A Versatile Eif3d in Translational Control of Stress Adaptation

Spotlight A Versatile eIF3d in Translational Control of Stress Adaptation

Longfei Jia1 and Shu-Bing Qian1,* 1Division of Nutritional Sciences, Cornell University, Ithaca, NY, USA *Correspondence: [email protected] https://doi.org/10.1016/j.molcel.2020.12.016

Lamper et al. (2020) reported that eIF3d-mediated cap-dependent translation is subject to regulation by phosphorylation during chronic glucose deprivation, providing a mechanism underlying selective translation of stress genes essential for cell survival.

A remarkable feature of all living organ- tion, however, overall translation is because eIF3d contains a natural HIV-1 isms is the ability to sense fluctuations reduced only by 60% (An et al., protease cleavage site. Remarkably, of environmental cues and respond to 2020). This does not necessarily mean glucose starvation led to an increased adverse conditions by adjusting cellular that the remaining 40% of translation cap-binding activity of eIF3d by 10- activities. Protein synthesis consumes solely relies on cap-independent mech- fold. Since phosphorylation is a common a lion’s share of cellular energy. Not sur- anisms. Notably, most mRNAs are still mechanism in stress signaling pathways, prisingly, many stress conditions sup- capped when the canonical eIF4F is in- the authors confirmed nutrient-depen- press global protein synthesis as part hibited. Although cap-dependent trans- dent phosphorylation of eIF3d. Using of the stress response. However, translation is widely believed to be driven biochemical and genetic approaches, lation of certain mRNAs needs to be through eIF4F, alternative mechanisms they not only mapped the phosphoryla- maintained or even upregulated to sus- likely exist to mediate cap-dependent tion sites but also identified CK2 as tain vital functions. Given the diverse mRNAtranslationintheabsenceof the responsible kinase for eIF3d phos- types of stressors, how stress-specific functional eIF4F. phorylation. During chronic glucose mRNAs are selected for translation Indeed, a previous study uncovered limitation, the inhibited CK2 leads to has been an area of great interest. eIF3d as another mRNA cap-binding pro- dephosphorylation of eIF3d. The subse- Even for the same type of stress, acute tein (Lee et al., 2016). eIF3d is a subunit of quent increased cap-binding and and chronic stress conditions induce eIF3, the largest and most complex initia- enhanced Jun translation indicate a tight different translational regulation. Stress tion factor that binds to the solvent- regulation of eIF3d-specialized mRNA adaptation is essential for cells to exposed side of the 40S ribosome. eIF3 translation via phosphorylation. Notably, restore cellular homeostasis, but the un- is involved in nearly every step of transla- CK2 has a broad range of phosphoryla- derlying mechanisms remain incom- tion initiation, including ribosome loading, tion substrates including eIF3b, whose pletely understood. Lamper et al., 2020 scanning, and start codon selection (Vala´ - functional coordination with eIF3d merits reported that eIF3d mediates a unique sek et al., 2017). Surprisingly, eIF3 has further investigation. translational program in mammalian been shown to regulate protein synthesis Apparently, Jun is not the only one cells during cellular adaptation to pro- in a selective and mRNA-specific manner mRNA preferentially bound by eIF3d. longed glucose starvation. (Lee et al., 2015). It appears that the cap- However, it is challenging to determine In eukaryotic cells, mRNA translation binding activity of eIF3d defines the role the scope of eIF3d-targeted mRNAs un- typically begins with the binding of of eIF3 in general versus specific mRNA der metabolic stress because eIF3d is eIF4F to the 7-methylguanylate (m7G) translation. Importantly, eIF3d-targeted an integral subunit of the eIF3 complex. cap found on the 50 end of mRNAs. mRNAs, as exemplified by Jun,possessa Lamper et al. (2020) developed an eIF4F is a heterotrimeric complex con- stem loop structure in 50 UTR to block ca- elegant approach by introducing a TEV sisting of eIF4E (cap-binding), eIF4G nonical eIF4F binding. However, whether protease cleavage site proximal to the (scaffold), and eIF4A (RNA helicase) the eIF3d-mediated cap-dependent trans- cap-binding domain of eIF3d. After (Hershey et al., 2019). Under lation is subject to regulation was unclear removing capped mRNAs undergoing the normal growth condition, activated until now (Figure 1). canonical translation, mRNAs with the mTORC1 phosphorylates eIF4E-binding The first clue came from the observa- 50 end cap protected by eIF3d were en- proteins (4E-BPs), permitting cap- tion that chronic glucose deprivation riched followed by deep sequencing. A dependent mRNA translation. A multi- increased translation of Jun. Lamper total of 668 transcripts were uncovered tude of stresses inhibit mTORC1 et al. (2020) measured the cap-binding from glucose deprived HEK293T cells, activity, resulting in hypophosphory- activity of endogenous eIF3d by sepa- suggesting a broad spectrum of eIF3d- lated 4E-BPs that sequester eIF4E rating the cap-binding domain from the programmed translation. Gene ontology (Shu et al., 2020). Upon mTORC1 inhibi- entire eIF3 complex. This is possible analysis revealed that eIF3d-specific

10 Molecular Cell 81, January 7, 2021 ª 2020 Elsevier Inc. ll Spotlight

phosphorylation-regulated eIF3d-mediated translation is specific to glucose deprivation but not serum or glutamine starvation. Does the cell have other types of switches to produce stimulus- specific translatome? Upon hypoxia, cells use different eIF4F variants to reprogram the translational output (Uni- acke et al., 2012). Blocking canonical eIF4F might be a common theme to re- direct mRNAs into specific translation pathways. Future studies will be required to uncover those pathways that enables the critical remodeling of the cellular proteome depending on environmental and physiological conditions of the cell. Given the multiple initi- Figure 1. eIF3d Acts as a Translational Switch between Canonical and Noncanonical Cap- ation factors, diverse RNA-binding pro- Dependent Translation teins, and rich sequence elements, it is Under the normal growth condition, mTORC1 phosphorylates 4EBP, permitting eIF4E-mediated cap- dependent translation. CK2 phosphorylates eIF3d, thereby inhibiting its cap-binding activity. In response likely that their interactions represent a to chronic glucose deprivation, hypophosphorylated 4EBP prevents cap-dependent translation, critical regulatory nexus that determines contributing to suppression of global protein synthesis. However, dephosphorylated eIF3d gains the cap- the translational priorities of mRNAs. binding activity, facilitating selective translation of mRNAs such as Jun and Raptor. This translational reprogramming is essential in metabolic stress adaptation. ACKNOWLEDGMENTS mRNAs are enriched in cell metabolism pands the functional diversity of eIF3. The research about translational control in the Qian and signaling pathways. One interesting Composed of 13 subunits, the 800-kDa laboratory is supported by US National Institutes of Raptor Health (R01GM1222814 and DP1GM142101) and target is , which encodes an eIF3 complex has long been viewed as HHMI Faculty Scholar (55108556) to S.-B.Q. essential component of mTORC1. Like a static entity by recruiting the 40S ribo- Jun Raptor , also bears conserved sec- some to mRNA via interaction with REFERENCES ondary structures in 50 UTR as evi- eIF4F. Accumulating evidence suggests denced by SHAPE, an RNA structure that eIF3 also participates in non-canoni- An, H., Ordureau, A., Ko¨ rner, M., Paulo, J.A., and analysis. Importantly, disruption of the cal translation mediated by the internal Harper, J.W. (2020). Systematic quantitative anal- Raptor ysis of ribosome inventory during nutrient stress. stem loop structure of abolished ribosome entry site (IRES) as well as Nature 583, 303–309. eIF3d cap binding. m6A(Meyer et al., 2015). The remarkable The eIF3d-mediated translational con- feature of eIF3d also reshapes our Guan, B.J., van Hoef, V., Jobava, R., Elroy-Stein, O., Valasek, L.S., Cargnello, M., Gao, X.H., Krokowski, trol of Raptor is quite interesting because current concept of cap-dependent trans- D., Merrick, W.C., Kimball, S.R., et al. (2017). A mTORC1 essentially controls the canonical lation. It is clear that cap-dependent Unique ISR Program Determines Cellular 68 eIF4F-mediated cap-dependent transla- mRNA translation is no longer equated Responses to Chronic Stress. Mol. Cell , 885–900.e6. tion. Although a transient inhibition of with the canonical eIF4F. The deployment mTORC1 might be beneficial for cells un- of eIF3d in cap-dependent translation Hershey, J.W.B., Sonenberg, N., and Mathews, M.B. (2019). Principles of Translational Control. der acute stress, prolonged stress requires enables cells to reprogram their transla- Cold Spring Harb. Perspect. Biol. 11, a032607. cellular adaptation via mTORC1 reactiva- tome, such that proteins that confer Lamper,A.M.,Fleming,R.H.,Ladd,K.M.,and tion. This phenomenon was initially adaptive benefits are preferentially syn- Lee, A.S.Y. (2020). A phosphorylation-regulated documented during chronic endoplasmic thesized. As this new paradigm of eIF3d translation switch mediates cellular reticulum stress (Guan et al., 2017). It is translatome remodeling emerges, a num- adaptation to metabolic stress. Science 370, 853–856. conceivable that eIF3d acts as a switch ber of outstanding questions remain. For via phosphorylation, permitting restoration instance, the eIF3d-specialized mRNAs Lee, A.S., Kranzusch, P.J., and Cate, J.H. (2015). of eIF4F-dependent canonical translation. require presence of conserved stem eIF3 targets cell-proliferation messenger RNAs 0 for translational activation or repression. Nature This translational reprogramming corre- loop structures in 5 UTR. It is unclear 522, 111–114. sponds to the transition from acute stress how those structures block the canonical Lee, A.S., Kranzusch, P.J., Doudna, J.A., and Cate, response to chronic stress adaptation. eIF4F binding. Structural studies revealed J.H. (2016). eIF3d is an mRNA cap-binding protein Failure of such translational switch, as is the cap-binding domain of eIF3d (Lee that is required for specialized translation initiation. 536 the case by introducing phosphomimetic et al., 2016), but how eIF3d recognizes Nature , 96–99. eIF3d, causes cell death under chronic the stem loop remains obscure. Meyer, K.D., Patil, D.P., Zhou, J., Zinoviev, A., glucose deprivation. Another fundamental question is how Skabkin, M.A., Elemento, O., Pestova, T.V., Qian, S.B., and Jaffrey, S.R. (2015). 5’UTR m(6)A The discovery of eIF3d cap-binding ac- cells reprogram translation in response Promotes Cap-Independent Translation. Cell 163, tivity and stress-induced regulation ex- to distinct stimuli. It appears that the 999–1010.

Molecular Cell 81, January 7, 2021 11 ll Spotlight

Shu, X.E., Swanda, R.V., and Qian, S.B. (2020). J., Holcik, M., Pause, A., and Lee, S. (2012). An ox- M.P., Hronova´ , V., Herrmannova´ ,A.,Hashem, Nutrient Control of mRNA Translation. Annu. Rev. ygen-regulated switch in the protein synthesis ma- Y., and Gunisova ´ , S. (2017). Embraced Nutr. 40, 51–75. chinery. Nature 486, 126–129. by eIF3: structural and functional insights into the roles of eIF3 across the trans- Uniacke, J., Holterman, C.E., Lachance, G., Vala´ sek, L.S., Zeman, J., Wagner, S., lation cycle. Nucleic Acids Res. 45, Franovic, A., Jacob, M.D., Fabian, M.R., Payette, Beznoskova´ ,P.,Pavlı´kova´ , Z., Mohammad, 10948–10968.

12 Molecular Cell 81, January 7, 2021