T and B white blood cells, central actors in the immune system’s response to microbes thanks to the Syk protein.
Simulations on Anton Supercomputer Hosted at PSC Show How Critical Immune Protein Interaction May Be Disrupted
The spleen tyrosine kinase (Syk) protein is a linchpin in the signals inside immune cells. It causes the cell to launch a response to an invading microbe. Drugs that disrupt Syk’s interactions with other proteins could be powerful treatments for autoimmune disease or some blood-cell cancers. Scientists from Purdue University used a second-generation Anton system developed by D. E. Shaw Research (DESRES) and hosted at the Pittsburgh Supercomputing Center to dissect how a network of electrical charges within the Syk protein underlies the unusually strong bond, and how a relatively small additional negative charge on Syk disrupts that bond and turns off the signal.
WHY IT’S IMPORTANT
Our immune system has to monitor a wilderness of microbes, viruses, and other incoming substances, reacting to and protecting us from the ones that will harm us and ignoring those that won’t. When it’s too laid back, we get sick. When it’s too aggressive, we have an allergic reaction — or worse, it attacks our own tissues, and we develop autoimmune disease.
A particularly important actor in the network of chemical signals that underlie immune response is the Syk protein. Syk sits inside the cell, waiting for one of the receptor proteins that stick out of the cell membrane to respond to an incoming signal from that an invading microbe is present. Syk binds to that receptor, triggering a cascade of events that leads to an immune response.
Because of its central position in the signaling network, Syk is an important target for medical therapies. A drug that deactivates Syk could be useful for treating autoimmune diseases. In some immune-cell cancers, deactivating Syk could help control the malignancy.
“Syk will bind to that phosphorylated receptor with very, very high affinity … much tighter interactions than most signaling pairs have in the cell. So the question is what helps to undo the interaction? Once a signal process is turned on, it also has to be turned off. You can’t just let it go forever… what we learned is how it’s turned off.”
— Carol Beth Post, Purdue University
Carol Beth Post, Distinguished Professor of Medicinal Chemistry and Molecular Pharmacology at Purdue University, knew from her team’s previous work that Syk sticks to receptors much more tightly than these protein-to-protein interactions usually involve. To turn the signal off when it’s no longer needed, the cell has to disrupt this interaction somehow. Knowing how it does so could be invaluable for designing drugs to perform the same trick.
To dissect Syk’s activity, Post and her team turned to Anton, a special-purpose supercomputer that was designed and constructed by DESRES for simulating biomolecules (like Syk) for long time periods. The second-generation Anton machine they used was made available to scientists without cost by DESRES and was hosted at PSC with operational funding support by the National Institutes of Health. DESRES recently replaced that Anton with a third-generation machine, which also receives operational funding from the NIH.
HOW PSC HELPED
Syk homes in on a receptor protein that stretches from the outside of the cell, through the cell membrane, and then into the cell interior. When the outside part of that protein responds to an immune signal, the signal transfers through the membrane to the inside part. That inside part then becomes phosphorylated. That’s when a negatively charged group of atoms of phosphorus and oxygen attaches to the receptor.
Syk’s own structure has a bunch of positively charged amino acids. Not surprisingly, the positives of Syk glom onto the negatively charged phosphate group on the receptor. But even so, the strength of the interaction was stronger than investigators might have expected.
Previous work had suggested that the Syk signal turns off when Syk in turn is phosphorylated. Scientists had a pretty good idea that the negative charge of the new phosphate would disrupt the positive-negative interactions that held the two proteins together, and that a degree of disorder in both proteins would also develop. But they weren’t sure exactly how all this would play out.
Anton, which was specifically designed to perform molecular dynamics simulations, proved vital for the simulations Post and her lab members needed to unravel what was happening. Additionally, the second-generation Anton system at PSC could perform these in roughly a tenth of the time that general-purpose supercomputers would have taken. Anton put multiple-microsecond simulations of the complex interactions between two large molecules within reach. That’s a small amount of time in everyday life — millionths of a second. But at the blistering speed of chemistry, it’s huge.
The Purdue scientists didn’t exactly know what to expect from the simulations. What they did see was a remarkable network of charged interactions on Syk that coordinated to keep the different regions of Syk well ordered, in a perfect form to hone in on the receptor. When Syk itself is phosphorylated, the scale of the single phosphate group attached to Syk seems small. But in the Anton simulations, it disrupted the entire charged network, like the fall of a single domino making a long row topple. Interestingly, the regions of Syk didn’t exactly fall apart — they just stuck together a lot more loosely, which was enough to disrupt interaction with the receptor and Syk’s signaling to the rest of the cell. The way these interactions formed and were disrupted matched exactly what scientists had seen in the lab, offering confidence that the simulation was getting it all right, and that the sim could be used to predict other aspects of Syk’s behavior correctly.
“What Anton allowed us to do was to run the simulations with Syk both unphosphorylated and phosphorylated. The unphosphorylated Syk system was very well ordered on the microsecond time scale, but phosphorylated Syk became very disordered on the microsecond time scale and unfavorable for interacting with the receptor. We wouldn’t have seen that change with shorter simulations.”
— Carol Beth Post, Purdue University
These initial simulations won’t be the last word on the Syk system. Amazingly, in the simulations the regions of Syk are still moving around in a way that indicates that the system hasn’t yet settled into all of its positions. Post estimates that about 10 times as long a simulation may be needed for the entire system to reach final equilibrium. Future work will involve extending the sims, possibly with the new, third-generation Anton at PSC, as well as further collaboration with laboratory research designed to test the simulations and identify phenomena for the sims to target.