Published 3 July 2013 by Kathleen Raven

Imaging the near invisible with TEM: a master class

Though nanometer-level imaging has come far with transmission electron microscopy, Nobel Laureate Dan Shechtman (Nobel Prize 2011, Chemistry) warned his master class audience on Tuesday that today’s images will seem primitive a few years in the future. For now, the five students—all but one of them female—presented research at the edge of what this writer can imagine.

Nobel Laureate Dan Shechtman opens the master class in TEM. Photo by Kathleen Raven

All researchers addressed two major conundrums of TEM and biological samples: How does a researcher avoid radiation damage while also capturing the highest detailed image with the microscope? How can one build a 3-D model of the images?

TEM images are, in fact, three-dimensional, Shechtman reminded us. But the human brain is not able to fill in the 3-D information contained in the image. “This is the same as if you are looking at something with one eye covered. You know that the object has three dimensions, but with only one eye, you can see only two dimensions,” he explained.

Lindsay Baker, now at Utrecht University in The Netherlands, gave an overview of the “power of TEM” using work she completed in a lab at the University of Toronto in Canada. She echoed Shechtman’s words when she asked her colleagues: “How do we go from 2-D images of carbon, oxygen, hydrogen or nitrogen molecules to 3-D?” She quickly translated the problem using an everyday example. A doctor may examine a patient’s broken bone in a X-ray image, but only part of the bone is visible at a time. His or her job would be much easier if the entire bone fragment could be seen from all angles at once. To do this with single particles, Baker and colleagues followed three steps: 1) they created a map, based on previous knowledge, of what they envisioned the 3-D image to be; next, 2) the group assigned projection directions for each image; and, finally, 3) they used this information to rebuild the map with each different TEM image they recorded. The team used hundreds of thousands of single particle images. “We did this over and over again, hundreds of times, until we had an accurate 3-D model,” Baker explained.

But what about the parts of the cell, for example, that are not accessible for TEM. How do we image those? Julia Mahamid of the Max Planck Institute for Biochemistry in Germany gave these string of steps: vitrification, correlative microscopy, a focused ion beam and cryo-electron tomography. “This sounds simple, but it is not,” Mahamid said. The audience murmured in agreement: nothing about it sounded simple. Mahamid balanced her intricately detailed presentation with gorgeous 3-D images spinning around like colorful dancers on a black screen. At the end, we were able to see an entire centrosome in a miotic HeLa cell.

A slide from Mehtap Özaslan’s presentation. She is with the Paul Scherrer Institute in Switzerland. Photo by Kathleen Raven

Shechtman reminded the young researchers that TEM is only as strong as the specimen, which, of course, the human must provide. “Your contribution is the specimen, so preparation is a number one priority,” he said.

Evelyn Auyeung, who joined Northwestern University’s International Institute for Nanotechnology in 2009, took the microphone next. “This is a story of how I was able to help modify a system for characterization by TEM,” she began. The system in question? DNA nanoparticle super-lattices. These super-lattices are usually operated on in a solution, but they must be in a solid state for successful TEM imaging. As recently as 2008, scientists thought it couldn’t be done. Auyeung highlighted a quote from Nykypanchuk et. al., in Nature: “Such open framework of a super-lattice makes the structure vulnerable to collapse upon solvent removal, which prevents accurate morphology visualization by electron microscopy,” he wrote. With a hint of triumph in her speech, Auyeung clicked the projector to show exactly these images. Her lab used silicon encapsulation and X-ray scatter patterns to image the lattice structures.

Thomas Lunkenbein, who studies with the Fritz Haber Institute of the Max Planck Society in Germany, then discussed TEM in catalysis. He drew ironic laughter from the crowd when a red arrow on his presentation appeared to highlight mere blackness on the screen. Imaging problems are everywhere!

And, finally, Mehtap Özaslan of the Paul Scherrer Institute in Switzerland, gave a riveting presentation that Shechtman highly praised. The fault of a lack of summary for these last two presentations lies entirely with the fault of the writer, who simply could not keep up with the technical notes.

At the end of the class, Shechtman asked: “What have we learned?” He paused for a moment while the room let the expert answer his question. “One thing we have learned is that many signals can be collected from specimens. We need to move towards better and better signal collections,” he said. With that, he charged all of the eager scientists in the room to go back to their home countries and help develop great laboratories that may one day provide the environment for Nobel Prize-worthy work.

The Master Class Group: From left to right: Evelyn Auyeung, Julia Mahamid, Mehtap Özaslan, Lindsay Baker, and Thomas Lunkenbein. Photo by Kathleen Raven

Kathleen Raven

Kathleen Raven reports on cutting-edge solid tumor cancer drug developments and clinical trials for BioPharm Insight, owned by The Financial Times Group, in New York City. She’s previously written for Reuters Health, Scientific American, MATTER, Nature Medicine and other U.S. publications. She has been a recipient of the following short-term reporting fellowships: National Academies Keck Futures Initiative, Goethe Institute, Fulbright Berlin Capital and Falling Walls. She has two master’s degrees from the University of Georgia in Ecology and Health & Medical Journalism.