Home Cellular science TACC’s HPC Stampede2 simulations reveal how cells regulate

TACC’s HPC Stampede2 simulations reveal how cells regulate



(Left to right) Po-Chao Wen, Nandan Haloi and Emad Tajkhorshid from the University of Illinois at Urbana-Champaign

Supercomputer simulations performed on the Texas Advanced Computing Center (TACC) Stampede2 system have helped scientists studying cell function understand how protein molecules dance with a partner to respond to signals and regulate the activities of others. .

The migration of a compound called adenosine triphosphate (ATP) out of the cell’s powerhouse, the mitochondria, is crucial in providing cells with vital energy. And the interaction between a protein enzyme called hexokinase-II (HKII) and voltage-gated anion channel (VDAC) proteins located on the outer membrane of mitochondria is essential for this flow to energy-consuming parts of the cell.

HPC simulations revealed for the first time how VDAC binds to HKII. The work was supported by grants awarded by the Extreme Science and Engineering Discovery Environment (XSEDE) to Stampede2. XSEDE is funded by the National Science Foundation.

This basic research into how proteins interact from the cell’s powerhouse, the mitochondria, will help researchers understand the molecular basis of diseases such as cancer.

“We had strong evidence that they bond, but we didn’t know how they bond to each other,” said Emad Tajkhorshid, J. Woodland Hastings Chair in Biochemistry at the University of Illinois at Urbana-Champaign. “That was the million dollar question.”

Tajkhorshid co-wrote a to study published in Nature Communications Biology, Jun 2021. The study found that when the enzyme and channel proteins bind to each other, the conduction of the channel changes and partially blocks the flow of ATP. Simulations on TACC’s Stampede2 system have revealed this connection.

In addition, the Ranch system allocated by XSEDE to TACC holds permanent offsite file storage for study data.

HPC simulations reveal how the cell’s mitochondrial voltage-dependent anion channel (VDAC) binds to the enzyme hexokinase-II (HKII). Artistic representation of the membrane binding of the cytosolic enzyme Hexokinase (light blue), followed by its complex formation with the integral membrane protein VDAC (dark blue), on the surface of the outer membrane of the mitochondria. ATP (red) is phosphorylated by HKII. Credit: Haloi, N., Wen, PC., Cheng, Q. et al.

“If it weren’t for XSEDE, we wouldn’t be studying many of these complex biological projects and systems because you just can’t afford to run 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 will have to go back to studying smaller systems, ”Tajkhorshid said.

This work has implications for a better understanding not only of healthy cells, but also of cancer cells.

Basically, a cell needs ATP to metabolize glucose. It uses the “P” to convert glucose into glucose phosphate, giving it a “handle” that the cell can use. Hexokinase-II enables conversion by binding to the mitochondrial channel to engulf ATP and phosphorylate it.

“We have shown how phosphorylation affects this binding process between the two proteins. This has also been verified experimentally, ”Tajkhorshid said.

The VDAC channel is essential for efficient delivery of ATP directly to hexokinase. “It can work like a double-edged sword. For a healthy cell this is good. For a cancer cell, it also helps the cell to grow and proliferate, ”he said.

The Tajkhorshid team has developed the most detailed and sophisticated model to date of the complex formed by the binding of HKII and VDAC, combining molecular dynamics simulations of all atoms at the highest resolution with techniques of coarser Brownian dynamics. The size of the VDAC-HKII complex system was approximately 700,000 atoms including the membrane. It is about a fifth of the diameter of the COVID-19 virus.

“What emerges from our approach is that we actually took into account the cell background of this interaction,” said Po-Chao Wen, post-doctoral research associate at NIH Center for Macromolecular Modeling and Bioinformatics, University of Illinois at Urbana-Champaign.

Wen explained that their simulation design began with the hypothesis that the VDAC protein in the outer membrane could interact with HKII, which is located in a different part of the cell called the cytosol. They hypothesized that HKII should first bind to the membrane, drifting over it until it reached a VDAC protein.

The membrane-seated VDAC has already been well modeled, and the researchers used this knowledge to decompose the modeling of the HKII-VDAC complex into three parts, initially focused on HKII.

To study how HKII binds to the mitochondrial outer membrane, they used the molecular dynamics of all atoms and a tool developed by their center called the Highly Mobile Membrane Model (HMMM), which deals with membrane interaction.

Stampede2 (left) and Ranch (right). Credit: TACC

They then used Brownian dynamics to study how HKII drifts across the membrane to meet VDAC, creating numerous encounter / collision events between a seated VDAC and an HKII adrift on a flat membrane.

“Then we used the molecular dynamics of all the atoms to get a more refined model and specific size of the interaction to look for that particular protein-protein interaction,” Wen added. This helped them find the more stable complex of the two proteins formed.

“It seemed almost impossible when we started this process, due to the long timescales of a few milliseconds to seconds of all-atom simulations,” said study co-author Nandan Haloi, also a doctoral student at the center.

Many other computer tools have been developed by the group, including the NAMD commonly used for molecular dynamics.

“These are really expensive calculations, which would require millions of dollars to set up independently. And you have to run on parallel supercomputers using our NAMD code, otherwise we might not be able to meet the times we needed, ”Tajkhorshid said.

“We are extremely satisfied with TACC and their support not only for this work, but also for most of our projects and also for the development of our software, optimizing the software and making it faster. TACC has been great in supporting us, ”said Tajkhorshid.

TACC scientists are working with the NIH Center for Macromolecular Modeling and Bioinformatics to continuously optimize the NAMD software, currently used by thousands of researchers.

The next steps in the research include more ambitious systems such as the fusion of two cells, important for understanding how neurons in the brain fire signals; and how a virus such as the novel coronavirus fuses with the host cell.

Tajkhorshid’s group has received a leadership resource allocation on the NSF-funded flagship Frontera supercomputer at TACC to investigate some of these ambitious projects.

Tajkhorshid said, “We like to think of our work as a computer microscope that examines molecular systems and processes, how molecules come together, how they move and how they change their structure to perform a particular function that people have. summer. indirect measurement experimentally. Supercomputers are essential in providing this level of detail, which we can use to understand the molecular basis of diseases, drug discovery, etc.

The study, “Structural basis of the complex formation between the mitochondrial anion channel VDAC1 and hexokinase-II”, was published in June 2021 in the journal Nature Communications Biology. The authors of the study are Nandan Haloi, Po-Chao Wen, and Emad Tajkhorshid of the University of Illinois at Urbana-Champaign; Qunli Cheng, Meiying Yang, Gayathri Natarajan, Amadou KS Camara, and Wai-Meng Kwok of the Medical College of Wisconsin.

This research was funded by grants R01-HL131673 and P41-GM104601. from the National Institutes of Health. The simulations for this study were performed using allocations to the National Science Foundation Supercomputing Centers (grant number XSEDE MCA06N060) and the Blue Waters Supercomputer from National Center for High-Performance Computing Applications at the University of Illinois at Urbana-Champaign.

source: Jorge Salazar, TACC