Anton 2 Helps Cornell Scientists Discover Unexpected Gate in Molecular “Credit Card Reader”

A protein that scrambles the contents of the inner and outer surfaces of the cell membrane has an unexpected gate at its center, discovered by Weill Cornell Medical College scientists using Anton 2 supercomputer simulations at PSC. The slot suspected of allowing membrane components to move between the membrane’s inner and outer surfaces works much like a credit card reader, as scientists had thought from previous work. But thanks to the uniquely long time periods and large molecular complexes that Anton 2 can simulate, the Weill group has discovered an unforeseen additional mechanism that could provide drug targets in cancer, heart disease, stroke and blood disorders.

Why It’s Important

Life would not be possible if the cells making up all organisms didn’t have the ability to separate themselves from the outside world. That separation allows them to regulate their life cycle. The membrane surrounding the cell makes it all possible, by taking advantage of the fact that oil and water don’t like to mix.

The cell membrane holds together because its components—called phospholipids—are dual-personality molecules that have an oil-loving tail and a water-loving head. To escape the water that both fills and surrounds living cells, the membrane forms as two sheets—leaflets—of phospholipids. The phospholipids stack side-to-side with their oil-loving tails facing inward toward each other and their water-loving heads pointed out toward the surrounding water.

“The membrane is composed of two leaflets. The leaflets are different from one another in terms of composition … There’s a compositional asymmetry, with different phospholipids in the outer leaflet versus the inner leaflet.”—Harel Weinstein, Weill Cornell Medical College

Cell membranes are so fundamental to life they’d be scientifically important even if that were the whole story. But life depends on conditions being different inside and outside the cell. So the cell membrane has different types of phospholipids in its inner and outer leaflets. The cell membrane must maintain that difference, and so it includes molecular machines to do so. This is important because the “wrong” phospholipids showing up on the outside leaflet of the membrane serve as a red flag that a cell has developed a serious problem. One of the first lines of defense against cancer, for example, is when cells are damaged and start to destroy themselves in a process known as “programmed cell death,” specialized neighboring cells recognize the “wrong” phospholipids and eliminate the damaged cell.

Understanding how the cell creates and maintains the mix of phospholipids in their inner and outer membrane leaflets is an important question for a number of human diseases. That’s why George Khelashvili and Harel Weinstein of Weill Cornell Medical College, collaborating with laboratory scientists there and elsewhere, wanted to better understand a family of scramblases. These membrane-embedded proteins scramble the proper mix of phospholipids by flipping them in and out. They’re an important part of the red-flag system that leads to programmed cell death, as well as playing a role in medical problems, including heart disease, stroke and bleeding disorders.

Cell membrane, showing the two leaflets, each made up of phospholipids with water-loving heads (yellow circles) facing out and fat-loving tails (gray zigzags) facing in. Jerome Walker. Licensed under the Creative Commons Attribution-Share Alike 3.0 Unported license.

How PSC Helped

Khelashvili and Weinstein turned Anton 2 to try to simulate how a family of scramblases called TMEM16 help phospholipids flip to the opposite side of the cell membrane. Anton 2 is a special-purpose supercomputer, hosted at PSC thanks to D. E. Shaw Research,  that only carries out such molecular dynamics simulations. This specialized power was important because the TMEM16 system is so big, with many hundreds of thousands of atoms in the protein, the membrane and surrounding water. The complex system needed a long time in the simulation to “settle down” before the scientists could fully sample their behavior.

“TMEM16 proteins are huge in size. We are talking about 600,000- to 700,000-atom systems … Plus the process we are studying is a long time-scale process, taking tens to hundreds of microseconds at least to observe the sampling of the system that’s required. These two considerations together render the Anton 2 system the only one nowadays that can conduct the calculations in a reasonable amount of time.”—George Khelashvili, Weill Cornell Medical College

Previous laboratory work by other scientists had given rise to the idea that two slots at either end of the TMEM16 scramblase helped the phospholipids flip through from one side to the other by hiding their water-loving heads from the oil-loving interior of the membrane. The idea is called the “credit card model,” because the head of the phospholipid slides through the slot like a credit card through a magnetic reader.

Working in step with their colleagues in the laboratory of Alessio Accardi at Weill, and with the help of PSC’s Marcela Madrid, they created a virtual TMEM16 surrounded by phospholipids and water. Then they watched as, in simulations on Anton 2, these virtual components moved around. They simulated the system over time periods of tens of microseconds—multiple hundred-thousandths of a second. That’s a tiny amount of time to us but extremely long for such computer simulations.

The team’s simulation agreed with the credit card model. But further analysis of the movements in the entire system revealed a surprise. At the top and middle of the slot, pairs of amino acids that make up part of the TMEM16 protein chain were interacting strongly with each other. A network of such interactions in the top formed a salt bridge. Together, these interactions served as locks, preventing even water molecules from entering the center of the slot. This “dry” slot prevents the phospholipid heads from sliding through.

“When we did the Anton 2 simulations, we found that yes, the slot is open overall, but that’s not sufficient to allow the head of a phospholipid to go through. It’s not simply open or closed. Something happens … that allows the first constriction [in the slot] to open up so that a phospholipid head group and water can penetrate this dry area. [It’s] a very specific set of interactions that opens a series of gates.”—Harel Weinstein, Weill Cornell Medical College

As revealed in the simulation, opening the lock in TMEM16’s slot took a coordinated series of movements. First, the phospholipid heads crowd in at specific places, causing the salt bridge to break. But it doesn’t just open up. Instead, because of the presence of the phospholipid headgroups in one place, the salt bridge is weakened. The network shifts to new salt bridges. These create an opening for another phospholipid headgroup to enter. In turn, this weakens the new lock so that eventually all the locks are broken and the gate opens completely. The whole slot is now exposed, so that a phospholipid head and lubricating water molecules can enter and slide through.

The Weill collaborators reported their results in the journal Nature Communications late last year. The detailed molecular description had allowed them to measure precisely many of the ideas that had emerged from previous lab experiments and revealed the gating mechanism. These insights drove them and their colleagues to design new lab experiments to validate the simulations and answer new questions that would not have arisen otherwise. One advantage of the simulation results is they offer a solid ground for asking specific “what if” questions regarding the function and mechanisms of modified forms of the scramblase. These include those resulting from disease-related mutations or changed cell environments. The idea is to target cancer, heart disease, stroke, bleeding disorders and other medical problems related to TMEM16 in future work.