Inhibition, escape, and attenuated growth of severe acute respiratory syndrome coronavirus treated with antisense morpholino oligomers

Abstract

The recently emerged severe acute respiratory syndrome coronavirus (SARS-CoV) is a potent pathogen of humans and is capable of rapid global spread. Peptide-conjugated antisense morpholino oligomers (P-PMO) were designed to bind by base pairing to specific sequences in the SARS-CoV (Tor2 strain) genome. The P-PMO were tested for their capacity to inhibit production of infectious virus as well as to probe the function of conserved viral RNA motifs and secondary structures. Several virus-targeted P-PMO and a random-sequence control P-PMO showed low inhibitory activity against SARS coronavirus. Certain other virus-targeted P-PMO reduced virus-induced cytopathology and cell-to-cell spread as a consequence of decreasing viral amplification. Active P-PMO were effective when administered at any time prior to peak viral synthesis and exerted sustained antiviral effects while present in culture medium. P-PMO showed low nonspecific inhibitory activity against translation of nontargeted RNA or growth of the arenavirus lymphocytic choriomeningitis virus. Two P-PMO targeting the viral transcription-regulatory sequence (TRS) region in the 5' untranslated region were the most effective inhibitors tested. After several viral passages in the presence of a TRS-targeted P-PMO, partially drug-resistant SARS-CoV mutants arose which contained three contiguous base point mutations at the binding site of a TRS-targeted P-PMO. Those partially resistant viruses grew more slowly and formed smaller plaques than wild-type SARS-CoV. These results suggest PMO compounds have powerful therapeutic and investigative potential toward coronavirus infection.

FIG. 1.

P-PMO targeting schematic. Relative positions of P-PMO targets on the genomic plus-strand of SARS CoV are indicated by grey boxes. Enlarged regions of the genome indicate the specific target sequences of P-PMO directed against the TRS and AUG regions and the relative proximity of the 1ABFS P-PMO to the ribosomal frameshift site. Nucleotide positions refer to the published sequence of the SARS-CoV-Tor2 strain.

FIG. 2.

Peptide-conjugated PMO specifically reduce cell-free and cell culture expression of plasmid-generated target sequences. (A) Cell-free translation assay demonstrates the inhibition of SARS AUG region/luciferase translation by AUG P-PMO relative to nonspecific activity of the DSCR control P-PMO. Data are expressed as percent inhibition relative to 12 untreated control values. (B) The converse assay, inhibiting translation from DSCR target/luciferase RNA specifically with the DSCR P-PMO demonstrates the lack of cross-reactivity of DSCR and AUG P-PMO in this system. (C) Luciferase expression from AUG region/luciferase mRNA generated from a transiently transfected plasmid in cell culture was inhibited by single-AUG P-PMO and combination AUG PMO treatments. Concentrations of combined P-PMO are expressed as the molarity of a 1:1 mixture of two nonoverlapping compounds. Error bars indicate standard error of the mean.

FIG. 3.

Peptide-conjugated PMO reduce SARS-CoV cytopathology and titer. Qualitative changes in cell morphology and density were gauged against untreated infected (upper right) and untreated uninfected (upper left) controls. (A) Representative images of cells pretreated for 6 h with selected R₉F₂-conjugated PMO and fixed 72 h after inoculation. (B) Dose-response of titer reduction. Triplicate wells of Vero-E6 cells were pretreated with P-PMO or vehicle-only at 6 h before inoculation with SARS-CoV at a multiplicity of 0.1 PFU/cell. Virus yield was quantified 24 h after. The limit of detection for the assay shown was 100 PFU/ml. Error bars indicate standard error of the mean. (C) Cells were pretreated with 20 μM TRS2 R₉F₂-PMO or mock treated 6 h before inoculation. Culture medium was collected at 15 h and 24 h and replaced with medium containing P-PMO or medium alone as indicated. Virus yield was measured at 15 h, 24 h, and 48 h.

FIG. 4.

Peptide-conjugated PMO and coiled-coil peptides inhibit the propagation of SARS-CoV infection. Plaque diameter on treated and mock-treated cells was visualized (A) and measured (B) 72 h after inoculation under the same experimental conditions as described for Fig. 3A. (C) Comparison of 72-h plaque diameter on cells treated with R₅F₂R₄-AUG1 P-PMO, R₅F₂R₄-DSCR P-PMO, hygromycin B (HygB), or coiled-coil SARS-HR2 and on mock-treated cells (untreated). Error bars indicate standard error of the mean. (D) Reverse transcription-25-cycle PCR comparison of viral subgenomic RNA 2 to 9 levels in an equivalent number of mock-treated or 20 μM TRS2 P-PMO-treated cells 24 h after inoculation. Sizes in bp are indicated to the right. Amplicons of 104, 127, 156, 188, 212, 259, 299, and 353 bp were expected, corresponding to viral subgenomic RNAs 2 through 9, respectively.

FIG. 5.

Characterization of P-PMO-resistant SARS-CoV. SARS-CoV was serially passaged on cells pretreated with 2 μM or 10 μM R₉F₂-PMO or mock-treated cells. (A) Virus yield over the first nine passages was quantified 24 h after inoculation at an initial multiplicity of ≈10 PFU/cell. (B) The diameters of 50 plaques were measured after 11 viral passages on untreated or 10 μM P-PMO-treated cells. (C) Growth kinetics of P-PMO-resistant plaque-purified SARS-CoV on untreated Vero-E6 cells are shown. Biologically cloned virus was cultured from plaque-purified stocks selected after 11 passages on untreated cells or cells treated with AUG1, AUG2, S2M, 3TERM, 3UTR, TRS2, or DSCR P-PMO. Growth curves for five median-growth partially TRS2-resistant SARS-CoV biological clones are shown. TCID₅₀ titrations were calculated for four fourfold replicates. Error bars indicate standard error of the mean. (D) The 5′-terminal regions of P-PMO-resistant and mock-treated clones were amplified and sequenced in the antigenomic orientation. A portion of the TRS2 P-PMO target region is shown, with the mutations in TRS2-resistant clones underlined.

FIG. 6.

Mechanisms of P-PMO efficacy and partial resistance. (A) Binding of P-PMO to the TRS region (nucleotide positions 51 to 79) was assessed in a cell-free translation inhibition assay. The relative binding strength of P-PMO to the wild-type SARS-CoV target (wt) and the three-mispair target (mut) is expressed as percent inhibition of luciferase expression. Zero percent inhibition was determined by the average level of luciferase expression from untreated control translations programmed with both wild-type and mutant mRNAs. (B) Comparison of inhibition of luciferase expression downstream of the entire SARS-CoV 5′-UTR sequence by three P-PMO. Error bars indicate standard error of the mean. (C) Low-energy secondary structures of the TRS hairpin of wild-type SARS-CoV (wt) and TRS2 P-PMO-selected SARS-CoV (TRS2-R) were generated using Mfold (49). The core TRS is near the top of the stem and shown in white circles; flanking sequences are depicted inside black circles. Nucleotides are numbered from the beginning of the predicted stem-loop.