PreviewsLeading Edge A Deeper Look into Translation Initiation Andrei A. Korostelev1,* 1RNA Therapeutics Institute, Department of Biochemistry and Molecular Pharmacology, University of Massachusetts Medical School, 368 Plantation Street, Worcester, MA 01605, USA *Correspondence: [email protected] http://dx.doi.org/10.1016/j.cell.2014.10.005 Eukaryotic translation initiation requires coordinated assembly of a remarkable array of initiation factors onto the small ribosomal subunit to select an appropriate mRNA start codon. Studies from Erzberger et al. and Hussain et al. bring new insights into this mechanism by looking at early and late initiation intermediates. Initiation is a key step of translation, re- particle scans the mRNA in the 50-to-30 To tackle this problem, Erzberger and sulting in the precise placement of an direction until it finds a start codon in colleagues employed a hybrid, multi-tech- initiation AUG codon into the ribosome, an appropriate sequence context. Met- nique approach. First, they determined thereby establishing the correct open tRNAi then base pairs with the AUG codon high-resolution crystal structures of iso- reading frame of an mRNA for protein in the P site of the small subunit, forming lated eIF3 subunits, namely eIF3a, the synthesis. Bacterial translation initiation the 48S initiation complex (Figure 1C). eIF3a,eIF3c dimer, the eIF3b core, and is conceptually simple, relying on a Joining with the 60S ribosomal subunit eIF3i bound with the terminal domains Shine-Dalgarno sequence upstream of completes the formation of the 80S ribo- of eIF3b and eiF3g. Next, using a com- an initiation codon to anchor the mRNA some that is ready to proceed with protein bination of negative-stain electron micro- to the small ribosomal subunit and three synthesis. scopy and crosslink mass-spectrometry, initiation factors to help recruit the initiator Two recent reports of molecular struc- the authors provided a detailed view of tRNA and fend off elongator tRNAs and tures of eukaryotic initiation complexes the interactions and organization of eIF3 the large subunit until the initiator tRNA help to unveil several important aspects in the 40S,eIF1,eIF3 complex. As an clicks into the P (peptidyl) site. In eukary- of the initiation mechanism (Erzberger internal control, intra-40S protein cross- otes, however, initiation is an intricate et al., 2014; Hussain et al., 2014). Erz- links revealed interactions consistent process that requires at least a dozen initi- berger and colleagues report a structure with those predicted from the 40S crystal ation factors (reviewed in Hinnebusch and of yeast eIF3 bound to the 40S ribosomal structure, demonstrating the robustness Lorsch, 2012; Jackson et al., 2010). Two subunit in a complex with eIF1, repre- of the crosslinking approach. Analysis new studies provide an unprecedented senting an early intermediate in transla- of 90,000 possible crosslink-restrained level of insight into this process (Erzberger tion initiation. In accordance with its large structural models of the 40S,eIF1,eIF3 et al., 2014; Hussain et al., 2014) size, eIF3 is considered to be a ‘‘master complex converged on a final cluster of The major challenges in studying trans- regulator’’ of initiation, mediating interac- solutions in which the position of each lation initiation structurally are the number tions with many initiation factors, promot- eIF3 subunit on the surface of the 40S of steps and the size of the complexes ing mRNA recruitment and scanning, subunit is assigned with high confidence involved and the intricate choreography and preventing premature association and agrees with crystallographic and of conformational changes required to of the initiation complex with the large electron-microscopy data. These ana- scan for the proper start codon and act ribosomal subunit (Hinnebusch and lyses reveal that eIF3 forms a horseshoe- on it. Some initiation factors are multisu- Lorsch, 2012; Jackson et al., 2010). In like structure, embracing the 40S subunit bunit complexes, such as the 750 kDa yeast, eIF3 contains six proteins—eIF3a, (Figure 1A). The eIF3 subunits intertwine, mammalian initiation factor 3 (eIF3) eIF3b, eIF3c, eIF3g, eIF3i, and eIF3j— bridging the mRNA entrance and exit composed of 13 proteins. Assembly of which are highly conserved and form the tunnels near the A and E sites, respec- the translation initiation complex on the core of eIF3 in all eukaryotes. Based on tively, agreeing with previous structural, small 40S ribosomal subunit occurs in a biochemical and cryo-EM studies, eIF3 genetic, and biochemical data (reviewed stepwise fashion. The 40S,eIF3 particle is associated with the solvent-exposed in Hinnebusch, 2014). The model provides (Figure 1A) bound with eukaryotic initia- surface (the ‘‘back’’) of the small ribo- a framework for future studies and repre- tion factors eIF1, eIF1A, and eIF5 recruits somal subunit (Hashem et al., 2013 and sents a major step toward elucidating the methionine initiator tRNA (Met-tRNAi) references therein). Due to the dynamic how eIF3 orchestrates many aspects of attached to GTP-bound eIF2. This 43S nature of eIF3 and the lack of high-resolu- initiation by reaching out to the binding pre-initiation complex (Figure 1B) then tion structural information for the com- sites of other initiation factors in the 43S associates with an mRNA whose 7-meth- plex, assigning the position of each eIF3 preinitiation complex as well as interacting ylguanosine 50 cap has been recognized subunit on the 40S subunit has been a with the mRNA at both the entrance and by the eIF4F multiprotein factor. The 43S challenge. exit tunnels in the 43S,mRNA complex. Cell 159, October 23, 2014 ª2014 Elsevier Inc. 475 ture, Lomakin and Steitz (2013) proposed that the nucleotide at position +4 in the A site is ‘‘sensed’’ by the N-terminal tail of eIF1A, although the details of this communication remained unresolved. Now the structure by Hussain et al. pro- vides a detailed view of mRNA interac- tions, including uS7 and a conserved loop of eIF2 contacting the mRNA at the position À3, and the N-terminal tail of eIF1A interacting directly with nucleotide +4, thus suggesting how initiation factors might examine the start-codon context during scanning. Figure 1. Schematic of 48S Initiation Complex Formation and Architecture In summary, complementary work by (A) Location of subunits of initiation factor eIF3 on the solvent surface of the 40S subunit, as revealed by Erzberger et al. (2014) and Hussain Erzberger et al. (2014). A (aminoacyl), P (peptidyl), and E (exit) sites of tRNA binding and the location of eIF1 et al., 2014 provides a deeper view into (brown) on the intersubunit surface are labeled. translation initiation by presenting a (B) Schematic view of the 43S preinitiation complex formed prior to mRNA recruitment, inferred from the structure of a partial mammalian 43S complex (Hashem et al., 2013). The schematic was rendered using detailed picture of the interplay between the structures of the 40S,eIF1,eIF1A complex (PDB ID 4BPE) and archaeal aIF2,GDPNP,Met-tRNAi initiation factors, tRNA, mRNA, and the ternary complex (PDB ID 3V11). 40S subunit in an early and late initiation (C) Partial 48S initiation complex with Met-tRNAi fully accommodated and base paired with the AUG start codon, as revealed in the 4 A˚ cryo-EM structure by Hussain et al. (2014). The putative location of eIF5 states. The next frontier is to reveal struc- (behind eIF2 in this view) is not shown. tural mechanisms of the intermediate steps, including recruitment of eIF4F- bound mRNA onto the 43S preinitiation Initiation involves scanning by the 43S however, is different between the two complex and scanning of the 50 untrans- complex to search for a start codon. structures, suggesting that eIF2 and likely lated region of the mRNA by the 43S A recent cryo-EM study of a partial other initiation factors properly orient the complex and associated helicases. mammalian 43S particle, formed in the tRNA in the 48S complex. Interestingly, absence of mRNA (Figure 1B), found eIF1 remains bound next to the P site REFERENCES that the initiator tRNA is located 7A˚ rather than being forced out by steric outside of the P site (Hashem et al., hindrance with tRNA, as expected from Erzberger, J.P., Stengel, F., Pellarin, R., Zhang, S., 2013), consistent with a state (POUT) that biochemical (Hinnebusch and Lorsch, Schaefer, T., Aylett, C.H., Cimermancic, P., Boeh- allows mRNA to load and/or bypass 2012; Jackson et al., 2010) and structural ringer, D., Sali, A., Aebersold, R., and Ban, N. 158 during scanning. In new work, Hussain work (Lomakin and Steitz, 2013; Rabl (2014). Cell , 1123–1135. and colleagues (2014) report a 4 A˚ cryo- et al., 2011). As eIF1 is slightly distorted Hashem, Y., des Georges, A., Dhote, V., Langlois, R., Liao, H.Y., Grassucci, R.A., Hellen, C.U., Pes- EM structure of a partial yeast 48S initia- and displaced by tRNA from its position, tova, T.V., and Frank, J. (2013). Cell 153, 1108– tion complex, formed downstream of the this new structure represents an inter- 1119. , 40S eIF3 and 43S complexes (Figures mediate state that is sampled just prior Hinnebusch, A.G. (2014). Annu. Rev. Biochem. 83, 1A and 1B). This structure reveals rear- to dissociation of eIF1. 779–812. rangements and interactions in the initia- The structure presented in Hussain Hinnebusch, A.G., and Lorsch, J.R. (2012). Cold tion complex upon arrival of an AUG et al. also provides new insight into the Spring Harb. Perspect. Biol. 4. Published online codon in the P site (Figure 1C). Although roles of eIF2 and eIF1A in start codon July 18, 2012. http://dx.doi.org/10.1101/cshper- eIF3 was not captured, the structure selection. The most efficient initiation spect.a011544. offers the highest-resolution view to site is an AUG codon in the context of Hussain, T., Lla´ cer, J.L., Ferna´ ndez, I.S., Munoz, A., Martin-Marcos, P., Savva, C.G., Lorsch, J.R., date of the mRNA, eIF2,GTP,Met-tRNAi the Kozak consensus sequence (Kozak, Hinnebusch, A.G., and Ramakrishnan, V.