A spherical region of illumination within an optical lattice microscope may contain thousands of points of light for massively parallel, rapid imaging over a large volume within a cell.

In a dark, high-ceilinged room near Grigorieff's laboratory, Carsten Sachse pours liquid nitrogen into a stainless steel sample holder on the side of an electron microscope, a foot-thick gray column taller than an NBA star. It hisses and boils off an icy steam.

He points to a computer monitor next to the microscope, to an image of gray fibers packed tightly side by side. They're lab-grown fibrils of amyloid beta peptide, the molecule in the brain suspected of causing Alzheimer's disease, magnified 59,000 times. The regular packing of the fibrils means that Sachse, a visiting graduate student who works with Marcus Fändrich at the Leibniz Institute for Age Research, in Jena, Germany, may well be able to use electron microscope images of similar preparations to determine the three-dimensional atomic structure of amyloid peptide packed into fibrils. Understanding that could be key to blocking tissue damage in Alzheimer's patients. "It looks very promising," he says.

Electron microscopy, Grigorieff says, can help elucidate the molecular structures of complexes too big to analyze by crystallography or nuclear magnetic resonance yet too small to see with a light microscope. "EM is a good technique to bridge the gap to high resolution," Grigorieff says.

But not just any electron microscope. Grigorieff's $2 million microscope contains a 300,000-volt field-emission electron gun to accelerate electrons through relatively thick samples, while ensuring they march in lockstep—a property needed to enhance contrast. It has specialized CCD cameras—tricked-out large-format digital cameras, essentially—to capture electron images and diffraction patterns. The researchers run recently developed algorithms on high-powered clusters of computers to turn huge amounts of electron microscopy data into three-dimensional images. And the darkened room where Sachse works contains a $400,000 climate-control system that prevents even the smallest drafts and shifts in temperature—all to keep their samples extraordinarily steady, which they must do to obtain good data.

Such attention to technology has paid dividends for Grigorieff. Working in collaboration with HHMI investigator Melissa J. Moore of Brandeis University and former postdoc Melissa Jurica, the group got the first-ever glimpse of the three-dimensional structure of the spliceosome, a molecular machine that splices newly formed RNA to form messenger RNA, which in turn encodes the correct amino acid sequence of the protein. The researchers obtained thousands of electron microscope images of single frozen spliceosomes and then reconstructed the complex's three-dimensional shape on a computer. The result: a cylinder on a hollow ovoid domain with an arm-like extension that seems optimized to perform the contortions necessary to cut and splice a thread-like RNA molecule.

Grigorieff's team has also determined the structures of other cellular machines, including, in collaboration with HHMI investigator Axel T. Brünger of Stanford University, a molecular machine called NSF that helps nerve cells export packets of molecules that enable them to signal their neighbors, and, with Thomas Walz of Harvard Medical School, a soccer-ball-shaped delivery structure called a clathrin coat. "We need to know what these molecules look like," Grigorieff says.

Illustration: Courtesy of Eric Betzig

	PAGE 3 OF 6	Go Back | Continue