This year’s Nobel Prize in Chemistry honors an interesting mix of developments. It honors three researchers who overcame an apparent physical limitation in our ability to image microscopic objects, in the process building microscopes that are proving to be incredibly useful for biology. But because the breakthroughs depended in part on our understanding of the behavior of individual molecules, the prize comes in chemistry.
The limit in question is the diffraction limit, first described back in the 1800s by Ernst Abbe. This limit means that the best resolution we can obtain in imaging an object is half the wavelength of the light we’re using to image it. If we’re using visible wavelengths, this means we can’t do much better than about 250nm (a distance that dwarfs viruses and individual proteins). Although lots of improvements in microscopy have been made since the 1800s, all of them kept running into diffraction-related problems.
At least, that was the case until recently. The Nobel Prize honors not one but two distinct ways of overcoming the limit. (Conveniently, we have coverage of both-see the sidebar.) In the case of one of the recipients, it honors an idea that came to him when he had given up on research and was working in the family business.
The view from (Stephan) Hell
The first of these is called STED microscopy, for STimulated Emission Depletion. It’s the brainchild of Stefan Hell, a Romanian who is now a German citizen and directs institutions in GÖttingen and Heidelberg. Hell started his work by using standard fluorescent microscopy, where a large population of molecules is stimulated to glow by a laser light. Imaging the resulting glow, however, ran into the diffraction limit, meaning that small objects could not be resolved.
Rather than try to resolve the population of molecules better, Hell decided to limit the number of molecules that are glowing at any given instant. To do so, he took advantage of the detailed behavior of fluorescent molecules. These have to absorb a photon to reach an excited state, after which they’ll emit another to return to the ground state-this second photon is what we see as fluorescence. Normally, this process is a bit stochastic-you can’t tell when the second photon will be emitted. But it’s also possible to hit the fluorescent molecule with a second photon of the right wavelength and get it to emit immediately, effectively draining the molecule back to its ground state. This is the “stimulated emission” of STED.
Now, it’s not possible to create a laser beam that only excites a tiny population of molecules-the diffraction limit bites you here too. But it is possible to shape a laser beam with a small, empty hole in the middle. Hell realized that, if you use this beam to drain the fluorescent molecules to their ground state, then the only things glowing will be in the empty hole in the middle.
Hell’s system uses a single exciting laser and a donut-shaped draining laser, both centered on the same point. Fluorescent molecules in the center of that point glow, and little outside of it does, since they’re all being drained. By scanning that point across a sample, it’s possible to build up an image of every area in the sample, all done at extremely high resolution. Since the size of the hole in the middle shrinks as the intensity of the light increases, it’s technically possible to shrink the glowing point to an arbitrarily small size. Of course, boosting the intensity enough will vaporize your sample, so there’s still some limit in operation here.
Hell himself is a rare example of someone who took an idea from start to finish. He published the first theoretical papers on the idea, helped build the first microscope, and eventually used it to image a bacterium at never-before seen resolution. He’s now involved in commercializing the technology as well.
Seeing single molecules
Fluorescent microscopy relies on looking at a population of molecules. Hell’s approach involves shaping the light field such that the population is as small and physically compact as possible. The other half of the prize goes to two scientists who started from the opposite direction-learning to image single molecules, and then building these images into a picture of a larger structure.
To do this, you had to demonstrate that it was possible to image a single molecule. And the first to do this was W. E. Moerner, then of IBM, now at Stanford. Moerner managed to measure the absorbance by a single molecule that was held in liquid helium in 1989. This development opened the floodgates; over the next few years, other individual molecules were imaged, some at room temperature. By 1995, people had imaged individual proteins with a specialized microscope.
By the late ’90s, Moerner was at UC San Diego, where Roger Tsien was doing his Nobel-winning work on the Green Fluorescent Protein. Moerner discovered a switchable form of GFP, one that was inactive until it was hit with photons of a specific wavelength. Once activated, photons of a second wavelength could cause it to fluoresce for a while before it spontaneously switched back to the ground state.
Eric Betzig saw this result and ran with it. A few years earlier, he was pursuing a different avenue to high-resolution imaging, one that was so frustratingly difficult that he eventually left academics and went to work in the family business. But with an idea in hand, Betzig began putting together a team to build a new microscope based on the idea of a switchable fluorescent molecule. (He’s now back in research full time at the Howard Hughes Medical Institute’s Janelia Farm facility.)
The resulting technique, called PALM (Photoactivated Localization Microscopy), involves using a very weak light pulse to activate a switchable fluorescent molecule. The weakness ensures that only a small number of individual molecules will fluoresce, spread randomly over whatever is being imaged. As a result, all of the light sources for a given bit of imaging are single molecules, rather than a population. The diffraction limit still cuts into our ability to figure out where they are, but it’s possible to get an accurate estimate by calculating something called a “point-spread function” for each source of light.
This gets you what is essentially an incomplete pointillist image of whatever it is you’re trying to view. But by repeating the process, you get a different set of molecules fluorescing and a few more details. Do it repeatedly, and you can build up a complete image of the object in question, all from single fluorescent molecules.
If yesterday’s honoring of the blue LED praised a technology that has reached the mainstream, today’s award to super-resolution microscopy focuses on future potential. These techniques have existed less than a decade, and they’re just now getting into the hands of individual research groups or facilities. They’ll undoubtedly give us a different view of life, but it’ll take a little time for the research community to figure out what questions are best answered with them.