Outpacing evolution: synthetic protein design for biomedical application

From the Stoddard Lab, Fred Hutch Basic Sciences Division, and the Baker Lab, University of Washington

Evolution is the world’s all-time greatest inventor. Over the epochs since life first formed on our planet, it has innovated proteins of all shapes, sizes, and functions, and orchestrated their interplay with biological machines of unparalleled complexity. Evolution has also been a key contributor to medical research advancements, providing us not only with the cognitive prowess to carry it out, but also the many naturally derived drugs used to treat various diseases (evolution is also responsible for most of what makes us sick, but I digress). For all its power, evolution has its limits. It is, to put it bluntly, slow and stupid, succeeding only by blind trial and error over countless iterations. It is constrained by the flexibility of the system in which it works to accommodate change. And it is limited, for the most part, to incremental changes that improve the fitness of the organism in which they are generated.

So, what do you do when evolution has not created the protein you need to achieve your research goals? For Drs. David Baker, Professor of Biochemistry and head of the Institute for Protein Design at the University of Washington and Fred Hutch/UW/Seattle Children’s Cancer Consortium member and Barry Stoddard, Professor in Fred Hutch’s Basic Sciences Division and Cancer Consortium member, the answer is, you make it yourself. A human inventor, after all, has the advantage of intelligence. Of learning from our mistakes. Of an audacious lack of constraint. By using computational algorithms based on the principles of protein design learned from evolution, and a bit of good old-fashioned biochemistry, Drs. Baker and Stoddard have a knack for creating proteins never before seen in nature. In a recent research article published in Proceedings of the National Academy of Sciences and led by Baker Lab member Dr. Derrick Hicks, the two groups report the design of thousands of new proteins with potentially important biomedical implications.

“Cyclic two-fold (C2) symmetric molecules are common in biology and medicine,” state the authors. These proteins can control various processes within a cell, and can, in some instances, serve as therapeutic agents, an example being the protease inhibitors used as anti-HIV drugs. Designing scaffold proteins that bind and encase C2 molecules can be valuable to controlling them for therapeutic use but designing scaffolds that fit the various shapes of C2 molecules is a challenging task. “Before this paper, we had shown that Circular Tandem Repeat Proteins (cTRPs) were highly stable designed proteins, that have substantial translational potential as novel signaling modules for synthetic biology and medicine,” explain Dr. Stoddard and graduate student Madison Kennedy, also an author on the paper. “Turning cTRPs into a variety of self-assembling constructs (such as dimers) with appropriately sized internal cavities is a first step in creating proteins that might be engineered to bind small molecule ligands in their interior.” Thus, the Baker lab used the Rosetta protein modeling software to model 100,000 curved repeat proteins that were predicted to dimerize and create circular ring-like structures with internal cavities of varied sizes and shapes to fit different C2 small molecules (see Figure). They then refined their computational modeling to optimize the protein sequences for improved binding properties and identified ~3,000 cTRP sequences whose dimerization was predicted to be the most biochemically favorable.

protein design strategy
Schematic of synthetic design strategy to generate curved repeat proteins (left), which assemble into circular symmetrical homodimers with variable internal cavities (middle) capable of fitting a range of C2 symmetric small molecules (right) Image provided by Dr. Barry Stoddard

From these sequences, the group selected 101 of their best modeled proteins, representing a wide range of sizes and shapes of the internal cavities, to synthesize and test their biochemical properties against the computational predictions. Of these, 39 were successfully synthesized and determined to be suitable for further testing. Based on X-ray scattering analysis, 31 of the 39 proteins were found to form stable dimers, consistent with computational predictions. Finally, to take  a closer look at these proteins, they determined crystal structures for 3 of them. 2 structures fit the models very well, with only minor deviations, while 1 shape deviated substantially from the model, serving as a useful reminder that, while computational modeling is often quite accurate, it is far from infallible.

This work established the groups’ ability to design cTRPs with a wide range of sizes and shapes. The next step in this project is to examine the ability of these synthetic proteins to bind C2 small molecules in their internal cavities. “In work going on right now, we and our colleagues are modifying these symmetric dimeric proteins to specifically bind a range of similarly symmetric small molecules, and to establish that a wide diversity of such molecules can be targeted. Beyond that, eventually we hope to create asymmetric, self-assembling cTRPs that can bind any type of small molecule, regardless of their chemical composition and structure,” said Stoddard and Kennedy.

This work was supported by the National Institutes of Health, the Howard Hughes Medical Institute, the Department of Energy, the National Science Foundation, the Audacious Project, and the Open Philanthropy Project at the Institute for Protein Design.

Fred Hutch/UW Cancer Consortium members Barry Stoddard and David Baker contributed to this work.

Hicks DR, Kennedy MA, Thompson KA, DeWitt M, Coventry B, Kang A, Bera AK, Brunette TJ, Sankaran B, Stoddard B, Baker D. 2022. De novo design of protein homodimers containing tunable symmetric protein pockets. Proc Natl Acad Sci U S A. 119(30):e2113400119.