As obligate intracellular parasites, viruses are completely dependent upon the trans-lational machinery resident in their host cells. It is not surprising, therefore, that a major host defense component centers on impeding viral mRNA translation. Indeed, one of at least four stress responsive eIF2a kinases present in mammalian cells, the double stranded RNA dependent protein kinase PKR is induced by the antiviral cytokines interferon a/p. It has been proposed that abundant dsRNA, a replicative intermediate formed in the replication of RNA viruses and a by-product of overlapping transcription units on opposite DNA strands of DNA viruses, is a signature of viral infection. PKR binds dsRNA and in the presence of this activating ligand forms a dimer whereupon each subunit phosphorylates the other. It is this activated, phosphorylated form of PKR that then goes on to phosphorylate other substrates, including eIF2a, the regulatory subunit of eIF2 (31).
eIF2 is a heterotrimeric G protein, forming a ternary complex composed of its a,p, and 7 subunits bound to GTP along with the initiator tRNA, that is responsible for chaperoning the initiator tRNA to the 40S ribosomal subunit, forming what is known as the 48S complex (see Fig. 3A) (32). Like many G proteins, a GTPase activating protein (eIF5) and a guanine nucleotide exchange factor (eIF2B) regulate its activity. Normally, eIF5 promotes GTP hydrolysis following the joining of the 60S ribosomal subunit to form an 80S ribosome, releasing eIF2 bound to GDP. To participate in subsequent rounds of polypeptide chain initiation, the GDP form of eIF2 requires the activity of eIF2B in order to exchange GDP for GTP. However, phosphorylation of eIF2a leads to a strong association between eIF2B and eIF2, effectively preventing eIF2B from catalyzing the nucleotide exchange reaction (see Fig. 3B). As the quantity of eIF2B present in cells is limiting, phosphorylation of small amounts of eIF2a by PKR can have relatively large effects on translation by sequestering eIF2B, hindering nucleotide exchange, and inhibiting translation. Unchecked, PKR activated by dsRNA in virus infected cells would therefore effectively deplete active eIF2, inhibit ternary complex formation, and prevent viral and cellular protein synthesis. Should cells initially infected succeed in inhibiting translation, the viral invader would effectively be stopped in its tracks, denied access to the cellular translational apparatus it needs to complete its life cycle. This arm of the innate host response then is designed to sacrifice the initially infected cells for the benefit of the larger population. However, numerous viruses, including HSV-1, have captured a variety of functions to counter this cellular response (32). This struggle for control of the translational machinery is often an integral component of viral pathogenesis (33-35).
Normally, the 7134.5 gene product prevents the accumulation of phosphorylated eIF2a by recruiting a cellular phosphatase, protein phosphatase 1a (PP1a), to remove phosphate from eIF2a (36). Interestingly, the domain of the 7^4.5 protein that contains this activity is homologous to a domain in the GADD34 protein, a cellular PP1a binding protein that promotes eIF2a dephosphorylation in response to other forms of cell stress (37). Besides their defect in protein synthesis, 7^4.5 mutant derivatives are hypersensitive to interferon a, and therefore more sensitive to this arm of innate host defenses that serve to limit viral replication in the host (38,39). This hypothesis is supported by reports demonstrating that 7134.5 mutant viruses, whereas neuroattenuated in normal mice and mice with deficiencies in their acquired immune response, exhibit restored neurovirulence in mice with deficiencies in innate immunity (33,34).
Ensuing genetic studies revealed that HSV-1 actually encodes multiple functions to control eIF2a phosphorylation, as the dsRNA binding protein specified by the true late Us11 gene prevents PKR activation (40-43). Analysis of a panel of 7134.5 and Us11 mutants established that both Us11 and 7134.5 gene products act at different times in the productive growth cycle to regulate eIF2a phosphorylation in infected cells (44) (see Fig. 3C). Importantly, Us11 expressed in its natural context as a late 72 protein is required to properly regulate PKR activation, eIF2a phosphorylation, and viral translation. Thus, because late proteins are not produced in nonpermissive tumor cells infected with a 7134.5 mutant virus, the 7134.5 mutant is actually doubly deficient in that it also fails to translate the 72 Us11 mRNA. Moreover, the interferon sensitivity of 7134.5 mutants, previously attributed solely to the absence of 7134.5 function, likewise results from the absence of the 7134.5 protein and the failure to synthesize the Us11 polypeptide (45).
Fig. 3. Regulation of translation by phosphorylation of eIF2, a critical translation initiation factor. (A) Composed of a,p, and y subunits, eukaryotic translation initiation factor 2 (eIF2) forms a ternary complex with GTP and the initiator tRNA (tRNAi). This complex associates with the 40S ribosomal subunit bound to eIF3, and recognizes the 5'-end of the mRNA through an association with eIF4F. Once the AUG codon in the mRNA has been identified by a unidirectional translocation process termed scanning, GTP hydroylsis stimulated by eIF5 and the subsequent release of the eIF2-GDP complex facilitate the joining of the 60S ribosome subunit and translation elongation commences. The guanine nucleotide exchange factor eIF2B is required to exchange the GDP bound to eIF2 and replace it with GTP, thus recycling the active form of eIF2. (B) After phosphorylation of eIF2 on its a subunit by PKR, an eIF2a kinase, eIF2B remains tightly bound and cannot exchange the GDP bound to eIF2 for GTP. This failure to recycle eIF2 to its active, GTP bound form inhibits the initiation of translation. (C) Regulation of eIF2a phosphorylation by different functions that act during discrete phases in the HSV-1 lifecycle. Early in the HSV-1 lifecycle, small quantities of dsRNA, or perhaps other effectors that remain to be identified activate the normally dormant cellular PKR kinase. After assembling a dimer of PKR on dsRNA, each subunit of the multimer phosphorylates the other (PKR-P). Because PKR is activated in cells infected with a Yi34.5 mutant virus, preventing the translation of y2 mRNAs, we propose that the Yi34.5 gene product, through its interaction with PP1a, is able to adequately dephosphorylate the quantities of phosphorylated eIF2a (eIF2a-P)
Was this article helpful?