7 September 2009. Pushing the limits of light and electron microscopy, two new studies promise to open up ever more detailed nano-vistas into cells and tissues. Writing in the August 27 issue of Neuron, Bernardo Sabatini and colleagues at Harvard Medical School describe a combination of fluorescence imaging and stimulated-emission depletion (STED) techniques to achieve nanometer-level microscopic resolution in living tissues. And in an advance in electron microscopy, Richard Leapman, Alioscka Sousa, and colleagues at the NIH National Institute of Biomedical Imaging and Bioengineering report the modification of scanning transmission EM tomography for use in thicker samples. They use the method to construct an image of an entire malaria-infected human erythrocyte at 5-10 nm resolution. That work appeared online August 30 in Nature Methods.
In the realm of light microscopy, the long-sought achievement of resolutions finer than the wavelength of light became a reality in the past few years, with the advent of methods such as stimulated emission depletion (STED) and others (see ARF related news story on nanoscopy, Nature magazine’s 2008 Method of the Year). Sabatini and coauthors Jun Ding and Kevin Takasaki employed STED to improve the resolution of two-photon laser-scanning fluorescence microscopy by a factor of three. By using infrared lasers that can penetrate intact tissue, they obtained dramatic images of dendritic spines on living neurons buried 100 microns deep in mouse brain slices (see image below). The researchers were able to quantitate spine morphology and get repeat images of the same spines over time.
A mouse brain dendrite imaged in an intact tissue slice using 2-photon laser scanning microscopy (2PLSM, left panel), or STED-enhanced 2PLSM (STED, right panel). The enhanced imaging resolves spine necks and heads in living neurons. Image credit: Bernardo Sabatini
For looking at fine structure of neurons, Sabatini told ARF, researchers previously had two choices. They could use EM, which gives extremely high resolution but requires fixed tissue, or they could image cells in culture. “This technique bridges that gap, opening the possibility of nanometer resolution in intact tissues,” he said. “This is crucial because neurodegenerative diseases are expressed at the organ level, and we need to be able to study brain tissue in context to understand disease.”
This work is just the first step, Sabatini said. Theoretically, it is possible to improve resolution another 10-fold, down to the 50-nm level, and the group is working out the practical challenges of that now. At that resolution, researchers should be able to make out the fine structure of cells and synapses, such as individual postsynaptic densities and intracellular organelles, all of which are affected in the course of neurodegenerative disease.
The second technique relies on bright-field scanning transmission EM (BF-STEM), which is capable of rendering three-dimensional images of organelles and macromolecules in cells. However, the technique is limited to samples thinner than 300 nm—thicker samples cause too much electron scattering. First authors Sousa and Martin Hohmann-Marriott tweaked the detection method to minimize blurring and boost the analyzable sample thickness to one micron. With the ability to achieve comparable resolution to thin sections, the researchers reconstructed a whole erythrocyte, including its major intracellular features, from just four 1-micron-thick slices (see image below).
Surface rendering of membrane structures inside a human erythrocyte infected with a malaria parasite (Plasmodium falciparum). The three-dimensional reconstruction was obtained using axial tomography in the scanning transmission electron microscope. Image credit: Afrouz Azari
“We expect that BF-STEM tomography will also prove useful for reconstructing micrometer-thick regions of brain tissue in three dimensions at a spatial resolution of a few nanometers,” Leapman wrote in an e-mail to ARF. “This will enable visualization of intact key structures such as dendrites, spines, synapses, and postsynaptic densities without the need to perform serial thin sectioning. A more complete picture could then emerge about structural differences between the normal and diseased states. It might also be possible to obtain more specific information about the 3D distributions of specific proteins in neuronal structures by using BF-STEM tomography to image heavy-atom clusters attached as labels to antibodies.”—Pat McCaffrey.
Ding JB, Takasaki KT, Sabatini BL. Supraresolution imaging in brain slices using stimulated-emission depletion two-photon laser scanning microscopy. Neuron. 2009 Aug 27;63(4):429-37. Abstract
Hohmann-Marriott MF, Sousa AA, Azari AA, Glushakova S, Zhang G, Zimmerberg J, Leapman RD. Nanoscale 3D cellular imaging by axial scanning transmission electron tomography. Nat Methods. 2009 Aug 30. Abstract