Reproductive Biology: Ovaries Exhibit Ongoing Power to Produce Eggs Biological Chemistry: Molecular Cowboy Seen Herding Actin Filaments Oncology: Angiogenesis Inhibitors Revived, Revealed in Progress Against Cancer Continuing Education Continuing Ed Takes Courses to the Web Inflammation Marker Tied to Type 2 Diabetes Brake-off, Gas-on Approach Drives Cell Regrowth Role Strengthened for Enzyme in Suppressing Breast Tumors   Jain Elected to National Academy of Engineering Grant Advances Clinical Research at Children's Two Innovator Awards Totaling $10m Received for Breast Cancer Loeffler Named Suit Proffesor in Radiation Oncology Bristol-Myers Squibb Grant Goes to Golub for Cancer Research Vlahakes Named Chief for Surgery at MGH Skin SPORE Offers Grants to Faculty Martin Wins Nesson Award Honors and Advances   Training Institute Aids LMA Workers, Employers   Letting Nurses Take the Lead in Teaching Hands-on Care   Front Page

BIOLOGICAL CHEMISTRY Molecular Cowboy Seen Herding Actin Filaments Formins Appear to Ride the Tip of Growing Polymers The lasso and hitching post have served cowhands for centuries. But today's hired hands may be surprised to hear that they were invented not by one of their own, but by Mother Nature. When two monomers face each other, the lasso of one (red) wraps around the post of the other (gray), and vice versa. The resulting hinged dimer is capable of "stepping" away from the growing end of an actin filament. The step is just the right length to allow an actin monomer to fill in behind (above). Once the actin is secured to the growing filament, another step is taken and the process repeated. (Images courtesy of Cell Press) Michael Eck, HMS associate professor of biological chemistry and molecular pharmacology at the Dana-Farber Cancer Institute, reports in the March 5 Cell that a proteinaceous post and lasso have been around for millennia and are essential for the activity of formins, proteins that do a little herding of their own. Formin Foremen First discovered in 1990 by HMS professor and chair of genetics Philip Leder, formins are a diverse group of extremely large proteins. Their function became clear recently when HMS associate professor of pediatrics David Pellman, also at Dana-Farber, discovered that they regulate actin filament assembly (see Focus, Sept. 13, 2002). Actin filaments form the backbone of the cytoskeleton, but in addition, their assembly and disassembly underlies many cellular remodeling events such as nerve axon growth, cell division, and the budding of yeast. Though it is well established that formins herd actin monomers into neat linear filaments, exactly how they accomplish this has been a mystery. Now, Eck, with collaborators Pellman and Bruce Goode at Brandeis University, demonstrate a model for formin activity based on the X-ray crystallographic structure of the yeast homolog BniP1. "The model resolves the paradox that formins both find the barbed end of actin filaments and promote assembly there." --David Pellman Because the large formins are notoriously difficult to crystallize in their entirety, Yingwu Xu, a postdoctoral fellow in Eck's lab, focused on that highly conserved actin-binding part of the molecule, the formin homology-2 domain. He made dozens of constructs before finding one that formed perfect crystals. Analysis of these reveals that the protein exists as a dimer, which is formed when monomers, shaped like a slanted L and having the post and lasso on opposite ends, come face to face to create a parallelogram. The crystal structure also shows that the lasso of one monomer wraps around the post of its partner (see figure). "This creates a structure that is extremely stable," said Eck. So how does the dimer direct actin assembly? The shape of the molecule suggested to Eck that the growing end of the actin filament, the so-called barbed end, could fit into the hollow formed by the conjoined monomers. But this fit is so snug that it does not leave room for further actin mon-omers to bind to the filament, indicating that the dimer might act more as a capping protein than a driver of assembly. The driver function is facilitated by another region of the protein called the linker. Stair-step Growth The linker is a short alpha helical rope that connects the lasso to the bulk of the protein. "This linker has some interesting properties," said Eck. X-ray data predict that it is not that well defined and may be easily stretched. "If it were to uncoil, it could easily reach about 27 angstroms," Eck said, "which is just about the distance between the two actin subunits at the barbed end of a filament." So Eck and colleagues began to focus their attention on the linker. Because the posts and lassos provide most of the glue that holds the monomers together, and because the linkers give the lassos some flexibility, he came to the conclusion that each monomer may be free to slide up and down relative to the other, and by about the length of an actin molecule. A good way to envision this is to imagine two dancers facing each other and holding hands. Though joined, they are each free to squat down or stand up. This predicted mobility has led Eck and colleagues to propose what they call the "stair-stepping" model for formin activity (see figure). Michael Eck, David Pellman, and colleagues have discovered that formin dimers are held together by proteinaceous posts and lassos. (Photo by Leah Gourley) In this model, one half of the dimer steps away from the growing actin filament, allowing an actin molecule to slide in and take its place. Once the actin binds to the filament, the opposite side of the dimer takes a step, making room for another actin monomer. This stepping action is repeated ad infinitum, creating a trail of bound actin in the formin's footsteps. "The model resolves the paradox that formins both find the barbed end of actin filaments and promote assembly there," said Pellman. "This is because the formins surf the growing end of the filaments," explained Eck. "Once they bind, they can direct assembly of the filaments without ever leaving." The model is also backed up by experimental evidence. Xu found that the linker is easily digested by proteolysis, confirming that it is readily exposed to solvent and therefore probably not in a tight alpha helix. Cutting the linker also results in the formation of two "hemidimers," monomers that have lost their own lasso but have their partner's still lashed to the post. These are incapable of guiding filament growth, and in fact, do the opposite, acting as capping proteins. This suggests that to be fully functional, monomers must remain lassoed to each other, as predicted by the stair-stepping model. --Tom Fagan