken cells but decreased in dAEC. Contrastingly, the induction of antiviral responses was stronger in dAEC than in chAEC, and chicken and duck fibroblasts. Taken together, these data demonstrate that although duck endothelial cells are permissive to HPAIV infection, they display markedly different innate immune responses than chAEC and embryonic fibroblasts. These differences may contribute to the species-dependent differences in endotheliotropism and consequently HPAIV pathogenesis.

Keywords: cell tropism; chicken; duck; endothelial cells; highly pathogenic avian influenza; influenza A virus; innate immunity; pathogenesis; primary cell culture.

Figure 1

Endothelial cell characteristics of chicken and duck aortic endothelial cells. ChAEC and dAEC were isolated and passaged as described in the Materials and Methods section. (A) Representative bright field microscopy pictures of a tube formation assay after 4 h performed with chAEC (p14) and dAEC (p15). Human endothelial cells (HUVEC, p10) and canine epithelial cells (MDCK, p54) were used as positive and negative controls, respectively. In each corner, a schematic representation of the formed structures is shown. The scale bar represents 500 µm. Representative immunofluorescence images (B) and flow cytometry histograms (C) showing uptake of Alexa Fluor^®488-conjugated ac-LDL by chAEC (p15) and dAEC (p16) after 4 h of treatment. Human endothelial cells (EA-hy, p17) and human epithelial cells (H441, p19) were used as positive and negative controls, respectively. The scale bar represents 100 µm. (D) RT-PCR for vWF expression on chAEC (p16) and dAEC (p17). The left lane is the DNA size marker; bp = base pairs. ‘-RT’ = samples where RNA was used as template as a control for genomic DNA detection.

Figure 2

Sialic acid expression on the surface of primary avian endothelial cells. HEK-293T cells (p29), chAEC (p15), and dAEC (p16) were incubated for 2 h in medium with (dashed line) or without (black solid line) sialidase. Cells were stained with MAL-II (α2,3-specific) or SNA (α2,6-specific) lectins and subjected to flow cytometry. The negative control (in grey) consisted of untreated cells from which the lectins were omitted during the staining process. Representative flow cytometry histograms of at least three independent experiments are shown.

Figure 3

Replication kinetics of H5N1 HPAIV in duck and chicken endothelial cells. (A) MDCK, CEF, DEF, chAEC, and dAEC were inoculated with A/Vietnam/1203/04 H5N1 virus isolate at an MOI of 1. Cells were harvested at 6 hpi, and infection percentages were determined by NP staining and flow cytometry. Bars indicate the mean of four biological replicates, and the error bars represent the SD. (B) CEF, DEF, chAEC, and dAEC were inoculated as described for panel A, and supernatants were harvested at the indicated time points. Viral copy numbers were quantified by RT-qPCR on the matrix gene segment (left), and infectious virus titers were determined by endpoint titration in MDCK cells and expressed as TCID₅₀/mL (right). Bars indicate the mean of three biological replicates, and the error bars represent the SD. Dotted line represents the limit of detection of the endpoint titration assay. (C) ChAEC and dAEC were inoculated with A/Vietnam/1203/04 H5N1 virus isolate at an MOI of 0.001. Supernatants were harvested at the indicated time points, and infectious virus titers were determined by endpoint titration in MDCK and expressed as TCID₅₀/mL. Data are presented as mean ± SD from three independent experiments. Dotted line represents the limit of detection of the endpoint titration assay. Statistically significant differences were determined by one-way ANOVA followed up with individual unpaired t-tests. * p < 0.05; ** p < 0.01.

Figure 4

Host innate immune responses of avian primary cells upon HPAIV H5N1 inoculation. Monolayers of CEF, DEF, chAEC, and dAEC were inoculated with A/Vietnam/1203/04 H5N1 virus isolate at an MOI of 1. The cells were harvested at 6 and 12 hpi and analyzed for gene expression differences as compared to mock-inoculated cells. Messenger RNA levels were determined by a nucleic acid dye-based (A,C) or primer/probe (B) RT-qPCR. Fold changes were calculated using the 2^−ΔΔCT method with GAPDH serving as a housekeeping gene for normalization. (C) contains data from the 12 hpi time point. Bars indicate three biological replicates, and the error bars represent the SD. Statistically significant differences were determined on log-transformed fold changes by one-way ANOVA, followed up with individual unpaired t-tests. Only intra-species or intra-cell type significances are depicted. * p < 0.05; ** p < 0.01.

Figure 5

Cell tropism of H5N1 HPAIV in lungs of chicken and duck embryonated eggs. Fourteen-day-old ECE and 18-day-old EDE were mock-inoculated with PBS (A) or inoculated with 10³ TCID₅₀ RG-A/turkey/Turkey/1/05 H5N1 (B) via the allantoic route (two embryos per condition). Embryos were harvested at 24 hpi and processed for immunohistochemistry. Top panels show HE staining. Bottom panels show influenza A virus NP staining as indicated by a red precipitate. Representative images of the lungs are shown. The mesenchyme (asterisk), epithelial cells from the premature parabronchi (arrow), and NP-positive endothelial cells (arrow head) are indicated. The scale bar represents 100 µm.