Electron microscopy (EM) is one of the most fundamental high resolution-imaging methods that uses a beam of electrons as a probe to investigate surfaces. The incredibly short wavelengths of electrons enable researchers to investigate samples at nanometer scales. Many disciplines of science: biology, chemistry, materials science, and physics are all predominant areas of activity for these devices.
Standard EMs utilize a single electron beam to image samples and consequently have inherent limitations when it comes to throughput and overall efficiency. Operators must maintain vigilance over the instruments while also conducting their own specialized analysis, which ultimately derails the fluidity of the project workflow.
The FAST-EM is a unique scanning transmission electron microscope that prioritizes automation, speed, and throughput to overcome the limitations of typical electron microscopes. With its 64 multi-beam-array, this instrument yields images 100 times quicker than standard EMs. Reliable and automated processes are at the core of FAST-EM; as such, researchers are equipped with the tools necessary to deal with a high number of sample sections and huge quantities of data. The reliability and automation of the FAST-EM ensures consistent data quality without user intervention.
One of the most vital and delicate steps in the workflow of electron microscopy involves sample preparation. Since FAST-EM acquires data at unprecedented speeds, it begs the question of whether sample preparation can keep up with imaging speed, or if it would create a bottleneck in the workflow.
This paper aims to demonstrate that FAST-EM is compatible with multiple highly-automated EM sample preparation techniques that support increasing throughput and prevent bottlenecks in the workflow.
Sample preparation in electron microscopy
Sample preparation for EM imaging is critical to obtaining high-quality images but can be labor-intensive and time-consuming for researchers and technicians. The delicate nature of sample preparation for a transmission electron microscope can be highlighted with four important aspects: vacuum compatibility, electron contrast, thinness, and conductivity. The reason these aspects arise in electron microscopy is simply because of the instrument itself – and the imaging method associated with it. Treatment is direly needed to ensure samples can withstand the environment within the microscope.
EMs require a vacuum chamber, so without dehydration (acetone / ethanol) and a subsequent embedment in a resin (acrylic / epoxy) the sample’s ultrastructure will be compromised. A high electron contrast is desirable too for the images yielded from EM, which means that – in the case of biological samples – heavy metal staining agents must be applied to emphasize visibility of certain parts. Furthermore, ultramicrotomy is used to slice samples into ultra-thin sections because thinness plays a significant role in preparing an EM sample. The final aspect of thorough sample prep is conductivity, and it is highly relevant due to the negative charge of the electron streams. The sample ought to remain neutral, so to prevent any excess charging there must be a conductive layer with the sole function of conducting charge around. In standard procedure, carbon is coated onto the sample in question due to its conductive properties.
Sample Pickup and Placement
To produce high quality images, the process of collecting sample sections must be rigorously controlled. However, manual collection of samples is frequently accompanied by mistakes. It’s understandably easy to overlook subtle details in the sectioning and collecting samples, leading to time spent rectifying mistakes instead of advancing research at hand. The unwanted costs may not come just economically, but temporally too. The problem is only exemplified during large-volume imaging projects usually undertaken in biological contexts.
Automated Sample Preparation
Contamination, degradation, displacement, and structural integrity are chief concerns for sample preparation, sectioning, pickup, and placement. The option of repeatedly shouldering laborious work is simply not viable for researchers. It is therefore natural to aim to automate the sample prep workflow so that all the issues and errors are circumvented as much as possible.
Several solutions have been developed, such as automated processing stations for fixation and embedding of samples that standardize the process and drastically reduce human intervention (1). Automation of sectioning is another recent development that enables collection of hundreds to thousands of sections (2-4). As far as high throughput is concerned, an automated pickup and placement system is ideal. By reducing the need for a diligent and well-trained operator, research teams could collect thousands of samples while minimizing the loss of sample records. But how exactly does FAST-EM integrate automated processes? Here we present a guide in illustrating FAST-EM’s usefulness when combined with automated sample prep techniques.
FAST-EM Sample Support and Handling
Samples for the FAST-EM are mounted on custom designed scintillator substrates. A single sample holder has a capacity of nine scintillators –arranged in a 3 x 3 array – where each rigid scintillator is 14 x 14 mm. Sometimes wrinkles, warpage, and inferior quality images are results of substrates lacking in rigidity. The FAST-EM prevents any such issues by implementing stalwart reliable substrates with unobstructed views.
The concern about sample conductivity is resolved with FAST-EM’s scintillators due to them being conductive by integration of a metal layer. Biological samples that otherwise require a carbon layer for conductivity can be placed directly on the substrate and subjected to analysis.
Loading sections onto the sample holder is usually a matter of choice for the research team. The FAST-EM remedies a viable solution with two automated collection methods pertinent to individual sections and ribbons, which will be elaborated on below.
Automated Sectioning and Collection of Ribbons
Ribbons of biological specimens are series of consecutive slices of the sample connected by a thin layer of paraffin. Depending upon the specific research demands of the project, multiple ribbons may be required. The ribbons are usually generated using an ultramicrotome and are collected by floating on water in a boat of the ultramicrotome. The scintillator of the FAST-EM is submerged in the boat before sectioning and sections are deposited upon them by draining water from the boat. This method can be used in any ultramicrotome with a boat large enough to hold a scintillator.
To bypass the need for diligent care in the collection process, autonomous methods have been developed that are commercially available (4). These systems are capable of autonomously sectioning and collecting multiple ribbons with little supervision. The perks of such an automated system lie in the fact that the processes are fully programmable and accommodate different substrates – including FAST-EM’s scintillators. Furthermore, the block-face dimensions can be tailored to suit a specific need, ranging from micrometers to millimeters.
Automated Sectioning and Collection of Individual Samples
Due to the typology and dimensions of the biological structures and organisms that are imaged using array tomography, sectioning then collecting individual sections for imaging is common.
A study was published recently on an automated, high-throughput technique that simplifies the process of collecting hundreds of individual sections (2). This method enables bulk staining with liquids and uninterrupted imaging in systems such as FAST-EM, as well as volumetric correlative light and electron microscopy.
In this innovative sample preparation technique, each cut section includes superparamagnetic particles, enabling remote magnetic actuation which agglomerates the sections as they float in the bath. The sections are gathered in the center of the bath where they are attached to a diamond knife before being deposited onto the silicon wafer (or scintillator) underneath. Once the sections are collected, their order can be computationally retrieved by correlating the locations of fluorescent polymer beads in the resin.
Example: Large Volume Sample Preparation and Acquisition
What would a typical workflow look like for volumetric imaging with FAST-EM?
Sectioning a 0.25 mm3 biological sample would require processing 6250 sections, assuming each section is 40 nm thick. Utilizing an automated sample preparation system (4) and FAST-EM, it would take roughly 4 days for scientists to process.
Thanks to their high level of automation, modern ultramicrotomes such as the one described in (4) can consistently fit 4 ribbons of 50 sections onto a scintillator without any user intervention. Initial setup of the system takes about 45 minutes, after which it can section 200 sections in roughly 15 minutes. Exchanging scintillators once they are full only takes about 10 minutes.
From there, the process continues without interruption, enabling up to 9 scintillators with up to 1800 sections to be filled in roughly 5 hours. Using the other sample preparation technique discussed in (2), results would be similar.
After all samples are prepared on the scintillators, they can be placed in the sample chamber. The FAST-EM promptly begins imaging once the stage is prepped and ready to go. In this example, images are yielded after about 19 hours due to an efficiency rate of 100 Mpix per second at 4 nm pixel resolution.
More scintillators can be prepared as the system is acquiring data, parallelizing the processes of preparing samples and imaging. Following this parallel workflow, image acquisition can be reduced to less than 4 days of work.
Post-processing is accelerated by FAST-EM as well—each section’s regions of interest are presented to the user already stitched together, at which point the final steps of 3D stacking and reconstruction can be completed.
In STEM imaging, preparing samples is a sensitive process that requires meticulous care and significant time investments. Often, researchers must find creative ways to streamline their projects’ workflows and decide which sample prep tools are the most suitable for their needs. With such a vast selection of preparation methods, it can be an arduous process. However, when FAST-EM was designed, careful attention was paid to ensuring its compatibility with these high-throughput sample preparation techniques. Automated sectioning and collecting perfectly complement the FAST-EM; and with simultaneous sample preparation and imaging, researchers can leverage their products with the full potential of a highly efficient workflow.
(1) Zechmann, B., & Zellnig, G. (2009). Microwave-assisted rapid plant sample preparation for transmission electron microscopy. Journal of Microscopy, 233(2), 258–268. https://doi.org/10.1111/j.1365-2818.2009.03116.x
(2) Templier, T. (2019). MagC, magnetic collection of ultrathin sections for volumetric correlative light and electron microscopy. ELife, 8, 1. https://doi.org/10.7554/elife.45696
(3) Schalek, R., Kasthuri, N., Hayworth, K., Berger, D., Tapia, J., Morgan, J., Turaga, S., Fagerholm, E., Seung, H., & Lichtman, J. (2011). Development of High-Throughput, High-Resolution 3D Reconstruction of Large-Volume Biological Tissue Using Automated Tape Collection Ultramicrotomy and Scanning Electron Microscopy. Microscopy and Microanalysis, 17(S2), 966–967. https://doi.org/10.1017/s1431927611005708
(4) Leica https://www.leica-microsystems.com