Supplementary figure 1S · Interactive

Computational workflow for HDAC8 inhibitor discovery

A three-phase pipeline integrating structure preparation, docking validation, and post-docking analysis, applied to nine novel spiro[indoline-tetrazolo[1,5-c]quinazoline] derivatives and five reference compounds against an inhibitor-bound HDAC8 structure (PDB 2V5X).

Preparation · A–C

Validation · D–F

Analysis · G–I

Ligand preparation

Novel set: 9 spiro derivatives (KB-268–272, KB-280–283)

3D generation and energy minimisation (RDKit); Gasteiger charges; rotatable bonds assigned automatically.

Protein preparation

Target: HDAC8–inhibitor complex (PDB 2V5X, 2.25 Å)

Waters removed, hydrogens added, side chains optimised; co-crystallized inhibitor retained during prep, removed before docking.

Reference compounds

Controls: 5 reference compounds

PCI-34051, NCC149, entinostat, octanediamide (the 2V5X inhibitor), and a p53-derived substrate peptide (sequence from the companion complex 2V5W).

▼

Docking validation

Redocking: 7 ligand–structure pairs (HDAC8 and related HDAC isoforms)

RMSD < 2.0 Å acceptance criterion. 5 of 6 independent redockings passed (only entinostat in 6UOC marginally exceeded, 2.14 Å).

CB-Dock2 blind docking

Cavity detection: 5 binding pockets

CurPocket curvature-based detection (≈4–5× faster than its predecessor); template (FitDock) and structure-based (AutoDock Vina) scoring.

Experimental validation

MTT assays: KB-268, KB-270, KB-292

IC₅₀ across four cell lines (MCF-7, HCT-116, Jurkat, J774.2); cytotoxicity as a preliminary comparison.

▼

Interaction analysis

Contacts: residue-level binding profile

Discovery Studio Visualizer 2017 R2; CASP contact criteria (van der Waals radii + 0.5 Å).

SAR development

Structure–activity: features vs affinity

Pharmacophore identification; correlation and regression against admetSAR 3.0 descriptors.

ADMET prediction

Drug-likeness: absorption to toxicity

admetSAR 3.0; physicochemical, transporter, metabolism, and toxicity profiles.

Data flow and key outputs

Phase	Input	Method	Key output	Decision point
A–C	Compound structures, PDB 2V5X, reference ligands	Structure preparation, energy minimisation	Docking-ready structures	Quality control passed
D–F	Prepared structures	CB-Dock2 blind docking; MTT cytotoxicity	Binding scores; IC₅₀ values	Validation criteria met
G–I	Docking poses, experimental data	Interaction analysis, SAR, ADMET prediction	Lead identification	KB-282, KB-281 selected

Methodology summary

Preparation (A–C)

Ligand library: 9 novel spiro compounds + 5 reference compounds
Target selection: inhibitor-bound HDAC8 (2V5X), a well-resolved ligand-occupied active site
Rationale: retains the Asp101 active-site contact characterised across this structural series, relevant to inhibitor design

Validation (D–F)

Protocol check: RMSD < 2.0 Å for redocked controls
Cross-validation: NCC149 (0.49, 0.32 Å), PCI-34051 (0.30, 0.36 Å), entinostat (1.25, 2.14 Å)
Internal check: octanediamide into parent 2V5X recovered exactly (0.00 Å), reported as a consistency check, not independent validation
Note: redocking set includes HDAC8 orthologue and other-isoform structures; PDB identities to be confirmed (see manuscript)

Analysis (G–I)

Interaction mapping: protein–ligand contact characterisation
SAR: structure–activity relationships by substituent and electronic class
ADMET: drug-likeness assessment and multi-criteria lead selection

Preparation Validation Analysis

Lead identification. KB-282 and KB-281 were selected as lead compounds on the basis of Pocket 1 binding scores (−12.3 and −11.0 kcal/mol) and predicted ADMET profiles. Key feature. Integration of computational prediction with a preliminary experimental comparison supports lead prioritisation. Reproducibility. Docking parameters and validation criteria are defined in Section 2.1 of the manuscript and Supplementary Material S1.