Nonstructural proteins of respiratory syncytial virus suppress premature apoptosis by an NF-kappaB-dependent, interferon-independent mechanism and facilitate virus growth

Abstract

The two nonstructural (NS) proteins NS1 and NS2 of respiratory syncytial virus (RSV) are abundantly expressed in the infected cell but are not packaged in mature progeny virions. We found that both proteins were expressed early in infection, whereas the infected cells underwent apoptosis much later. Coincident with NS protein expression, a number of cellular antiapoptotic factors were expressed or activated at early stages, which included NF-kappaB and phosphorylated forms of protein kinases AKT, phosphoinositide-dependent protein kinase, and glycogen synthase kinase. Using specific short interfering RNAs (siRNAs), we achieved significant knockdown of one or both NS proteins in the infected cell, which resulted in abrogation of the antiapoptotic functions and led to early apoptosis. NS-dependent suppression of apoptosis was observed in Vero cells that are naturally devoid of type I interferons (IFN). The siRNA-based results were confirmed by the use of NS-deleted RSV mutants. Early activation of epidermal growth factor receptor (EGFR) in the RSV-infected cell did not require NS proteins. Premature apoptosis triggered by the loss of NS or by apoptosis-promoting drugs caused a severe reduction of RSV growth. Finally, recombinantly expressed NS1 and NS2, individually and together, reduced apoptosis by tumor necrosis factor alpha, suggesting an intrinsic antiapoptotic property of both. We conclude that the early-expressed nonstructural proteins of RSV boost viral replication by delaying the apoptosis of the infected cell via a novel IFN- and EGFR-independent pathway.

FIG. 1.

Kinetics of NS expression and apoptosis. (A) Total protein of RSV-infected A549 cells at the indicated times postinfection was probed in immunoblot with sera against viral NS proteins and F. C, sham-infected cells at 48 h. Note the early expression of NS1 and NS2. (B) At the same time points, apoptotic DNA fragmentation was quantitated by cell death detection ELISA (Materials and Methods). White bars, sham infected; gray bars, RSV infected. Most data points are averages of three experiments with errors shown as bars.

FIG. 2.

Effect of anti-NS siRNA on NS proteins and apoptosis. The design and transfection of anti-NS siRNAs have been described in Materials and Methods. (A) Immunoblot and reverse transcription-PCR analyses show siRNA-dependent reduction of NS proteins (P) and RNA (R), respectively, measured at 18 hpi. (B) Immunoblot showing no effect of scrambled siRNA. U, uninfected cells, treated with 20 nM scrambled siRNA. (C) DNA fragmentation in these cells was assayed as described in the legend to Fig. 1. Cells were transfected with anti-NS1 (white bars) or anti-NS2 (gray bars) siRNA 6 h prior to the addition of the virus. Each data point represents an average of three experiments with error bars (not shown for small values). C, sham-infected cells at 60 h.

FIG. 3.

Triparametric staining of A549 cells. All nuclei were stained by 4′,6′-diamidino-2-phenylindole (DAPI) (blue), apoptotic cells were stained by TUNEL assay (green), and RSV-infected cells were stained by indirect immunofluorescence (IF) using a mixture of antibodies against RSV proteins (red) at 12 or 30 hpi as described in Materials and Methods. U, sham-infected cells. Panels A, B, and C show RSV-infected cells (MOI = 3) transfected, respectively, with no siRNA, 80 nM anti-NS1 siRNA, and 20 nM anti-NS2 siRNA. The two bottom panels show low-MOI infections and apoptosis of a proportionately smaller number of cells.

FIG. 4.

Prosurvival pathways in RSV-infected A549 cells. (A) Phosphorylated PDK, AKT, and GSK were assayed by immunoblot at the indicated times after infection (or sham infected [U]) with or without transfection by anti-NS siRNAs. Control cells received 50 nM of scrambled (SC) siRNA. (B) Similar immunoblot assay of phosphorylated IκBα in RSV-infected (or sham-infected [U]) cells at 5 hpi. (C) Assay of NF-κB activation by reporter luciferase assay in HEK293 cells treated with the indicated concentrations of anti-NS siRNAs (or 50 nM of scrambled siRNA [SC]) at 18 hpi as described before (9). Luciferase activity in RSV-infected HEK293 cells not treated with siRNA was taken as 100, and the other values were expressed as a percentage of this. In both panels A and B, immunoblots of the unphosphorylated proteins and Sp1 serve as unchanged controls.

FIG. 5.

Triparametric staining of Vero cells. The procedure is essentially identical to that described in the legend to Fig. 3, except that Vero cells, deleted of type I IFN genes, were used instead of A549 cells. Vero cells, with or without siRNA pretreatment, were infected with RSV and processed for staining at indicated times postinfection (12 or 30 h). The two bottom panels (−RSV) show “control” uninfected cells either untreated (no siRNA) or treated (+ NS siRNA) with a mixture of 80 nM anti-NS1 plus 20 nM anti-NS2 siRNA and stained 40 h later; note the lack of apoptosis in both panels. DAPI, 4′,6′-diamidino-2-phenylindole.

FIG. 6.

Measurement of RSV growth. (A) Intracellular replication of recombinant RSV expressing green fluorescent protein (rgRSV) (18) in A549 cells. When used, all inhibitors were added to the culture 2 h prior to infection and maintained throughout infection. The various treatments were the following: (a) pan-caspase inhibitor (Z-VAD-FMK); (b) caspase-3 inhibitor (Ac-DEVD-CHO); (c) none; (d) 80 nM anti-NS1 siRNA; (e) 20 nM anti-NS2 siRNA; (f) 25 μM PI3K inhibitor (LY294002); and (g) 15 μM NF-κB inhibitor (SN50). All inhibitors were from Calbiochem. (B) Specific reduction of progeny RSV by anti-NS siRNA. A549 cells were infected with RSV or human parainfluenza virus type 3, and infectious progeny virions in the culture media at 72 hpi were titrated as described in Materials and Methods. Cells were treated with the indicated concentrations of siRNAs (written above the bars) 6 h prior to the addition of the virus. Results are averages of three experiments with error bars shown.

FIG. 7.

Early apoptosis in A549 and NHBE cells infected with NS-deleted (ΔNS) RSV. Apoptosis was quantified in A549 (A) and NHBE (B) cells at 12 h after infection with wild-type and NS-deleted RSV using an ELISA-based assay as described in Materials and Methods. All values were expressed as fold of the wild-type RSV value taken as 1. Data with siRNA are added for comparison. (C) The phosphorylated form and total protein of the indicated prosurvival factors were detected using specific antibodies (Materials and Methods). A trial immunoblot was first performed to check the relative band intensities. Based on that result, the volume of each sample was recalculated to run a second gel such that the RSV N protein is present in equal amounts in all lanes, and this result is presented. The lowest panel confirms the loss of the corresponding NS protein in the deletions. Note the activation of phosphorylation of all three proteins in wild-type RSV infection (confirming the results shown in Fig. 4) and its inhibition in the ΔNS mutants. (D) Immunoblot detection of unphosphorylated (subpanel a) and phosphorylated (subpanel b) EGFR in A549 cells infected with wild-type RSV (lanes W) or ΔNS1 (lanes 1), ΔNS2 (lanes 2), and ΔNS1,2 (lanes 12) mutant and harvested at 1, 6, 18, or 24 h postinfection. U represents uninfected control cells. A portion of the membrane, stained with Ponceau S, is shown in subpanel c to document similar total protein loads. Note the complete switch of the unphosphorylated EGFR to the phosphorylated form between 6 and 18 h of RSV infection. WT, wild type.

FIG. 8.

Antiapoptotic properties of recombinant NS proteins. The coding sequences and the 5′-untranslated region of the NS mRNAs were cloned in pcDNA3 (Invitrogen, Carlsbad, CA) between the unique HindIII and BamHI sites. Indicated amounts (in micrograms on the x axis) of the plasmid DNAs were transfected into A549 cells in 6-well plates using Lipofectin (Invitrogen). Control cells received 4 μg of vector with no insert. At 24 h posttransfection, TNF-α (R&D Systems, Minneapolis, MN) was added to a final concentration of 60 ng/ml, and cells were harvested 6 h later for ELISA-based apoptosis assay (upper panel, NS2, speckled bars; NS1, striped bars; both, gray bars) and immunoblot analysis (lower panels). Apoptosis by TNF-α alone was taken as 100, and other values are expressed as its percentage. Results with >4 μg plasmid were not presented due to visible cytopathic effect and cell loss. The immunoblot for actin (8) shows equal protein loading.

FIG. 9.

Working model for NS action. The model is based on past and present results as detailed in Discussion. In brief, stimulation of RSV growth may result from (i) subversion of IFN signaling by NS proteins and (ii and iii) suppression of premature apoptosis by the NS-dependent PI3K→AKT→NF-κB pathway and NS-independent EGFR→ERK pathway. These pathways are incomplete, and multiple crossovers are possible but not shown to avoid complexity. Notable among them are sphingosine kinase, which mediates the activation of ERK and AKT (30), and multiple protein kinase C isoforms that play an important role in activating ERK and NF-κB (5, 29).