Skip Internal Navigation
We study the mechanism and regulation of protein synthesis, focusing on GTPases and protein kinases that control this fundamental cellular process. We use molecular-genetic and biochemical studies to dissect the structure-function properties of the translation initiation factors eIF2, a GTPase that binds methionyl-tRNA to the ribosome, and eIF5B, a second GTPase that catalyzes ribosomal subunit joining in the final step of translation initiation. Our studies have revealed a critical role for eIF2 in start codon selection and have defined a functionally important contact between eIF5B and the ribosome. We also investigate stress-responsive protein kinases that phosphorylate eIF2alpha. Recent studies revealed fast evolution of the antiviral kinase PKR in vertebrates and linked this with altered sensitivity to poxvirus inhibitors of the kinase. Finally, our studies on the factor eIF5A revealed an unanticipated role in the elongation phase of protein synthesis.
Analysis of eIF2 binding to Met-tRNAiMet and the ribosome
The gamma subunit of eIF2 is a GTPase that, based on sequence and the structure of the archaeal homolog aIF2alpha, resembles the bacterial translation elongation factor EF-Tu. However, in contrast to EF-Tu, which binds to the A-site of the 70S ribosome, eIF2 binds Met-tRNAiMet to the P-site of the 40S subunit. To gain insights into how eIF2 binds to Met-tRNAiMet and then associates with the 40S ribosome, we used directed hydroxyl radical probing to identify eIF2 contacts within the 40S–eIF1–eIF1A–eIF2–GTP–Met-tRNAi–mRNA (48S) complex (1). We generated a cysteine-deficient version of eIF2 and then introduced single Cys residues at predicted surface exposed sites (based on the aIF2 structure) on eIF2alpha, eIF2beta, and eIF2gamma. The mutant proteins were purified from yeast and then derivatized with Fe(II)-BABE on the Cys residue. Following addition of hydrogen peroxide, hydroxyl radicals formed in the vicinity of the ferrous iron will diffuse and cleave nucleic acid and protein backbones. Based on the structure of the EF-Tu ternary complex, we predicted that linkage of Fe(II)-BABE to domain III of eIF2gamma would result in cleavage of Met-tRNAiMet in the T-stem. However, derivatization of Fe(II)-BABE to domain III resulted in cleavage of the D-stem of Met-tRNAiMet and of 18S rRNA at the top of helix h44, a prominent landmark on the intersubunit surface of the 40S subunit. Based on the results of these and other cleavage experiments, and the fact that Met-tRNAiMet is bound to the P-site of the 40S subunit, we generated a model of the 48S complex in which domain III of eIF2gamma binds near 18S rRNA helix h44 and in which eIF2gamma–Met-tRNAiMet contacts are restricted to the acceptor stem of the tRNA. In this model of the eIF2 ternary complex, the Met-tRNAiMet is rotated nearly 180° relative to the position of the tRNA in the EF-Tu ternary complex. Consistent with the alternate models of the eIF2 and EF-Tu ternary complexes, we found that the EF-Tu-T394C mutation in domain III severely impaired Phe-tRNA binding, whereas the corresponding eIF2gamma-K507C mutation did not impair Met-tRNAiMet binding to eIF2. Thus, despite their structural similarity, eIF2 and EF-Tu bind to tRNAs in substantially different manners, and we propose that the tRNA-binding domain III of EF-Tu has acquired a new function in eIF2gamma to bind to the ribosome (1).
Structure-function analysis of the universally conserved translational GTPase eIF5B/IF2
During the final step of translation initiation, the large 60S ribosomal subunit joins with the 40S subunit, already bound to an mRNA, to form an 80S ribosome competent for protein synthesis. We previously discovered the translation initiation factor eIF5B, an ortholog of the bacterial translation factor IF2, and showed that it catalyzes ribosomal subunit joining. The eIF5B is a GTPase that binds to GTP and hydrolyzes the nucleotide in the presence of 80S ribosomes. Moreover, we showed that GTP hydrolysis by eIF5B is a regulatory switch governing the release of eIF5B from the ribosome following subunit joining, and we identified a functionally important contact between domain II of eIF5B and helix h5 of 18S rRNA in the 40S ribosome. Our current efforts aim to elucidate eIF5B's structure-function properties and to understand the role played by eIF5B in GTP binding and hydrolysis.
The eIF5B factor resembles a chalice, with the alpha-helix H12 forming the stem connecting the GTP-binding domain cup to the domain IV base. Helix H12 has been proposed to function as a rigid lever arm governing domain IV movements in response to nucleotide binding and as a molecular ruler fixing the distance between domain IV and the G domain of the factor. To investigate its function, we altered the length and rigidity of helix H12 (2). Whereas helix H12 mutations had minimal impacts on GTP binding and on eIF5B ribosome binding and GTPase activities, shortening the helix impaired the rate of subunit joining in vitro. Moreover, both shortening the helix and increasing its flexibility impaired the stability of Met-tRNA bound to the 80S product of subunit joining. These data support the notion that helix H12 functions as a ruler connecting the GTPase center of the ribosome to the P site where Met-tRNAiMet is bound and that helix H12 rigidity is required to stabilize Met-tRNAiMet binding (2).
e>Molecular analysis of eIF2alpha protein kinase substrate recognition and viral regulation
Phosphorylation of eIF2alpha is a common mechanism for downregulating protein synthesis under stress conditions. Four distinct kinases phosphorylate eIF2alpha on Ser51 under different cellular stress conditions. GCN2 responds to amino acid limitation, HRI to heme deprivation, PERK to ER stress, and PKR to viral infection. Consistent with their common activity to phosphorylate eIF2alpha on Ser51, the kinases show strong sequence similarity in their kinase domains. Phosphorylation of eIF2alpha converts eIF2 from a substrate to an inhibitor of its guanine-nucleotide exchange factor eIF2B. The inhibition of eIF2B impairs general translation, slowing the growth of yeast cells and, paradoxically, enhancing the translation of the GCN4 mRNA required for yeast cells to grow under amino-acid starvation conditions. We used structural, molecular, and biochemical studies to define how the eIF2alpha kinases recognize their substrate.
In collaboration with Frank Sicheri, we obtained the X-ray structure of eIF2alpha bound to the catalytic domain of PKR. Back-to-back dimerization enables each PKR protomer to engage a molecule of eIF2alpha in the crystal structure. Given that all four eIF2alpha kinases share the PKR residues mediating kinase domain dimerization and eIF2alpha recognition, we propose that all four kinases similarly dimerize and recognize eIF2alpha. Based on our results, we propose an ordered mechanism of PKR activation in which catalytic domain dimerization triggers autophosphorylation on Thr446, which in turn is required for specific eIF2alpha substrate recognition.
Our previous crystallography and accompanying molecular-genetic analyses revealed that PKR helix alphaG contacts eIF2alpha on a face remote from the Ser51 phosphorylation site; however, the helix alphaG contact is critical for eIF2alpha phosphorylation. Moreover, when the structure of free eIF2alpha, in which the position of Ser51 is resolved, was docked on the structure of the PKR–eIF2alpha complex, Ser51 was about 20 Å from the kinase active site. Mutation of Thr487 in PKR helix alphaG cripples the kinase and prevents eIF2alpha phosphorylation. To gain new insights into how Ser51 gains access to the PKR active site, we randomly mutated eIF2alpha and screened for mutants that would restore Ser51 phosphorylation by the PKR-T487A mutant kinase (3). An eIF2alpha-L47I mutation was found to restore the ability of PKR-T487A to inhibit yeast cell growth. Examination of the eIF2alpha structure revealed that L47 forms a hydrophobic network with L50, I58, and I62—a network that restricts the position of Ser51. We proposed that disruption of this hydrophobic network was necessary to expose Ser51 to the kinase active site. To test this hypothesis, we mutated the hydrophobic residues and examined Ser51 phosphorylation by wild-type and mutant forms of PKR both in yeast and in vitro. We found that substitution of Ser or Pro for Leu50 restored Ser51 phosphorylation by PKR-T487A (3). Notably, whereas WT eIF2alpha is not a substrate for phosphorylation by protein kinase C (PKCalpha), eIF2alpha-L50P can be phosphorylated by PKCalpha. NMR analyses by our collaborators Frank Sicheri and Lewis Kay revealed that the L50S mutation enhanced the mobility of the loop containing Ser51. Moreover, limited proteolysis studies revealed that binding of PKR to eIF2alpha enhanced the protease sensitivity of the Ser51 loop. Finally, kinetic studies of eIF2alpha phosphorylation revealed that the L50S mutation did not affect the Km for Ser51 phosphorylation but substantially improved the kcat for the reaction, consistent with the notion that exposure of Ser51 is a rate-limiting step for eIF2alpha phosphorylation. Based on these studies and the structure of the PKR–eIF2alpha complex, we propose that docking of eIF2alpha on PKR induces a conformational change, greater than the spontaneous "breathing" of the Ser51 loop, which enables Ser51 to engage the phospho-acceptor binding site of the kinase. Finally, we propose that the protected state of Ser51 in free eIF2alpha prevents promiscuous phosphorylation and the attendant translational regulation by heterologous kinases (3).
As part of the mammalian cell's innate immune response, the double-stranded RNA–activated protein kinase PKR phosphorylates the translation initiation factor eIF2alpha to inhibit protein synthesis and thus block viral replication. To subvert this host cell defense mechanism, viruses produce inhibitors of PKR. Several members of the poxvirus family express two distinct types of PKR inhibitor: a pseudosubstrate inhibitor (such as the vaccinia virus K3L protein that resembles the N-terminal third of eIF2alpha) and a double-stranded RNA-binding protein called E3L. High-level expression of human PKR inhibited the growth of yeast, and co-expression of the vaccinia virus K3L or E3L protein [or the related variola (smallpox) virus C3L or E3L protein, respectively] restored yeast cell growth. We previously identified PKR mutations that confer resistance to K3L inhibition but do not affect eIF2alpha phosphorylation. We proposed that these paradoxical effects on pseudosubstrate versus substrate interactions reflect differences between the rigid K3L protein and the plastic nature of eIF2alpha around the Ser51 phosphorylation site (3). We are currently characterizing PKR mutations that confer resistance to E3L inhibition. We have also characterized an eIF2alpha ortholog, termed vIF2alpha, found in ranaviruses that infect lower vertebrates including fish, frogs and salamanders. Our studies revealed that vIF2alpha does not functionally substitute for eIF2alpha in yeast. However, expression of vIF2alpha suppresses the toxic effects of PKR in yeast. We conclude that vIF2alpha functions as an inhibitor of PKR probably as a pseudosubstrate like vaccinia virus K3L (4).
Molecular analysis of the hypusine-containing protein eIF5A
The translation factor eIF5A, the sole protein containing the unusual amino acid hypusine [Ne-(4-amino-2-hydroxybutyl)lysine], was originally identified based on its ability to stimulate the yield (endpoint) of methionyl-puromycin synthesis, a model assay for first peptide bond synthesis. However, the precise cellular role of eIF5A is unknown. Using molecular-genetic and biochemical studies, we recently showed that eIF5A promotes translation elongation and that this activity is dependent on the hypusine modification. Given that eIF5A is a structural homolog of the bacterial protein EF-P, we propose that eIF5A/EF-P is a universally conserved translation elongation factor (5).