Year of Award

2026

Document Type

Dissertation

Degree Type

Doctor of Philosophy (PhD)

Degree Name

Biochemistry & Biophysics

Department or School/College

Department of Chemistry and Biochemistry

Committee Chair

Stephen Lodmell

Commitee Members

Klára Briknarová, Mark Grimes, Kent Sugden

Keywords

11FN3, EmCAST, Fibronectin, Rational stabilization, Structural preference, β-sheet domain

Abstract

The EmCAST webserver (www.EmCAST.org) provides a mechanism where mutations that stabilize a protein can be predicted based on overlapping structural preference with amino acid sequence. Previously, it had been thought that surface mutations to proteins would offer only minimal stabilization. Many stabilization methods available rely on multiple amino acid substitutions, often with poor results. We hypothesize that stabilization can occur with a minimal number of surface exposed residue substitutions, without compromising the structure of the protein of interest. Previous validation of EmCAST has been done with α-helical domains. This dissertation provides important validation of the ability of EmCAST to stabilize a β-sheet protein, a more difficult task due to the naturally lower solvent accessibility of surface exposed residues and the multiple long-range interactions within β-sheets. We show in this work that with only three surface-level mutations, we can significantly stabilize a β-sheet domain protein, the eleventh domain of fibronectin type 3 (11FN3), without perturbing the structural integrity. We use chemical denaturation with guanidine hydrochloride to probe stability and X-ray crystallography to determine structural integrity. When denatured with guanidine hydrochloride (GdnHCl), wild type (WT) 11FN3 stability, Δ𝐺𝐻2𝑂°′, was measured to be 2.34 ± 0.07 kcal mol-1. The triple 11FN3 mutant variant, D21P/T71G/S77T, had a stability of 5.75 ± 0.06 kcal mol-1: a difference of 3.40 ± 0.09 kcal mol-1. The crystal structure of this same mutant variant reveals that the overall structure is not affected by these mutation additions. We also looked closer at the dynamic β-strand G. The stability prediction, A94P, for this region was completely opposite – predicted stability of 0.39 kcal mol-1, actual stability was -1.14 kcal mol-1. This instability was highlighted by the two different WT 11FN3 structures available in the PDB, and the change in strand register along with a single amino acid residue shift. Further exploration into structural changes will be needed for this mutation. We will present a single structure containing the A94P mutation, the D21P/T71G/S77T/A94P mutant variant, which also reveals no overall structural perturbations. More structural studies need to be carried out in the context of fewer stabilizing mutations to properly assess how A94P affects 11FN3 structure.

Download

Share

COinS
 

© Copyright 2026 Kassandra Lee Boshae