Proteolytic activation of the porcine epidemic diarrhea coronavirus spike fusion protein by trypsin in cell culture

Abstract

Isolation of porcine epidemic diarrhea coronavirus (PEDV) from clinical material in cell culture requires supplementation of trypsin. This may relate to the confinement of PEDV natural infection to the protease-rich small intestine of pigs. Our study focused on the role of protease activity on infection by investigating the spike protein of a PEDV isolate (wtPEDV) using a reverse genetics system based on the trypsin-independent cell culture-adapted strain DR13 (caPEDV). We demonstrate that trypsin acts on the wtPEDV spike protein after receptor binding. We mapped the genetic determinant for trypsin-dependent cell entry to the N-terminal region of the fusion subunit of this class I fusion protein, revealing a conserved arginine just upstream of the putative fusion peptide as the potential cleavage site. Whereas coronaviruses are typically processed by endogenous proteases of the producer or target cell, PEDV S protein activation strictly required supplementation of a protease, enabling us to study mechanistic details of proteolytic processing. Importance: Recurring PEDV epidemics constitute a serious animal health threat and an economic burden, particularly in Asia but, as of recently, also on the North-American subcontinent. Understanding the biology of PEDV is critical for combatting the infection. Here, we provide new insight into the protease-dependent cell entry of PEDV.

FIG 1

Infection with the wild-type PEDV isolate but not with cell culture-adapted PEDV benefits from trypsin activity. Inocula were supplemented with soy bean trypsin inhibitor (SBTI), trypsin, or a combination of both and applied to Vero cells. After 15 h of incubation, infected cells were detected by immunofluorescence staining against the nucleocapsid protein (green). Nuclei were stained with DAPI (blue).

FIG 2

The S protein determines trypsin dependency of PEDV propagation. (A) Schematic representation of the recombinant PEDV genomes carrying the PEDV-CV777 or the PEDV-caDR13 S gene in the isogenic background of caDR13 (PEDV-Swt and PEDV-Sca, respectively). The ORF3 gene was substituted by a GFP sequence. (B) Vero cells were inoculated in the presence or absence of trypsin or soy bean trypsin inhibitor (SBTI). After 2 h, the inoculum was removed and incubation continued in the presence of SBTI to prevent syncytium formation. At 11 h postinfection (p.i.), infected cells were examined by GFP expression using fluorescence microscopy (green). Nuclei were stained with DAPI (blue). (C) The percentage of infected cells was determined by quantifying GFP-expressing cells using flow cytometry. The averages with standard deviations (SD) from 4 experiments are displayed relative to the inoculation in the presence of trypsin. (D) To assess the effect of trypsin on the release of infectious PEDV particles from producer cells, inoculations were performed for 2 h before the medium was refreshed, and incubation was continued in the absence or presence of trypsin. At 14 to 16 h p.i., supernatants were collected. Trypsin was added to all samples 1 h before infectious virus titers were determined by endpoint dilution. The ratios of infectivity in samples obtained in the presence/absence of trypsin were calculated and displayed (*, P value of 0.026 in paired-sample t test). Seven independent experiments with multiple replicates were carried out, and each dot represents the ratio obtained from one pair of samples. (E) The effect of trypsin on the release of viral RNA (vRNA) from PEDV-infected cells was quantified. Vero cells were infected with wtPEDV and caPEDV and cultured from 2 to 16 h p.i. in the presence or absence of trypsin. RNA was subsequently purified from supernatants or cells, and vRNA levels were quantified by qRT-PCR. The relative amounts of vRNA of the samples treated with trypsin compared to the samples treated with SBTI are displayed.

FIG 3

Characterization of S protein activation. (A) An entry assay was performed as described in Fig. 2B. Vero cells were inoculated for 2 h with PEDV-Sca or PEDV-Swt, and infection was quantified by flow cytometry, showing the averages with SD from 3 experiments displayed relative to the inoculation in the presence of trypsin. For pretreatment, the inoculum or the target cells were exposed to trypsin for 1 h at 37°C. SBTI was supplemented to quench the trypsin activity at the indicated steps. (B) Trypsin treatment of S protein on purified virions. Recombinant viruses carrying a FLAG tag at the C terminus of the PEDV-S protein (PEDV-Sca_flag and PEDV-Swt_flag) were produced in the absence of trypsin and purified by pelleting through 20% sucrose. A similar purification procedure was done with culture medium from PEDV-Sca and mock-infected Vero cells (mock). Samples were exposed to 15 μg/ml trypsin or left untreated for 30 min. Infectivity was determined by endpoint dilution (50% tissue culture infective dose [TCID₅₀]/ml), and virus particles were subjected to Western blot analysis. Proteins were detected by a mouse monoclonal anti-FLAG antibody conjugated with horseradish peroxidase or rabbit anti-PEDV-S1 serum. (C) Virus supernatant was pretreated as described above with SBTI (filled bars) or trypsin (open bars). After addition of an excess SBTI, virus was added to cells and attachment was allowed for 1 h at 8°C in the absence of trypsin activity. After binding, an entry assay was performed as described for Fig. 2B. Infection was quantified by flow cytometry, showing the averages with SD from 4 experiments displayed relative to inoculation in the presence of trypsin without pretreatment.

FIG 4

Mapping the genetic determinant for trypsin-enhanced PEDV entry. (A) Putative organization of class I fusion protein features in PEDV S protein. SP, signal peptide; FP, fusion peptide (S891-V910); HR1 and HR2, heptad repeat regions; TM, transmembrane domain; S1/S2 junction, region of the furin cleavage site in MHV-A59; S2′, location of putative cleavage site within the S2 subunit in SARS-CoV and IBV S protein; drawn to scale. (B) A schematic overview shows the chimeric S proteins of the recombinant PEDV variants. Red and green regions are derived from Swt or Sca protein, respectively. (C) An entry assay was performed as described in Fig. 2 to compare the trypsin-dependent entry of PEDV variants. (D) The alignment of the amino acid sequence N-terminal of the fusion peptide (FP) depicts a conserved arginine or a glycine substitution (bold italic). Putative S2′ cleavage site is indicated with an arrow.

FIG 5

Substitution of arginine at position 890 results in reduced syncytium formation capacity of transiently expressed S protein. (A) After overnight incubation with PEDV-Swt, infected cells were treated with trypsin for 1 h while live images were obtained. Representative images are shown. (B) Vero cells were transiently transfected with expression plasmids encoding Swt and its point mutant Swt_R890G for 48 h. Cells were treated with trypsin or trypsin and SBTI for 1 h and subsequently examined by immunostaining against S protein (green). Nuclei were stained with DAPI (blue). Representative images are shown. (C) The numbers of nuclei per focus were quantified and displayed as binned frequency distribution histogram (four independent experiments, Swt n = 390, Swt_R890G mutant n = 330). Syncytia containing 1 to 4 nuclei were small, 5 to 8 nuclei were medium, and more than 8 nuclei were large. The average syncytium sizes induced by Swt and Swt_R890G mutant significantly differ from each other (P value of <0.0001, nonparametric t test).