We study the mechanism and regulation of protein synthesis in eukaryotic cells focusing on regulation by GTP-binding (G) proteins and protein phosphorylation. The first step of protein synthesis is binding the initiator Met-tRNA to the small ribosomal subunit by the factor eIF2. The eIF2 is a GTP-binding protein and during the course of translation initiation the GTP is hydrolyzed to GDP. The eIF2 is released from the ribosome in complex with GDP and requires the guanine-nucleotide exchange factor eIF2B to convert eIF2-GDP to eIF2-GTP. This exchange reaction is regulated by a family of stress-responsive protein kinases that specifically phosphorylate the alpha subunit of eIF2 on serine at residue 51, and thereby covert eIF2 into an inhibitor of eIF2B. Among the family of eIF2alpha kinases are GCN2, which is activated under conditions of amino acid starvation, PKR, which is activated by double-stranded RNA and downregulates protein synthesis in virally infected cells, and PERK, activated under conditions of ER stress. Previously, in collaboration with Frank Sicheri, we determined the structure of the PKR kinase domain in complex with eIF2alpha. This structural analysis revealed that eIF2alpha binds to the C-terminal lobe making intimate contact with helix alphaG, while catalytic domain dimerization is mediated by a back-to-back orientation of the kinase N-terminal lobes. ? ? In order to subvert the anti-viral defense mediated by PKR viruses produce inhibitors of the kinase. Several members of the poxvirus family express two different types of PKR inhibitors: a pseudosubstrate inhibitor and a double-stranded RNA binding protein called E3L. The vaccinia virus K3L protein resembles the N-terminal third of eIF2alpha with both proteins containing a beta-barrel fold of the OB-fold family. Whereas high-level expression of human PKR was toxic in yeast, this growth inhibition was suppressed by co-expression of the vaccinia virus K3L protein or the related variola (smallpox) virus C3L protein. We used this yeast assay to screen for PKR mutants that are resistant to K3L inhibition, and we identified twelve mutations mapping to the C-terminal lobe of the PKR kinase domain in the vicinity of the eIF2alpha-binding site. The PKR mutations specifically conferred resistance to the K3L protein, but not the E3L protein, both in yeast and in vitro. In vitro studies revealed that WT PKR and the PKR mutants phosphorylated eIF2alpha with the same kinetics; however, the mutant kinase was less sensitive to inhibition by K3L. Consistently, the PKR-D486V mutation led to nearly a 15-fold decrease in K3L binding affinity. Our results support the identification of the eIF2alpha binding site on an extensive face of the C-terminal lobe of the kinase domain and they indicate that subtle changes to the PKR kinase domain can drastically impact pseudosubstrate inhibition while leaving substrate phosphorylation intact. ? ? Continuing our collaborative studies with Frank Sicheri on stress-responsive protein kinases, we characterized the unique kinase-endonuclease IRE1. Accumulation of misfolded proteins in the endoplasmic reticulum (ER) under stress conditions activates a stress response pathway referred to as the unfolded protein response (UPR). The primary sensor of ER stress in the UPR is the protein IRE1. The IRE1 is a transmembrane protein with an N-terminal domain situated in the ER lumen and a cytoplasmic domain consisting of a protein kinase domain and C-terminal kinase-associated nuclease (KEN) domain. Dimerization of IRE1 under ER stress conditions activates kinase autophosphorylation and nuclease activity. The KEN domain then splices, in a spliceosome-independent manner, mRNAs encoding transcriptional regulators of the UPR: HAC1 mRNA in yeast and Xbp1 mRNA in mammals. The Sicheri lab obtained the crystal structure of the cytoplasmic catalytic domain of IRE1 revealing the structure of both the kinase and KEN domain (Lee et al., 2008; reference 1). Back-to-back dimerization of the kinase domain in the IRE1 crystal structure juxtaposes the KEN domains and activates the ribonuclease. Four autophosphorylation sites were identified in the IRE1 kinase domain, and mutational and biochemical studies revealed that autophosphorylation facilitates ATP binding and the accompanying dimerization of the kinase domain. We showed that yeast cells expressing IRE1 mutants with mutations in the dimer contact residues are unable to grow in medium containing tunicamycin, an inhibitor of protein glycosylation and inducer of ER stress. Likewise, the dimerization mutants were unable to splice the Xbp1 mRNA in vitro. Comparison of the structure of the IRE1 KEN domain with the structure of the structurally distinct tRNA endonuclease identified putative nuclease active site residues. Mutation of these residues blocked Xbp1 mRNA splicing in vitro and yeast cell growth under ER stress conditions. Taken together, these data reveal an unexpected convergent evolution of tRNA endonuclease and IRE1 KEN domain catalytic mechanism (Lee et al., 2008). We propose that dimerization of IRE1 lumenal domains in response to ER stress promotes kinase domain trans-autophosphorylation, which in turn facilitates nucleotide binding and back-to-back dimerization of the kinase domains. Dimerization of the KEN domains in the resultant structure enables recognition and splicing of the HAC1/Xbp1 mRNA.? ? The translation initiation factor eIF2 binds the initiator Met-tRNA to the small ribosomal subunit. To gain further insights into the role of GTP binding and hydrolysis by eIF2, we mutated the conserved Asn135 residue in the eIF2gamma GTP-binding domain Switch I element to Asp. The N135D mutation impaired Met-tRNA binding to eIF2 and caused a Sui- phenotype, enhancing initiation from a non-canonical UUG codon. Previous studies in the Donahue laboratory correlated a Sui- phenotype with decreased Met-tRNA binding affinity, suggesting that premature release of Met-tRNA from eIF2 led to initiation at the UUG codon. Consistently, an A208V mutation restored Met-tRNA binding affinity and suppressed the slow-growth and Sui- phenotypes of the eIF2gamma-N135D mutant. In contrast, an A382V mutation restored Met-tRNA binding and suppressed the slow-growth, but not the Sui-, phenotype. Moreover, an eIF2gamma-A219T mutation impaired Met-tRNA binding but unexpectedly enhanced the fidelity of initiation, suppressing the Sui- phenotype associated with the eIF2gamma-N135D,A382V mutant. This uncoupling of start codon selection and Met-tRNA binding affinity to eIF2 indicates a more direct role for eIF2 in start site recognition. Interestingly, overexpression of eIF1, which is thought to monitor codon-anticodon interaction during translation initiation, likewise suppressed the Sui- phenotype of the eIF2gamma mutants. We propose that structural alterations in eIF2gamma subtly alter the conformation of Met-tRNA on the 40S subunit and thereby affect the fidelity of start codon recognition independent of Met-tRNA binding affinity.? ? The GTP-binding protein eIF5B catalyzes ribosomal subunit joining in the final step of translation initiation. The eIF5B is an ortholog of prokaryotic translation initiation factor IF2. Previous studies revealed that eIF5B consists of four domains that structurally assemble to form a chalice-shaped molecule. The G domain plus domains II and III form the cup of the chalice, a long alpha helix forms the stem, and domain IV is the base of the chalice. Our previous studies revealed that GTP hydrolysis by eIF5B activates a regulatory switch required for eIF5B release from the ribosome following subunit joining. Consistently, we have recently identified mutations in the ribosome that suppress the growth and translation defects associated with mutations that impair eIF5B GTPase activity.
Showing the most recent 10 out of 38 publications
