INTRODUCTION INTO THE TRANSLATION INITIATION PATHWAY IN EUKARYOTES
Canonical eukaryotic translation initiation ensures timely and spatially coordinated formation of the trimeric complex between the 40S small ribosomal subunit (40S), initiator Met-tRNAiMet and an mRNA at its extreme 5’ end, and concludes with the assembly of an elongation-competent 80S ribosome at the authentic AUG start codon (summarized in Fig. 1).
The entire process is orchestrated by numerous individual proteins and three protein complexes commonly called eukaryotic initiation factors (eIFs). To begin a new translational cycle, first a pool of separated ribosomal subunits has to be generated from those that have just finished (terminated) translation of a given mRNA in the „previous“ cycle. The ultimate product of translation termination that has to be recycled into individual components is called a post-termination ribosomal complex (post-TC). It consists of an 80S couple still bound to mRNA, P-site deacylated tRNA and eukaryotic release factors (eRFs) 1 and 3 (or at least eRF1). Based on experiments carried out with purified mammalian factors in in vitro reconstituted systems it was originally proposed that, at low Mg2+ concentration (Mg2+ ions promote ribosomal subunit association), recycling can be mediated solely by eIFs 1, 1A and 3. In detail, eIF3, together with its loosely associated eIF3j subunit, eIF1 and eIF1A, first dissociates 60S subunits from the post-TCs. Subsequently, eIF1 promotes release of the tRNA from the P-site of the liberated 40S subunits. Finally, eIF3j significantly enhances eIF3’s mRNA dissociation activity to complete the recycling reaction. However, later on it was reported that with elevated Mg2+ concentrations, recycling strictly depends on ABCE1 (ATP-binding cassette subfamily E member 1), which ensures splitting of post-TCs into free 60S subunits and tRNA•mRNA•40S complexes in the first recycling reaction (Fig. 1). Consistently, RLI1, a yeast homologue of ABCE1 was also shown to be involved in termination per se in S. cerevisiae, however, its precise role remains to be elucidated. Preliminary results likewise indicate that yeast eIF3 directly participates in translation termination and/or recycling. Hence, it appears that the role of ABCE1/RLI1 and several canonical initiation factors as “terminators” of the translational cycle is conserved among eukaryotes. It should be noted that the way how eRFs 1 and 3 are ejected from post-TCs is still unclear. Finally, eIF6, a protein associated with the 60S subunit, is believed to prevent ribosomal subunit re-association.
In the first step of a new translational cycle, Met-tRNAiMet is bound by the trimeric eIF2 complex in its GTP form to produce the Met-tRNAiMet•eIF2•GTP ternary complex (TC). Subsequently, the multiprotein eIF3 complex, together with eIFs 1, 1A and 5, promotes recruitment of the TC to the small ribosomal subunit (40S), producing the 43S pre-initiation complex (PIC). In fact, numerous studies carried out over the last two decades suggest that there are two major ways of how eIFs associate with ribosomes to form the 43S PIC: i) the “stochastic – prokaryotic-like” pathway with eIFs binding to the small subunit on individual basis; and ii) the “higher order – eukaryotic” pathway, where eIFs 1, 3, 5 and the TC assemble into a large multifactor complex (MFC) and approach the 40S ribosome as a pre-organized unit (Fig. 1 and 2). Importantly, recent data from plant and human cells provide evidence that these two pathways are also evolutionary conserved among all eukaryotes. By definition, the MFC-driven pathway is generally considered to ensure the efficiency of the whole initiation process especially under conditions permissive for growth. In any case, upon initial binding of the aforementioned factors, eIFs 1 and 1A serve to stabilize a specific conformation of the 40S head relative to its body that opens the mRNA binding channel for mRNA loading. That requires dissolving the latch formed by helices 18 (h18) and 34 (h34) of 18S rRNA and establishing a new interaction between RPS3 and h16.
In the next step, in the current text book view, eIF3 and the eIF4F complex promote recruitment of mRNA to thus “activated” 43S PIC with help of the poly(A)-binding protein (PABP) forming the 48S PIC. eIF4F comprises the cap-binding protein eIF4E, the DEAD-box RNA helicase eIF4A and eIF4G, which functions as a “scaffold“. It binds eIF4E, eIF4A, PABP and in mammals also eIF3, through which the connection between the eIF4F•mRNA and the 43S PIC could be bridged (Fig. 1 – “M” dashed line). In budding yeast, direct eIF3-eIF4G interaction has not been detected, and the eIF3-binding domain is not evident in yeast eIF4G. Instead it was proposed that eIF5 might bridge the contact between eIF4G and eIF3 in the 48S PIC, as it was shown to be capable of simultaneous binding to both factors in vitro (Fig. 1 – “Y” dashed line). Taking into account that yeast eIF3 is also considered to be more critical factor for mRNA recruitment than eIF4G, it could be that the molecular mechanism of this particular initiation step differs in certain aspects between lower and higher eukaryotes. Alternatively, in the light of the recent in vivo studies carried out in yeast and mammalian cells, it seems also plausible that the mRNA recruitment step is, in general, less dependent on the direct eIF4G–eIF3 contact than it has been believed so far (see below). Importantly, stable binding of the 43S PIC near the 7-methylguanosine cap of natural mRNAs requires melting the secondary structures that often occur in their 5’ untranslated region (UTR) and the eIF4A helicase, as part of the eIF4F complex, is the prime candidate for this role. It should also be mentioned that formation of an interaction between the cap-binding protein eIF4E and eIF4G has been shown to serve as one of the two major targets for the general translational control, especially in mammalian cells (Fig. 1). In yeast the global controls that feed off this regulatory step have not been clearly identified as yet indicating that they might not be so robust.
In contrast to prokaryotic cells, the mRNAs of which posses a Shine-Dalgarno sequence ensuring a direct placement of the start codon into the ribosomal P-site, eukaryotic ribosomes have to search the 5' UTR of an mRNA for usually the first AUG codon by a successive movement called scanning. This is accompanied by unwinding secondary structures in an ATP-dependent reaction stimulated by helicases eIF4A (with its co-activators eIF4B or eIF4H), DHX29 and DED1 (Fig. 1). The mechanism of scanning per se is still largely unexplored. Besides the requirement for helicases, it is also known that in the absence of secondary structures, the presence of the TC and eIFs 1, 1A, and 3 in 48S PICs suffices for locating the AUG start in the mammalian reconstituted systems.
Most importantly, during scanning ribosomes have to read, integrate and respond to a variety of poorly understood signals that orchestrate the AUG recognition. These signals originate from mutual molecular and functional interactions between mRNA and the 40S ribosome with a number of initiation factors such as eIF1, eIF1A, eIF2 (TC), and eIF5. In the open conformation of the 40S ribosome that is induced by eIFs 1 and 1A, as mentioned above, and that is conducive for scanning, the anticodon of Met-tRNAiMet is not fully engaged in the ribosomal P-site in order to prevent premature engagement with putative start codons. eIF2 partially hydrolyzes its GTP with the help of the GTPase accelerating factor (GAP) eIF5; however, prior to start codon recognition, the “gate-keeping” function of eIF1 prevents the release of the resultant phosphate ion, producing GTP- and GDP•Pi-bound 2 states of the factor, possibly in equilibrium. Encounter of the AUG start codon induces a reciprocal conformational switch of the 48S PIC to the closed/scanning arrested form, stabilized by a functional interaction between eIF1A and eIF5, with the initiator Met-tRNAiMet fully accommodated in the P-site. This irreversible reaction serves as the decisive step stalling the entire machinery at the AUG start codon and is triggered by displacement or dissociation of eIF1, possibly promoted by eIF1A and eIF5, and subsequent release of free Pi (Fig. 1). In short, eIF1 and eIF1A (via its C-terminal tail) antagonize the codon-anticodon interactions in the P-site by blocking the full accommodation of initiator tRNA in the P-site in a manner that is overcome efficiently by the action of the N-terminal tail of eIF1A and eIF5 upon establishment of a perfect AUG-anticodon duplex in an optimal Kozak AUG context. As will be discussed later, besides the aforementioned factors, there is an increasing number of reports suggesting that also the multifunctional eIF3 complex significantly contributes to the regulation of AUG recognition.
The scanning-arrested 48S PIC can now join the large ribosomal subunit with the help of GTP-bound eIF5B, upon which most of the interface-side-based eIFs are ejected with the exception of eIF1A and most likely also eIF3 and eIF4F. Finally, GTP-hydrolysis on eIF5B stimulated by the GTP-ase activating center (GAC) of the 60S subunit triggers the release of eIF1A and eIF5B producing an active 80S ribosome poised for elongation (Fig. 1).
To enter a new initiation cycle, “discharged” eIF2•GDP must interact with the pentameric eIF2B, which acts as the GTP/GDP exchange factor (GEF) for eIF2 and exchanges its GDP for a GTP nucleotide. Only this „charged“ form of eIF2 can stably bind Met-tRNAiMet to form a new ternary complex. According to the recent reports, eIF2•GDP leaves the PICs in the binary complex with eIF5 that antagonizes eIF2B-promoted guanine nucleotide exchange (see below) (Fig. 1). It is important to note that the step of the ternary complex formation is the other of the two major targets of the general translational control (Fig. 1). Several kinases phosphorylate the α-subunit of eIF2 upon various cellular stress conditions turning it form a substrate to an inhibitor of an exchange reaction, which leads to a global translational shut down.
eIF3 – the GIRL FOR EVERYTHING
eIF3 has been demonstrated to promote or at least fine tune nearly every single step of translation initiation and now it seems that its influence reaches even beyond that. In budding yeast, eIF3 comprises five core essential subunits (a/TIF32, b/PRT1, c/NIP1, i/TIF34, and g/TIF35) and one non-core subunit (j/HCR1) (Fig. 2). These all have corresponding orthologs in the more complex mammalian eIF3, which contains seven additional non-conserved subunits (eIF3d, e, f, h, k, l, and m). Despite recent progress, the true composition of the core of mammalian eIF3 remains somewhat obscure. Whereas there is only very limited information on the subunit–subunit interaction web of mammalian eIF3, the labyrinth of mutual contacts among the yeast subunits has been mapped in great detail (Fig. 2). Systematic effort was also devoted to mapping the binding site of eIF3 on the 40S. Results obtained from several labs clearly suggested that eIF3 associates with the head and beak regions of the upper body of the solvent-exposed side of the 40S ribosome.
YEAST MODEL SYSTEM
The ultimate aim of our work is indeed to gain understanding of processes relevant to human health. As such, the regulatory mechanisms underlying eukaryotic translation should be studied in the corresponding biological systems. However, it is no surprise that there is only a handful of reports studying translation in humans in the living cells, owing to bigger complexity of higher eukaryotic cells as well as a lack of feasible approaches. In contrast, the relative ease of the work with the budding yeast Saccharomyces cerevisiae stands behind the fact that yeast genetics has been a powerful tool for dissecting the mechanism of eukaryotic translation initiation during the past two decades. Given the high conservation of this fundamental pathway in all eukaryotes, grandly speaking from yeast to humans it should be only one step up, when the advances in methodology allow, since the basic mechanism will be known and the questions to be asked will concern mainly the apparent differences between the lower and higher-order systems.
Hence, significant part of our work makes a great use of the budding yeast as a model organism and the following paragraph brings a brief overview of yeast genetics relevant to this what we do. In addition to the slow growth (Slg-) and temperature sensitive (Ts-) phenotypes, several specific phenotypes are being investigated. Mutations that reduce the fidelity of start codon cognition, allowing increased utilization of near-cognate codons (UUG or AUU), produce the Sui- phenotype (suppressor of initiation codon mutation). Sui- mutations are isolated by selecting for growth on medium lacking histidine of a yeast strain in which the start codon of the HIS4 gene has been changed to AUU (the his4-303 allele). Two other important phenotypes affect the ability of the cell to respond to amino acid starvation by increasing translation of the transcription factor GCN4. Translation of GCN4 mRNA is regulated by four short, upstream open reading frames (uORFs) (Fig. 3).
In unstarved cells, where TC levels are high, after the first uORF is translated the 40S subunit regains a TC and reinitiates on one of the downstream uORFs. This prevents complexes from reaching the GCN4 start codon and initiating there. When cells are starved for amino acids, the kinase GCN2 is activated to phosphorylate the α-subunit of eIF2. Phosphorylation of eIF2α leads to inhibition of TC formation. The low TC levels allow the 40S subunit to scan past the start codons of uORFs 2-4 before reacquiring a TC, thus increasing the frequency of initiation at the GCN4 start codon. Mutations that constitutively activate GCN4 translation, allowing survival of amino acid starvation even in gcn2Δ cells, produce the Gcd- phenotype (general control derepressed). Mutations that prevent activation of GCN4 translation by amino acid starvation in GCN2+ cells produce the Gcn- phenotype (general control nonderepressible). We and others have shown that many of the Gcd- mutations reduce the efficiency of TC binding to the 40S subunit. Translation-born Gcn- phenotypes can arise from a variety of reasons such as for example: 1) reduced rate of mRNA recruitment; 2) reduced initiation at uORF1 (leaky scanning phenotype; PIC complexes miss the correct start codon with increased frequency and continue scanning downstream); 3) failure of the 40S ribosome to resume scanning after terminating at uORF1; 4) reduced rate of scanning (slow scanning phenotype); 5) instability of re-scanning ribosomes on the mRNA (defect in processivity of scanning).
REINITIATION (REI)
