On the discovery of interferon-inducible, double-stranded RNA activated enzymes: the 2'-5'oligoadenylate synthetases and the protein kinase PKR

The demonstration that double-stranded (ds) RNA inhibits protein synthesis in cell-free systems prepared from interferon-treated cells, lead to the discovery of the two interferon-induced, dsRNA-dependent enzymes: the serine/threonine protein kinase that is referred to as PKR and the 2′,5′-oligoadenylate synthetase (2′,5′-OAS), which converts ATP to 2′,5′-linked oligoadenylates with the unusual 2′-5′ instead of 3′-5′ phosphodiesterase bond. We raised monoclonal and polyclonal antibodies against human PKR and the two larger forms of the 2′,5′-OAS. Such specific antibodies proved to be indispensable for the detailed characterization of these enzyme and the cloning of cDNAs corresponding to the human PKR and the 69–71 and 100 kDa forms of the 2′,5′-OAS. When activated by dsRNA, PKR becomes autophosphorylated and catalyzes phosphorylation of the protein synthesis initiation factor eIF2, whereas the 2′-5′OAS forms 2′,5′-oligoadenylates that activate the latent endoribonuclease, the RNAse L. By inhibiting initiation of protein synthesis or by degrading RNA, these enzymes play key roles in two independent pathways that regulate overall protein synthesis and the mechanism of the antiviral action of interferon. In addition, these enzymes are now shown to regulate other cellular events, such as gene induction, normal control of cell growth, differentiation and apoptosis.     

Inhibition of HeLa cell protein synthesis following poliovirus infection correlates with the proteolysis of a 220,000-dalton polypeptide associated with eucaryotic initiation factor 3 and a cap binding protein complex

Following poliovirus infection of HeLa cells, the synthesis of cellular proteins is inhibited but translation of poliovirus mRNA proceeds. The defect in the recognition of host cell mRNA may be due to a change in a cap recognition complex which, when added to an infected cell lysate, restores the ability to translate capped mRNAs. We employed immunoblotting techniques to examine initiation factors in crude lysates from uninfected and poliovirus-infected HeLa cells. Using an antiserum against eucaryotic initiation factor 3, we detected an antigen of approximate molecular weight 220,000 in uninfected cell lysates but not in infected cell lysates. Antigenically related polypeptides of 100,000 to 130,000 daltons, presumably degradation products, were detected in the infected cell lysate. The time course for degradation of the 220,000-dalton polypeptide correlates with that for inhibition of cellular protein synthesis in vivo. A portion of the population of 220,000-dalton polypeptides apparently associates with initiation factor eIF3 but is readily dissociated in buffers containing high salt. Affinity-purified antibodies against the polypeptide recognize a protein of the same size in a purified preparation of a cap binding protein complex obtained by cap-affinity chromatography. We postulate that the 220,000-dalton polypeptide is an essential component of the cap recognition complex and that its degradation in poliovirus-infected cells results in the inhibition of host cell translation. These results are in the first demonstration of a specific structural defect in an initiation factor resulting from poliovirus infection.     

Monensin and nigericin prevent the inhibition of host translation by poliovirus, without affecting p220 cleavage.

Addition of monensin or nigericin after poliovirus entry into HeLa cells prevents the inhibition of host protein synthesis by poliovirus. The infected cells continue to synthesize cellular proteins at control levels for at least 8 h after infection in the presence of the ionophore. Cleavage of p220 (gamma subunit of eukaryotic initiation factor 4 [eIF-4 gamma]), a component of the translation initiation factor eIF-4F, occurs to the same extent in poliovirus-infected cells whether or not they are treated with monensin. Two hours after infection there is no detectable intact p220, but the cells continue to translate cellular mRNAs for several hours at levels similar to those in uninfected cells. Nigericin or monensin prevented the arrest of host translation at all the multiplicities of poliovirus infection tested. At high multiplicities of infection, an unprecedented situation was found: cells synthesized poliovirus and cellular proteins simultaneously. Superinfection of vesicular stomatitis virus-infected HeLa cells with poliovirus led to a profound inhibition of vesicular stomatitis virus protein synthesis, while nigericin partially prevented this blockade. Drastic inhibition of translation also took place in influenza virus-infected Vero cells treated with nigericin and infected with poliovirus. These findings suggest that the translation of newly synthesized mRNAs is dependent on the integrity of p220, while ongoing cellular protein synthesis does not require an intact p220. The target of ionophore action during the poliovirus life cycle was also investigated. Addition of nigericin at any time postinfection profoundly blocked the synthesis of virus RNA, whereas viral protein synthesis was not affected if nigericin was added at 4 h postinfection. These results agree well with previous findings indicating that inhibitors of phospholipid synthesis or vesicular traffic interfere with poliovirus genome replication. Therefore, the action of nigericin on the vesicular system may affect poliovirus RNA synthesis. In conclusion, monensin and nigericin are potent inhibitors of poliovirus genome replication that prevent the shutoff of host translation by poliovirus while still permitting cleavage of p220.     

Isolation and structural characterization of cap-binding proteins from poliovirus-infected HeLa cells.

In poliovirus-infected HeLa cells, poliovirus RNA is translated at times when cellular mRNA translation is strongly inhibited. It is thought that this translational control mechanism is mediated by inactivation of a cap-binding protein complex (comprising polypeptides of 24 [24-kilodalton cap-binding protein], 50, and approximately 220 kilodaltons). This complex can restore the translation of capped mRNAs in extracts from poliovirus-infected cells. We have previously shown that the virally induced defect prevents interaction between cap recognition factors and mRNA. Here, we show that the cap-binding protein complex (and not the 24-kilodalton cap-binding protein) has activity that restores the cap-specific mRNA-protein interaction when added to initiation factors from poliovirus-infected cells. Thus, the activity that restores the cap-specific mRNA-protein interaction and that which restores the translation of capped mRNAs in extracts from poliovirus-infected cells, copurify. The results also indicate, by an alternative assay, that the cap-binding protein complex is the only factor inactivated by poliovirus. We also purified cap-binding proteins from uninfected and poliovirus-infected HeLa cells. By various criteria, the 24-kilodalton cap-binding protein is not structurally modified as a result of infection. However, the 220-kilodalton polypeptide of the cap-binding protein complex is apparently cleaved by a putative viral (or induced) protease. By in vivo labeling and m7GDP affinity chromatography, we isolated a modified cap-binding protein complex from poliovirus-infected cells, containing proteolytic cleavage fragments of the 220-kilodalton polypeptide.     

Poliovirus mutant that does not selectively inhibit host cell protein synthesis.

A poliovirus type I (Mahoney strain) mutant was obtained by inserting three base pairs into an infectious cDNA clone. The extra amino acid encoded by the insertion was in the amino-terminal (protein 8) portion of the P2 segment of the polyprotein. The mutant virus makes small plaques on HeLa and monkey kidney (CV-1) cells at all temperatures. It lost the ability to mediate the selective inhibition of host cell translation which ordinarily occurs in the first few hours after infection. As an apparent consequence, the mutant synthesizes far less protein than does wild-type virus. In mutant-infected CV-1 cells enough protein was produced to permit a normal course of RNA replication, but the yield of progeny virus was very low. In mutant-infected HeLa cells there was a premature cessation of both cellular and viral protein synthesis followed by a premature halt of viral RNA synthesis. This nonspecific translational inhibition was distinguishable from wild-type-mediated inhibition and did not appear to be part of an interferon or heat shock response. Because the mutant is recessive, our results imply that (at least in HeLa cells) wild-type poliovirus not only actively inhibits translation of cellular mRNAs, but also avoids early inhibition of its own protein synthesis. Cleavage of the cap-binding complex protein P220, which has been associated with the selective inhibition of capped mRNA translation, did not occur in mutant-infected cells. This result supports the hypothesis that cleavage of P220 plays an important role in normal poliovirus-mediated translational inhibition.     

The cap-binding protein complex in uninfected and poliovirus-infected HeLa cells

In poliovirus-infected HeLa cells, the mechanism of protein synthesis initiation factor recognition of m7G cap groups on mRNA is impaired. Translation of capped host cell mRNAs is inhibited, whereas translation of uncapped poliovirus mRNA proceeds exclusively. The site of this defect has been localized to the cap-binding protein complex (CBPC). To elucidate the specific structural and functional defects of the CBPC following poliovirus infection, the CBPC and/or its polypeptide components were purified from uninfected and poliovirus-infected HeLa cells. The CBPC from uninfected cells consisted of tightly associated 24- and 220-kDa polypeptides; minor amounts of polypeptides of 40, 44, and 80 kDa also consistently co-purified with the p24/p220 cores. No evidence of a 50-kDa, eIF-4A-related polypeptide subunit of the CBPC was obtained. The CBPC from poliovirus-infected cells had undergone major structural alterations. The 220-kDa component was absent; antigenically related (100-130 kDa) degradation products were present instead. The 24-kDa component co-purified with the p220 degradation products, but other components were missing. The association of the infected cell CBPC components was quite labile compared with that demonstrated by the components of CBPC from uninfected cells. Differential stimulation of capped, but not uncapped mRNAs in a cell-free translation assay was demonstrated by unmodified CBPC. Conversely, modified CBPC from poliovirus-infected cells differentially stimulated in vitro translation of uncapped poliovirus mRNA but not capped mRNAs. The implications of these results for the mechanism of cap-independent translation are briefly discussed.     

Functional characterization of eukaryotic mRNA cap binding protein complex: effects on translation of capped and naturally uncapped RNAs

We examined the effects of a eukaryotic mRNA cap binding protein (CBP) complex purified by cap analogue affinity chromatography [Edery, I., Humebelin, M., Darveau, A., Lee, K.A. W., Milburn, S., Hershey, J.W.B., Trachsel, H., & Sonenberg, N. (1983) J. Biol. Chem. 258, 11398 11403], on translation of several capped and naturally uncapped mRNAs in extracts prepared from poliovirus-infected or mock-infected HeLa cells. The CBP complex has activity that restores capped mRNA (globin, tobacco mosaic virus, and others) function in extracts from poliovirus-infected HeLa cells. Translation of two naturally uncapped RNAs (poliovirus and mengovirus RNAs), the translation of which is not restricted in extracts from poliovirus-infected cells, is also not stimulated by the CBP complex. Translation of several capped eukaryotic mRNAs (vesicular stomatitis virus, reovirus, and tobacco mosaic virus) in extracts from mock-infected cells is inhibited when the potassium ion concentration is increased. However, translation of capped AMV-4 RNA, which has negligible secondary structure at its 5' end, is resistant to this inhibition. Furthermore, the CBP complex reverses the high salt induced inhibition of translation of the former mRNAs. Since mRNA secondary structure is more stable at elevated salt concentrations, these data are consistent with a model in which the CBP complex has a role in melting mRNA secondary structure involving 5'-proximal sequences, to facilitate ribosome binding.     

Association of cap-binding protein with eucaryotic initiation factor 3 in initiation factor preparations from uninfected and poliovirus-infected HeLa cells.

Extracts from poliovirus-infected HeLa cells are unable to translate vesicular stomatitis virus or cellular mRNAs in vitro, probably reflecting the poliovirus-induced inhibition of host cell protein synthesis which occurs in vivo. Crude initiation factors from uninfected HeLa cells are able to restore translation of vesicular stomatitis virus mRNA in infected cell lysates. This restoring activity separates into the 0 to 40% ammonium sulfate fractional precipitate of ribosomal salt wash. Restoring activity is completely lacking in the analogous fractions prepared from poliovirus-infected cells. The 0 to 40% ammonium sulfate precipitates from both uninfected and infected cells contain eucaryotic initiation factor 3 (eIF-3), eIf-4B, and the cap-binding protein (CBP), which is detected by means of a cross-linking assay, as well as other proteins. The association of eIF-3 and cap binding protein was examined. The 0 to 40% ammonium sulfate precipitate of ribosomal salt wash from uninfected and infected cells was sedimented in sucrose gradients. Each fraction was examined for the presence of eIF-3 antigens by an antibody blot technique and for the presence of the CBP by cross-linking to cap-labeled mRNAs. From uninfected cells, a major proportion of the CBP cosedimented with eIF-3; however, none of the CBP from infected cells sedimented with eIF-3. The results suggest that the association of the CBP with eIF-3 into a functional complex may have been disrupted during the course of poliovirus infection.     

Structural difference between the 5'' termini of viral and cellular mRNA in poliovirus-infected cells: possible basis for the inhibition of host protein synthesis.

Host protein synthesis in poliovirus-infected HeLa cells is interrupted, but the host mRNA appears to remain completely intact and unmodified. The average size and poly (A) content of host mRNA was previously known to be unchanged (Koschel, 1974; Leibowitz and Penman, 1971), and this was confirmed. In addition, the 5' terminal methylated "cap" structures remained intact, and no further base modifications at the level of 1 base in 1,000 could be detected. Poliovirus RNA from viruses was previously shown not to have "caps" (Wimmer, 1972), and in this work poliovirus RNA from polyribosomes was found to have pUp at its 5' end. Since, initiation of protein synthesis is probably the basis for the inhibition of cellular protein synthesis in infected cells, the difference in the 5' ends of the host cell and viral RNA could be the basis of selective translation of viral RNA during infection.     

Protein synthesis in extracts from interferon-treated Mengo virus infected cells

We previously showed in intact L cells that interferon treatment did not modify the shut-off of cellular RNA and protein synthesis induced by infection with Mengo virus although viral replication is inhibited (1,2). We have also demonstrated that inhibition of host protein synthesis was not due to degradation of messengers since cellular mRNA could be extracted from interferon-treated infected cells and efficiently translated in a reticulocyte lysate(2). Cellular mRNA was not degraded although 2–5A was present as reported here. We prepared cell-free systems from such cells at a time when cellular shut-off was fully established. The undegraded messengers remained untranslated under cell-free protein synthesis conditions and almost no polysomes were detected. The decreased amount of [³⁵S]Met-tRNA-40S complex observed in these lysates might account for the inhibition of protein synthesis at the level of initiation.     

The 3′-untranslated region of brome mosaic virus RNA does not enhancetranslation of capped mRNAs in vitro

The translation enhancing ability of cis-acting 3′-terminal untranslated region (3′-UTR) of brome mosaic virus (BMV) was examined. Two chimeric mRNA constructs translated in rabbit reticulocyte lysates contained the BMV coat protein (CP) gene and NPTI gene, respectively. It was shown that the 3′-UTR of BMV RNA enhanced the translational efficiency of uncapped but not capped messages.     

Capped mRNAs with reduced secondary structure can function in extracts from poliovirus-infected cells

Extracts from poliovirus-infected HeLa cells were used to study ribosome binding of native and denatured reovirus mRNAs and translation of capped mRNAs with different degrees of secondary structure. Here, we demonstrate that ribosomes in extracts from poliovirus-infected cells could form initiation complexes with denatured reovirus mRNA, in contrast to their inability to bind native reovirus mRNA. Furthermore, the capped alfalfa mosaic virus 4 RNA, which is most probably devoid of stable secondary structure at its 5' end, could be translated at much higher efficiency than could other capped mRNAs in extracts from poliovirus-infected cells.     

Absence of non-specific effects of RNA interference triggered by long double-stranded RNA in mouse oocytes

RNA interference (RNAi) is a conserved eukaryotic mechanism by which double-stranded RNA (dsRNA) triggers the sequence-specific degradation of homologous mRNAs. Recent concerns have arisen in mammalian systems about off-target effects of RNAi, as well as an interferon response. Most mammalian cells respond to long dsRNAs by inducing an antiviral response mediated by interferon that leads to general inhibition of protein synthesis and nonspecific degradation of mRNAs. Moreover, recent reports demonstrate that under certain conditions, short interfering RNAs (siRNAs, 21-25 bp) may activate the interferon system. Mouse oocytes and preimplantation embryos apparently lack this response, as potent and specific inhibition of gene expression triggered by long dsRNA is observed in these cells. In the present study, we analyzed the global pattern of gene expression by microarray analysis in transgenic mouse oocytes expressing long dsRNA and find no evidence of off-targeting. We also report that genes involved in the interferon response pathway are not expressed in mouse oocytes, even after exposure for an extended period of time to long dsRNA.     

Effect of interferon treatment on blockade of protein synthesis induced by poliovirus infection

Treatment of HeLa cells with lymphoblastoid interferon leads to a drastic inhibition of infective poliovirus. Even relatively high concentrations of human lymphoblastoid interferon HuIFN-alpha (Ly) (400 IU/ml) do not prevent destruction of the cell monolayer after most of the cells have been infected with poliovirus. Analysis of macromolecular synthesis in a single step growth cycle of poliovirus in interferon-treated cells detected no viral protein synthesis. In spite of this inhibition of viral translation, the shut-off of host protein synthesis in interferon-treated cells is apparent when they are infected both at low and high multiplicities. Although viral RNA synthesis is inhibited considerably in cells treated with interferon, a certain amount is detected, suggesting that some viral replication takes place. Analysis of membrane permeability after poliovirus infection shows a leakage to 86Rb+ ions and modification of membrane permeability to the translation inhibitor hygromycin B at the moment when the bulk of virus protein synthesis occurs. These changes are delayed and even prevented if cells are pretreated with interferon. A situation is described in which host protein synthesis is shut-down with no major changes in membrane permeability, as studied by the two tests mentioned above. Prevention of viral gene expression by inactivation with ultraviolet light of the input virus or by treatment with cycloheximide blocks the shut-off of protein synthesis. This does not occur in the presence of 3 mM guanidine. These observations are in agreement with the idea that some poliovirus protein synthesis takes place in interferon-treated cells and this early gene expression is necessary to block cellular protein synthesis.     

Adenovirus late protein synthesis is resistant to the inhibition of translation induced by poliovirus

Inhibition of host protein synthesis after poliovirus infection has been suggested to be a consequence of the proteolytic degradation of a p220 polypeptide necessary to translate capped mRNAs. However, the synthesis of several adenovirus late proteins on capped mRNAs was resistant to poliovirus inhibition. Thus, the hexon protein was still made 8 h after poliovirus superinfection. The synthesis of other adenovirus proteins such as the fiber was much more sensitive to poliovirus-induced inhibition than the hexon, either in the absence or in the presence of guanidine. Detailed densitometric analyses clearly showed the differential behavior of several adenovirus late mRNAs to poliovirus shut-off of translation. This is striking in view of the fact that a common leader sequence in the 5' termini is present in the adenovirus late mRNAs. The use of 3-methyl quercetin, an inhibitor of poliovirus RNA synthesis (Castrillo, J. L., Vanden Berghe, D., and Carrasco, L. (1986) Virology 152, 219-227), showed that translation of several capped adenovirus mRNAs took place in poliovirus-infected cells after the synthesis of host proteins had ceased. The poliovirus mRNA and the adenovirus mRNA coding for the hexon protein are very efficient mRNAs and have a leader sequence of more than 740 and 250 nucleotides, respectively, with very rich secondary structures making it difficult to predict how the scanning model will operate on these two mRNAs.