KRAS-G12D Binding Affinity Selectivity
The 1-oxa-3,7-diazaspiro[4.5]decan-2-one scaffold has been validated as a selective KRAS-G12D inhibitor class, with lead compounds HDB-2 and HDB-3 achieving strong protein binding affinities (Kd of 28.29, 48.17, and 85.17 nM) and nanomolar cellular IC50 values [1]. While the specific Kd of 3-(2-Methylpropyl)-1-oxa-3,7-diazaspiro[4.5]decan-2-one for KRAS-G12D has not been reported in the public domain, its branched isobutyl side chain introduces critical steric volume at the N-3 position. This substitution is structurally distinct from the 3-butyl linear isomer, which has a different shape complementarity profile for the Switch-II hydrophobic pocket. The position of the second nitrogen also defines functional divergence; the 3,7-diazaspiro arrangement creates a different electronic distribution on the piperidine ring compared to the 3,8-diazaspiro isomer, potentially altering hydrogen bond acceptor/donor properties for target engagement.
Class ref. Kd range: 28–85 nM
| Evidence Dimension | KRAS-G12D Protein Binding Affinity (Kd) |
|---|---|
| Target Compound Data | Not publicly reported for this specific compound |
| Comparator Or Baseline | Class-leading derivative HDB-2 (1-oxa-3,7-diazaspiro[4.5]decan-2-one derivative): Kd = 28.29 nM; HDB-3: Kd = 48.17 nM; other derivative: Kd = 85.17 nM [1] |
| Quantified Difference | Not calculable; class-level inference only |
| Conditions | In vitro protein binding assay (specific method not detailed in abstract); cell-based assays in KRAS-G12D-driven cancer cell lines [1] |
Why This Matters
Demonstrates that the 1-oxa-3,7-diazaspiro[4.5]decan-2-one core supports nanomolar target engagement, and the specific 3-(2-methylpropyl) substituent can be rationally designed to occupy the hydrophobic pocket, making it a non-interchangeable building block for structure-activity relationship (SAR) exploration.
- [1] Ryu S, Baig MH, Panwar U, et al. 1-oxa-3,7-diazaspiro[4.5]decan-2-one derivatives as potent KRAS-G12D inhibitors: A multidisciplinary approach integrating machine learning, synthesis, and biological evaluation. Comput Biol Med. 2026;202:111437. View Source
