Major advances in the imaging of biomolecules—everything from the needles that bacteria use to attack cells to the structure of Zika virus—have garnered three scientists the 2017 Nobel Prize in Chemistry. The award goes to three pioneers of a technique called cryo–electron microscopy (cryo-EM): Jacques Dubochet of the University of Lausanne in Switzerland, Joachim Frank of Columbia University, and Richard Henderson of the Medical Research Council Laboratory of Molecular Biology (LMB) in Cambridge, U.K.
Over the past century, more than a dozen Nobel prizes have been awarded for X-raying crystals of proteins and other complex molecules to picture their atomic structure. This year, the prize went to a technique that can take similarly close snapshots of some of the large, wriggly structures that X-rays can’t see. “It has totally revolutionized structural biology,” says Venki Ramakrishnan, a structural biologist at the LMB who shared the 2009 Nobel Prize in Chemistry for his X-ray crystallography work and who now counts himself as an cryo-EM evangelist.
Holger Stark, a cryo-EM expert who heads the Max Planck Institute for Biophysical Chemistry in Gottingen, Germany, says the method's resolution has improved dramatically in recent years, enough to not only see the outline of a protein’s backbone, but also the amino acid arms that extend to the sides as well – details that are essential for working out how the protein operates chemically, and critical for drug discovery. The award is well deserved, he says. “It was rather quick. But it wasn’t too surprising, was it?”
The electron microscope has been around for a while—there were prototypes as early as 1931—but it had serious limitations for biologists. Samples must be contained in a vacuum, which dries out biological molecules and warps molecular structures. Then the samples are pelted by a beam of radiation that can fry sensitive biomolecules. Beginning in the 1950s, X-ray crystallography allowed biologists to create static images of the structure of proteins. But it requires obtaining copious amounts of purified protein and then coaxing all the identical copies to pack into a regular crystalline orientation—impossible for many large proteins and multi-protein complexes. In the 1980s, nuclear magnetic resonance began providing protein structures, but mostly for small proteins in solution.
It has totally revolutionized structural biology
Henderson took a major step forward out of frustration with X-ray crystallography: He couldn't coax proteins embedded in a cell membrane to crystallize. So he placed a bacterial cell membrane containing a protein called bacteriorhodopsin into an electron microscope and covered it in a glucose solution to keep it moist. He then lowered the amount of energy in the beam, creating a low-contrast picture. Because the molecules were embedded in the membrane in an orderly fashion, Henderson got a diffraction pattern that he could turn into a higher resolution image. By 1975, he had created a 3D picture of the protein. It was, at the time, the finest ever portrait made of a protein with an electron microscope. It was a bit of a special case because of the ordered arrangement of the proteins in the cell membrane, but in principle, this approach could be used for any molecule found in cells.
A challenge remained: how to use low-energy radiation to see less organized molecules. The problem is that these collections of molecules yield a confusing assortment of images, as they are oriented randomly. In work between 1975 and 1986, Frank came up with a way to make sense of the data, by using algorithms to sort the images into related groups of shapes and then averaging each group. This not only sharpened the 2D images but also allowed them to be combined into a 3D structure.
Another significant advance came from Dubochet in the early 1980s. The problem with freezing biological samples is that ice crystals diffract the electron beam, blurring the image. Dubochet realized that if the water was frozen quickly enough—vitrified, like glass—ice crystals wouldn't form. By cooling samples with liquid nitrogen to –196°C, Dubochet created breathtakingly sharp images. Because molecules are flash-frozen, they are caught in a variety of states, allowing researchers to assemble these pictures into movies that recreate their motion—capturing how the shape of the protein changes as it does its work in the cell.
Taken together, these approaches pushed the resolution of cryo-EM to 0.5 nanometers. That still wasn’t good enough to match X-ray crystallography, which today achieves a resolution of around 0.2 nanometers, and sometimes even better. “We kind of got stuck,” Frank says. But the replacement of film-based detectors with digital ones—a development pushed by Henderson-- has changed the landscape for cryo-EM, enabling resolution as low as 0.25 nanometers. “That brought us into the realm of X-ray crystallography,” he says.
Cryo-EM techniques are elucidating the structures of ever larger complexes including the ribosome, the cellular factory that produces proteins, and the spliceosome, the complex that trims non-coding material out of RNA. Henderson says the vast world of membrane-bound proteins and protein complexes that can’t be crystallized is now in reach. “In a few years, perhaps 5 years, we might know most of the structures, at least in humans and pathogenic bacteria,” Henderson says. “It’s really quite an exciting time.”
And the revolution isn’t over yet. The National Institutes of Health and other national biomedical funding agencies are spending tens of millions of dollars to build new cryo-EM centers to expand access to the machines. One upshot, says Ramakrishnan, is that structural biologists like himself are getting out of the business of X-ray crystallography. “Nobody in my lab wanted to set up crystallization trays anymore and pray our proteins would crystallize.”
The exodus could have an impact on the billion-dollar, stadium-sized synchrotrons that generate the X-ray light for crystallography experiments. With so many structural biologists turning to cryo-EM, “there might soon be too many synchrotrons around”, Stark says. “That could happen.” Some synchrotrons in Europe, he adds, are already adding cryo-EM facilities. For the moment, X-ray crystallography remains the gold standard for imaging proteins that can be crystallized. But if this week’s Nobel prize is any sign, a new standard bearer is on its way.