Researchers at Harvard Medical School in Boston have developed a method for studying the orientation of septin filaments in yeast buds. The technique, which involves GFP and polarization microscopy, may have applications in studying the structures of a wide variety of proteins. Yeasts reproduce by forming buds, tiny bumps on the cell wall that eventually grow into a full-fledged yeast cell. The bud separates from the mother in a process called cytokinesis, in which fibers made of proteins called septins play a central role. The four main septin proteins were discovered in the famous Lee Hartwell screen and described by him in a 1971 Experimental Cell Research paper. This yeast cell was imaged with the mother-daughter axis parallel to the light path. Images are acquired via the X-X and Y-Y polarizer configurations, and the subtracted image is shown pseudocolored. “Septins localize at the junction between the two resulting cells in close apposition to the cytoplasmic face of the membrane,” explained Alina M. Vrabioiu, a postdoctoral fellow at the school. They are thought to spatially and temporally coordinate the many subprocesses of cytokinesis, she said. Temperature-sensitive mutations in any of the four genes for the septin proteins lead to cytokinesis failure. The Cdc3 and Cdc12 septin proteins are particularly interesting for research. They have a C-terminus that is thought to form an alpha-helix coil, which makes them good candidates for attaching GFP in a relatively rigid position. This is necessary for the new technique, in which the researchers used a specially designed microscope stage that let them rotate the sample in the X-Y plane. They could position a yeast cell in two orientations, with the mother-daughter axis either parallel or perpendicular to the light path, so that they could study a single yeast cell from three points of view. In this color combined image (the transmitted light is blue, the X-X polarizer axis is red and the Y-Y polarizer axis is green), the top yeast (out of focus) is positioned with the mother-daughter axis in the stage plane, and the bottom one has the mother-daughter axis almost perpendicular to the stage (parallel to the light path), so it is the same ring as in the figure on the facing page. As detailed in the Sept. 28 issue of Nature, their setup included a Nikon microscope with two filter wheels (each with two orthogonal polarizers) in the emission and excitation paths. The filters in the excitation wheel were mounted so that they transmitted horizontally polarized light (parallel to the X-axis) or vertically polarized light (parallel to the Y-axis.) The emission polarizers were mounted so that they duplicated the orientation of the excitation polarizers. A Hamamatsu camera collected images. Because the GFP was bound rigidly, Vrabioiu and her adviser, Timothy Mitchison, could use the polarization setup to ascertain the orientation of the fluorescent molecules and, consequently, the septins within the overall structure of the yeast’s cell wall. “We can determine what we call the average dipole direction,” Vrabioiu explained. “If the filaments were random or the GFP were random, we would expect no difference in the X-X and Y-Y intensities [the intensities on the polarizers’ axes.]” To determine the filaments’ alignment, they calculated the P-value for each orientation. This value is the difference in intensity between the horizontal and vertical polarizations divided by the sum of the intensities. For perfect alignment, P approaches 1.0. For Cdc3, Vrabioiu and Mitchison found the value to be 0.55. Vrabioiu said that these results mean that there is some flexibility in the septin-to-GFP linkage, there is some flexibility of the septin C-terminus relative to the filament, or there is some distribution of filaments around the average. For Cdc12, the researchers calculated a P-value of 0.8, which led them to conclude that, in the early stages of yeast budding, when the mother and daughter cells form an hourglass shape, the septin filaments are parallel to the long axis of the hourglass. When the researchers took into account the fact that the measurements were averaged over a curved surface by acquiring a Z-stack and deconvolving it or mathematically estimating the effect, P for Cdc12 approached 1.0, Vrabioiu said. “Since Cdc3 and Cdc12 are together in the complex, this suggests that, in Cdc12, the fluorophores are more constrained than in Cdc3 and that the majority of the filaments have a tight distribution around the average, which is the Y-axis.” Interestingly, as the yeast bud moves to form a separate cell, the septin filaments shift 90° and form rings that clearly show anisotropic intensity when observed from one end of the yeast as cross sections through the bud neck. The researchers considered several possible scenarios before concluding that the filaments simply rotate 90°, rather than dissociate and repolymerize. “We do not understand the mechanism of this process,” Vrabioiu said. “[The filaments] clear the middle [of the neck] and turn 90° at the location of the rings.” What happens next also is not completely clear. The filaments are thought to direct the formation of the next bud, but by the time the next hourglass is formed, they are mostly gone. Vrabioiu said that they plan to continue testing the new imaging method. “We hope that our approach of rigidly attaching the GFP to the C-terminal alpha-helix could be used for in vivo and in vitro studies of other proteins — myosins, syntaxins, kinesins. It could be used for orientation studies such as ours or for tracking single protein internal motion.” They also would like to continue studying septins. “Given that the septins are conserved and essential for cytokinesis in the majority of eukaryotes, including humans, it is possible that they play a mechanical role in the cell division of those organisms, too,” she said. She acknowledged that testing mammalian cells or tissue in culture will be more difficult. Yeast may need a highly ordered scaffold to maintain the inward bud-neck membrane curvature for most of the cell cycle. “For the mammalian cells, the cleavage furrow’s inward curvature is a transient thing, so it could be that the yeast mechanism applies only to yeast.”