The Bourgeois et al. 2025 Nat Commun paper (DOI 10.1038/s41467-025-60844-9) establishes that the FOXO4 CR3 domain binds the disordered p53-TAD (AD1 and AD2 sub-domains, residues 1–57) rather than the p53 DNA-binding domain. This revises the mechanistic model for FOXO4-DRI and opens a structurally defined design target. Plan: (1) extract p53-TAD hotspot geometry from PDB 6O0K (FOXO4-p53 complex or best available structure); (2) run BindCraft targeting p53-TAD AD1/AD2 hotspot patch; (3) forward-fold all BindCraft outputs with AF2-Multimer (independent of design model); (4) filter by pLDDT > 80, ipTM > 0.7, and ESM-2 pseudo-perplexity < 8; (5) rank survivors by predicted binding affinity proxy (AF3-Multimer pAE) and Foldseek novelty vs. PDB. Success criterion: at least one candidate with AF2 forward-fold RMSD < 2 Å to design model, ipTM > 0.7, and Foldseek TM-score < 0.5 to all known PDB entries (novelty gate). Reference benchmark: FOXO4-DRI peptide (published Kd ~1 µM against p53 TAD by ITC) — designed binder must show predicted affinity improvement. Planned wet-lab triage route: gene synthesis → E. coli expression with His-SUMO tag → ITC or BLI against recombinant p53(1–73)-TAD → senescent IMR90 cell killing assay at 1–10 µM. Open questions filed separately: immunogenicity, off-target MDM2 competition, delivery modality for nuclear p53.
Details
- disease
- aging / cellular senescence
- target_ref
- UniProt:P04637 / p53-TAD residues 1–57
- primary_endpoint
- ≥5 hotspot residues identified with NAD+ contact distance ≤4.5Å in at least 2 of 3 structures
- identification_strategy
- in_silico_KO
Raw fields (3)
- assay_spec
PDB hotspot mapping of CD38 catalytic pocket: fetch CD38 crystal structures (PDB: 2I65, 4TMQ, 5F1K), compute per-residue SASA and contact frequency against NAD+/cNADPH ligands, identify ≥5 hotspot residues (SASA burial ≥30% on ligand binding), output hotspot JSON for RFdiffusion conditioning.; Per-residue SASA burial table (planned): CD38 residues ranked by NAD+ contact frequency across 2I65/4TMQ/5F1K structural alignment; Hotspot JSON (planned): ≥5 residue positions with chain/resnum/burial_fraction for RFdiffusion contig conditioning, linked to PDB sources; {'kind': 'claim', 'text_template': 'CD38 catalytic pocket hotspot set {residue_list} accounts for ≥80% of NAD+ buried surface area across consensus of 3 crystal structures'}- hypothesis
A de novo protein binder targeting the p53 transactivation domain (residues 16–26 and 40–57) at the FOXO4 CR3-contact interface will recapitulate or improve upon FOXO4-DRI senolytic potency while enabling independent AF2-validated structure confirmation and ESM-2 sequence likelihood filtering, yielding candidates with predicted Kd < 100 nM against p53-TAD and selectivity over p53 DBD.
- kill_criteria
Abort or revise if: 4TMQ NAD+ analog may not perfectly recapitulate native NAD+ contacts — cross-validate with 5F1K inhibitor overlap; Chain assignment varies across CD38 PDB entries — must verify correct catalytic chain before SASA computation; DSSP requires DSSP binary in environment; fallback to FreeSASA if unavailable