Computational Protein Design
The manipulation of protein sequences represents one of the most powerful engineering approaches that can be applied to problems across a wide range of scientific disciplines and industrial processes. Proteins are increasingly serving as drugs and drug delivery devices in medicine (e.g. antibodies, insulin, erythropoetin), while others are enzymes for chemical reactions (e.g. cellulases, amylases, proteases). Organisms have evolved proteins to serve a very specific function under a specific set of biological conditions giving the host a competitive advantage; however when these proteins are isolated their activities and stabilities are typically negatively affected.
Through the application of protein engineering methodologies, it is possible to regain and at times surpass the activities and stabilities of the wild-type protein in a non biological environment. The Mayo laboratory has been developing computational protein design (CPD) software and coupling it with state of the art experimental approaches to identify engineered variants. CPD has a number of advantages over traditional protein engineering techniques: a) algorithims can quickly identifiy promising candidate sequences from large sequence pools (>1080) based upon physical principles, b) specific regions of the protein (e.g. antibody-antigen interface) can be targeted and multiple simultaneous mutations evaluated, c) protein surfaces can be screened and redesigned in silico to improve areas prone to aggregation, d) de novo functionality can be engineered into a scaffold that had not been observed in the native protein and e) experimental data from CPD designs can be input into the design algorithims to influence the next set of predictions.
We are actively applying our CPD software in the areas of red fluorescent proteins, broadly neutralizing antibodies against influenza and methane activation.