Joachim Frank has written "the" book—or at least many of the major chapters—on the ribosome, the cell's factory for synthesizing proteins. His studies have provided increasingly detailed pictures of what this large, multicomponent molecule looks like, as well as how it functions. But when Frank first started to work on the ribosome's structure, almost 30 years ago, he saw it primarily as a test model for a new technique for constructing three-dimensional (3D) images of molecules by using electron microscopy.
An electron microscope focuses a beam of electrons on an object. The way the electrons are scattered by the object and then combined by the lens provides two-dimensional (2D) images, or projections, of surface and internal structures. During his graduate studies, which he completed in 1970, Frank started thinking about how to turn these 2D images into 3D ones. By 1975, when he was offered a position to head an image-processing group at a new electron microscopy facility of the Wadsworth Center in Albany, New York, he had a clear idea of how to do it.
To obtain the 3D image of a molecule viewed through a microscope, researchers would gather many 2D images as the molecule was tilted into various views and then fit these images together to reconstruct a 3D map. However, multiple exposures to electrons, in an electron microscope, would quickly burn the molecule to a cinder. Frank reasoned that instead of tilting a single molecule, he could take advantage of the fact that molecules exist in thousands of copies, all lying in different orientations on the microscope's specimen grid. Thus, the microscope's field of vision could yield a large set of single molecule images with random "tilts." The actual angles of the molecules would have to be determined by a computer program … and that was the hard part!
To accomplish the averaging and 3D reconstruction, Frank developed an image-processing program called SPIDER (short for System for Processing Image Data from Electron microscopy and Related fields) during his first few years at Wadsworth. Since its unveiling in 1978, SPIDER has been refined and updated and has become a very popular, freely available program.
At around the same time, Miloslav Boublik of Rockefeller University showed Frank some electron microscopy images he had obtained of the human ribosome. "They were very crisp images. I realized how steady and constant the ribosome was as a molecule and a beautiful subject of study," says Frank. He decided to use the ribosome as a test model for his averaging and reconstruction techniques.
Ribosomes consist of two subunits, each made of proteins and RNA, that together bind to messenger RNA (mRNA) and use it as a template to build chains from amino acids delivered to the ribosome by transfer RNA (tRNA). These chains then fold into proteins, the workhorses of a cell's metabolism. Frank produced the first averaged image of the small, 40S, subunit of the human ribosome at a resolution of 20 Angstroms (Ǻ); the structure was published in Science in 1981. "That publication generated a lot of excitement," recalls Frank. It also propelled his scientific career.
When cryoelectron microscopy—a technique for "snap freezing" samples so that they don't have to be stained or fixed before viewing them, showing them in their unaltered, native form—became available, Frank quickly applied it to the ribosome. In 1990, he obtained the 3D structure of the complete ribosome at 45-Ǻ resolution. "That was the first clear structure that showed two subunits," he recalls. In 1995, he homed in even closer. He obtained the 3D structure of the ribosome at 25-Ǻ resolution, making it possible to infer how the mRNA and tRNA interact with the ribosome as well as where the chain of amino acids, or polypeptide, emerges. "That was our first foray into serious functional work," says Frank.
To gain further insights into function, Frank started to take 3D "snapshots" of ribosomes at different stages during the process of protein synthesis, or translation. This study led to the discovery, in 2000, of a ratchet-like rotation of the subunits relative to one another each time the mRNA advances by one unit of the genetic code, or three nucleotides.
Further insights into the dynamics of translation were made possible by a collaboration with Mans Ehrenberg, a biochemist at Uppsala University in Sweden who had been using antibiotics and other compounds to freeze translation at particular steps. "Before, we often just did hit-and-miss experiments to understand function. After getting together with Ehrenberg, it became more of a coherent line of inquiry that we could follow," says Frank.
After moving to Columbia University in 2008, Frank continues to delve into the structure of the ribosome. But his studies, in various new collaborations, have also started to branch out to include other important molecules, such as group II introns and the nuclear pore. And when he is not writing papers that describe the intricacies of these structures, Frank immerses himself in writing fiction. "I have a bank of short stories, poems, and three novels. Most of them are unpublished," he says, adding that he has published more than a dozen short stories and prose poems under his name. He is also an avid reader of fiction, although he admits he does not always find the time to devote to this passion. "And now I have started to need glasses to read—and that's really depressing," he laughs.