Skip to Main Content
Skip Nav Destination

Dubochet, Frank, and Henderson share 2017 chemistry Nobel for cryo-EM

4 October 2017

The laureates turned the electron microscope into a powerful tool for studying biologically important molecules.

Dubochet, Frank, Henderson
From left: Jacques Dubochet, Joachim Frank, and Richard Henderson. Credits: Keystone/Jean-Christophe Bott, Columbia University Medical Center, Cambridge University

Jacques Dubochet, Joachim Frank, and Richard Henderson are to be awarded the 2017 Nobel Prize in Chemistry for developing cryoelectron microscopy (cryo-EM) to examine the structure of biomolecules, the Royal Swedish Academy of Sciences announced on Wednesday. The three researchers, all of whom started out as physicists, methodically developed a biomolecular imaging technique that currently supplements and sometimes even supplants x-ray crystallography. They will share the 9 million Swedish krona (roughly $1.1 million) prize equally.

Henderson, of the MRC Laboratory of Molecular Biology in Cambridge, UK, was the first to use electron microscopy to obtain an atomic model of a protein. Frank, of Columbia University, developed computer algorithms that pieced together blurry two-dimensional electron micrographs of biomolecules into useful three-dimensional structures. And Dubochet, of the University of Lausanne in Switzerland, embedded biomolecular samples in amorphous water, which allowed them to retain their shape even when exposed to vacuum conditions.

Today, spurred by recent advances in detector technology, researchers are racing to analyze the structures of various biomolecules, particularly large and flexible ones, that had evaded classification with x-ray crystallography. Recent studies have probed the Zika virus, molecules governing circadian rhythms, and proteins involved in the spread of cancer cells (see Physics Today, August 2016, page 13).

Methodical progress

For all the success of x-ray crystallography in exposing the secrets of DNA, RNA, and thousands of proteins, the technique has limitations. Some proteins, including those entrenched in cell membranes, don’t crystallize easily; others resist organizing into crystals large enough to extract useful structural data.

Zika virus
This atomic-structural model of the Zika virus was informed by cryo-EM imaging. Credit: Johan Jarnestad/The Royal Swedish Academy of Sciences

Henderson faced both those challenges in the 1970s when he attempted to analyze bacteriorhodopsin, a proton-shuttling protein that resides in the membranes of archaea. He and colleague Nigel Unwin decided to see whether they could do better with transmission EM. Due to electrons’ charge and minuscule wavelength, the particles scatter strongly off atoms and are capable of providing subatomic resolution (see Physics Today, March 1999, page 21). But there’s a price to pay. The samples must be placed in vacuum, which would instantly sap them of water. The desiccated molecule would then get bombarded by thousands of electrons per square nanometer, a concentration sure to break bonds. Despite some previous work by chemistry Nobel laureate Aaron Klug and others, EM was thought to have little use in biochemistry beyond imaging some viruses and stained tissues.

Henderson and Unwin found workarounds by keeping the bacteriorhodopsin inside a glucose-coated membrane for protection and reducing the electron intensity. Although the researchers could not observe individual molecules, they oriented the proteins at different angles under the microscope and analyzed the resulting diffraction patterns. In 1975 they managed to produce a rudimentary 3D structural model of the protein, with a resolution of 7 Å. Henderson would continue his quest to produce a sharper electron-derived structure of bacteriorhodopsin for the next 15 years.

If EM was going to become a bigger player in biomolecular imaging, researchers needed to develop surefire methods for treating and analyzing a wide variety of samples. In 1982 Dubochet made a major advance in protecting samples from the harsh vacuum conditions. Rather than immersing the samples in liquid water, which would evaporate, or ice, which would strongly diffract the electron beams, he flash-froze proteins in solution into a glassy phase. Dubochet and colleagues succeeded in realizing such an amorphous state, known as vitrification, by thinning the samples and immediately exposing them to liquid nitrogen–cooled ethane (see the box in the article by Bob Glaeser, Physics Today, January 2008, page 48).

The use of vitrified water enabled the study of molecules in their natural, hydrated state. The embedded molecules’ frozen shield also allowed them to withstand electron exposures several times greater than they could at room temperature. In 1984 in Nature, Dubochet published electron images of viruses encased in vitrified water. The cryo in cryo-EM was born.

As Dubochet was working out how to obtain useful images under the microscope, Frank was devising analytical techniques to transform 2D images—which were inherently noisy due to the limitations on beam intensity—into 3D structural data. In 1977 Frank and colleagues used cross-correlation functions to derive the position and orientation of molecule substructures. Four years later they incorporated their image-analysis algorithms into a software package they called SPIDER. Frank applied the method to create models of ribosomes.

Coming of age

In 1990 Henderson succeeded in generating the first truly high-resolution biomolecular structure via an EM technique. Using several of the world’s best electron microscopes at the time, his team produced a 3D density map of bacteriorhodopsin that could be fit to an atomic model.

Henderson’s dogged pursuit kept cryo-EM on the map; the work of Frank and Dubochet made the technique more practical and applicable to a larger suite of molecules. But recent technological advances have now pushed cryo-EM over the top. Around 2013 new electron detector technology began offering an unprecedented combination of contrast and speed. Imaged molecules that had looked like fuzzy blobs suddenly came into focus at near-atomic resolution.

Over the past four years, the structural biology community has taken notice of cryo-EM’s enhancements. In fact, researchers are beginning to approach the limits set in 1995 by Henderson, who published calculations indicating that in principle the technique could achieve 3 Å resolution of molecules as small as 38 kilodaltons (1 Da equals one atomic mass unit). At this year’s meeting of the American Crystallographic Association, scientists announced that they had used cryo-EM to determine the structure of human hemoglobin, a 64 kDa protein, with 3.2 Å resolution (see Physics Today, September 2017, page 22). For larger molecules, researchers have achieved an image resolution of 1.8 Å. Through the imaging of molecules that have long held on to their structural secrets, scientists can now work on basic research, drug discovery, and the prevention of disease transmission.

Close Modal

or Create an Account

Close Modal
Close Modal