ATF4 (Activating Transcription Factor 4)

ATF4 is the master transcriptional effector of the integrated-stress-response (ISR): a 351-amino-acid bZIP transcription factor whose translation is selectively upregulated when global cap-dependent translation is suppressed by eif2alpha phosphorylation. Chronic or dysregulated ATF4 activity in aged brain is implicated in cognitive decline, and ATF4 hyperactivation in cancer underlies hypoxia tolerance and therapy resistance.

Identity

UniProt: P18848 (ATF4_HUMAN)
NCBI Gene: 468
HGNC symbol: ATF4 (HGNC:786)
Ensembl: ENSG00000128272
Mouse ortholog: Atf4 (one-to-one ortholog)
Length: 351 amino acids (canonical isoform)
Other names: CREB-2 (historical); TaxREB67 (HTLV-1 Tax-responsive element binding protein 67)
GenAge entry: none identified (#gap/needs-canonical-id — confirm absence via GenAge human search)

Key functional domains

bZIP domain (residues ~278–341) — mediates sequence-specific DNA binding at CRE-like sites (consensus: 5’-GTGACGT[AC][AG]-3’) and homo/heterodimerization with other bZIP family members (ATF3, CHOP/DDIT3, CEBPB, FOS, JUN)
Basic motif (~280–300) — direct DNA contact
Leucine zipper (~306–334) — dimerization interface
No canonical activation domain of the acidic-blob type; transcriptional activation is context- and partner-dependent

Translational regulation: the uORF mechanism

ATF4 mRNA carries an unusually structured 5’ UTR with two upstream open reading frames (uORFs):

uORF1 (short, ~3 codons) — permits ribosome re-initiation at the downstream ORF after translation; acts as a positive regulatory element
uORF2 (overlapping with the ATF4 main ORF in frame position) — under normal ternary-complex-abundant conditions, ribosomes that translate uORF1, reload a new ternary complex quickly, and re-initiate at uORF2 instead of the ATF4 ORF; ATF4 protein is not produced

When eif2alpha is phosphorylated (by perk, GCN2, HRI, or PKR), the ternary complex (eIF2-GTP-Met-tRNA) becomes limiting. Ribosomes that translate uORF1 then scan forward but cannot re-initiate at uORF2 because ternary complex reloading is slow; they bypass uORF2 and re-initiate at the ATF4 main ORF instead. The net result: global protein synthesis falls, but ATF4 translation rises ¹ ².

This mechanism is conserved from yeast (where the ortholog is GCN4, regulated by a four-uORF mechanism and GCN2) to mammals, and is the molecular basis for ISR selectivity.

Dimension	Status
Pathway conserved in humans?	yes
Phenotype conserved in humans?	yes
Replicated in humans?	yes

ATF4 as ISR master effector

Once translated, ATF4 is a broadly acting transcription factor driving a gene-expression programme that — in the acute phase — is adaptive and pro-survival:

atf3 — a feedback repressor; ATF3 heterodimerizes with ATF4 and dampens ISR gene expression over time
chop (DDIT3) — a bZIP/death domain TF; ATF4-CHOP heterodimer shifts the response toward apoptosis / terminal UPR when stress is unresolvable
gadd34 (PPP1R15A) — the p-eIF2α phosphatase regulatory subunit; ATF4-induced GADD34 recruits PP1 to dephosphorylate eIF2α, providing negative feedback and allowing translational recovery
Asparagine synthetase (ASNS) — amino acid metabolic adaptation
Amino acid transporters — SLC7A11 (xCT), SLC7A5 (LAT1), and related solute carriers; upregulated to increase amino acid import under nutrient stress
NRF2 (NFE2L2) targets — indirect: ATF4 and NRF2 share target loci and interact at the AR-like element; oxidative stress coregulation ³
Autophagy genes — ATF4 has been reported to induce MAP1LC3B (LC3B) and other ATGs under ER stress, linking the ISR to autophagy needs-replication

The ATF4 transcriptional programme is transient by design: GADD34-mediated eIF2α dephosphorylation and ATF3-mediated repression together shut down ATF4 activity once stress resolves.

Post-translational regulation

ATF4 protein stability is tightly controlled:

SCF(β-TrCP) E3 ubiquitin ligase ubiquitinates ATF4 at a phosphodegron requiring sequential phosphorylation at Ser219 (by RSK) and Ser215 (by GSK3); proteasomal degradation under basal conditions keeps ATF4 low ³
PHD3 (EGLN3) hydroxylates ATF4 at Pro>XX (exact site under study), linking ATF4 stability to oxygen sensing; may contribute to hypoxia-induced ATF4 accumulation needs-replication
p300/EP300 acetylates Lys311, affecting transcriptional potency in a context-dependent manner
SUMOylation at Lys53, Lys259, Lys267, Lys272 (multiple SUMO sites; functional significance of individual sites not fully resolved) unsourced — sites from UniProt P18848; primary citation for functional significance not confirmed

Role in aging

Chronic ISR activation and cognitive aging

In aged mammals, multiple stressors — accumulated unfolded proteins, mitochondrial dysfunction, oxidative stress, reduced proteostasis capacity — sustain low-grade eIF2α phosphorylation and thus chronically elevated ATF4 activity in post-mitotic neurons. This constitutive ISR activation is thought to impair long-term potentiation (LTP) and memory consolidation by suppressing overall protein synthesis required for synaptic plasticity.

Evidence (mouse):

Krukowski et al. 2020 administered ISRIB (ISR inhibitor, which reactivates eIF2B to restore ternary-complex levels despite p-eIF2α) to aged mice and observed reversal of age-related spatial and working memory deficits, with concurrent reduction of elevated ATF4 levels in aged brain tissue ⁴. needs-replication — single study; ISRIB mechanism also involves eIF2B subunit stabilization independently of ATF4 per se.

Dimension	Status
Pathway conserved in humans?	yes — ISR is highly conserved
Phenotype conserved in humans?	partial — age-related cognitive decline is universal, but role of ATF4 specifically is not directly shown in humans
Replicated in humans?	no — only mouse pharmacology; no human ATF4-cognitive-aging trial data

needs-human-replication — Whether suppressing ATF4/ISR activity improves cognitive aging in humans is untested. ClinicalTrials.gov search for ISRIB or ISR inhibitors in aging/cognitive decline returned no completed trials as of 2026-05.

ATF4 knockdown in aged neurons

ATF4 knockdown (shRNA/antisense in aged-brain mouse models) produces partial rescue of synaptic and memory phenotypes in some reports unsourced — specific DOI not confirmed; tag for verification. The claim is mechanistically coherent but needs a primary citation.

Role in disease

Cancer: drug resistance and hypoxia tolerance

ATF4 is hyperactivated in many solid tumors via:

Hypoxia — tumor hypoxia activates HIF-1α-independent eIF2α phosphorylation (HRI branch) and may stabilize ATF4 via PHD3 inhibition
Nutrient deprivation — GCN2-mediated eIF2α phosphorylation under amino acid scarcity (common in nutrient-poor tumor microenvironments) drives ATF4-dependent amino acid transporter induction (SLC7A11, SLC7A5), enabling survival
Chemotherapy stress — many genotoxic agents trigger ISR via PERK or HRI, transiently inducing ATF4 and its pro-survival gene targets

ATF4-driven upregulation of ASNS confers asparagine self-sufficiency, enabling resistance to L-asparaginase in ALL and other malignancies unsourced.

UPR-PERK arm coordination

ATF4 is the sole transcriptional output of the perk arm of the unfolded-protein-response. The three UPR branches (PERK/ATF4, IRE1/XBP1, ATF6) converge on overlapping but distinct target genes; ATF4 specifically drives the CHOP-dependent apoptotic output when ER stress is unresolvable. Cells with hypomorphic PERK show enhanced ER stress sensitivity, confirming the ATF4 axis is required for adaptive translational regulation ⁵.

Pathway membership

integrated-stress-response — central effector; sole transcription factor selectively translated under p-eIF2α
unfolded-protein-response — PERK arm output
perk — kinase upstream of p-eIF2α → ATF4 axis (PERK protein page)
autophagy — indirect; ATF4 may induce autophagy genes under prolonged ISR

Key interactors

eif2alpha — upstream; when phosphorylated, licenses ATF4 translation via uORF bypass
perk — upstream eIF2α kinase for ER stress; one of four mammalian eIF2α kinases
chop — heterodimer partner; ATF4-CHOP drives apoptotic gene expression
atf3 — heterodimer partner and feedback repressor; dampens ATF4 targets
gadd34 — ATF4 transcriptional target; GADD34-PP1 dephosphorylates p-eIF2α (negative feedback)
nrf2 — co-regulator at overlapping oxidative-stress response elements

Pharmacological context

There is no approved drug targeting ATF4 directly. Indirect modulators include:

ISRIB (integrated stress response inhibitor) — research tool; stabilizes eIF2B, restores ternary complex and suppresses ATF4 induction; demonstrates cognitive rescue in aged mice ⁴; no human trial data
Salubrinal / Guanabenz / Sephin1 — pharmacological eIF2α phosphatase inhibitors (block GADD34-PP1); prolong ATF4 induction as a strategy for certain diseases (not aging)
GSK3 inhibitors — block ATF4 Ser215 phosphorylation and thus delay SCF(β-TrCP) degradation; may modulate ATF4 levels in neurodegeneration contexts dose-response-unclear

Limitations and gaps

#gap/needs-human-replication — ATF4’s role in human cognitive aging is entirely inferential from mouse ISR pharmacology
#gap/needs-replication — ATF4 knockdown rescue of aged-neuron phenotypes: no primary DOI confirmed; needs sourcing
#gap/unsourced — ATF4-induced autophagy gene upregulation: mechanistically plausible, primary citation not confirmed
#gap/unsourced — SUMOylation functional significance: sites from UniProt P18848; functional data not confirmed
#gap/unsourced — ATF4/ASNS in L-asparaginase resistance: claim needs primary citation
#gap/no-mechanism — PHD3 hydroxylation site and oxygen-sensing-to-ATF4-stability link is incompletely characterized
#gap/needs-canonical-id — GenAge ID: no entry found; confirm absence against GenAge-human and GenAge-models databases
#gap/needs-canonical-id — HGNC ID 786 assumed from standard nomenclature; not independently verified against HGNC database; cross-check on next lint pass

Footnotes

doi:10.1016/s1097-2765(00)00108-8 · Harding HP, Novoa I, Zhang Y et al. · Molecular Cell 2000 · n=N/A · in-vitro + in-vivo (mouse) · design: mechanistic · model: mammalian cells / mouse; 3199 citations; shows eIF2 kinases selectively enhance ATF4 production under stress ↩
doi:10.1073/pnas.0400541101 · Vattem KM, Wek RC · PNAS 2004 · n=N/A · in-vitro · design: mechanistic · model: rabbit reticulocyte lysate + mammalian cell culture; 1669 citations; defines uORF1/uORF2 re-initiation mechanism controlling ATF4 translation ↩
doi:10.1083/jcb.200408003 · Lu PD, Harding HP, Ron D · J Cell Biology 2004 · n=N/A · in-vitro + in-vivo · design: mechanistic · model: MEFs and mouse; 980 citations; ISR gene expression regulation via uORF re-initiation; ATF4 as ISR integration point ↩ ↩²
doi:10.7554/eLife.62048 · Krukowski K et al. · eLife 2020 · n=N/A (multiple cohorts, ~10-20/group typical for this design) · in-vivo · design: randomized (mouse, aged C57BL/6) · model: aged mice; 164 citations; ISRIB reverses age-related memory decline with reduced ATF4 in aged brain ↩ ↩²
doi:10.1016/s1097-2765(00)80330-5 · Harding HP, Zhang Y, Bertolotti A et al. · Molecular Cell 2000 · n=N/A · in-vitro + in-vivo (mouse) · design: mechanistic · model: PERK-knockout cells; 1988 citations; establishes PERK as required for translational regulation during UPR ↩

🧬 Aging Wiki

Explorer

atf4