Foldit players demonstrate that human spatial awareness can outsmart computers in modeling simian AIDS-causing virus.
It took the online gaming community just three weeks to solve the structure of a retroviral protease that has eluded scientists for many years. Players of the protein folding game Foldit were challenged to produce three-dimensional models of the simian AIDS-causing Mason-Pfizer monkey virus’ (M-PMV) retroviral protease (PR). Their solutions were good enough to allow scientists to make a few fine-tuning molecular tweaks and solve the crystal structure of the protein. They claim the results show the M-PMV protein carries surface features that could represent drug targets for retroviruses including HIV.
Scientists at the University of Washington in Seattle, and colleagues at Mickiewicz University in Poznan, the Polish Academy of Sciences, and the Academy of Sciences of the Czech Republic, report on the challenge, and the results, in Nature Structural Biology. The paper is titled Crystal structure of a monomeric retroviral protease solved by protein folding game players.”
Previous work had suggested that biologically active retroviral proteases exist in a homodimeric active form, and preventing PR dimerization has been proposed as a mechanism for disrupting enzyme activity, explain the University of Washington’s Firas Khatib, Ph.D., David Baker, Ph.D., and colleagues. The retroviral protease of M-PMV, a simian AIDS-causing monkey virus, crystallizes as a monomer, but despite the availability of several crystal forms of M-PMV, researchers have for over a decade been unable to solve the structure by molecular replacement (MR) using either homodimer-derived models or an NMR structure of the protein monomer.
“We wanted to see if human intuition could succeed where automated methods had failed,” remarks Dr. Khatib, a researcher in the protein structure lab of Dr. Baker. The Washington team therefore threw down the gauntlet to players of Foldit, an online game created by computer scientists at the University of Washington Center for Game Science in collaboration with the Baker lab.
Played by thousands of people worldwide, Foldit requires no real knowledge of proteins or structural biochemistry, but instead relies on human spatial awareness and three-dimensional pattern matching skills. Players can pull, twist, and manipulate a protein to generate the most efficiently “folded” molecule that obeys all the basic laws of physics in terms of the placement of hydrophobic and hydrophilic moieties and charged ions, adhering to the rotational limits of bonds, and the need to fill space (i.e., reduce voids within the molecule). Players collaborate with teammates, while competing with other players to obtain the highest-scoring (lowest-energy) models. A model generated by one player is passed to another player in the same team for further refinement, and then to another team player, until a model of sufficient accuracy is generated.
“People have spatial reasoning skills, something computers are not yet good at,” adds co-author Seth Cooper, Ph.D., at the UW Department of Computing Science and Engineering, who is co-creator of Foldit and its lead designer and developer. “Games provide a framework for bringing together the strengths of computers and humans.”
Foldit players were challenged to build accurate models of M-PMV PR starting from the NMR coordinates (which had failed in MR). Ten NRM templates were provided, along with a new alignment tool. Remarkably, the authors report, and despite the complete failure of all previous approaches, it took just three weeks for the players to generate models that were accurate enough to fine-tune using molecular replacement and solve the actual crystal structure of the protein.
Importantly, the results indicated that the molecular surface that normally forms the M-PMV PR dimer interface displays features which provide opportunities antiretroviral drug design, potentially including anti-HIV drugs, the researchers state. “Compounds that bind to the surface formed in the monomer, but not the dimer, should shift the equilibrium in favor of the former, which is catalytically inactive.”
The achievement marks the first instance—as far as the researchers are aware—in which gamers have solved a longstanding scientific problem. “The critical role of Foldit players in the solution of the M-PMV PR structure shows the power of online games to channel human intuition and three-dimensional pattern-matching skills to solve challenging scientific problems,” they conclude. “These results indicate the potential for integrating video games into the real-world scientific process: the ingenuity of game players is a formidable force that, if properly directed, can be used to solve a wide range of scientific problems.”