Cryo-imaging gives deeper view of thick biological materials

Electron microscopy is an exceptional tool for peering deep into the structure of isolated molecules. But when it comes to imaging thicker biological samples to understand how those molecules function in their cellular environments, the technology gets a little murky.

Cornell researchers devised a new method, called tilt-corrected bright-field scanning transmission electron microscopy (tcBF-STEM), to image thick samples with higher contrast and a fivefold increase in efficiency.

The Sept. 23 publication of the findings, in Nature Methods, arrives two years after the death of co-author Lena Kourkoutis, M.S. '06, Ph.D. '09, associate professor in applied and engineering physics in Cornell Engineering, whose work in cryo-electron microscopy drove much of the nearly 10-year effort.

The project began when David Muller, the Samuel B. Eckert Professor of Engineering, realized that the electron microscope pixel array detector (EMPAD) that his group co-developed, which is capable of seeing atoms at record resolution, might also be useful for imaging biological samples. To do that requires cryo-electron microscopy in which a liquid-containing sample is snap frozen at cryogenic temperatures so it can survive in the vacuum of an electron microscope. In that process, a beam of electrons is fired at the molecule, but the electrons damage the sample in the process.

"You kind of cook them," Muller said. "A lot of this imaging technique is a bit like playing dodgeball. Once you've thrown the ball out and you hit someone, then you know where they were, but they're out of the game, and you can't throw the ball there again."

Efficiency, therefore, is paramount.

"Our detectors were very good at detecting individual electrons, and it could collect all of them that went through the sample, but we were used to working with images with millions of electrons per pixel, not single electrons," Muller said. "Lena was already developing cryogenic methods with very low numbers of electrons. So we started working with her, and her group took the lead."

The new method they created, tcBF-STEM, removes a drawback of traditional transmission electron microscopy, in which a camera records the pattern of electrons, as a sort of shadow image, after the beam has passed through the sample. Because energy is lost through electron scattering in the sample, the resulting image can blur - a problem that only worsens in thicker samples.

The new approach places the imaging optics before the sample, so it doesn't matter if the electrons lose energy in the sample - there are no imaging lenses afterwards. Instead, a high-speed EMPAD detector captures the angular spread of the ricocheting electrons from a small spot, then repeats the process at another point, and another, much like sweeping a searchlight across a landscape.

"STEM is traditionally very inefficient and only uses about about one in 100 of the transmitted electrons," Muller said. "Now we know how to use almost every electron."

The new method, which was conducted at cryogenic temperatures but is not limited to them, can create images inside intact bacterial cells and large organelle up to 500-800 nanometers thick - an improvement of roughly a factor of five. Potential applications could range from discovering the elusive function of proteins to imaging lithium-ion batteries, which, like biological materials, are extremely sensitive to radiation.

'Beautiful results'

By 2017, the team was getting "beautiful results," but student turnover slowed their progress, until Kourkoutis picked Yue Yu, Ph.D '23 to take on the project for her doctoral thesis.

"Lena was both a brilliant scientist and a mentor who fully supported her students to grow," said Yu, the paper's lead author and currently a researcher at the Chan Zuckerberg Institute for Advanced Biological Imaging. "The tcBF-STEM project ended up spanning my entire Ph.D. and beyond, and throughout it all, Lena showed nothing but absolute encouragement. She sat with me many times to search through books and texted me back immediately with great excitement when we first saw apoferritin with tcBF-STEM. I would not be the scientist I am today without her."

In 2021, Kourkoutis was diagnosed with cancer, and it sapped her energy, said Muller, who was both Kourkoutis's colleague and had also served as her graduate adviser more than a decade earlier.

"The last day Lena came to campus was for Yue Yu's Ph.D. thesis defense," he said. "It was on Friday, and on Monday she was in hospice."

By then, the team had begun presenting their results at conferences, including that same week, and even some of the most skeptical biologists were excited, according to Muller.

"Things were starting to work experimentally. It was this thing she'd worked on for many years," he said. "And at the point where it was getting the recognition, it was just a little too late for her."

A number of researchers have already begun using the team's technique and building off the work. Now, with the publication of the paper, they can finally cite it.

"What's interesting is the cryo-EM community had been following her work, and was kind of waiting for, 'When are you going to publish?' I think, in hindsight, we always thought of Lena as indestructible," Muller said. "A lot of what was driving this was sort of making sure that her work got remembered."

Co-authors include research associate Steven Zeltmann; Katherine Spoth, Ph.D. '18, Kayla Nguyen, M.S. '14, Ph.D. '18 and Xiyue Zhang, M.S. '21, Ph.D. '24; doctoral student Michael Colletta; and researchers from Chan Zuckerberg Institute for Advanced Biological Imaging and the New York Structural Biology Center.

The research was supported by the National Science Foundation (NSF), the Packard Foundation and Chan Zuckerberg Institute for Advanced Biological Imaging.

This researchers made use of the Chan Zuckerberg Institute for Advanced Biological Imaging, the Cornell Center for Materials Research (CCMR) Shared Facilities and PARADIM, all of which are supported by the NSF.