EIF2D

eIF2D is a noncanonical translation factor that acts at the interface of 40S ribosomal subunit recycling and reinitiation, governing the post-termination and noncanonical-initiation fates of ribosomes (PMID:30146315, PMID:34016977). Structurally it pairs an eIF1-like domain (linked to scanning and initiation codon selection) with a SWIB/MDM2 fold not previously seen among initiation factors, providing the architecture for its dual roles (PMID:28736176). In yeast and in 40S footprinting assays, eIF2D (Tma64) recycles 40S subunits stranded at stop codons, preventing aberrant reinitiation at downstream AUG and 3' UTR codons, although the MCT-1/DENR heterodimer carries the bulk of canonical stop-codon recycling and eIF2D plays a comparatively minor role there (PMID:30146315, PMID:34016977). eIF2D's mechanistic specialization is its winged helix domain, absent from MCTS1-DENR, which enhances 40S binding and directs eIF2D toward recycling during intrinsic ribosome destabilization at nascent N-terminal acidic stretches, a context where it would clash with ABCE1 in conventional termination-coupled recycling (PMID:41335470). Through these activities eIF2D supports specialized translation programs: it mediates ATF4 translational induction during the integrated stress response via its RNA-binding motif and 5' leader uORFs (PMID:32938929), drives repeat-associated non-AUG (RAN) translation of C9orf72 GGGGCC expansions into toxic dipeptide-repeat proteins (PMID:34654821), and enables IRES-independent translation of the enterovirus genome when the eIF2-dependent pathway is inactivated (PMID:36689548). Its reinitiation function at specific uORFs is mechanistically distinct from MCTS1-DENR, and eIF2D loss causes widespread gene deregulation beyond uORF control (PMID:39748120). eIF2D is dispensable for HCV IRES-driven and Sindbis virus subgenomic mRNA translation (PMID:28240315, PMID:29487587). Translation of eIF2D's own mRNA is downregulated during hyperosmotic stress through a uORF stop-codon-dependent mechanism (PMID:30419262).

1. 2017 High

   Establishing the domain architecture of eIF2D was needed to rationalize how a single factor could engage the ribosome like a canonical initiation factor yet carry an unusual module; the structure revealed an eIF1-like domain plus a SWIB/MDM2 fold.

   Evidence X-ray crystallography of the human eIF2D C-terminal domains at 1.4 A

   PMID:28736176

   Open questions at the time

   • Full-length structure and the role of the N-terminal/winged helix regions not resolved
   • No ribosome-bound structure to show how the domains contact the 40S
   • Functional consequences of inter-domain interactions untested
2. 2018 High

   It was unknown whether eIF2D acts physiologically after termination; deletion studies showed it (with MCT-1/DENR) recycles 40S subunits at stop codons and suppresses aberrant downstream reinitiation.

   Evidence Ribosome profiling of yeast deletion strains, 3' UTR reporters, in vitro translation

   PMID:30146315

   Open questions at the time

   • Relative contribution of eIF2D vs the MCT-1/DENR heterodimer not quantified
   • Direct biochemical demonstration of 40S release by eIF2D alone lacking
3. 2018 Medium

   To define the boundaries of eIF2D function in noncanonical viral translation, knockouts tested its requirement for IRES- and stress-driven viral synthesis; eIF2D proved dispensable for HCV IRES and Sindbis subgenomic mRNA translation.

   Evidence CRISPR/Cas9 and siRNA depletion in HAP1 cells with IRES-luciferase and viral protein synthesis assays

   PMID:28240315 PMID:29487587

   Open questions at the time

   • Negative results do not exclude redundancy with other factors
   • Single-lab findings without orthogonal confirmation
4. 2018 Medium

   How eIF2D's own expression is tuned under stress was unknown; reporter dissection revealed selective downregulation during hyperosmotic stress via a uORF stop-codon-dependent mechanism distinct from delayed reinitiation or stalling.

   Evidence Reporter mRNA assays and the FLERT technique with uORF mutational analysis

   PMID:30419262

   Open questions at the time

   • Trans-acting factors mediating this autoregulation not identified
   • Physiological consequence of lowered eIF2D during stress not measured
5. 2020 High

   Whether eIF2D contributes to stress-responsive gene expression was open; genetic and domain analyses placed eIF2D/DENR as critical for ATF4 translational induction during the integrated stress response, requiring its RNA-binding motif.

   Evidence Drosophila loss-of-function epistasis, RNA-binding motif mutagenesis, human cell ATF4 induction under ER stress

   PMID:32938929

   Open questions at the time

   • Direct RNA target sequences bound by the RNA-binding motif not mapped
   • Step in the ATF4 5' leader reinitiation cycle that eIF2D acts on not pinpointed
6. 2021 High

   The minor stop-codon recycling role of eIF2D was clarified by direct intermediate detection; 40S footprinting confirmed unrecycled 40S accumulation upon TMA64 loss but assigned the bulk of recycling to MCT-1/DENR.

   Evidence 40S selective ribosome footprinting in yeast deletion strains

   PMID:34016977

   Open questions at the time

   • Why eIF2D contributes little at canonical stop codons left unexplained until later structural reasoning
   • Substrate contexts where eIF2D dominates not defined here
7. 2021 High

   It was unknown whether eIF2D participates in pathogenic noncanonical initiation; loss-of-function across species showed eIF2D promotes RAN translation of C9orf72 repeats, reducing dipeptide-repeat toxicity.

   Evidence C. elegans loss-of-function genetics, mammalian cell assays, lifespan readouts measuring poly-GA/poly-GP

   PMID:34654821

   Open questions at the time

   • Mechanism by which eIF2D initiates at the repeat (codon/frame selection) not resolved
   • Whether this is direct or via its recycling activity unclear
8. 2023 Medium

   The scope of eIF2D in viral noncanonical initiation was extended; depletion showed eIF2D with eIF2A is required for IRES-independent enterovirus genome translation when eIF2 is inactive, enabling recombination.

   Evidence Genetic depletion of eIF2A/eIF2D, viral replication/recombination assays, translation reporters

   PMID:36689548

   Open questions at the time

   • Individual contribution of eIF2D vs eIF2A not separated
   • Single-lab finding with viral readout rather than direct biochemistry
9. 2025 High

   The basis for eIF2D's mechanistic divergence from MCTS1-DENR was unresolved; TCP-seq and domain analysis showed eIF2D specializes in 40S recycling during intrinsic ribosome destabilization via its unique winged helix domain, which clashes with ABCE1 at canonical termination.

   Evidence TCP-seq selective translation complex profiling, eIF2D-deficient cells, domain structure-function analysis

   PMID:41335470

   Open questions at the time

   • Proposed ABCE1 clash inferred rather than structurally demonstrated
   • Full set of IRD-prone substrates in vivo not enumerated
10. 2025 Medium

    Whether eIF2D's reinitiation role overlaps with MCTS1-DENR was tested; cell-free assays and knockdown profiling distinguished eIF2D's uORF reinitiation activity and revealed widespread gene deregulation beyond uORF control.

    Evidence Cell-free reinitiation assay in HeLa lysates, ribosome profiling in knockdown cells, uORF reporters

    PMID:39748120

    Open questions at the time

    • Direct mechanism linking eIF2D loss to non-uORF gene deregulation unknown
    • Single-lab in vitro reconstitution

• A unified structural and biochemical account of how eIF2D selects which post-termination/noncanonical initiation events to act on, and how its RNA-binding motif and winged helix domain coordinate recycling versus reinitiation in vivo, remains open.
• No ribosome-bound structure of eIF2D capturing recycling or reinitiation
• RNA target specificity of the RNA-binding motif not defined
• Rules dictating when eIF2D versus MCTS1-DENR is deployed not established