Senior Thesis: Student Simulates Behavior of Miniproteins
For his senior thesis project, Ben Levine ‘26 focused on an emerging class of disease treatments using computationally designed miniproteins that can overcome many of the limitations faced by current medications. Drugs based on miniproteins—which are small, stable proteins—can be designed to precisely fit a disease target, improving their potency and reducing potential side effects.
Proteins are complex molecules that are essential for the structure and function of cells in the body. Many human diseases are caused by malfunctions in interactions between proteins. In some neurodegenerative diseases like Alzheimer’s disease, for example, proteins will unfold and stick together, creating build-ups called plaques that can disrupt communication between nerve cells in the brain. Levine said some recent miniprotein design structures may be able to bind to these proteins to disrupt plaque formation.
“There's a lot of potential, but it's really important that before you start creating a whole new class of drugs, you know exactly how they'll behave in the body,” said Levine, a biology and neuroscience and behavior double major.
Questions remain regarding the reliability of how designed miniproteins will behave and move within the body. Levine’s research aims to help answer these questions by demonstrating the accuracy of computer simulations to predict miniprotein behavior in comparison to measurements made through similar in-lab experiments. His ultimate goal is to create a universal method of accurately simulating miniprotein behavior, since simulations are cheaper, less time-consuming, and more manipulatable than full-scale, in-lab experiments.
“This can save you a lot of time and help improve the safety and effectiveness of many protein drugs in the future,” Levine said. “Overall, in terms of dynamics, our simulations seem to be quite accurate when compared to protein dynamics from in-lab experiments.”
Levine focused on the dynamics of miniproteins at high temperatures through an advanced simulation technique called replica exchange molecular dynamics. He found that higher temperature simulations can illuminate movement patterns that standard simulations cannot.
This approach challenges previous suggestions in the field that high temperature simulations would be inaccurate because proteins at higher temperatures move more rapidly, which could lead to more unpredictable behavior and potential unfolding. While only investigating six miniprotein structures, he was able to study a broad sample of protein movements because he simulated each structure at 48 different temperatures.
“Our findings suggest that if we had more miniprotein structures to sample, this method would be pretty viable for predicting experimental data for new miniprotein drugs of varying stability and behavior,” Levine said.
If researchers can predict the movements of a protein drug, they can then more confidently use it to treat diseases driven by malfunctions in interactions between two or more proteins. Current medications use molecules that are too small to bind in this way and cannot interact stably with the flat protein surfaces. Once bound, the protein drug can enact whichever functions it is designated to do, helping to treat a disease. “You can design a miniprotein to bind to many different proteins by providing the design model with [the correct] inputs,” Levine said.
“It's really important to know exactly how they will behave,” Levine said. “Because you don't want the drug to be binding to something else and then causing a bad side effect or just not binding to what you want at all.”
Currently, Levine is preparing to submit his research for publication in an academic journal, in collaboration with his thesis advisor and lab primary investigator Colin Smith, associate professor of chemistry. Once he graduates this spring, he will study biomedical sciences in a Ph.D. program at the University of Virginia.
