Two Ways to Merge
Computer simulations suggest both existing theories of membrane fusion are valid
Cell membranes fuse with other membranes to allow material in and out. If incoming material includes invading viruses, that can be bad news for the cells. Until recently, the process of membrane fusion has been poorly understood. Now, a powerful computer model shows that neighboring membranes can merge in two distinct ways—a fact that was previously unknown. The work could help researchers clarify how viruses invade cells, and possibly lead to ways to fight viral infections.
“Ultimately we’d like to be able to control fusion in biological systems and induce or inhibit it for therapeutic purposes,” says Peter Kasson, PhD, a chemistry postdoctoral scholar at Stanford University and leader of the study. This paper takes steps in that direction. “Our model helps provide an explanation for how you get these two fusion processes,” he says.
Previously, scientists debated whether membranes fuse by joining directly from their initial contact or by going through a “hemifused” halfway point, where the outside layers have merged but the insides remain separate. “We show that both could happen,” says Kasson. The work was published in the August 8, 2006, issue of the Proceedings of the National Academy of Sciences.
Observing two membranes combining in a lab is difficult because it happens so quickly—on the order of microseconds. Earlier models haven’t represented membranes in as much detail, over such long timescales, or with as many simulations as this one does, says Kasson.
The team ran 10,000 separate simulations of membrane fusion using a distributed computing network called Folding@Home, in which people around the world donate screensaver time to biological research. In each simulation, the fusing membranes began with different starting conditions and evolved based on laws of physics and chemistry. The result: The simulated membranes merged through either of the two routes rather than exclusively through one or the other.
Erik Lindahl, PhD, professor of bioinformatics at Stockholm University, Sweden, thinks the project sets the pace for future work in the field. “The key thing is that they’re not doing one simulation, they’re doing many,” he says. “In ten years nobody will publish a single simulation anymore.”
Siewert-Jan Marrink, PhD, head of the molecular dynamics group at the University of Groningen, the Netherlands, and creator of a previous model of membrane fusion, agrees. “I do consider this work to be a significant step forward,” he says. “In my original publication of the fusion process of the same system I was only able to look at a few events, but I could not tell how relevant these were. In Kasson’s work this has become possible. Only by comparing many independent instances of the process can global assessments be made.”
Kasson is delighted that his model explains experimental observations and can help in planning new experiments. “That’s the most exciting part,” he says, “when we can come full circle.”