The FOXO4–p53 axis maintains senescent cell viability by sequestering p53’s TAD, blocking pro-apoptotic transcription. Three 2025 structural papers (Bourgeois et al. Nat Commun doi:10.1038/s41467-025-60844-9; Kohoutova et al. Nat Commun doi:10.1038/s41467-025-59106-5; Mandal et al. Protein Sci doi:10.1002/pro.4287) now resolve the disordered p53-TAD interaction with FOXO4-DBD at atomic detail, providing hotspot residues for binder conditioning. Plan: (1) Extract interface hotspots from PDB structure 8PNJ (FOXO4-DBD:p53-TAD) via pdb_hotspot_profile. (2) Run BindCraft conditioned on FOXO4-DBD hotspot surface to design 30–50 de novo binders that sterically occlude p53-TAD docking. (3) Forward-fold top 10 candidates with AF2-Multimer (independent oracle; not RFdiff/BindCraft internal). (4) Rank by ipTM, pAE, and ipSAE score; threshold ipTM > 0.7. (5) Filter by ESM-2 sequence likelihood (perplexity < 10). (6) Run Foldseek against PDB to assess novelty (max TM-score < 0.5 to known structures). (7) Select lead candidate; predict KD range via Rosetta ddG. Success criteria: lead binder ipTM > 0.7 on AF2-Multimer, Rosetta ddG < -10 REU relative to FOXO4-DRI reference, Foldseek TM-score < 0.5 (novel scaffold). Planned output: protein_design artifact per candidate, finding artifact for lead, open_question artifact for off-target / immunogenicity risks.
Details
- disease
- aging, cellular senescence
- target_ref
- UniProt:P98177 (FOXO4) / UniProt:P04637 (TP53)
- primary_endpoint
- At least 1 BindCraft design passes ESM-2 perplexity filter (pseudo-perplexity < 10 on ESM-2 650M)
- identification_strategy
- in_silico_KO
Raw fields (3)
- assay_spec
Fetch PDB structure of FOXO4-DBD:p53-TAD complex (DOI 10.1038/s41467-025-59106-5, PDB accession TBD from Kohoutova 2025), extract hotspot residues at the FOXO4-DBD:p53-TAD interface, run BindCraft binder design targeting those hotspots, filter by ESM-2 perplexity, forward-fold top 3 with AF2-Multimer; Interface contact map of FOXO4-DBD:p53-TAD showing hotspot shell residues colored by buried SASA — planned; BindCraft output table: ≤5 candidate sequences with pLDDT, ipTM, iPAE per design — planned; AF2-Multimer predicted complex for top 3 candidates, superimposed on native FOXO4-DBD:p53-TAD reference — planned; {'kind': 'claim', 'text_template': 'Candidate FOXO4-disruptor-v1 achieves ipTM ≥ 0.7 and iPAE ≤ 10 Å against p53-TAD epitope in AF2-Multimer forward-fold — planned'}; protein_design substrate per passing candidate linked to research_plan:5e4168c3 — planned- hypothesis
A de novo protein binder targeting the FOXO4 DNA-binding domain (DBD) at the p53 transactivation domain (TAD) interface — informed by the 2025 NMR/crystallographic structures of the FOXO4-DBD:p53-TAD complex — can outperform FOXO4-DRI on binding affinity and selectivity, selectively eliminating senescent cells while sparing healthy tissue.
- kill_criteria
Abort or revise if: Kohoutova 2025 PDB deposition may not be live yet — fallback is AF2-Multimer of FOXO4-DBD + p53-TAD1-67 which is lower-resolution for hotspot assignment; p53-TAD is intrinsically disordered — BindCraft may produce low-confidence poses; may need to constrain binder design to FOXO4-DBD surface instead of TAD surface; FOXO4-DRI cell-penetrance relies on retro-inverso chemistry; a designed miniprotein will need explicit delivery assessment flagged as open_question