PAGE 4 OF 6
A closer view within an optical lattice microscope reveals individual, tightly focused excitation points (red) arranged in a periodic array.
For particularly complex molecular machines, it's often necessary to combine methods. David A. Agard, an HHMI investigator at the University of California, San Francisco, uses a technique he's optimized called cryoelectron tomography that takes images of particles from different angles and then assembles those images into a three-dimensional model. He's used the method to visualize the centrosome, an organelle that manages the cell's internal skeleton. To show how the centrosome carries out that task, he uses single-particle electron microscopy and x-ray crystallography to view its components at atomic resolution.
Grigorieff and others also use electron microscopy to visualize two-dimensional crystals of membrane proteins in a lipid bilayer, much like that in the cell. He and others have even used electron microscopy to visualize membrane proteins in atomic detail. In December, Walz reported using this technique, called electron crystallography, to determine the three-dimensional structure of a cellular water-pore protein called aquaporin; the resolution was high enough to spot discrete lipid molecules clinging to the side of the protein. Electron microscope images of single protein complexes (as opposed to crystals) have lower resolution, but they can distinguish parts of proteins, such as helices and loops. At Janelia Farm, Grigorieff says, he'll try to push single-particle cryoelectron microscopy methods to "routinely get to such high resolution that you can build an atomic model."
At Janelia Farm, other scientists will press light-microscopy methods to observe cells or even entire brains at high resolution. Janelia Farm group leader Eugene W. Meyers, who wrote programs that dramatically sped the sequencing of the human genome, will design new computing methods to assemble three-dimensional reconstructions of working brains from two-dimensional microscope images. Group leader Karel Svoboda, currently an HHMI investigator at Cold Spring Harbor Laboratory, will use two-photon excitation microscopy, a form of fluorescence microscopy developed in the 1990s, to observe living neurons in mouse brains. "We need new imaging methods to figure out how the brain works," Rubin says.
Janelia Farm biologists will make use of more than a decade of extraordinary advances in light microscopy. Until the early 1990s, some biologists examined cell shape and behavior by using light microscopy, while others tried to understand their protein of interest by localizing it in cells that had been killed and fixed, says Doug Murphy, a cell biologist at Johns Hopkins School of Medicine in Baltimore who will move to Janelia Farm to direct its shared light microscopy facility. "Now we don't just want to see where it is, but how it's behaving."
New fluorescent probes make that possible. In the mid-1990s, HHMI investigator Roger Y. Tsien of the University of California, San Diego, began working on a naturally fluorescent jellyfish molecule called the green fluorescent protein (GFP). Since then, he and others have developed GFP variants that glow red, yellow, and many other colors. Biologists quickly learned to tag proteins by fusing them with GFP or its cousins. Today biologists can follow several differently colored proteins simultaneously in live cells and in real time. "To see molecules zip around inside living cells and tissues—that's been huge," Svoboda says.
Illustration: Courtesy of Eric Betzig