May Serve as Tool for Illuminating Numerous Molecular Pathways

Life at the surface of the cell is anything but placid. Islands of receptor proteins bob on a swirling lipid sea, waiting to attract the attention of a messenger. Once matched, some receptor–messenger pairs disappear as the lipid membrane suddenly invaginates and swallows them. Endocytosis, this process by which the cell engulfs proteins, microbes, and other molecules, has captured the attention of scientists for hundreds of years. In 1974, researchers discovered that certain endocytic pathways depend on a remarkable three-legged protein, clathrin. For decades, Tomas Kirchhausen, HMS professor of cell biology, and colleagues have been helping to piece together an understanding of how clathrin works, but there are gaps in the picture.

Tom Kirchhausen and his colleagues are using dynasore in the lab to explore puzzles such as how membrane traffic sends signals that regulate cell size. They are also looking at dynamin’s role in bacterial invasion.

Caddies in Question
It appears, for example, that clathrin molecules, aided by helper proteins, approach the cell membrane from below and, through an astonishingly swift and graceful sequence, mold it into a bubble-shaped vesicle. Yet the roles of many helpers are still poorly defined. One such protein, dynamin, is thought to play an especially important part, coming in at the end and essentially pinching off the completed vesicle. Still, a clear picture of its comings and goings has been lacking.

Eric Macia, Marcello Ehrlich, Ramiro Massol, Kirchhausen, and their colleagues have stopped the protein in its tracks and report in the June 6 Developmental Cell that dynamin plays a dual role: it detaches the completed vesicle from the cell membrane, but it also comes into play earlier in the process, at the point of invagination.

What may be most exciting is the way the researchers made their traffic-stopping discovery. Macia, Ehrlich, and Massol, HMS research fellows in cell biology, working with Kirchhausen and colleagues, screened a library of 16,000 compounds and found one with the ability to block dynamin activity. They added the compound, dynasore, to cultured human cells. Two minutes later, the cells exhibited a complete block of endocytic traffic along the clathrin pathway. What is more, the endocytic vesicles were frozen in two positions—either fully formed but still attached to the plasma membrane by a small tether or shaped like a U, representing the kinds of half-formed pits one might see just after invagination (see figure page 1). “That was not expected,” said Kirchhausen, who is also a senior investigator at the CBR Institute for Biomedical Research. “Perhaps dynamin is necessary to go beyond the point of invagination.”

Even more surprising was how effectively and quickly dynasore worked. “It’s a cool reagent because you can put it in cells and, within a few minutes, there is a nice block on the entry pathway,” Kirchhausen said. He and his colleagues found that cells treated with dynasore rebuffed the advances of a variety of molecules, including transferrin, low-density lipoprotein, and cholera toxin. When the dynamin-blocking agent was washed out, the substances were able to enter.

Dynamin plays a dual role. During endocytosis, the cell membrane invaginates (top left), forming a vesicle that breaks free and travels to the cytoplasm (bottom left). Both of these steps are blocked by the dynamin-inhibiting agent dynasore. Vesicles do not detach (top right). Some fail to develop past the point of invagination (bottom right).

“This is indeed a terrific tool,” said Venkatesh Murthy, the Morris Kahn associate professor of molecular and cellular biology at Harvard University, who was not an author on the paper. “Since the compound can rapidly and reversibly block endocytosis, one can do experiments that may not be possible with knockouts or RNAi.”

An even more tantalizing approach would be to use dynasore to keep out certain disease agents, such as cholera toxin. “There is a problem—you would need a way to deliver this to specific cells. You might do that topically,” said Kirchhausen. “In my dreams, I would have a spray with dynasore that I would use to just spritz myself if I had a flu infection. In fact, the influenza virus uses two paths and one of them is dependent on dynamin.”

Magic Bullet
Catching—and stopping—dynamin in the act of vesicle formation was something of a pipe dream until recently. Clathrin-coated pits take a mere 20 to 60 seconds to form. Some researchers suspected dynamin might play a role at more than one point in the process, but they had no way to perturb, and visualize, dynamin’s activities in real time. Two lucky events would bring those goals within Kirchhausen’s reach.

The first occurred when Timothy Mitchison, the Hasib Sabbagh professor of systems biology, sent over a postdoctoral candidate, Christopher Brunner, who happened to be interested in membrane biology. Working with the Institute of Chemistry and Cell Biology (ICCB), Brunner screened the nearly 16,000 compounds and found one that blocked dynamin activity. Macia, currently at the Centre National de la Recherche Scientifique in Valbonne, France, characterized the protein and found that it prevented dynamin from carrying out its main activity, the hydrolysis of GTP.

Kirchhausen mentioned to Stephen Harrison, HMS professor of biological chemistry and molecular pharmacology, that he was looking to name the new protein. “Steve said, ‘Why don’t you call it dynasore?’ I said, ‘Dynasore?’ said Kirchhausen. “‘Sore to dynamin—painful for dynamin.’ The name just clicked.”

To test dynasore’s mettle, they decided to see whether it could prevent the entry of dynamin-dependent proteins in actual cells. They began with two proteins commonly found in the body, transferrin, used for iron transport, and low-density lipoprotein (ldl), used to carry cholesterol. Macia; Ehrlich, currently at Tel Aviv University in Israel; and colleagues added the proteins, fluorescently labeled, to human cultured cells pretreated with dynasore. Two minutes later, the cells were washed and stained. The slides showed that the cells rejected the transferrin and ldl proteins. Cholera toxin also enters for the most part through clathrin-coated vesicles. Again, dynasore-treated cells repelled the toxin’s incursion, though not completely. “It might be taking another route,” Kirchhausen said.

The researchers had perturbed the nimble dynamin, but the question was how? At what stage of pit formation had it been vulnerable to dynasore? Over the past few years, thanks to the funding of a private donor, Kirchhausen had garnered the resources to develop a method for producing time-lapse images that could be assembled into the form of molecular movies (Focus, March 7, 2003). “That was the other lucky accident,” he said. Using the technique, the researchers watched what happened when dynasore was added to cells with two fluorescently labeled vesicle proteins, clathrin and a helper, AP-2. Normally, the fluorescent spots can be seen to undergo a complete life cycle—from initial gathering of clathrin molecules to the formation of the clathrin coat to its disintegration—in 20 to 60 seconds. Movies of the dynasore-treated cells revealed a very different situation: specks of fluorescence became locked at the cell membrane.

Judging by the degree of fluorescence, which intensifies as coat formation proceeds, the dynamin—dependent vesicles appeared to be arrested at two different moments—late and early. Electron micrographs confirmed that the vesicles were stuck at two different stages—fully formed but attached to the membrane and U-shaped as though arrested just at the point of invagination.

How dynamin acts at each of these two points is not clear. “There are so many models thrown out without facts,” Kirchhausen said. “We need to go back to the molecular snapshots and understand what’s going on. There is a whole network of interactions that we simply do not understand.”