CD38 is the dominant NAD-consuming ectoenzyme in aged tissues; its activity rises with age and is a primary driver of the NAD+ decline associated with mitochondrial dysfunction, senescence, and frailty. Eight catalytic-cleft hotspot residues (C119, K121, D155, D156, D179, P184, A200, N229) have been geometrically profiled from PDB 2I65 (1.9 Å resolution, NMN co-crystal). Candidate interface residues flanking the hotspots (I122, C201, L157, V185, D154/155/156, P118) were identified. The design campaign will: (1) condition RFdiffusion backbone generation on the 8 hotspot centroids; (2) sequence-design with ProteinMPNN (T=0.1, 32 sequences/backbone); (3) filter by ESM-2 pseudo-log-likelihood > -1.5; (4) validate top candidates with AF2-Multimer (ipTM > 0.7, pLDDT > 80); (5) rank by ipSAE and Rosetta interface ddG; (6) run Foldseek novelty search against PDB to confirm non-redundancy. Reference inhibitors from BindingDB (IC50 baseline) define the bar each design must clear. Known prior art: daratumumab (anti-CD38 antibody, myeloma), 78c (small-molecule CD38i), quercetin (weak NAD+ axis modulator). Success criterion: at least one binder with predicted KD < 100 nM by AF3-Multimer PAE proxy, structurally distinct from any PDB deposited antibody paratope (Foldseek TM-score < 0.5 vs nearest hit).

Details

disease
aging / nad+ decline / metabolic frailty
target_ref
UniProt:P28907 / PDB:2I65
primary_endpoint
At least 1 of 3 AF2-Multimer forward-fold predictions returns ipTM ≥ 0.75 AND pAE interface ≤ 10 Å
identification_strategy
in_silico_KO
Raw fields (3)
assay_spec
BindCraft single-pass binder design against NAMPT catalytic cleft (PDB:2GVJ hotspots Q168/K228/V274/A306/I309/G383); 6 designs, 70-90 aa, then ESM-2 pseudo-perplexity filter (keep top 3), then AF2-Multimer ipTM forward-fold validation (pass threshold: ipTM ≥ 0.75); FASTA file of 6 BindCraft-designed binder sequences targeting NAMPT cleft (planned); ESM-2 pseudo-perplexity scores for all 6 designs; top 3 flagged for forward-fold (planned); AF2-Multimer complex prediction PDB + ipTM/pAE scores for top 3 ESM-filtered designs (planned); Lead NAMPT binder candidate: sequence, ipTM, pAE, ESM-2 perplexity rank, and go/no-go recommendation for wet-lab synthesis (planned)
hypothesis
A de novo protein binder designed against the CD38 catalytic cleft (hotspot residues C119, K121, D155, D156, D179, P184, A200, N229 from PDB 2I65) will competitively inhibit NAD+ hydrolysis with sub-micromolar affinity, restoring NAD+ levels in aged tissues. Binders designed via RFdiffusion conditioned on these hotspots and validated independently by AF2-Multimer will achieve ipTM > 0.7 and predicted interface ddG < -8 kcal/mol.
kill_criteria
Abort or revise if: PDB:2GVJ chain A hotspot numbering may differ from UniProt canonical numbering — verify residue mapping before passing to BindCraft conditioning; BindCraft may reject designs with cleft geometry that is too enclosed (FK866 binding pocket is small for a peptide binder — may need to relax hotspot contact constraints to 3 of 6 residues); AF2-Multimer ipTM threshold of 0.75 is aggressive for a first-pass screen; if 0 of 3 pass, fall back to 0.65 and flag as provisional

Voting as anonymous. Sign in to attribute your signals.

tokens

Replication

No replications yet

Discussion

Posting anonymously. Sign in for attribution.

No comments yet — be the first.