Adenosine triphosphate (ATP) is a compound used to funnel energy from mitochondria to other parts of the cell, enabling energy-driven functions like muscle contractions. For ATP to flow, though, the interaction between the hexokinase-II (HKII) enzyme and the proteins found in a specific channel on the mitochondria’s outer membrane. Now, simulations conducted on supercomputers at the Texas Advanced Computing Center (TACC) have simulated this protein-protein binding for the first time.
“We had strong evidence that they bind, but we didn’t know how they bind to each other,” explained Emad Tajkhorshid, a professor of biochemistry at the University of Illinois at Urbana-Champaign and lead author of the study, in an interview with TACC’s Jorge Salazar. “That was the million-dollar question.”
To get to the heart of it, the researchers employed a couple of supercomputing resources at TACC. First, the 10.7-Linpack petaflops Stampede2 system. On Stampede2, Tajkhorshid and his colleagues ran a sophisticated simulation of the binding between the two proteins (and the host membrane), including major components from all-atom simulations. Overall, the simulation comprised some 700,000 atoms.
“We used all-atom molecular dynamics to get a more refined model and specific size of the interaction to look for this particular protein-protein interaction,” said Po-Chao Wen, a postdoctoral research associate at the University of Illinois at Urbana-Champaign. “It seemed almost impossible when we started this process, because of the long timescales of milliseconds to seconds of the all-atom simulations.”
“These are really expensive calculations, which would require millions of dollars to set up independently,” added Tajkhorshid. “And you need to run on parallel supercomputers using our NAMD [molecular dynamics] code, otherwise we could not reach the time scales that we needed.”
Luckily, the team was backed by supercomputing allocations from XSEDE, which allowed them access to TACC’s infrastructure. Beyond Stampede2, they also made use of TACC’s Ranch system, which serves as a massive, high-performance archival file system that now hosts the data from the protein simulations conducted by the team.
“If it wasn’t for XSEDE, we wouldn’t be studying many of these complex projects and biological systems because you simply can’t afford running the simulation. They usually require long simulations, and we need multiple copies of these simulations to be scientifically convincing. Without XSEDE it is impossible. We would have to go back to studying smaller systems,” Tajkhorshid said.
With the TACC resources in-hand, the researchers were able to see the binding process in all its detail. “We showed how the phosphorylation affects this process of binding between the two proteins,” Tajkhorshid said. “That was also verified experimentally.”
Next, they’ll expand the research, examining other kinds of protein binding – such as how SARS-CoV-2 binds to host cells.
“We like to look at our work as a computational microscope that allows one to look at molecular systems and processes,” Tajkhorshid said. “How molecules come together, how they move and how they change their structure to accomplish a particular function that people have been indirectly measuring experimentally. Supercomputers are essential in providing this level of detail, which we can use to understand the molecular basis of diseases, drug discovery and more.”
To learn more, read the reporting from TACC’s Jorge Salazar here.