De novo Design and Synthesis of Selective Inhibitors for Carbonic Anhydrase Isoforms II, IX and XII
Project goal
Establish a de novo, structure-based workflow to design and synthesize small-molecule inhibitors selective for carbonic anhydrase isoforms II, IX, and XII, achieving nanomolar binding affinity and a novel mechanism distinct from known references.
Description of Activities (Stages)
- 1.
Preparation of Structural Models: Collected and refined X-ray or homology models of CA-II, IX, and XII active sites. Prepared protein structures for in silico design.
- 2.
De Novo Design Using Rosetta Libraries: Employed Rosetta fragment libraries to propose initial scaffolds targeting the zinc-containing active site, ensuring isoform-specific interactions.
- 3.
DFT Calculations on Selected Candidates: For top designs, performed density functional theory calculations to assess electronic properties and optimize substituents for binding and stability.
- 4.
Scoring and ML-Based Binding Energy Estimation: Applied custom scoring in Rosetta and used a machine learning model to predict free energy of binding across isoforms, prioritizing selectivity for II, IX, and XII.
- 5.
Synthesis of Selected Compounds: Synthesized a library of 20 derivatives based on in silico prioritization, varying functional groups to tune affinity and selectivity profiles.
- 6.
In Vitro Affinity Measurements: Measured dissociation constants (Kd) via microscale thermophoresis, identifying compounds with Kd around 5 nM for target isoforms and diversified affinities across isoforms
- 7.
Mechanistic Studies and Comparative Analysis: Evaluated how binding mode differs from reference inhibitors, planning crystallographic analysis (XRD of protein–ligand complex) to confirm predicted interactions.
- 8.
Data Management and Workflow Automation: Developed Python scripts to automate Rosetta workflows, handle format conversions, and populate an internal database of designs and assay results.
Resources used
Data
Public structural data for CA isoforms. Software: Rosetta suite, DFT packages, ML frameworks for binding energy prediction, custom Python scripts.
Computational Infrastructure
Single PC (20 cores, 32 GB) for Rosetta runs, DFT computations, and ML-prediction.
In Vitro Confirmation
Microscale thermophoresis assays confirmed nanomolar binding affinity (Kd ≈ 5 nM) for selected inhibitors against target isoforms, with differentiated affinity profiles across isoforms.
Publication status
The manuscript is in the final production stage.