Introduction
Human respiratory syncytial virus (RSV) is recognised globally as the main cause of acute lower respiratory infection in young children. Respiratory syncytial virus is an enveloped, non-segmented, negative sense, single-strand RNA virus in the Pneumovirdae family [
1,
2] and, like all viruses, is reliant on the host cell for replication and release. Many viruses use cellular translation factors to ensure their own proteins are produced, and to stifle the innate host defence mechanisms in order to rapidly proliferate [
3]. For example, several RNA viruses including RSV have subverted cellular eEF1A for viral replication and genome synthesis [
4,
5]. RNA viruses may utilize eEF1A because it is an abundant protein in the eukaryotic cell, second only to actin, which has multiple cellular functions besides its well-recognised role in protein translation. These non-canonical roles of eEF1A in cellular functions include nucleocytoplasmic trafficking, protein degradation, apoptosis, heat shock response and actin remodelling [
5‐
7]. We previously confirmed that eEF1A is important in RSV replication, by demonstrating that downregulation of eEF1A using siRNAs in RSV-infected cells restricted the replication of viral genomic RNA and release of infectious virions [
8]. We further demonstrated that eEF1A interacts directly with the RSV N protein, which suggests that eEF1A can engage and stabilise the viral genome replication complex. A similar function has been reported for the HIV-1 reverse transcription complex [
9,
10].
One of the alternative functions of eEF1A is as an actin binding protein that can regulate actin stress fiber formation in a Rho/Rho kinase pathway dependent manner [
11]. RSV requires actin for virus replication, and F-actin is reported to be important for viral transcription and morphogenesis. This is dependent on cellular profilin activity, which regulates F-actin formation in cells [
12]. The seven subunit actin-related protein (ARP) 2/3 complex promotes the assembly of branched F-actin required for various cellular processes such as vesicle trafficking and plasma membrane protrusion during cell migration [
13]. More recently, the ARP2 was identified in a genome-wide siRNA screen as important for RSV gene expression [
14]. Further analysis showed that ARP2 specifically supported viral spread during filopodia formation that could shuttle virus from infected to uninfected cells [
14,
15]. Filopodia are slender cellular cytoplasmic projections that contain F-actin cross-linked into bundles by actin-binding proteins [
14,
16]. Given the association between eEF1A and actin, and the dependence of RSV on actin for virus replication, we investigated whether the reduced RSV replication by down-regulated eEF1A, as we have reported previously [
8] was, in part, due to the effect of change in actin filament structure, in addition to the effect on RSV genome replication.
In this study we aimed to investigate if an eEF1A actin-related role is important for RSV replication. Two approaches were used to investigate the association between eEF1A, cellular actin and RSV replication: downregulation of eEF1A expression using siRNA, and treatment with an eEF1A inhibitor, Did B. Did B, a cyclic depsipeptide produced by
Trididemnum solidum, binds to eEF1A and inhibits the eEF1A function in protein translation [
4,
17]. The results demonstrated that down-regulating eEF1A using siRNA disrupted actin stress fiber formation and reduced RSV egress. Similarly, treatment with Did B at low concentrations, which do not inhibit cellular protein translation or induce cytotoxic effects, led to similar disruption of actin stress fiber formation and reduced RSV production. This study demonstrates the important roles of eEF1A in regulation of actin organization, and in RSV infectious virus production.
Materials and methods
Cell culture and virus stock preparation
Human epithelial carcinoma cells (HEp-2, ATCC) cells were grown in Opti-MEM media (Gibco) supplemented with 5% heat-inactivated foetal bovine serum and 1% penicillin-streptomycin (10,000 U/ml) (Gibco), incubated at 37 °C with 5% CO2.
A stock of RSV A2 (ATCC) was propagated in HEp-2 cells and purified through a 30/60%
w/
v sucrose cushion as described previously [
18] and stored at − 80 °C until required. The titre of the resultant viral stock was quantified by standard immune-plaque assay. Briefly, virus suspension was diluted in a 10-fold series and used to infect HEp-2 cell monolayers, then incubated at 37 °C for two hours and overlaid with Opti-MEM /60% methyl cellulose/2% FBS/1% penicillin/streptomycin. After seven days incubation at 37 °C with 5% CO
2, monolayers were fixed with 60% methanol/40% acetone, blocked with 5% skim milk in PBS and probed with goat-anti RSV polyclonal antibody (Virostat). RSV positive plaques were visualised with HRP-conjugated secondary antibody (Life technologies) and DAB colour developer (Sigma-Aldrich). Viral titre was calculated as plaque forming units (pfu)/ml.
Treatment of HEp-2 cells with Did B
Didemnin B (Did B, kindly provided by the Natural Product Branch, NCI, USA) was dissolved in DMSO as a 10 mM stock and stored at − 80 °C. Confluent HEp-2 cells were treated with a range of concentrations from 0 nM to 16 nM. The same volume of DMSO was used as vehicle control. For effect of Did B on translation, HEp-2 cells were transfected with pCMV-Gluc2 plasmid and re-seeded into 24 well plates containing various concentration of Did B as described. The luciferase activity in culture supernatant was measured after 24 and 48 h of treatment using Biolux Gaussia luciferase Flex Assay kit (NEB). Subsequent experiments were conducted with a final concentration of 2.5 nM of Did B. Cell proliferation was monitored using CellTiter 96® Aqueous One Solution Cell Proliferation assay (Promega), as instructed by the manufacturer.
RSV infection
HEp-2 cells were infected with RSV A2 at a multiplicity of infection (MOI) of 1 pfu/cell at 37 °C for four hours. Inoculum was then removed and the cells were washed with 1x PBS prior to treatment with Did B. Cells were then incubated at 37 °C and sampled at 24 and 48 h post infection (p.i.) with RSV.
Downregulation of eEF1A using siRNA
siRNA targeting the eEF1A mRNA transcript (ID: SASI_Hs02_00331772 and SASI_Hs02_00331773, Sigma-Aldrich) or a universal negative control (siMM: SIC001) were transfected into HEp-2 cells using Lipofectamine RNAiMAX (Invitrogen, USA) according to the manufacturer’s protocol. The efficiencies of eEF1A down-regulation were examined at 24, 48, 72 and 96 h post-transfection (p.t.). For the effect of downregulation of eEF1A on RSV replication, RSV infection was performed at 48 h p.t..
Fluorescent staining and confocal microscopy
Cells either treated with Did B, or transfected with siRNAs or infected with RSV were washed with PBS, fixed with 4% formaldehyde and permeabilised with 3% BSA/1% Triton X-100 in PBS. The cells were then blocked with 5% BSA in PBS. Actin was stained with Phalloidin (Sigma Aldrich), eEF1A was identified using anti-eEF1A (Santa Cruz) and RSV was identified using an anti-N antibody (Virostat). Fluorescent images were captured using a Leica TCS SP2 confocal scanning microscope (Leica Microsystems) with 100 × objective lenses. All images shown are maximum intensity projections. ZEN Blue Black software was used to quantify pixel colocalization of actin and eEF1A at the cortex from z-stack confocal images using identical parameters. At least 12 cells treated with 2.5 mM Did B, DMSO or untreated were analysed. The colocalization coefficient was calculated as previously described [
19].
Lactase dehydrogenase (LDH) release assay
The cellular release of lactate dehydrogenase into supernatant was measured by Promega’s CytoTox-ONE™ assay, performed as per manufacture’s protocol, with the following changes: 50 μl of cell culture supernatant, or OptiMEM™ as blank control, was added to individual wells of a 96 well plate. The same volume 50 μl of CytoTox-ONE™ reagent was added to each sample and incubated at 22 °C for 10 min, the reaction stopped with 25 μl CytoTox-ONE™ stop solution and quantified using a Clariostar ELISA plate reader at 450 nm.
RT-qPCR for viral transcription and translation
Total RNA was extracted from cell pellets or 1 ml culture supernatant using TRIzol™ reagent (Life Technologies). The RNA was then reverse transcribed using Superscript III First-Strand Synthesis SuperMix™ kit (Life Technologies), as per manufactures protocol. Three separate reactions using different primers to generate cDNA were used for each sample; gene specific primers specific to RSV N mRNA, random hexamer or oligo-dT primers. Quantitative PCR was performed on the Qiagen RotorGene using primers specific for N cDNA or genomic cDNA of the RSV, and primers for cellular β-actin as an internal control as described previously [
8].
SDS-PAGE and Western blot analysis
Cell lysates were boiled in SDS-PAGE sample buffer and separated by 10% sodium dodecylsulfate – polyacrylamide gel electrophoresis (SDS-PAGE). Gels were electro-blotted onto a polyvinylidene fluoride (PVDF) membrane (Pall) using a semi-dry transfer system (Bio-Rad Laboratories). The membranes were blocked with 5% milk in PBS and probed with either anti-eEF1A antibody (Santa Cruz Biotechnology) or anti-β-tubulin antibody (Sigma Aldrich).
Statistical analysis
Statistical analysis were performed using an unpaired one way ANOVA test or a Student’s T-test as indicated. Statistical significance was set at: * = P < 0.05, ** = P < 0.01, *** = P < 0.001 and **** = P < 0.0001.
Discussion
Subversion of various host cell apparatus to promote viral replication and spread is a strategy used by all viruses. As the most abundant cellular protein, actin is a key player in many physiological processes [
5]. Its arrangement within the cytoskeleton is vital to a plethora of dynamic cellular and structural functions including shape, cytoplasmic transportation, stabilization of organelles, cell movement and intracellular interactions. Consequently, viruses have evolved strategies to undermine the host cell’s actin-regulated functions to enable virus trafficking and egress [
23]. Viruses in the family
Pneumoviridae, such as RSV, require assembly within the cytoplasm, and dispersal of RSV virions via budding of the plasma. These activities are achieved through actin dynamics [
11].
eEF1A is a major F-actin binding protein that has the ability to link F-actin into bundles forming structures such as stress fibers [
5]. Given its abundance in the cell and its function in stabilising F-actin, binding to and facilitating F-actin bundling into stress fibers, it is not surprising that many viruses have evolved mechanisms to redirect cellular eEF1A as cofactors in viral transcription, translation and assembly [
4,
7]. Previous studies have found eEF1A in highly purified fractions of a number of viruses, indicating that eEF1A may in fact be a critical component in the viral replication complex [
4,
24]. Former research into the RNA virus, tomato bushy stunt virus (TBSV), demonstrated that a chemical inhibitor of eEF1A, Did B, effectively inhibited replication of TBSV in vitro [
25].
Having previously demonstrated that RSV viral replication requires eEF1A dependant N-protein binding, we implicated eEF1A as a potential antiviral target for RSV [
8]. Our aim in this current study was to investigate whether Did B could also mitigate RSV replication in vitro. By binding eEF1A, Did B blocks eEF1A’s role in translation reportedly by inducing a conformational change that blocks its interaction with eEF2 [
17]. Our analysis has shown that treatment of HEp-2 cells with Did B at or below 4 nM does not significantly reduce the cell’s translation ability or viability. This is not entirely surprising as the estimated cellular concentrations of total eEF1A is high; approximately 40–60 μM, and ~ 80% of cellular eEF1A is bound to F-actin [
26], which could mitigate an effect of a 2.5 nM concentration of Did B on translation but still affect its ability to stabilise F-actin stress fibers. When we treated these HEp-2 cells with either sieEF1A or 2.5 nM of Did B, a distinct loss of structure and organisation of actin stress fibers is evident within the cytoplasm. eEF1A knockdown and Did B affected F-actin and eEF1A distribution where increased accumulation near the plasma membrane was observed. This localization suggested a redistribution of F-actin to the cortex in cells that were treated with sieEF1A or 2.5 nM Did B.
Actin is known to play a significant role in viral entry for many viruses, particularly those targeting epithelial cells such as RSV [
14,
27]. Actin is similarly essential for RSV egress, conventionally known to induce cell surface membrane budding in order to proliferate, which requires changes to the cytoskeleton [
28,
29]. We have shown that knock-down of eEF1A disrupts cellular actin function in HEp-2 cells, as did Did B. Confocal images of HEp-2 cells treated with a concentration of 2.5 nM Did B and inoculated with RSV-A2 at an MOI of 1, present accumulated peripheral RSV particles and far less extracellular virions compared to control treated, RSV inoculated cells (Fig.
2). This suggests that the ability of RSV to modify the host cytoskeleton for release of newly replicated progeny is disrupted when eEF1A function in stress fiber formation is inhibited.
Our images of control treated HEp-2 cells display long, actin extensions in RSV infected cells and many filaments stained with an anti-RSV antibody; whereas cells treated with Did B appeared to have fewer extensions. This is in accord with filament formation observed in RSV induced A549 cells, as well as HMPV induced human bronchial epithelial cells (BEAS-2B) [
14,
30]. These corresponding results indicate that pneumoviruses are capable of inducing actin-dependent cellular extensions in vitro in a range of cell lines.
We verified that cellular stress fibers are critical for RSV egress through a quantifiable reduction of extracellular RSV in HEp-2 cells treated with Did B. The number of infectious RSV particles released by Did B treated cells (shed virus) is considerably reduced at 24 h p.i., while the level of RSV genomic RNA in infected cells is similar to the cells with no treatment or DSMO treatment, indicating that the release of shed virus was affected in cells treated with Did B. The amount of cell death induced by RSV infection, quantified by LDH release, was significantly reduced in Did B treated cells at both 24 and 48 h p.i., compared to untreated cells. Taken together, these data suggests that by 48 h after inoculation with RSV, the quantity of viral particles released into the supernatant from Did B treated HEp-2 cells is reduced (Fig.
4), most likely as a consequence of the host cell’s inability to form actin structures formed through crosslinking of F-actin. The significantly elevated level of non-infectious RSV genomic RNA released from control cells (RSV infected, but not treated with Did B) at 48 h p.i., is likely due to more rapid cell death and the resultant release of a greater amount of RSV genomic RNA into the supernatant from necrotic cells. We propose that treatment with 2.5 nM of Did B, not only mitigates the release of RSV through budding or extracellular extension via disruption of actin bundling, but also delays cell death, in turn delaying the release of infectious RSV and reducing spread of progeny. Treatment with Did B at 2.5 nM did not affect transcription of RSV nucleocapsid mRNA or genomic RNA replication significantly compared to the control treatment up to 48 h p.i..
In light of our results, selective inhibitor of actin crosslinking by eEF1A could be a part of an antiviral strategy for pneumoviruses. Although complete inhibition of RSV was not observed in this study, reduced RSV-induced cell death was a noted consequence of actin stress fiber dysregulation in RSV infected cells. This Did B effect would be worth testing with other viral species.