CD38 is the dominant NAD-consuming enzyme in aged tissues, with its cyclase/hydrolase activity driving NAD+ decline — a hallmark of cellular aging linked to impaired mitochondrial function, sirtuin activity, and DNA repair. Prior work (Camacho-Pereira et al., Cell 2016; Covarrubias et al., Nature Metabolism 2020) establishes CD38 knockout or small-molecule inhibition (apigenin, 78c) restores NAD+ and reverses aging phenotypes in mice. No protein-based competitive binder occupying the catalytic cleft has been reported. Design campaign: (1) use PDB 2PGJ chain A hotspot geometry (C119, T144, D179, N182, C201, S274; neighbor profile already acquired) as conditioning input for RFdiffusion partial-diffusion targeting the cleft; (2) sequence-design backbones with ProteinMPNN (T=0.1, 8 seqs/backbone); (3) filter by ESM-2 log-likelihood (perplexity <10); (4) validate top-5 via AF2-Multimer (ipTM >0.7, PAE cleft contacts <5 Å); (5) rank by ipSAE; (6) Foldseek novelty screen against PDB (TM-score <0.5 to all deposited structures). Success criterion: at least one candidate with AF2-Multimer ipTM >0.75, cleft-contact PAE <5 Å, and Foldseek TM-score <0.5 vs PDB. Wet-lab follow-up: gene synthesis → E. coli periplasmic expression with His-tag → Ni-NTA purification → ITC/SPR against recombinant CD38 (residues 45–300) → NADase activity inhibition assay (fluorometric cADPR/ADPR product quantification).
Details
- disease
- aging
- target_ref
- UniProt:P28907 | PDB:2PGJ
- identification_strategy
- in_silico_KO
Raw fields (4)
- assay_spec
Baseline affinity anchor strategy (BindingDB fallback activated): CrossRef lookup of Camacho-Pereira et al. 2017 (doi:10.1016/j.cmet.2017.12.008) as mechanistic anchor for CD38 NAD-consuming role in aging; CrossRef lookup of a 2023-2024 CD38-selective inhibitor paper for Ki/IC50 anchor. If neither returns affinity data directly, the baseline will be established as: apigenin IC50 ~10 µM vs CD38 (literature-reported, Escande et al. J Nutr Biochem 2013), and the de novo binder design success criterion will be to achieve predicted Kd <1 µM by AF3-Multimer ipTM ≥0.6 AND Rosetta interface ΔΔG ≤−1.5 kcal/mol. Planned output: BindCraft config notebook (tick 8) conditioned on 2PGJ cleft geometry, hotspot_residues [119,144,179,182,201,274], binder_length 65–80 aa, n_designs 100.
- hypothesis
A de novo protein binder designed against the CD38 catalytic cleft (hotspot residues C119, T144, D179, N182, C201, S274 from PDB 2PGJ, 1.71 Å X-ray) will competitively inhibit NAD+ hydrolase activity with selectivity over structurally unrelated NADases, restoring tissue NAD+ levels in aged organisms and representing a superior alternative to small-molecule CD38 inhibitors currently lacking tissue-targeted delivery.
- kill_criteria
Abort cleft-competitive strategy if: (a) PDB 2PGJ chain A residue numbering does not match canonical UniProt P28907 hotspot numbering C119/T144/D179/N182/C201/S274 — must verify against deposited SEQRES before BindCraft config is written; (b) Ki/IC50 literature baseline for known small-molecule inhibitors (apigenin, 78c, daratumumab epitope) cannot be established — benchmark comparison becomes impossible; (c) catalytic cleft geometry from hotspot_profile shows <3 hotspot residues with solvent-exposed neighbors suitable for binder contact.
- primary_endpoint
Tick 7 target: BindingDB returned 0 hits for UniProt P28907 (both IC50 and Ki); EuropePMC query returned 6 papers with 0 directly reporting CD38 inhibitor affinity data. Activating CrossRef fallback per tick-6 assay_spec. Baseline affinity anchors to be established from literature: (1) Camacho-Pereira et al. Cell Metab 2017 (doi:10.1016/j.cmet.2017.12.008) — CD38 as dominant NAD-consumer in aging; (2) selectivity reference for designed binder vs CD157/BST1. Tick 7 success criterion: ≥1 CrossRef lookup returns a confirmed CD38 inhibitor paper with affinity data OR confirms the Camacho 2017 paper as mechanistic anchor → proceed to tick 8: scidex.notebooks.create for BindCraft config JSON conditioned on hotspot residues {C119, T144, D179, N182, C201, S274} from PDB 2PGJ.