Computational Methods for Automated Generation of Enzyme Mutants - PowerPoint PPT Presentation

Computational Methods for Automated Generation of Enzyme Mutants Kasper Primdal Lauritzen Department of Chemistry, University of Copenhagen June 25, 2012 — Slide 1/10

Outline 1 Backgound 2 Methods 3 Results 4 Outlook Kasper Primdal Lauritzen (Dept. of Chemistry) — Automated Enzyme Mutation — June 25, 2012 — Slide 2/10

Background Goal Get from WT structure to activity estimate of mutants in an automated process. Model Candida Antarctica lipase B (CalB). An esterase, tested for amidase activity. Evaluation Interpolation between ES and TI states. Rough estimate, points to promising candidates. Kasper Primdal Lauritzen (Dept. of Chemistry) — Automated Enzyme Mutation — June 25, 2012 — Slide 3/10

Methods 3 central steps • Structure Preparation • Mutating • Interpolation / Evaluation Kasper Primdal Lauritzen (Dept. of Chemistry) — Automated Enzyme Mutation — June 25, 2012 — Slide 4/10

Methods Structure Preparation Crystal Structure obtained from PDB (1LBS) Preparation steps • Crystal Waters • Protonation • Substrate Placement • Initial Optimization Kasper Primdal Lauritzen (Dept. of Chemistry) — Automated Enzyme Mutation — June 25, 2012 — Slide 5/10

Methods Substrate Placement His224 His224 His224 Ser105 Ser105 N ε2 H O γ N ε2 H HN δ1 N ε2 Ser105 HN δ1 H O γ HN δ1 O γ R 2 O R 2 O R 2 O Substrate N Substrate Substrate N R 1 H N R 1 H TI R 1 ES H TS Kasper Primdal Lauritzen (Dept. of Chemistry) — Automated Enzyme Mutation — June 25, 2012 — Slide 6/10

Methods Mutation • PyMOL has a library of rotamers for each amino acid. • Loop over each rotamer, optimize with PyMOL and evaluate energy with MOPAC. • Choose rotamer with lowest energy as canonical mutant. • Place mutant in TI structure and optimize • Generate ES from optimized TI Kasper Primdal Lauritzen (Dept. of Chemistry) — Automated Enzyme Mutation — June 25, 2012 — Slide 7/10

Methods Interpolation • 10 interpolation frames to represent the reaction scheme. • Constrain the position of Ser105 O γ and the substrate carbonyl carbon in each frame. • Optimize the remaining structure • Evaluate the energy of each frame. Kasper Primdal Lauritzen (Dept. of Chemistry) — Automated Enzyme Mutation — June 25, 2012 — Slide 8/10

Results Time Requirements • A more strict initial optimization does not result in faster mutant optimizations. • A single mutant optimization can be done in roughly 24 hours. Optimization time WT vs. mutants 16 14 12 GNORM=5 10 s GNORM=2 y a 8 d n i e 6 m i T 4 2 0 Wildtype L140K L140R L140N L140Q Kasper Primdal Lauritzen (Dept. of Chemistry) — Automated Enzyme Mutation — June 25, 2012 — Slide 9/10

Outlook • Link MOPAC with each script (!) • Try every mutation • Multi-fold mutations • Automatic interpolation evaluation Kasper Primdal Lauritzen (Dept. of Chemistry) — Automated Enzyme Mutation — June 25, 2012 — Slide 10/10