Imagine a form of microscopy of whole cells where not just the usual three or four organelle types are labeled with different fluorescent colors, but where dozens of internal components can be simultaneously defined. Furthermore, imagine resolving them not with diffraction-limited resolution typical of traditional light microscopy, but with the top end of super-resolution on a scale of 10 nanometers or less—a factor of 20 finer than the optical diffraction limit. Finally imagine what you can see is not just a single two-dimensional image but a full three-dimensional dataset that can be sliced and viewed in any angle or perspective with equal fidelity. More importantly, these data are of entire whole cells and not just a few sections.
Of course, such powerful capability comes with one notable compromise—the cells are not observed live but are instead in a fixed state, either by chemical fixation or by cryo-vitrification. Nevertheless, this detailed frozen-in-time view of whole cells is what a relatively newly adopted technique, focused ion beam scanning electron microscopy (FIB-SEM), can bring to the biological community.
The focused ion beam (FIB) technology was first reported at the Symposium on Electron, Ion, and Photon Beam Technology in 1975. It has since been integrated with scanning electron microscopy (SEM) to become a versatile technique for the semiconductor industry and materials science. FIB-SEM consists of two beams: a focused ion beam for ultra-fine ablation and a scanning electron beam for imaging a focused ion beam for ultra-fine.
The electron beam raster scans the top surface of a specimen, while the backscattered electrons along with secondary electrons are recorded to construct a detailed image of that surface at the resolution of a few nanometers. After imaging, the FIB, typically comprising 30 keV gallium ions, strafes across the top of the specimen and ablates a few nanometers of material from the imaged surface, exposing a new and slightly deeper surface for subsequent imaging. Cycles of ablating and imaging gradually erode away the entire specimen while acquiring a stack of consecutive 2D images that form the 3D volume representation.
FIB-SEM technology applications in biological imaging started surprisingly recently, just a little over a decade ago. It could adopt with minor optimizations the existing heavy metal staining and resin embedding protocols that have been developed by biologists for transmission electron microscopy. Imaging native frozen specimens with small volume was later reported in 2013, which opens the possibility of directly visualizing biological ultrastructure in its most native state.
Imaging with 3D FIB-SEM offers unique benefits, such as high isotropic resolution (< 10 nm in x, y, and z), robust image alignment, and minimal artifacts. Together, these benefits allow better interpretation of the acquired images along all three dimensions and are important for the subsequent data analysis. Encouraging results of small imaged volumes have been reported for bacteria, yeast cells, parasites, cultured mammalian cells, worms, zebrafish, brain, liver, pancreas, kidney, muscle, bone, etc., providing a new way to comprehend living systems with nanometer resolution in three dimensions.
While FIB-SEM transcends the z-axis resolution limits imposed by microtome section thickness in serial section TEM, traditional TEM tomography still offers a complementary finer-resolved local view in
individual sections that are several hundred nanometers thick. Further advances over conventional FIB-SEM technology have been made in recent years to substantially improve both speed and reliability. As a result, the imageable volume of these customized systems has been expanded by more than four orders of magnitude from 103 µm3 to 3 x 107 µm3 while maintaining isotropic 8 x 8 x 8 nm3 voxels. Moreover, by trading off against imaging speed, the system can achieve even higher resolution appropriate for 4 x 4 x 4 nm3 voxels.
The expanded volumes enabled by such FIB-SEM capability open up a new regime in scientific discovery, where nano-scale resolution coupled with meso- and even macro-scale volumes is critical. Much of the initial impetus for development of FIB-SEM came from the requirements to image and reconstruct wiring diagrams of brain tissues. The fly connectome has been generated at 8 x 8 x 8 nm3 voxels using such a FIB-SEM platform. With the benefit of superior z resolution and fewer artifacts, the neural circuit reconstruction is more easily automated and more complete. Additionally, and as an example, an isotropic 3D electron microscopy reference library of selected whole cells and tissues (https://openorganelle.janelia.org) was established at 4 x 4 x 4 nm3 voxels. Combined with super-resolution light microscopy, correlative light-electron microscopy (CLEM) applications allow unambiguous protein identifications in the otherwise black and white EM micrograph. Researchers can now possibly build an extensive atlas of wild-type healthy cells, in cultured or native tissue environment, with complete details of features (e.g., ER morphology near a particular organelle or microtubule proximity to certain vesicles), and compare them against the perturbations of disease, mutations, development, or environment, shedding light on the ultimate relationship between structure and function.
It is awe-inspiring to see how FIB-SEM technology, complementing other volume EM modalities, enables fruitful discoveries for life science: from 3D volumes containing ultra-structures of a complete single cell to the most detailed connectome to date. Such data are rich in detail, enormous—giga to tera bytes in size—and fascinating to browse yet daunting to analyze, which introduces a new paradigm of how to observe and extract meaningful biology. Fortunately, now they can be computationally mined with improved data storage and processing and evolving analysis tools that are enabled by advances in computer technology. With tremendous confluence of capabilities, so much still needs to be defined and developed. We envision this as the beginning of a new transformative field and a holistic way of looking into the cellular world, presenting many new challenges and research opportunities across biology.
Related Resources from Cell Bio Virtual 2020
Recordings of two presentations at Cell Bio Virtual 2020–An Online ASCB|EMBO Meeting will be available online to ASCB members in mid to late January 2021.
Jennifer Lippincott-Schwartz discussed data of this type in her E.B. Wilson Lecture.
The challenges of sharing large datasets with the community were the subject of the panel discussion “Making the Most of New Cell Visualization Tools: Why Not Let Anyone Take a Look?” chaired by Eva Nogales.
About the Author:
C. Shan Xu is Senior Scientist and Associate Director of FIB-SEM Technology.
Harald Hess is Senior Group Leader at Janelia Research Campus, Howard Hughes Medical Institute.
