Inhibition of vaccinia virus L1 N-myristoylation by the host N-myristoyltransferase inhibitor IMP-1088 generates non-infectious virions defective in cell entry

Abstract

We have recently shown that the replication of rhinovirus, poliovirus and foot-and-mouth disease virus requires the co-translational N-myristoylation of viral proteins by human host cell N-myristoyltransferases (NMTs), and is inhibited by treatment with IMP-1088, an ultrapotent small molecule NMT inhibitor. Here, we examine the importance of N-myristoylation during vaccinia virus (VACV) infection in primate cells and demonstrate the anti-poxviral effects of IMP-1088. N-myristoylated proteins from VACV and the host were metabolically labelled with myristic acid alkyne during infection using quantitative chemical proteomics. We identified VACV proteins A16, G9 and L1 to be N-myristoylated. Treatment with NMT inhibitor IMP-1088 potently abrogated VACV infection, while VACV gene expression, DNA replication, morphogenesis and EV formation remained unaffected. Importantly, we observed that loss of N-myristoylation resulted in greatly reduced infectivity of assembled mature virus particles, characterized by significantly reduced host cell entry and a decline in membrane fusion activity of progeny virus. While the N-myristoylation of VACV entry proteins L1, A16 and G9 was inhibited by IMP-1088, mutational and genetic studies demonstrated that the N-myristoylation of L1 was the most critical for VACV entry. Given the significant genetic identity between VACV, monkeypox virus and variola virus L1 homologs, our data provides a basis for further investigating the role of N-myristoylation in poxviral infections as well as the potential of selective NMT inhibitors like IMP-1088 as broad-spectrum poxvirus inhibitors.

Fig 1. IMP-1088 inhibits VACV spread and virus yield.

(A) Step-wise illustration of virus yield, spread and cytotoxicity assays. (B) Quantification of VACV yield in the presence of IMP-1088. HeLa cells were infected with VACV WR at increasing concentrations of IMP-1088. The cells were harvested 24 hpi, lysed by freeze-thaw and virus yield was determined by plaque assay. Dotted line indicates 50% virus yield. (C) Measuring viral spread in the presence of IMP-1088 based on GFP expression. Cells were infected with VACV WR-GFP in the presence of different concentrations of IMP-1088. The concentration of IMP-1088 required to reduce viral spread by 50% (EC₅₀) was determined. (D) Cytotoxic effects of IMP-1088 determined by LDH assay. All experiments were performed twice with two replicates in each experiment. Values represent means +/- SEM.

Fig 2. Chemical proteomic identification and quantification of host and viral proteins after VACV infection, enrichment, and mass spectrometric detection.

(A) Identification of VACV proteins after YnMyr enrichment. 3 N-myristoylated VACV proteins depicted in purple, 115 VACV proteins in pink, 2738 human proteins in gray. (B) Label free quantification intensity of N-myristoylated VACV proteins L1, A16, G9, as determined in background, after metabolic tagging with YnMyr and after NMT inhibition with 2 μM IMP-1088. Average of 3 replicates, error bars depict standard deviation, significance tested by ANOVA. (C) Effect of VACV infection on 32 known co-translationally N-myristoylated protein levels of the host (purple). Other proteins in gray. For volcano plots shown in (A) and (C), the horizontal and vertical axes show the difference in quantified protein levels and the significance of the quantified difference, respectively. The labels above the right and left arrow indicate the two conditions being compared within the plot. The dashed vertical lines depict -0.5 Log₂ and +0.5 fold change; horizontal dashed line depicts significance cut-off (p = 0.05).

Fig 3. IMP-1088 does not affect VACV gene expression and morphogenesis.

(A) Schematic representation of early and late protein detection in the presence of IMP-1088. (B) BSC-40 cells were infected with purified VACV WR-pE/L LUC virus in the absence and presence of 2 μM IMP-1088 and 40 μg/mL AraC. The level of secreted luciferase from early promoter was determined at 2 hpi. (C) BSC-40 cells were infected with purified VACV WR-pF17R LUC in the absence and presence of 2 μM IMP-1088 and 40 μg/mL AraC. Luciferase levels were measured 24 hpi. RLU values of virus control, IMP-1088 and AraC treatments in (B) and (C) were compared using a one-way ANOVA followed by a Tukey’s multiple comparisons test. Ns indicates no significant difference and **** signifies a p<0.0001. (D) Transmission electron micrographs of VACV-infected cells in the absence and presence of IMP-1088. The various morphogenic forms of VACV are seen in both treatments; crescent (C), immature virus (IV), mature virus (MV), wrapped virus (WV) and extracellular virus (EV). Scale bar corresponds to 1 μm.

Fig 4. IMP-1088 decreases infectivity of progeny VACV particles.

(A) Coomassie stained SDS-PAGE gel showing difference in protein levels between IMP-1088-treated and untreated (control) viruses. BSC-40 cells were infected with VACV WR pE/L-LUC for 24 h in the absence and presence of 2 μM IMP-1088. Cells were harvested and lysed, and virus particles were purified by sucrose density gradient centrifugation. Either 7.5 or 15 μl of purified virus was run in a 4–12% SDS PAGE and stained with Coomassie blue. (B) Equivalent viral particles based on Coomassie staining were subjected to DNA isolation followed by real time PCR using VACV-specific primers. Ct values at different dilutions of purified DNA from the two treatments are shown. Data analyzed using Multiple t tests and statistical significance determined using the Holm-Sidak method, with alpha = 5%. For all 6 dilutions, p > 0.05 (C) Equivalent viral particles from untreated and IMP-1088 treated virus were tested by plaque assay to determine virus yields. Yields plotted for both treatments as pfu/mL. Statistical significance determined using an unpaired t test; p > 0.05.

Fig 5. IMP-1088 reduces EV infectivity without directly affecting yield.

(A) Schematic representation of assays used to assess VACV EV production after treatment with IMP-1088. (B) BSC-40 cells were infected with VACV IHD-J in presence of inhibitors 40 μg/mL AraC, 2 μM ST-246 and 2 μM IMP-1088 for 24 h. The culture media was collected, spun at low speed to remove debris and cells, and tested by plaque assay to determine virus yield (pfu/mL). Three replicates were tested per viral dilution for each treatment, and the mean values +/- SD are shown. (C) Total viral DNA in the culture media for all four treatments was quantified by DNA isolation followed by qPCR. Four replicates per dilution were tested for every treatment, and the means +/- SD are shown. In both B and C, a one-way ANOVA was performed to determine statistical significance, followed by Tukey’s multiple comparisons test. Ns = not significant, ** = p≤0.005 and **** = p<0.0001.

Fig 6. Decrease in VACV infectivity linked to defect in viral entry.

(A) IMP-1088 treated virus exhibits lower early gene expression compared to control virus. BSC-40 cells were infected with VACV propagated in the absence or presence of 2 μM IMP-1088, and luciferase levels were determined 2 hpi as a surrogate of early protein synthesis. Dotted line indicates level of detection. (B) Lower membrane fusion between cells and IMP-1088 treated virus compared to control virus. Virus grown in the presence or absence of IMP-1088 was purified and labelled with fluorescent dye DiO (same virus as Fig 4). Cells were infected with DiO-labeled virus at RT (to determine background signal) and at 37°C. Transfer of fluorescent dye from viral membrane to cellular membrane was measured by flow cytometry. (C) Western blot to confirm pulldown of L1 protein after metabolic labelling with YnMyr, chemical modification and precipitation using streptavidin resin. The input, supernatant and eluted fractions from uninfected and infected cells in the presence and absence of IMP-1088 were tested for presence of L1 using anti-L1 polyclonal antibody (R180).

Fig 7. Genomes and infectivity of G9 and A16 viruses with G2A mutations.

(A) Ratios of genomes to infectious units. WR, VACV WR-G9(2GA), and VACV WR-A16(G2A) mature virions were purified from infected cells and the infectivity determined by plaque assay. DNA was extracted from the purified virions and genome copies were quantified by ddPCR. (B) Plaque sizes. The areas of plaques formed by purified VACV WR, VACV WR-G9(G2A) and VACV WR-A16(G2A) from a representative experiment are shown. The areas of WR-A16(G2A) plaques were smaller than VACV WR in each of three independent experiments and those of WR-G9(G2A) were similar to those of VACV WR in one experiment and slightly smaller in two others.