The past two years have been exciting—some say revolutionary—for cryo-electron microscopy (cryo-EM). By the end of 2014, there were two 3.2 Å-resolution structures, the highest yet, determined using the method: β-galactosidase and a large subunit of the yeast mitochondrial ribosome (1,2).Reaching 3 Å-resolution with cryo-EM is in itself a big accomplishment thought to mark a new era in structural biology. Researchers who have been using X-ray crystallography to achieve atomic resolution have begun to flock to cryo-EM for its ability to resolve molecules that are not amenable to crystallization.

But the field as a whole believed that big technological advancements would be needed to go beyond 3 Å, a resolution that holds significance for drug design because it allows for visualization of water molecules, salt bridges and hydrogen bonding.About a year ago, working with the 700-kDa Thermoplasma acidophilum 20S proteasome (an enzyme important for ridding the cell of misfolded or excess proteins), Bridget Carragher and Clint Potter, then at the Scripps Research Institute, set out to see whether there was a resolution barrier. “You don’t even know, at the beginning, what counts. What exactly is important, where are you losing resolution, and why are you losing resolution?” Carragher said.

Carragher, Potter and their team spent time optimizing their instruments and getting everything “just right.” They used a protocol that was slow but certain not to cost them any resolution. Of course, the scientists also had top-of-the-line equipment: the FEI Titan Krios microscope paired with a Gatan K2 Summit direct detector.

They were one of the first research groups to break the 3 Å barrier, resolving the proteasome’s structure at 2.8 Å resolution in March (3). Near that same time researchers from Max-Planck-Institute for Biophysical Chemistry in Germany determined the structure of the E. coli 70S ribosome in complex with elongation factor Tu at a resolution of 2.65-2.9 Å resolution using spherical aberration-corrected cryo-EM (4), and Hong Zhou’s group at UCLA solved the structure of the anthrax protective antigen pore, a therapeutic target for anthrax infections, at a resolution of 2.9 Å (5).

Reasons for the Revolution

In single-particle cryo-EM, protein suspensions are frozen at liquid nitrogen temperature and imaged at high resolution in the electron microscope. The method has been around for more than 30 years, with its potential to solve larger structures at resolutions comparable to X-ray crystallization recognized for many of those years (6).

Although cryo-EM comes with the benefit of not requiring crystallization, its main challenge is that biological specimens tolerate only low electron doses before they are destroyed. Until a few years ago, the best resolution you could get with conventional means of detection such as photographic film and CCD cameras was (in most cases) 10-15 Å.

“It’s like taking a picture at night without a flash in your camera,” said Howard Hughes Medical Institute Investigator Eva Nogales, a biophysicist at the Lawrence Berkeley National Laboratory in California. “The way we have to go about it is to collect many many images of the same object.” Noisy images are difficult to align, however.

That has all changed with new direct detector technology (which detects electrons directly instead of converting them to light and then back to electrons). The detectors are less noisy and much faster, allowing researchers to collect enough frames of an image to make small videos. These make it possible to correct for tiny movements caused by the interactions of the electrons with the specimen.

The result is many more pictures. More data equals a larger computational burden, but software tools have developed in parallel, and methods such as maximum likelihood image processing are making it easier to select the molecules to include in analysis and to align the images.

As a result, said Nogales, the images are now 100 times better with direct detectors than they were with CCD cameras. Nogales, who as a postdoc solved the structure of tubulin, goes on to state that what used to take years to complete can now, in best cases, be done in a day or two.

Thus, with a more sophisticated toolbox, researchers have in short order eviscerated the 3 Å resolution barrier.

Last year, after Sriram Subramaniam of the National Cancer Institute and his group obtained their β-galactosidase with 3.2 Å resolution(2), they were not satisfied that they had done all they could to achieve atomic resolution. “We would make specimens but there was no real way of assessing which one was good and which was not,” he said.

He and his team rifled through the specimens, finding the occasional example in which that they would get close to 2.5 Å. “We then said, ‘Okay, let’s see if we can routinely hit images that go to the highest possible,’” Subramaniam said.

The hardest part was to understand where changes were needed. Working with the same equipment as before, the strategy came down to making improvements in specimen preparation, imaging, and data processing. The team collected data as best they could and systematically evaluated which portion of the data was most useful. The result, published in Science in May, was the best resolution given by cryo-EM to date: 2.2 Å (7).

What’s Ahead

Structural biologists are now getting new microscopes or bulking up the ones they already have. Now directing the Simons Electron Microscopy Center at the New York Structural Biology Center, Carragher and Potter have welcomed a new Krios with a direct detector to their line-up. They expect to have it running by the end of the year. The Purdue Cryo-EM Facility just received a new K2 Summit Direct Electron Detector to go with their Titan Krios microscope.

Before acquiring the K2, postdoctoral researcher Lei Sun in Michael Rossmann’s lab at Purdue University spent the better part of 2 years trying to obtain a high-resolution structure of the bacteriophage T4 portal protein (a protein important in genome packaging and initiating capsid assembly) using crystallography. The best resolution she could get was 6.5 Å.

After spending a week with a direct detector at UCLA, Sun was able to get the data the group needed for a 3.6 Å-resolution structure. Their work, published July in Nature Communications (8), revealed why the protein had been so difficult to crystallize in the first place: it had a variable number of molecular building blocks. Although the complex was supposed to have 12 subunits, a few had 11 whereas others had 13.

With the new detector, one of the first things Rossmann plans to do is to determine the structure of a recent isolate of enterovirus D68 (EV-D68), the cause of the 2014 outbreaks of respiratory illness in children in the United States. In January, his group described the crystal structures of an older isolate of EV-D68 bound to pleconaril, an anti-rhinovirus drug (9). Because EV-D68 has proven difficult to crystallize, the group is hoping to have better luck with cryo-EM.

Cryo-EM is still technically challenging, and the technology for achieving this resolution is in the hands of relatively few scientists. “Much of what we’re seeing now has been in the pipeline for many years,” said Werner Kühlbrandt, a structural biologist at the Max Planck Institute of Biophysics, Germany. These longstanding projects, in which people have worked long and hard to purify and stabilize their molecules, are ripe for cryo-EM. But an increasing number of projects will take on more unfamiliar structures that will require optimization before they see the likes of a Krios.

What’s more, the structures that have been resolved thus far are large, rigid, and for the most part symmetrical. In contrast, membrane protein complexes and other macromolecular complexes that are relevant for drug design are flexible.

Even for proteins as familiar as β-galactosidase, which scientists had already solved using crystallography, cryo-EM protocols are not as refined as they could be, Subramaniam said. His group hopes to push the resolution further, by focusing on sample preparation in particular. They also want to understand what aspects of their new methods will apply more generally. “We need to understand the landscape of how different proteins behave much better to be able generalize this,” he said.

In the meantime, optimism is high. “The fact that we can now tackle these problems [at sub-3 Å-resolution], is a like dream come true,” Kühlbrandt said. “Even 2 years ago, few would have believed that this was possible.”

References

1. Bartesaghi A, Matthies D, Banerjee S, Merk A, Subramaniam S. Structure of β-galactosidase at 3.2-Å resolution obtained by cryo-electron microscopy. Proc Natl Acad Sci 2014 Aug 12;111(32):11709-14.

2. Brown A, Amunts A, Bai XC, Sugimoto Y, Edwards PC, Murshudov G, Scheres SH, Ramakrishnan V. Structure of the large ribosomal subunit from human mitochondria. Science. 2014 Nov 7;346(6210):718-22.

3. Campbell MG, Veesler D, Cheng A, Potter CS, Carragher B. 2.8 Å resolution reconstruction of the Thermoplasma acidophilum 20S proteasome using cryo-electron microscopy. Elife. 2015 Mar 11;4. doi: 10.7554/eLife.06380.

4. Fischer N, Neumann P, Konevega AL, Bock LV, Ficner R, Rodnina MV, Stark H. Structure of the E. coli ribosome-EF-Tu complex at <3 Å resolution by Cs-corrected cryo-EM. Nature. 2015 Apr 23;520(7548):567-70.

5. Jiang J, Pentelute BL, Collier RJ, Zhou ZH. Atomic structure of anthrax protective antigen pore elucidates toxin translocation. Nature. 2015 May 28;521(7553):545-9. doi: 10.1038/nature14247. Epub 2015 Mar 16.

6. Henderson R. The potential and limitations of neutrons, electrons and X-rays for atomic resolution microscopy of unstained biological molecules. Q Rev Biophys. 1995 May;28(2):171-93.

7. Bartesaghi A, Merk A, Banerjee S, Matthies D, Wu X, Milne JL, Subramaniam S. Electron microscopy. 2.2 Å resolution cryo-EM structure of β-galactosidase in complex with a cell-permeant inhibitor. Science. 2015 Jun 5;348(6239):1147-51.

8. Sun L, Zhang X, Gao S, Rao PA, Padilla-Sanchez V, Chen Z, Sun S, Xiang Y, Subramaniam S, Rao VB, Rossmann MG. Cryo-EM structure of the bacteriophage T4 portal protein assembly at near-atomic resolution. Nat Commun. 2015 Jul 6;6:7548.

9. Liu Y, Sheng J, Fokine A, Meng G, Shin WH, Long F, Kuhn RJ, Kihara D, Rossmann MG. Structure and inhibition of EV-D68, a virus that causes respiratory illness in children. Science. 2015 Jan 2;347(6217):71-4.