Seeing Is Believing
Improving the view inside cells
Biophysicist Niels de Jonge longs to see what interests him, no matter how small. That includes individual proteins or receptors inside mammalian cells. After all, who knows what surprises a good look at the real thing will yield? This drive, and an affection for tough problems, led the Oak Ridge National Laboratory and Vanderbilt University researcher to find a means to image whole cells in liquid with a scanning transmission electron microscope. The technique improves the view inside a cell by a factor of more than 10 compared with ultrahigh-resolution optical imaging with quick processing times. To do this, de Jonge needed to overcome a sizable obstacle: Scanning transmission electron microscopy requires a high vacuum, a hostile environment for samples in liquid, including, of course, whole cells. Associate Editor Catherine Clabby interviewed de Jonge about the innovation.
American Scientist: Tell us more about the limitations you were trying to beat.
De Jonge: Since the origin in the 1930s of the electron microscope, there’s always been a wish to image cells just like you can do with light microscopy. It’s always been difficult due to the liquid. There have been some solutions, for example with the transmission electron microscope (TEM). But those only work with thin samples and not whole cells. If you go thicker, the contrast mechanism of the TEM prevents high resolution. Other approaches used the scanning electron microscope (SEM). There have been some good examples of a vacuum chamber with water vapor and another with special capsules. But SEM is a surface technique. You can look a little bit under the skin, but not much farther than 50 nanometers.
My vision has been to image a whole eukaryotic cell mostly in its native state. Those are rather thick: 5 to 10 micrometers. There has been huge progress with optical microscopy where you can look at whole cells. But the resolution is not so good. You can see regions where tagged receptors are but you cannot always see individual proteins or receptors. That’s really what you’re aiming for.
American Scientist: How did you overcome all that?
De Jonge: I came into contact with a very good scanning transmission electron microscope (STEM) team at Oak Ridge National Laboratory where I tried to apply STEM to this particular problem. Only a couple of groups use STEM in biology studies. Traditionally TEM is used. The reason has been that TEM is best for biological samples because it has good resolution with carbon. All biological materials have carbon. But we want to look through thick samples. If the sample produces too much signal, you can’t image it. I was looking for a technique with low contrast on carbon, which STEM has.
In the samples, you can use high-contrast labels—we’ve used gold—to tag molecules. The idea is that you can see very small labels of high atomic number inside the thicker layer of material of lower atomic number. To protect the cells, we used a special microfluidic device made from silicon chips. They have very thin silicon nitride windows that are electron transparent. You can make a sandwich with two of those chips. Liquid can be enclosed between the windows, separated from the vacuum, while the electron beams can go through the whole thing. And we needed a holder for placing the liquid enclosure inside the electron microscope.
American Scientist: What improvements in resolution did you achieve?
De Jonge: We have a resolution of 4 nanometers, which you cannot achieve optically. And STEM is faster than the finest-resolution optical imaging. If you go beyond a resolution of 50 nanometers with optical imaging, it’s very slow. It’s a very complicated matter for data processing. STEM can take an image in a matter of seconds. But there are limitations: The electron beam creates radiation damage.
I believe if you combine optical imaging and STEM you have a very viable imaging technology. You can look first in optical and see what processes are happening and in which regions certain receptors are located. Then you can zoom in with the electron microscope on a whole cell in the sample. You can actually see which proteins are interacting with which receptors.
American Scientist: In a Proceedings of the National Academy of Sciences article this year, you and collaborators Diana Peckys, Gert-Jan Kremers and David Piston described how you imaged single, gold-tagged epidermal growth factor molecules bound to receptors inside fixed African green monkey fibroblast cells. In this case, the resolution reached 4 nanometers. In what other kinds of studies do you envision this new tool would be useful?
De Jonge: It’s such a large field. What I am trying to do is find collaborations with biomedical researchers. I want to learn what kinds of problems they have and how I can help. We continue to work, for example, with epidermal growth factor. I’m also interested in the ways viruses interact with cells.
American Scientist: Will you attempt to keep improving this technology?
De Jonge: I’m working on developing a new way to get 3-D images from electron microscopes. The traditional technology involves tilting the sample. With the fluid holder, I don’t want to have to tilt anymore. Cells are 3-D structures. You want 3-D images of them.
In Sightings, American Scientist publishes examples of innovative scientific imaging from diverse research fields.