Bringing EM Back

AIP's News Director reflects on a revolution in Electron Microscopy.
Influenza virus

Colorized transmission electron micrograph showing H1N1 influenza virus particles

Media credits
Jason Socrates Bardi, Editor

Inside Science Minds presents an ongoing series of guest columnists and personal perspectives presented by scientists, engineers, mathematicians, and others in the science community showcasing some of the most interesting ideas in science today. 

(Inside Science) -- Twenty years ago, when I was a young graduate student just beginning to learn about structural biology, I became quite familiar with electron microscopy (EM) as a widely used tool in the field. Doing a rotation in an EM laboratory was one option available to me, and I could have chosen to spend my entire graduate career working with EM. The technique basically uses electrons rather than light to image materials, and it was generally seen as a versatile workhorse back then.

Laboratories all over the world were using it as a cornerstone technique of basic biomedical research, solving the unseen atomic arrangements in important biological molecules, uncovering how they shape human health and disease and helping to guide drug design.

And yet, when weighed in my mind against the other, "high resolution" structural tools like nuclear magnetic resonance (NMR) and X-ray crystallography, the gravitational pull of EM was somehow weaker. So I went a different direction, opting to pursue crystallography and other techniques -- and then leaving the bench altogether to study science journalism.

Now, after reading a new perspective by Caltech professor and Nobel laureate Ahmed Zewail and Caltech senior scientist Dmitry Shorokhov describing a new modality in electron microscopy called 4D ultrafast EM, I fear I missed the boat some twenty years ago.

(Conflict of interest disclosure: their perspective was published this month in The Journal of Chemical Physics, which is produced by AIP Publishing, a wholly owned subsidiary of my employer, the American Institute of Physics. This commentary was not solicited by AIP Publishing, but a journal manager there did send me the article to read).

I was right about NMR and crystallography being truly exciting techniques in the 1990s. Both had undergone revolutions in the 80s and 90s, which by then were beginning to bear serious fruit. More powerful magnets, dedicated X-ray beam lines at national accelerator facilities, enhanced techniques for data collection and analysis, and computer processor speeds that had been growing by leaps and bounds were all emboldening scientists to take on larger and larger structures -- a fact that made those two techniques dazzling to my gaggle of glassy-eyed graduate students.

Many of us gravitated towards laboratories working on NMR or crystallography, and in the coming two decades even some of us who left the field have been delighted from time to time to see how the incredibly complex structures of important molecular machines have been subsequently solved: the ribosome, the polymerase, the nuclear pore complex, and even whole virus particles -- projects that seemed all but unthinkable a few years before.

Looking back, I realize the advances in NMR and crystallography and the attempts to tackle these complex molecular structures made EM seem more important, even if they overshadowed it. The approach to solving huge structures involved breaking a molecule into chunks, getting high-resolution structures of those individual pieces with NMR or crystallography, and then using a map provided by EM to stitch them together into a coherent overall structure.

But that's exactly why EM fundamental lacked appeal for me a generation ago. It seemed like a way to get the picture on the box that helped you piece the puzzle together. Take away the puzzle pieces, and all you are left with is an empty box.

Was I ever wrong about that! I completely failed to see then how EM was on the verge of its own revolution, how like NMR and crystallography, it would also benefit from advances in high-performance computing, how new innovations like aberration correction would come about and allow researchers to push the envelope for resolution. Today we can get equally high-resolution structures in true atomic detail that rival or surpass those obtained with other methods, and EM has reached the point today where single particle imaging is even possible.

In their new perspective, Shorokhov and Zewail detail this slow moving revolution, starting with the early history of the electron microscopy in the 1930s, through Zewail's own theoretical work in the early 1990s, and on to the breakthrough technique of 4D ultrafast EM today, which promises to give us the ability to watch proteins folding and unfolding in real time on ultrafast timescales -- less than a trillionth of a second.

This is a game changer for many fields, not just structural biology, because EM is a tool applied broadly to study inorganic materials as well. But I can only speak to why this is exciting to my biologically trained eye: it promises to be fast enough to watch conformation changes as they occur in proteins, which could uncover previously unseen structures known as "transient intermediate states." These are structures that, for instance, a biological molecule involved in a disease like cancer, Alzheimer's or AIDS may adopt for a fleeting moment as it exerts its deleterious physiological effect. We may be able to target these transient states with new drugs -- something that would be impossible to even consider without an imaging tool because they come and go so in such fast, unseen moments.

Having reached my own proper appreciation of EM and its abilities twenty years later, now that I have read this perspective, I am left wondering what today's average graduate student must think. How cutting edge does it seem now? Will large numbers of graduate students gravitate to EM projects today? Shorokhov and Zewail seem confident it will.

"We hope this perspective will inspire the community, especially young scientists, to join this burgeoning field of 4D visualization of matter and to take part in the race -- not only against time but also for the highest resolution in space!" the authors write at the end of their paper.

Note to graduate students: Shorokhov and Zewail include a partial list of laboratories around the world that are already using 4D ultrafast EM in their paper.

This commentary is an editorial and reflects the personal perspective of AIP News Director Jason Socrates Bardi. It does not necessarily reflect the opinions of his employers or anyone else at AIP, its wholly owned subsidiary AIP Publishing, or the AIP Publishing title, "The Journal of Chemical Physics," which published the article that served as inspiration for this piece. 

Author Bio & Story Archive

Jason Socrates Bardi is the former News Director of the American Institute of Physics and a longtime science writer.