Attenuation of replication by a 29 nucleotide deletion in SARS-coronavirus acquired during the early stages of human-to-human transmission

Abstract

A 29 nucleotide deletion in open reading frame 8 (ORF8) is the most obvious genetic change in severe acute respiratory syndrome coronavirus (SARS-CoV) during its emergence in humans. In spite of intense study, it remains unclear whether the deletion actually reflects adaptation to humans. Here we engineered full, partially deleted (-29 nt), and fully deleted ORF8 into a SARS-CoV infectious cDNA clone, strain Frankfurt-1. Replication of the resulting viruses was compared in primate cell cultures as well as Rhinolophus bat cells made permissive for SARS-CoV replication by lentiviral transduction of the human angiotensin-converting enzyme 2 receptor. Cells from cotton rat, goat, and sheep provided control scenarios that represent host systems in which SARS-CoV is neither endemic nor epidemic. Independent of the cell system, the truncation of ORF8 (29 nt deletion) decreased replication up to 23-fold. The effect was independent of the type I interferon response. The 29 nt deletion in SARS-CoV is a deleterious mutation acquired along the initial human-to-human transmission chain. The resulting loss of fitness may be due to a founder effect, which has rarely been documented in processes of viral emergence. These results have important implications for the retrospective assessment of the threat posed by SARS.

Figure 1

Generation and evaluation of ORF8 variant recombinant SARS-CoV. Variants of the open reading frame 8 (ORF8) were designed in accordance to their appearance in nature. (a) rSCV_8full represents a single ORF8 as found in reservoir bats and amplification hosts in China, as well as in the early phase of the SARS epidemic. The middle phase of the epidemic was dominated by a virus variant carrying a 29 nt deletion resulting in the disruption of ORF8 in 2 reading frames, 8a and 8b, as seen in rSCV_epi. The absence of ORF8 is a genomic feature of reservoir bats in Europe and the epidemic virus in the late phase. Two deletion variants were constructed. Variant 1, rSCV_del8–1, perfectly misses ORF8. In variant 2, rSCV_del8–2, ORF8 is replaced by the short substitutional sequence AATAA in accordance to the upstream region of the nucleocapsid gene of the European SARS-related bat-CoV BtCoV/BM48–31/Rhi bla/Bulgaria/2008 (NC_014470). (b) Plaque morphology of rSCV_8full and rSCV_epi were very similar, while rSCV_del8–1 produced only diffuse and rSCV_del8–2 reduced plaques. Western Blot analysis revealed that only rSCV_8full, rSCV_epi, and rSCV_del8–2 infected cells expressed the nucleocapsid protein, while none could be detected in cells infected with rSCV_del8–1. Detection of β actin served as loading control. Virus replication of the three ORF8 variants, (rSCV_del8 = rSCV_del8–2) was monitored by plaque titration after infection of VeroFM cells at two different multiplicities of infection, 1 (c) and 0.001 (d). Virus growth was determined in at least 3 independent experiments in triplicates. Shown is one representative experiment. Error bars represent standard deviation of the mean.

Figure 2

Replication of ORF8 variants in primate cell cultures with and without IFN pre-treatment. (a) VeroFM and MA104 cells were infected with three ORF8 variant viruses at MOI = 0.001. Supernatants were harvested at 24 hpi and virus titers determined by plaque titration. The experiment was done in triplicates. Significance of replication differences between virus variants was determined by t-Test (***p < 0.001, *p < 0.05). (b) VeroFM cells were incubated with universal type I IFN at indicated concentrations 16 h prior to infection with rSCV_8full and rSCV_del8 at MOI = 0.001. At 24 hpi supernatants were plaque titered. Experiments were done at least twice in triplicates. Shown is one representative experiment each. Error bars represent standard deviation of the mean.

Figure 3

Generation of an hACE2-transgenic bat cell line for infection experiments with ORF8 variant rSCV. (a) Picture of an African Rhinolophus alcyone bat (copyright Victor Max Corman). Primary cell culture and immortalization of R. alcyone embryonic lung cells (b) was done as described in the Material and Methods section. To determine the RhiLu cell type the epithelial protein marker cytokeratin (c) and the fibroblast marker S-100A4 (d) were detected by immunofluorescence assay using mouse-anti cytokeratin or rabbit-anti S-100A4 Ig. Secondary detection was performed by incubating with goat-anti-mouse cyanin 2- or goat-anti-rabbit cyanin 3-labeled Igs. The bars represent 20 µm. (e) Integration of hACE2 into the genome of RhiLu cells was verified by PCR. The vector containing the hACE2-puromycin resistance gene construct was used as a PCR positive control. (f) Expression of hACE2 was confirmed by Western blot analysis using mouse anti-hACE2 Ig (1:1,000). In addition, β actin protein was detected using rabbit anti-β-actin Ig (1:2,000) to ensure that similar protein amounts were applied. MA104 cells, expressing hACE2 naturally, served as positive control. (g) Virus replication of rSCV_8full, rSCV_epi and rSCV_del8 was observed by titration of supernatants sampled at 24, 48, and 72 hpi. Cells were infected in triplicates at an MOI of 0.001. Error bars represent standard deviation of the mean.

Figure 4

Replication of ORF8 variants in different “non-SARS-CoV-host” cell lines and differentiated human airway epithelial cells. (a) Cotton rat, goat, and sheep cells were transduced with lentiviruses to transiently express the SARS-CoV receptor hACE2 for at least 72 h. Expression of hACE2 was verified by Western Blot analysis, detection of β actin served as a loading control. To improve clarity blots were cropped. Full length blots are presented in Supplementary Fig. S1. (b) Cells were infected in triplicates with different ORF8 variants at MOI 0.001 24 h after lentiviral transduction, 24 hpi supernatants were sampled and virus replication determined by plaque titration. (c) Differentiated human airway epithelial cells were infected in triplicates at MOI = 0.1, supernatants were sampled at 48 and 72 hpi for plaque titration. The experiment was done twice. Shown is one representative experiment. Error bars represent standard deviation of the mean. Significance of replication differences between virus variants was determined by t-Test (***p ≤ 0.001, **p ≤ 0.001, *p ≤ 0.05, n.s., not significant p > 0.05).