EIF2AK4

Summary: Conserved tRNA sensing supports the transferred binding function.

Reason: The exact mouse donor has an IDA tRNA-binding record. PMID16601681 demonstrates uncharged-tRNA-dependent kinase activation and a required HisRS-like region; its direct Northwestern assay uses viral RNA, not tRNA. Together with the conserved regulatory architecture, this supports the ISS claim without presenting that assay as purified human tRNA binding. The curated UniProt FUNCTION and DOMAIN statements support uncharged-tRNA binding and the HisRS-like region by similarity to mouse Q9QZ05 (ECO:0000250); they record the same ortholog transfer rather than supply independent evidence or a human binding measurement.

Summary: Fungal Gcn2 checkpoint participation is supported; transfer of that specific role to human EIF2AK4 remains unresolved.

Reason: IPR016255 matches EIF2AK4 and supplies this mapping. PMID17369398 reports Gcn2-dependent eIF2alpha phosphorylation, translation attenuation and delayed pre-replicative-complex formation after UV in fission yeast. PMID19033384 also reports a Gcn2-dependent G1 delay after hydrogen peroxide, using Mcm6 loading to distinguish that delay from the later cell-cycle delay remaining in the mutant. Its stimulus specificity does not refute checkpoint participation: GO:0000077 requires contribution to a DNA damage checkpoint, not a response to every damaging stimulus. The evaluated mammalian studies establish conserved kinase and stress-translation functions but do not resolve conservation of this particular checkpoint role. UNDECIDED records that biological transfer uncertainty, without asserting that the human function is false.

Summary: GCN2 contributes to antiviral defense in the mouse donor system.

Reason: The Q9QZ05 donor cites PMID16601681. Mouse loss-of-function and kinase-dependent antiviral assays support a positive defense contribution; this is an infection-dependent deployment of the translation-control mechanism, not evidence for direct human experiments against every virus.

Summary: GCN2-dependent dendritic-cell function supports an antigen-specific T-cell response.

Reason: The mouse donor is grounded in PMID24310610. Dendritic-cell GCN2 loss impairs cross-presentation and vaccine-induced T-cell responses. Retain that contextual immune contribution without interpreting it as proof of a T-cell-intrinsic GCN2 activation mechanism.

Summary: Adaptive immune responses depend on GCN2 in the tested vaccination context.

Reason: PMID24310610 supports the mouse ortholog transfer through dendritic-cell and vaccination experiments. The result concerns antigen-specific adaptive responses; it is not a universal claim that GCN2 promotes every immune response.

Summary: The established substrate-specific kinase activity is more informative than protein kinase activity.

Reason: PMID25329545 identifies GCN2 as the eIF2α kinase inhibited by p58IPK, but its operative body was not recovered. The specific chemistry is independently supported by the mouse-kinase study PMID10504407 and human GCN2-dependent eIF2α phosphorylation in PMID26102367. This refinement preserves the original experimental assertion without inventing the inaccessible assay details.

Proposed replacements: eukaryotic translation initiation factor 2alpha kinase activity

Summary: The active-site signature supports kinase activity, with a more informative established substrate class.

Reason: IPR008271 is a serine/threonine kinase active-site signature rather than an EIF2AK4-specific substrate predictor. Canonical human Q9P2K8-1 has the functional kinase domain and catalytic Asp848. Independent GCN2 evidence supplies the eIF2α specificity, so the proposed term does not pretend that the motif alone determines the substrate.

Proposed replacements: eukaryotic translation initiation factor 2alpha kinase activity

Summary: The EC mapping is correct but can be refined to eIF2α kinase activity.

Reason: EC 2.7.11.1 describes a broad serine/threonine protein kinase class. EIF2AK4 has an established eIF2α substrate preference; the specific term captures its principal catalytic activity without transferring the physiological effects of phosphorylation into the molecular-function slot.

Proposed replacements: eukaryotic translation initiation factor 2alpha kinase activity

Summary: The mouse kinase evidence supports the specific eIF2α kinase class.

Reason: The exact Q9QZ05 donor includes PMID10504407. That paper assays mouse GCN2 with eIF2α, including the human WT/S51A substrate comparison. Retain the ortholog transfer while distinguishing a human substrate from a directly tested full-length human kinase.

Proposed replacements: eukaryotic translation initiation factor 2alpha kinase activity

Summary: The phylogenetic placement supports inherited eIF2α kinase activity.

Reason: The exact target sequence lies below PTN000867953, which carries the literal positive IBD for this term. The complete selected target path contains no same-term loss assertion. Human target evidence is a legitimate descendant input to this ancestral placement, not circularity; the conserved activity is also independently demonstrated by mammalian experiments.

Summary: The GCN2-family mapping identifies the principal kinase activity.

Reason: IPR016255 is the GCN2 family entry and explicitly maps to eIF2α kinase activity. The human kinase architecture and experimental signaling evidence agree with this mapping; uncertainty about the entry’s separate DNA-checkpoint process does not invalidate the catalytic-family assignment.

Summary: GCN2’s eIF2α kinase activity is a central, independently supported function.

Reason: The original PMID25329545 assertion is consistent with the available abstract and the independently established GCN2 mechanism. Its inaccessible operative details remain an access limitation rather than grounds to overturn this well-supported activity. Human perturbation evidence and the mouse biochemical donor are kept distinct.

Summary: The yeast ortholog supports conserved substrate-specific kinase chemistry.

Reason: The P15442 donor has experimental eIF2α kinase records including PMID1739968. Its substrate-phosphorylation and translational-control evidence is orthologous support, not a human assay. The retained human target kinase domain and independent mammalian evidence make the transfer appropriate.

Summary: The mouse ortholog supplies experimental evidence for the defining catalytic activity.

Reason: Current Q9QZ05 records include direct biochemical evidence for eIF2α phosphorylation. The ISS source remains an ortholog transfer; human cellular studies independently connect EIF2AK4 to the same substrate response. No natural human product is assigned to the mouse construct.

Summary: The citation’s mouse-kinase experiment matches the imported ISS evidence scope.

Reason: PMID10504407 tests mouse GCN2, whereas the human material includes an eIF2α substrate and an incomplete GCN2 EST-derived sequence. Its Ser51 substrate experiment supports this conserved activity. The review does not reclassify it as direct biochemical evidence from full-length human EIF2AK4.

Summary: ATP binding is a supported cofactor property of the catalytic domain.

Reason: The IPR000719/IPR017441 mappings and target ATP-binding features support nucleotide binding. This necessary kinase property adds little functional specificity beyond eIF2α kinase activity and is retained as non-core, without a claim that every alternative product retains equivalent catalysis.

Summary: Positive ancestral placement supports a non-core nuclear-localization inference.

Reason: PAINT places a positive nuclear IBD at PTN000113601, an ancestor of the exact human GCN2 target PTN002475950. This curated ancestral placement is the positive support for the IBA; no selected same-term loss occurs on the target lineage. Its deep placement and sibling-clade donors do not themselves demonstrate erroneous transfer or loss in GCN2. The curated UniProt cytoplasmic location and the evaluated cytosolic human experiments are compatible with, but do not independently demonstrate, a nuclear pool. Retain the nuclear assertion as non-core phylogenetic inference, without presenting it as target-specific microscopy or inferring exclusive cytoplasmic residence.

Summary: Cytoplasmic activity is consistent with the inherited localization and GCN2 mechanism.

Reason: The literal cytoplasmic IBD at PTN000113601 lies on the exact target lineage. The placement agrees with the human protein’s official localization and cytoplasmic translation-control role. This does not imply exclusive residence in the cytoplasm.

Summary: The official subcellular-location mapping supports a cytoplasmic pool.

Reason: UniProt location SL-0086 maps to cytoplasm. The target record and its translation-initiation mechanism support that broad location; the mapping does not determine a single organelle or exclude a condition-dependent nuclear pool.

Summary: The mouse localization transfer is consistent with a cytoplasmic translation-control protein.

Reason: The exact mouse donor join cites PMID23447528. Its Polysome Profiles Methods and Figure 5C–F explicitly assay GCN2 in sucrose-gradient fractions from undifferentiated and differentiated mouse N2a cells, separately from IMPACT. Figure 5E normalizes GCN2 to its input; Figure 5F normalizes it to polysomal rpS6. This supports a cytoplasmic pool without presenting the mouse fractionation experiment as direct human localization.

Summary: The conserved cytosolic pool is appropriate for GCN2’s translation-control activity.

Reason: The cytosol IBD at PTN000867953 lies on the target lineage and includes the experimentally annotated yeast ortholog. This is a phylogenetic localization inference; it is compatible with, and more specific than, the broad cytoplasmic location.

Summary: The mechanism-specific eIF2α-phosphorylation term improves the broad translation-initiation assertion.

Reason: The yeast donor records include PMID1739968. GCN2 regulates initiation through phosphorylation of eIF2α, and this mechanism is supported in human cells by PMID26102367. The replacement captures that step without asserting that global and transcript-selective translation always change in the same direction.

Proposed replacements: regulation of translational initiation by eIF2 alpha phosphorylation

Summary: Protein phosphorylation is correctly assigned but its translation-control mechanism is more informative.

Reason: The P15442 donor’s experimentally supported substrate is eIF2α. Regulation of initiation through this phosphorylation is the relevant process refinement, rather than a claim that EIF2AK4 phosphorylates all proteins or performs the downstream transcriptional program itself.

Proposed replacements: regulation of translational initiation by eIF2 alpha phosphorylation

Summary: The mammalian phosphorylation evidence supports regulation of initiation by eIF2α phosphorylation.

Reason: The current mouse donor includes biochemical and cellular studies of the GCN2–eIF2α axis. The proposed process names the established mechanism while preserving the ortholog source and avoiding an unsupported human construct assignment.

Proposed replacements: regulation of translational initiation by eIF2 alpha phosphorylation

Summary: Learning effects are supported in a defined mouse behavioral model.

Reason: PMID16121183 compares Gcn2-null and wild-type mice. Training strength changes the direction of the memory phenotype; this supports a contextual learning role rather than a uniform enhancement of learning by GCN2 or a direct human behavioral experiment.

Summary: The mouse memory phenotype is retained as a contextual physiological function.

Reason: The exact donor cites PMID16121183. Its long-term behavioral and synaptic observations support the transferred memory role, with dependence on training conditions. They do not identify a separate memory-specific biochemical activity.

Summary: Human GCN2 perturbation supports initiation control through eIF2α phosphorylation.

Reason: PMID26102367 links human SQ20B GCN2 depletion to reduced starvation-induced eIF2α phosphorylation and p21 induction. Its detailed p21 reporter and polysome experiments include mouse systems. The source supports the mechanism without making every assay human or assigning autonomous RNA-binding activity to GCN2.

Summary: The transferred process is the conserved central GCN2 mechanism.

Reason: The Q9QZ05 donor has multiple experimental records for this process, including the biochemical and UV-response studies. Independent human data support the same eIF2α-dependent control of translation initiation. Positive transcript-selective output can coexist with reduced bulk initiation.

Summary: GCN2 inhibits viral translation rather than carrying out viral protein synthesis.

Reason: PMID16601681 reports kinase-dependent restriction of early Sindbis translation. The negative-regulatory term preserves this positive antiviral evidence and its mouse source while distinguishing control of translation from the ribosomal synthesis process itself.

Proposed replacements: negative regulation of viral translation

Summary: Positive yeast and mouse association data support the non-core ribosome-localization inference.

Reason: PMID2038314 supports salt-sensitive association of the yeast donor with ribosomal subunits and polysomes. Independently, PMID23447528 Figure 5C–F detects mouse GCN2 in N2a polysomal fractions and quantifies it separately from IMPACT, relative to input and polysomal rpS6. These positive observations support the transferred, context-dependent association. They do not establish obligatory structural membership or a direct human molecular-contact assay. The preparation-specific failures to capture stable GCN2 association in PMID32610081 and PMID39947182 do not negate that positive evidence. Retain KEEP_AS_NON_CORE for ribosome association without inferring a new binding activity.

Summary: The PAINT placement supports conserved stress-dependent repression of initiation.

Reason: The literal positive IBD at PTN008315168 lies on the target path and is grounded in mouse and fission-yeast evidence. No selected same-term loss is present. Human GCN2-dependent eIF2α signaling supports this conserved regulatory mechanism.

Summary: GCN2 actively represses translation initiation in the tested stress response.

Reason: The exact mouse donor cites PMID12176355. Knockout and eIF2α substrate controls link UV exposure to GCN2-dependent translational attenuation; the experiment supports a stress-regulatory mechanism rather than direct sensing or repair of DNA lesions.

Summary: The amino-acid-starvation response has sound phylogenetic and target grounding.

Reason: The target descends from the positive IBD at PTN008315168, with no selected same-term loss. Human starvation experiments independently support a GCN2-dependent response. Target self-evidence in the donor set is legitimate evidence used to place the ancestral function.

Summary: Amino-acid-starvation signaling is a well-supported cellular GCN2 function.

Reason: PMID25329545’s available abstract is consistent with its imported starvation annotation, although the operative body remains unavailable. The claim is independently supported by human GCN2 perturbation under leucine deprivation in PMID26102367, so the limited cache does not warrant overturning the established function.

Summary: The mouse donor establishes a UV-triggered GCN2 response.

Reason: PMID12176355 uses wild-type and Gcn2-deficient mouse fibroblasts and phospho-eIF2α/translation readouts. Retain the stress-response context without converting it into direct UV detection, DNA-repair chemistry, or demonstrated human DNA-checkpoint signaling.

Summary: GCN2-mediated translational restriction suppresses the tested viral replication.

Reason: The mouse donor PMID16601681 supports host antiviral restriction through kinase activity. GCN2 is a host regulator, not a viral polymerase or a universal inhibitor of all viral infections.

Summary: Negative regulation of neurite development is more precise than broad neuronal differentiation.

Reason: PMID23447528 shows enhanced spontaneous neurite outgrowth after GCN2 depletion or loss in mouse models. NeuN did not change in the primary-neuron comparison. The replacement captures the measured inhibitory morphological role without asserting suppression of every aspect of neuronal differentiation.

Proposed replacements: negative regulation of neuron projection development

Summary: The stress-dependent initiation-repression term specifies the demonstrated context.

Reason: The mouse donor PMID16601681 connects GCN2 kinase activity to translational inhibition during infection. The narrower stress-response term distinguishes this regulated activity from an unconditional repression of translation in every state.

Proposed replacements: negative regulation of translational initiation in response to stress

Summary: Autophosphorylation is part of the supported kinase activation mechanism.

Reason: The exact mouse donor includes biochemical phosphorylation experiments, including PMID16601681. Retain the conserved autophosphorylation process without assigning every observed phosphosite to a particular human alternative product or assuming identical activation thresholds. UniProt PTM assigns the Thr899 activation-related autophosphorylation behavior by similarity to mouse Q9QZ05 (ECO:0000250), not as an independently measured human phosphosite in this review.

Summary: The ARBA prediction is valid but much broader than the established translation-control mechanism.

Reason: Three literal ARBA00028884 condition sets match the captured target InterPro/FunFam fields. The rule predicts regulation of protein metabolism, not a unique substrate. Independent evidence justifies refining this broad output to regulation of initiation by eIF2α phosphorylation.

Proposed replacements: regulation of translational initiation by eIF2 alpha phosphorylation

Summary: Mouse feeding behavior links neuronal GCN2 to amino-acid homeostasis.

Reason: The exact donor PMID16054071 reports impaired aversion to amino-acid-imbalanced diets after brain-specific GCN2 inactivation. The inspected original summary and beginning of Results support this context; complete operative Methods were not recovered. Retain the curator-supported ortholog phenotype without adding human behavioral or circuit-specific claims.

Summary: The reported cold-response context cannot yet be evaluated at operative assay resolution.

Reason: PMID25329545 remains abstract/selected-caption only after bounded original-retrieval attempts. Its positive GCN2/p58IPK report is preserved, but the cold-treatment conditions and target-specific causal evidence were not recovered. General stress-kinase biology alone does not verify this distinct stimulus annotation. UniProt SUBUNIT also describes DNAJC3-mediated inhibition during hypothermic stress, explicitly by similarity (ECO:0000250). This positive curated context is retained, but it does not recover the unavailable stimulus-specific experiment or resolve the human cold-response assignment.

Summary: Selective translation can increase during starvation while bulk initiation decreases.

Reason: The exact mouse donor has positive-initiation records. Independently, PMID26102367 demonstrates a GCN2-dependent starvation response with selective p21 translation, with reporter/polysome work in mouse cells and human signaling evidence. The positive sign is transcript-selective and is not a claim of global translational stimulation.

Summary: The Rhea serine-phosphorylation reaction can be refined to the established eIF2α substrate.

Reason: RHEA:17989 supplies a generic protein-serine kinase mapping. GCN2’s well-supported eIF2α specificity is more informative for this target; the refinement does not imply that the generic reaction itself is incorrect.

Proposed replacements: eukaryotic translation initiation factor 2alpha kinase activity

Summary: The transferred serine kinase activity is most usefully represented as eIF2α kinase activity.

Reason: The exact mouse donor includes substrate-specific biochemical evidence. The refined class retains that chemistry and ortholog provenance while avoiding the implication that a mouse enzyme preparation was a human isoform assay.

Proposed replacements: eukaryotic translation initiation factor 2alpha kinase activity

Summary: GCN2-mediated signaling is the appropriate central process.

Reason: The available PMID25329545 abstract identifies regulation of GCN2 by p58IPK. The target signaling role is independently established in human cells by PMID26102367 and PMID39947182. Detailed cold-treatment or localization conclusions are not borrowed from those independent studies.

Summary: Human GCN2 catalytic signaling participates in the tested ribosome-surveillance response.

Reason: PMID39947182 uses GCN2 loss, WT versus kinase-dead rescue, and an eIF2α phosphomimetic. These support GCN2 signaling through initiation control. RNF10 performs the reported ubiquitination, and failure to capture stable GCN2–ribosome binding does not invalidate the signaling result.

Summary: The human starvation perturbation directly supports a GCN2-mediated signaling response.

Reason: PMID26102367 links GCN2 depletion to altered eIF2α phosphorylation and starvation-responsive p21 induction. The process annotation is appropriate; it does not assign GCN2 the downstream transcription-factor or cyclin-inhibitor molecular activities.

Summary: The mouse ortholog supports the conserved GCN2 signaling pathway.

Reason: The exact donor has experimental pathway annotations across nutrient and stress contexts. Independent human evidence supports the same central kinase-dependent pathway. Context-dependent outputs are not treated as separate catalytic functions.

Summary: Positive support for long-term potentiation is conditional on stimulation strength.

Reason: PMID16121183 shows loss of sustained potentiation after strong stimulation in Gcn2-null slices, despite enhanced responses to a weaker protocol. The positive-regulatory annotation is retained for the former context; it is not generalized into a uniformly positive effect.

Summary: The source supports inhibition of projection development, not execution of projection extension.

Reason: In PMID23447528, GCN2 loss or depletion increases spontaneous neurite outgrowth. The negative-regulatory replacement preserves this demonstrated role and the mouse-to-human inference while avoiding a positive growth-machinery interpretation of GCN2.

Proposed replacements: negative regulation of neuron projection development

Summary: The cellular amino-acid-starvation response is the appropriate specific process.

Reason: The original broad response annotation is consistent with the available PMID25329545 report, while independent human cell evidence establishes a cellular GCN2 response to amino-acid deprivation. The proposed refinement preserves the original source and does not supply unrecovered experimental details.

Proposed replacements: cellular response to amino acid starvation