Protein Synthesis Shut-Off Induced by Influenza Virus Infection Is Independent of PKR Activity

Abstract

The role of PKR activity in influenza virus-induced cell shut-off was studied by infection of PKR⁺ or PKR⁻ cell cultures and metabolic labeling in vivo. No differences in the synthesis of viral proteins or the decay of cellular protein synthesis were observed. To investigate the relevance of the inhibition of cellular pre-mRNA polyadenylation and nucleocytoplasmic transport in virus-induced shut-off, we carried out similar experiments with mutant viruses lacking C-terminal sequences of NS1 protein. No differences in the shut-off induced by mutant versus wild-type viruses were observed, indicating that these nuclear events are not relevant for shut-off. The analysis of cytoplasmic mRNA stability indicated that the accumulation of viral mRNA during the infection correlated with the progressive decay of cellular mRNA, in both the wild type and an NS1 deletion mutant.

The influenza virus infection of permissive cells leads to a progressive decline in the synthesis of cellular proteins, a phenomenon known as cell shut-off. Several steps in the gene expression program of the cell are altered by influenza virus infection. Thus, the transport of cellular mRNA from the nucleus to the cytoplasm is inhibited (18). This result was first interpreted as a consequence of the cap-snatching activity of the viral polymerase in the nucleus (18), and it was later explained as the outcome of the inhibition of hnRNA polyadenylation mediated by NS1 protein (24, 27). In influenza virus infection, the translation machinery is utilized essentially to produce viral proteins. Several observations may be relevant to explain this fact. The cellular mRNAs present in the cytoplasm before the infection are degraded during the infection cycle (2, 16) (see below). Translation of cellular mRNAs is inhibited at both the initiation and elongation steps (17), and viral mRNAs are preferentially translated (12, 13), presumably as a consequence of the activity of NS1 protein (6, 7, 22, 26).

The protein kinase PKR is one of the effectors of the interferon (IFN) response. Its expression is activated by IFN, and its activity is induced by double-stranded RNA (dsRNA). Activation of PKR involves its autophosphorylation and leads to the phosphorylation of the α subunit of eukaryotic initiation factor 2 (eIF-2α) and the subsequent inhibition of protein synthesis (for reviews, see references 5 and 10). It has been argued that influenza virus mRNAs are intrinsically “better translators” than cellular ones, and the observed effect of NS1 in this regard might be simply the consequence of blocking the activity of PKR (21), since NS1 protein is also a dsRNA-binding protein. To directly test this proposal and to evaluate the relevance of the above-mentioned alterations of cell metabolism in the induction of cell shut-off, we have studied the expression of cellular and viral genes in cells with a PKR gene knockout mutation and hence devoid of PKR activity.

Cultures of NIH 3T3 cells (PKR⁺) or the corresponding cells mutated at the pkr gene (PKR⁻) (28) were infected with influenza virus (WSN strain) at a multiplicity of infection of 5 to 10 PFU per cell and labeled by incorporation of [³⁵S]Met-Cys for 1 h at various times after infection. Total cell extracts were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and the proteins were visualized by autoradiography. The results are presented in Fig. 1. No obvious differences in the amounts of the viral proteins synthesized during the infection in PKR⁺ and PKR⁻ cells were observed (see bands labeled NP, M1, NS1, and NS2 in Fig. 1A). Likewise, the synthesis of cellular proteins was inhibited progressively during the infection to an extent and with a kinetics similar to those in both PKR⁺ and PKR⁻ cells (see bands indicated by stars in Fig. 1A and their quantification in Fig. 1B). To verify that the PKR⁻ cells are devoid of PKR protein, we carried out Western blot analyses with a PKR-specific antiserum and extracts obtained from either PKR⁺ or PKR⁻ cells. The results are presented in Fig. 1C and confirm the correct phenotype of either culture. These results indicate that whatever mechanism is responsible for the induction of cell shut-off in influenza virus-infected cells, it is independent of the activity of PKR.

FIG. 1.

Effect of PKR expression on influenza virus-induced cell shut-off. (A) Cultures of PKR⁺ or PKR⁻ NIH 3T3 cells were infected with the WSN strain of influenza virus at a multiplicity of infection of 5 to 10 PFU per cell. At the times indicated at the top of each lane, the cultures were labeled with [³⁵S]Met-Cys (200 μCi/ml) for 1 h and processed by polyacrylamide gel electrophoresis and autoradiography. A selection of the virus-encoded proteins is indicated to the right of each panel. The stars point to prominent cellular proteins whose synthesis declined during the infection. (B) The bands indicated by stars in panel A were quantified by microdensitometry. The results have been normalized with respect to the initial pulse period. (C) Total cell extracts were prepared from cultures of PKR⁺ or PKR⁻ NIH 3T3 cells and analyzed by Western blotting with anti-PKR serum.

As indicated above, the expression of NS1 protein leads to the inhibition of cellular pre-mRNA cleavage and polyadenylation by interaction with the 30-kDa subunit of CPSF (24) and to the accumulation of pre-mRNA in the nucleus (27). In addition, NS1 protein inhibits poly(A) elongation of mRNAs in the nucleus and their export as a consequence of its interaction with poly(A)-binding protein II (4). To investigate whether such alterations play a role in the influenza virus-induced shut-off, we took advantage of recently prepared mutant viruses that express versions of NS1 protein containing deletions. These viruses express short NS1 proteins containing either the 81 or 156 N-terminal amino acids, and therefore they miss the C-terminal effector domain entirely (NS1-81) or partially (NS1-156). These two mutants grow in single-cycle infections as efficiently as wild-type virus, although the expression of late viral proteins is reduced due to a low accumulation of late mRNAs. Their generation and phenotype will be described in detail elsewhere. Since the interaction of NS1 protein with the 30-kDa subunit of CPSF takes place through the C-terminal sequences of the former (24), if such interaction is relevant for the inhibition of cellular protein synthesis, then the C-terminally deleted NS1 proteins would be predicted to show an affected shut-off phenotype. Cultures of PKR⁺ or PKR⁻ NIH 3T3 cells were infected with NS1-81 or NS1-156 virus, and the infected cells were labeled with [³⁵S]-Met-Cys as indicated above. The patterns of labeled proteins obtained are shown in Fig. 2. As expected, wild-type NS1 protein is absent (compare with Fig. 1A), and mutant NS1-81 or NS1-156 proteins are expressed. Their identity was confirmed by Western blot analysis with anti-NS1 serum (data not shown). Despite lacking the C-terminal domain of NS1 protein, mutant viruses NS1-81 and NS1-156 induced a shut-off similar to that observed after wild-type influenza virus infection (compare with Fig. 1A), as determined by the diminished labeling in the extracts obtained at late times after infection and the disappearance of prominent cellular bands, indicated with stars in Fig. 2A, whose quantification is presented in Fig. 2B. Therefore, we can conclude that the interference with cell pre-mRNA polyadenylation and nucleocytoplasmic transport induced by influenza virus infection plays no role in virus-induced shut-off. This is in line with the fact that most of the cell mRNAs engaged in protein synthesis during the infection have been synthesized and transported to the cytoplasm prior to the infection and would not be affected by viral interference with nuclear events in gene expression.

FIG. 2.

Cell shut-off is not affected by C-terminal deletions of NS1 protein. (A) Cultures of PKR⁺ or PKR⁻ NIH 3T3 cells were infected with the WSN mutant NS1-81 or NS1-156 at a multiplicity of infection of 5 to 10 PFU per cell. The cultures were labeled and processed as indicated in the legend to Fig. 1. The numbers at the top of each lane indicate the time (in hours) after infection at which the cultures were labeled. The main influenza virus proteins are indicated to the right. The stars point to prominent cellular proteins whose synthesis declined during the infection. (B) The bands indicated with stars in panel A were quantified by microdensitometry. The results have been normalized with respect to the initial pulse period.

In view of these results, we examined the integrity of the cellular mRNAs present in the cytoplasm of cells infected with either wild-type influenza virus or the NS1-81 mutant. Cultures of COS cells were infected with either virus, and at various times after infection, total cytoplasmic RNA was isolated and analyzed by Northern blotting with an NP or a β-tubulin cDNA probe, as previously described (2). Along with the progressive accumulation of viral mRNAs, as depicted by NP mRNA (Fig. 3A), the accumulation of β-tubulin mRNA was gradually decreased during the infection (Fig. 3A), both in wild-type influenza virus-infected cells and in cells infected with mutant NS1-81. This is in contrast to the high stability of β-tubulin mRNA in uninfected cells, measured by a classical actinomycin D chase (2) (Fig. 3A). The quantification of these results is presented in Fig. 3B and confirms previous reports that documented the instability of cytoplasmic cellular mRNA after influenza virus infection (2, 16). Furthermore, the results obtained with NS1-81 mutant virus indicate that the C-terminal domain of NS1 protein does not play a role in such a phenomenon.

FIG. 3.

Stability of cellular mRNA in cells infected with wild-type (wt) or NS1-81 influenza viruses. (A) Cultures of COS-1 cells were infected with either wild-type or NS1-81 viruses as indicated in the legend to Fig. 1 or left uninfected. Some of the uninfected cultures were treated with actinomycin D (100 μg/ml) (Act-D). At the times indicated at the top of each lane, the cultures were fractionated into nucleus and cytoplasm as described previously (25). Total cytoplasmic RNA was extracted and analyzed by Northern blotting with either NP or β-tubulin probes labelled by random priming of the corresponding cDNAs, as described previously (2). (B) The intensity of the bands was determined by microdensitometry and is represented as normalized with respect to the uninfected cell extract (for β-tubulin [β-Tub]) or the extract obtained at 8 h postinfection (for NP).

It has been proposed that influenza virus, like many other viruses, has evolved mechanisms to avoid the activation of cellular PKR and hence a total block of protein synthesis (19). These would include the induction of expression of a cellular protein, p58, which inhibits the autophosphorylation and activity of PKR (20), and the expression of NS1 protein, a viral protein that has dsRNA-binding activity (14) and would sequester dsRNA during the infection (15, 21). In fact, it has been elegantly demonstrated that NS1 protein plays an essential role in the anti-IFN response mediated by influenza virus, because an NS1-null mutant virus can replicate in cellular systems deficient in the IFN pathway, but not in normal cells (11). Moreover, the anti-IFN action of NS1 protein is directed to overcome the PKR response, since the NS1-null mutant virus can replicate in PKR knockout mice (3). In addition to such a blockade of PKR activation and/or activity, influenza virus takes over the cellular protein synthesis machinery in such a way that at late times in the infection, essentially only viral mRNAs are translated. We have addressed the possible role of PKR activity in the latter issue by single-cycle infections in normal versus PKR-defective cells. Our results demonstrate that PKR activity is not relevant for the preferential translation of viral mRNAs in the infection (Fig. 1). Furthermore, the use of C-terminally-deleted NS1 mutants indicates that nuclear events in the cellular gene expression pathway do not play a role in virus-induced shut-off (Fig. 2). Although it is clear that cytoplasmic cellular mRNAs are degraded during the infection (2, 16) (Fig. 3), the interpretation of these results is not clear at present. In influenza virus-infected cells, viral mRNAs are translated preferentially over cellular ones, a fact that may be related to alterations in the cell translation apparatus (9) and the activity of NS1 protein (6, 7) through its association with the human homologue of Staufen protein (8, 23) and the eIF-4GI component of eIF-4F (1). It is conceivable that such inhibition of cellular mRNA translation would induce their destabilization, but we cannot exclude that virus infection induces the selective degradation of cellular mRNAs and, as a consequence, the inhibition of translation of cellular proteins.