Serial block face scanning electron microscopy (SBFSEM) is a powerful technique originally introduced by Leighton [1], substantially improved by Denk [2] and subsequently commercialized (Gatan Inc., Pleasanton, CA.). SBFSEM allows for the automated image acquisition of relatively large volumes of tissue at near nanometer-scale resolution, using a dry cutting ultramicrotome fitted into an SEM. In an automated process, a low voltage backscatter electron (BSE) image is obtained from the surface of an epoxy embedded tissue block face. The ultramicrotome then removes an ultra-thin section of tissue with a specially designed oscillating diamond knife (Diatome AG, Switzerland), and a block face image from the corresponding region is again obtained. This sequence is repeated over and over until the desired volume of tissue has been imaged. Although SBFSEM overcomes many obstacles routinely encountered with serial section TEM reconstruction, until recently there was a significant limitation to the resolution obtainable by this method compared to conventional TEM. This was due primarily to difficulties encountered using BSE imaging at low accelerating voltages. To overcome this we have developed a protocol for vastly increasing the heavy metal staining of specimens to improve BSE yield. This is accomplished by combining a variety of preexisting heavy metal staining methodologies not normally used together, including ferrocyanide-reduced osmium tetroxide, thiocarbohydrazide-osmium tetroxide (OTO), prolonged uranyl acetate treatment and en bloc lead aspartate staining. Using this approach, we demonstrate a dramatic improvement in image contrast and resolution from existing methods in a variety of specimens (Fig. 1).We have also combined this approach with a number of selective labeling methods such as Golgi impregnation to allow the reconstruction of whole cells in the nervous system (Fig. 2), as well as fluorescence photoconversion to label specifically targeted proteins [3,4]. Additionally, a powerful application of SBFSEM is its use in conjunction with a newly developed genetically encoded fluorescent reporter termed miniSOG (for mini singlet oxygen generator). MiniSOG is a small (106-residue) singlet oxygen generating protein engineered from a flavin-binding, blue light phototropin from Arabidopsis thaliana. MiniSOG has quantum yields for fluorescence and singlet oxygen of 0.30 and 0.47 respectively. It can be genetically fused to the target protein of interest for both fluorescence imaging and efficient photooxidation of diaminobenzidine into an osmiophilic polymer for subsequent electron microscopy [3]. Since miniSOG is genetically encoded and all other reactants (O 2 , diaminobenzidine, OsO 4 ) are permeant small molecules, there is no need to compromise chemical fixation to preserve protein epitopes or to permeabilize with detergents, which further degrade cellular ultrastructure. We have combined this approach with the intense heavy metal staining procedure outlined above for SBFSEM to enable 3D localization of gen...

A grand goal in neuroscience research is to understand how the interplay of structural, chemical and electrical signals in and between cells of nervous tissue gives rise to behavior. We are rapidly approaching this horizon as neuroscientists make use of an increasingly powerful arsenal of tools and technologies for obtaining data, from the level of molecules to nervous systems, and engage in the arduous and challenging process of adapting and assembling neuroscience data at all scales of resolution and across disciplines into computerized databases. This talk will highlight projects where development and application of new contrasting methods and imaging tools have allowed us to see otherwise hidden relationships between cellular, subcellular and molecular constituents of nervous systems. New chemistries for carrying out correlated light and electron microscopy will be described, as well as recent advances in large-scale high-resolution 3D reconstruction with TEM and SEM based methods. The Whole Brain Catalog (WBC), a Google Earth-like open-source virtual model of the mouse brain, will also be described. The WBC is as an example of an informatics framework and web-based tool whose purpose is partly to facilitate integration of 3D image data from multiple microscopy methods and to enable the linking of information derived from other analytical approaches to imaging data shared in the publically accessible catalog.

Abstract:Introduction: A host of new technological tools for data acquisition and processing are beginning to deliver new syntheses of information about biological systems, including new revelations regarding the complexities of the brain. Brain researchers are now beginning to be able to explore across the full range of scales, from genomics and molecular structure to networks of neuronal systems. Gaps in knowledge and limited abilities to span scales in tissues highlight the need for tools and methods that will allow the acquisition of high fidelity 3D image information at high resolution, but over very large expanses. Thus, despite rapid progress in development of new experimental methods, our ability to simultaneously study the brain across many key scales remains quite limited. To state the problem simply and specifically: experimental methodologies available today reveal only limited views of nervous system organization at the tissue and subcellular-supramolecular levels. Consequently many research groups are now working to expand data acquisition and computational methods to address the grand challenge of understanding how the structure of the brain gives rise to complex functions. Our group is deeply involved in this activity and has developed many tools that accelerate the delivery of large and detailed views of elements of the nervous system, extending to the nanometer scale. Several of our ongoing projects related to these multiscale challenges will be highlighted in this presentation. These include work on advanced extreme-scale and high quality image recording systems for electron tomography; development of specimens and procedures to increase the image quality, resolution and field of view for 3D volumes acquired by serial block face scanning electron microscopy. Advanced Detection Systems for TEM:Film has long been regarded as the gold standard for image recording. When scanning to achieve a (modest) dynamic range of 3 orders of magnitude, a true format of approximately 8k x 6.5k can be achieved. This still far exceeds the performance of current digital detection systems for TEM, though is well short of theoretical resolution based on grains size alone owing to the need to expose many grains to achieve wide dynamic range. While film provides excellent modulation transfer function (MTF), especially as compared with commercially available CCD cameras, it requires several post-acquisition steps like development and digitizing that are cumbersome and time-consuming. We have developed and tested a lenscoupled system that matches the resolution and format of film with a far superior dynamic range. Our goal was to build a large format high-resolution wide-field imaging detector with single electron sensitivity. Key to this goal is recording unique information in every pixel of the image by ensuring modulation transfer of at least 10% to the Nyquist limit.We have built an 8k x 8k lens-coupled camera system, where the electron image produced by the microscope is first converted to a photon image with the u...

Extended abstract of a paper presented at Microscopy and Microanalysis 2012 in Phoenix, Arizona, USA, July 29 – August 2, 2012.