Multiscale Computational Workflow to Determine Stability of Disulfide Bridges in Proteins─Application to IgG Antibodies

An antibody or immunoglobulin is a Y-shaped assembly of four peptide chains, stabilized by several disulfide bridges that are critical for the structure and function. Failure to build or any damage to these disulfide bridges makes the antibody unable to assume the native and working three-dimensional structure. Based on our previous developments, we present a multiscale computational framework to assess the thermodynamic as well as kinetic stability of each of the disulfide bonds in the structure of the antibody. The computational strategy is based on generating free energy surfaces for the nonenzymatic reduction of the disulfide bridges by glutathione. These data are interpreted in terms of rate-limiting energy barrier, which makes it possible to compare the stabilities of the different disulfide bridges within a single antibody or the stabilities of several different antibodies. Also, the application of efficient quantum chemistry methods and of enhanced sampling methods makes the assessment computationally inexpensive. We showcase the method on a study of the stability of rituximab, a monoclonal antibody with a widespread clinical use. We show in this case how the interchain disulfide bridges are more prone to reduction than the intrachain ones, quantify and explain the differences in stability.