Molecule of the Month: Designed Proteins and Citizen Science

The Protein Data Bank (PDB) has been the central archive of biomolecular structures since 1971. These structures, of proteins, nucleic acids, and their various complexes have helped us understand structure-function relations in biology and learn the rules about how polypeptide chains fold into functional proteins. Using this knowledge, scientists have recently attempted de novo design of custom proteins to solve some 21st century problems like breaking down plastic or efficiently building specific drug molecules. Protein folding and design programs like AlphaFold and Rosetta can predict structures of the designed proteins, but these proteins may not always be stable when synthesized to test function. In these scenarios scientists have turned to gamers and citizen scientists for help in improving the protein designs.

Foldit is an interactive computer game that enables players to design protein sequences and predict their structures. Its easy-to-use interface enables players to manipulate structures based on spatial intuition, while following rules for stabilizing interactions such as hydrogen bonds and hydrophobic interactions, as set by the Rosetta program, to score the highest points in the game. To start a challenge, scientists upload protein folding puzzles with varied levels of difficulty - ranging from simple modifications of an already folded structure, to structure prediction of an entire de novo designed protein. Players use their 3D problem-solving skills to explore a unique range of structures that might be missed by purely computational approaches. Afterwards, when these proteins are synthesized by scientists and tested in laboratory experiments to confirm intended properties and function(s), users can see their predicted structures as real proteins.

New protein folds have been a cornerstone of de novo design - for example, to engineer new enzymes and other functional proteins. Scientific groups have used algorithms like Rosetta to design new protein folds from scratch, such as Top7 (1qys, not shown). These new folds are the basis from which new enzymes are created, such as proteins that help fight off viral infections (7jzl and 3r2x, not shown) or the creation of molecules that can track specific chemicals in cells. Foldit players created 56 unique proteins from scratch, one of them even had a newly-discovered protein fold! Experimental structures of four of these new proteins are shown here (PDB ID 6mrr, 6nuk, 6msp and 6mrs). Thanks to new computational design tools, there are now more possibilities for de novo designed proteins than ever before.

Tools developed for de novo protein design have also been very effective for structure prediction and optimization. Three examples are listed here. Using structure optimization, scientists engineered an efficient PET depolymerase (6tht, not shown) that performs the difficult chemical task of breaking down polyester plastics. After 10 years of trying, scientists were finally able to determine the structure of the Mason-Pfizer monkey virus (M-PMV) protease, with help from Foldit players (3sqf, not shown). The structure is ready to be used in targeted drug discovery. Finally, Foldit players were able to improve the function of an existing protein by increasing its catalytic activity more than 10-fold. The starting point was an engineered protein that performs an unusual Diels-Alder reaction (PDB ID 3i1c). Foldit players were guided by scientists to build a “lid” for the enzyme to hold the substrate more tightly for more efficient catalysis (PDB ID 3u0s).