Review: Emerging roles of the signaling network of the protein kinase GCN2 in the plant stress response

Abstract

The pan-eukaryotic protein kinase GCN2 (General Control Nonderepressible2) regulates the translation of mRNAs in response to external and metabolic conditions. Although GCN2 and its substrate, translation initiation factor 2 (eIF2) α, and several partner proteins are substantially conserved in plants, this kinase has assumed novel functions in plants, including in innate immunity and retrograde signaling between the chloroplast and cytosol. How exactly some of the biochemical paradigms of the GCN2 system have diverged in the green plant lineage is only partially resolved. Specifically, conflicting data underscore and cast doubt on whether GCN2 regulates amino acid biosynthesis; also whether phosphorylation of eIF2α can in fact repress global translation or activate mRNA specific translation via upstream open reading frames; and whether GCN2 is controlled in vivo by the level of uncharged tRNA. This review examines the status of research on the eIF2α kinase, GCN2, its function in the response to xenobiotics, pathogens, and abiotic stress conditions, and its rather tenuous role in the translational control of mRNAs.

Keywords: Amino acids; Innate immunity; Reactive oxygen; Ribosome stalling; Translation initiation; tRNA.

Figure 1.

(A) Domain architecture of GCN2. At, Arabidopsis thaliana; Sc, Saccharomyces cerevisiae; Hs, Homo sapiens; Dm, Drosophila melanogaster; Ce, Caenorhabditis elegans. Numbers denote amino acid positions. RWD, a domain common in Ring-finger and WD40-domain proteins and yeast DEAD (DEXD)-like helicases; interacts with GCN1. Linker (+/−), Region with mostly charged amino acids (Arg, Lys, Gln and Asp) found between the RWD and pseudokinase domain. Pseudokinase (ΨPK), a domain similar to the eIF2α kinase domain that lacks crucial residues for catalytic activity and may function as a regulatory domain. EIF2α kinase, catalyzes the phosphorylation of eIF2α; double asterisks (**) symbols denote two conserved threonine autophosphorylation sites within the activation sub-domain. HisRS, histidinyl-tRNA synthetase (HisRS)-like domain. CTD, carboxy-terminal domain, functions in dimerization and ribosome-binding and together with the HisRS-like domain, inhibits kinase activity in yeast [105]. The m2 motif (pink rectangle) within the HisRS domain is a conserved YR dipeptide motif required for tRNA binding [19]. Also shown (*) is the TOR kinase phosphorylation site at Serine (Ser) 577 in ScGCN2, which is not conserved in AtGCN2. The amino acid coordinates of the individual domains are assigned based on [15, 22, 72, 106, 107] and NCBI protein domain prediction. CTD for DmGCN2 and CeGCN2 is predicted based on the presence of amino acids at the C-terminal region corresponding to the ScGCN2. AtGCN2 does not have any amino acids beyond the HisRS domain. (B) Model for the activation of ScGCN2. In non-starved yeast cells, interactions between the CTD, HisRS and kinase domains of each monomer keep GCN2 in an enzymatically inactive antiparallel dimer. Consequently, eIF2 continues to support translation initiation by metabolizing GTP and supplying the initiator tRNA^Met to the 40S ribosome. Upon amino acid starvation, uncharged tRNA accumulates, which is bound by the HisRS domain of GCN2, evoking a conformational change, thereby allowing the kinase domain to dimerize in the active back-to-back conformation. This structural rearrangement, in combination with the interactions of the RWD domain with GCN1/GCN20, allows the kinase domain to bind and phosphorylate eIF2α, resulting in suppression of translation initiation. Adapted and modified from [50,105]. One of the monomers of the GCN2 dimer is shown in transparent background for simplicity. PK_N and PK_c are the N- and C-terminal lobe of the kinase domain respectively.

Figure 2.

Translational regulation of the mammalian ATF4 mRNA by availability of active ternary complex (eIF2~GTP~Methionyl-tRNA). Under normal conditions, when active eIF2~GTP is plentiful, the ribosome that resumes scanning after having translated uORF1 (the ‘lure’), quickly acquires a fresh ternary complex, and therefore translates uORF2 (the ‘trap’). Because uORF2 overlaps the main ATF4 mORF, the ribosome bypasses that ATF4 AUG codon by leaky scanning and ATF4 is not translated. Under stress conditions, when ternary complex is in low abundance because of phosphorylation by eIF2 kinases, the ribosome that resumes scanning after uORF1 bypasses the uORF2 AUG before acquiring a fresh ternary complex. This allows the main ORF to be translated (after [26, 108]).

Figure 3.

GCN2 mediated translational repression is part of the integrated stress response in fungi, mammals and plants. MV, methyl viologen. For details see text.

Figure 4.

The GCN signaling network. GCN2 is activated by uncharged tRNA as well as by other signals that may or may not work through uncharged tRNA. The signals include pathogens and associated signaling molecules (the ethylene precursor ACC, salicylic acid [SA] and methyl jasmonate [MeJA]), xenobiotic agents (e.g., chlorosulfuron [CSF], glufosinate ammonium [BASTA], methyl viologen [MV]), dithiothreitol [DTT]), and abiotic stresses (e.g., ultraviolet light). Some of these triggers are known to affect the plastid reactive oxygen (ROS) levels (e.g., excess light, salt, cold, and some herbicides) which may directly or indirectly activate the cytosolic GCN2. GCN2 phosphorylates eIF2α but may potentially have other targets. The kinase activity of GCN2 requires a cofactor, GCN1, which functions with the ATP binding cassette protein ABCF3, known as Gcn20 in yeast and SCORD5 in Arabidopsis [95]. Many of the signals that activate GCN2 also repress translation and may do so through GCN2-dependent (shown) or independent channels (not shown). PRO and SERAT represent genes for amino acid biosynthetic enzymes. 40S and 60S denote the plant cytosolic ribosomal subunits. Single arrows denote influences; double arrows denote equilibria; ---o--- denotes physical interactions; stippled arrows denote events that lack direct evidence in plants.