|
||||||||
|
CELL BIOLOGY Protein Seen to Animate Cell SkeletonFindings on Actin Filament Assembly Have Bearing on Cancer Metastasis How cells assume different shapes, move, and divide is governed by their actin cytoskeleton and remains largely a mystery. Recently, HMS researchers have cast light on this dynamic structure, illuminating a mechanism by which actin filaments assemble.
Isabelle Sagot, David Pellman (l to r), and colleagues have found that formin proteins initiate the assembly of actin filaments in the cell. The research could help develop new therapies for cancer since the actin cytoskeleton governs cell division and movement. (Photo by Steve Gilbert) The cytoskeleton is made up of arrays of actin filaments that are arranged into widely different structures--parallel arrays that mediate muscle contraction, networks of branched filaments at the leading edge of migratory cells, and parallel bundles required to pinch cells apart at the end of cell division. The cytoskeleton can morph in a split second into different structures during cell division or in response to external signals. The critical juncture of filament growth is the process of nucleation, in which two or three actin monomers come together to form a "seed" for actin polymerization. Afterwards, the assembly of filamentous actin takes off. Because nucleation is the slowest step in filament assembly, it is the nodal point at which many signals converge. "When a cell wants to make some actin, one way to do it is to promote nucleation and shorten the lag before filament assembly," said David Pellman, HMS professor of pediatrics at the Dana-Farber Cancer Institute, who led the study. He and his colleagues discovered a mechanism for nucleating actin filaments that is mediated by conserved formin proteins in conjunction with another conserved protein, profilin. The research, a collaboration with the lab of Bruce Goode at Brandeis University, was published on July 22 online in Nature Cell Biology and appears in print in the August issue. Previously, researchers knew only one actin nucleator, the Arp2/3 complex, which assembles branched actin filaments. Tim Mitchison, the Hasib Sabbagh professor of cell biology at HMS, had shown that Arp2/3 drives processes like the rocketing of the bacterium Listeria monocytogenes in infected cells. But many processes like cytokinesis require linear rather than branched filaments. Researchers did not know if Arp2/3 did it all or if there was another mechanism for nucleation.
Working in yeast, researchers at HMS and Brandeis showed that formins and the protein profilin are necessary for the formation of actin cables (bottom) though not actin patches (top). Image courtesy of Isabelle Sagot Philip Leder, the John Emory Andrus professor of genetics at HMS, discovered formins in 1990. His first study showed that the proteins were crucial for limb development, and subsequent research showed that they were crucial for other aspects of normal development. Studies had also linked formins to signaling by Rho family GTPases, microtubules, and actin, but their molecular job remained obscure. "They were clearly interesting and had lots of interaction partners," said Isabelle Sagot, a postdoctoral fellow in Pellman's lab and the lead author of the study. "But no one knew what they were really doing." Among other models, Pellman's laboratory works on budding yeast, which has only two formins. "In yeast you can make mutations and manipulate genes very easily," Pellman said. In 1999, his group ran a genetic screen and found that formins regulate a specific aspect of cell polarity--the positioning of the mitotic spindle. This again linked formins to an interesting process but left unanswered the question of what exactly they do. Besides having only two formins, budding yeast has another advantage--it has just two actin structures. Early in the cell cycle, round actin patches are found in the daughter cells, and long, thin actin cables serve as highways for the transport of cell materials to construct the daughter cell. The cables are crucial for establishing cell polarity. In 1999 Rong Li, an HMS associate professor of cell biology, found that Arp2/3 was necessary for assembling actin patches. Her work also suggested that actin cables did not require Arp2/3. A clue that actin cables may be linked to formins came from a study showing that actin cables were required for spindle positioning. To uncover what formins do in cells, Pellman's group conducted a simple experiment. They created yeast in which all formin function could be removed through a temperature shift. They observed yeast cells at 24 degrees centigrade and saw that cables were being made. Raising the temperature to 34 degrees then inactivated the formins. Four minutes after the shift, the actin cables were completely eliminated while the actin patches were still untouched. Cells had lost their polarity since the cables act like highways directing material to the yeast bud. In their absence, the mother cell became bloated when material was deposited in the wrong place--the mother cell itself rather than the budding daughter cell. The experiment demonstrated that formins are required for making actin cables. This earlier study was published in the January 2002 Nature Cell Biology. Growing Actin SeedsStill, the molecular mechanism by which formins make the cables remained unknown, the main question being whether the proteins nucleate actin to make cables the way Arp2/3 does to make patches. To find out, the researchers faced some technical hurdles. Formins are enormous proteins and, so far, it has not been possible to work with them in their entirety. Using engineered yeast, Pellman's group set out to identify the smallest possible functional portion of the protein. They cut down the yeast formin--Bni1--and put it into cells to identify the smallest portion that could make actin cables. The resulting functional fragment was highly conserved; Sagot dubbed it Bni1 mini.She purified Bni1 mini and found that on its own, it accelerates actin polymerization in a test tube. She also saw that the lag phase before polymerization grew shorter as more Bni1 mini was added. Based on these findings, Pellman's group proposes a single model for generating differently shaped actin structures in cells: different nucleators generate differently shaped building blocks (linear or branched filaments) that are assembled by other factors into larger structures. The researchers also pieced together another puzzle. Profilin is an actin-binding protein that controls several steps in actin assembly. Although it is well known that the protein inhibits actin nucleation in a test tube, much research suggested that profilin may promote actin assembly in cells. Sagot found that purified profilin stimulated the nucleation activity of Bni1 mini through formation of a complex of the two. This provides biochemical evidence that when profilin binds to formins, it promotes actin nucleation. Besides clarifying basic cell processes, the research has potential therapeutic applications. A kind of hereditary deafness and one cause of female sterility are due to formin mutations. Understanding how these proteins work will shed light on the pathogenesis of these diseases. "It is early days yet," said Pellman. His main interest is cancer, which spreads when malignant cells escape from a primary tumor and crawl around the body. It is this metastasis that tends to kill patients. Since the cytoskeleton is essential for cancer cell invasion, Pellman hopes that understanding the process better will lead to drugs that inhibit this deadly dissemination. --Sena Desai |
