FOXO4-DRI is a D-retro-inverso peptide that disrupts the FOXO4–p53 transactivation domain interaction, selectively driving apoptosis in senescent cells. The published proof-of-concept (Baar et al. 2017, Cell) shows in vivo senolytic activity in mice but the original peptide has suboptimal pharmacokinetics and moderate potency. This plan uses PDB structural data for the FOXO4 forkhead domain (3L2C geometry retrieved this tick), literature-mined FOXO4–p53 interface residues, and BindCraft or RFdiffusion + ProteinMPNN to design improved L-amino-acid stapled peptides or miniprotein binders. Success criteria: planned BindCraft run producing ≥20 candidates → AF2-Multimer forward-fold validation (pLDDT > 80, ipTM > 0.7) → ESM-2 log-likelihood filter → Foldseek novelty screen vs PDB → top 3 candidates recommended for SPR synthesis. Prior art: FOXO4-DRI (Baar 2017), navitoclax comparator (BCL-XL senolytic). Hotspot residues on FOXO4 DBD confirmed present: W97, G98, N99, Q100, S101, Y102, A103, L105 (pdb_hotspot_geometry retrieved in prior tick). Note: 3L2C is the FOXO4 DBD–DNA complex; the FOXO4–p53 co-complex structure is the primary design target and literature search this tick will clarify the best structural template for conditioning.
Details
- disease
- cellular senescence / aging
- target_ref
- UniProt:O43524
- primary_endpoint
- ≥ 1 of 5 trajectories returns a design with ipTM ≥ 0.7 and pLDDT ≥ 80
- identification_strategy
- in_silico_KO
Raw fields (3)
- assay_spec
Fetch PDB:3L2C (FOXO4-DBD apo crystal); extract chain A hotspot residues W97, G98, N99, Q100, S101, Y102, A103, L105, K135; run BindCraft (50-residue binder, 5 trajectories) conditioned on that patch; filter outputs by ipTM ≥ 0.7 and pLDDT ≥ 80; rank survivors by ipSAE; planned — BindCraft output PDB(s) for top-ranked FOXO4 binder(s) passing ipTM ≥ 0.7 and pLDDT ≥ 80; planned — ranked table of BindCraft trajectories with per-design ipTM, pLDDT, ipSAE, and FOXO4-hotspot contact count; top candidate flagged as lead; {'kind': 'claim', 'text_template': 'Lead binder achieves ipTM ≥ 0.7 against FOXO4-DBD hotspot patch (W97–K135) with pLDDT ≥ 80 across ≥ 1 of 5 BindCraft trajectories'}- hypothesis
A de novo peptide binder designed against the FOXO4 forkhead domain interaction surface with p53-TAD will outperform FOXO4-DRI on at least one in-silico metric (predicted Kd proxy, ipTM, or pLDDT) while retaining selectivity for senescent cells over non-senescent cell types.
- kill_criteria
Abort or revise if: 3L2C is an apo DBD crystal — no p53-TAD co-structure in PDB; hotspot patch inferred from FOXO4-DRI mutagenesis data, not direct co-crystal contacts; may misplace binder epitope; 5 trajectories is a small batch; success probability depends on BindCraft's internal MPNN diversity — if all 5 fail QC, next gate should increase to 20 trajectories or switch to RFdiffusion partial-diffusion seeded from DRI helix geometry; ipSAE reference baseline for FOXO4-DRI requires AF2-Multimer prediction of DRI peptide + 3L2C chain A; that call is not yet in substrate and must be run in same tick or flagged as a blocker