CryoEM Status Update and Innovative Developments

Even before the 2017 Nobel Prize was awarded for the cryo-electron microscopy (cryoEM) of biological molecules, the field was generating excitement. The resolution revolution was well underway, and structural biology was looking to cryoEM as a characterization method approaching the gold standard of x-ray crystallography¹. Early on, the field of cryoEM primarily encompassed data generated by cryo transmission electron microscopes (cryoTEMs). Now, as microscopes, sample preparation instruments, and sample preparation methods have advanced, the types of data generated have become more diverse. CryoEM is no longer reserved for protein characterization by cryoTEM. The field has grown and now comprises data generated from cryo-scanning electron microscopes and focused ion beam (cryoSEM/FIB). The impact is more broadly felt across multiple disciplines, including imaging research on molecules, cells, tissues, and synthetic materials such as hydrogels.

Advances in imaging capabilities have been a driving force in cryoEM development. Improvements in cameras, detectors and image processing have paved the way for imaging with less energy input, which allows delicate biological samples to survive the imaging process. In TEMs, the low-dose beam exposure has translated to the application of tomography in which a single sample is imaged repeatedly at multiple angles to generate a 3D reconstruction^2,3. Advances in SEM/FIBs have also resulted in improved resolution while using less energy. Importantly, this has led to the improved imaging of nonconductive samples, which translates to better volume imaging in SEM/FIBs^4,5. These improvements have advanced all biological electron microscopy imaging, but they have especially improved cryo imaging, where samples are delicate and have no additional heavy metals added to improve the contrast.

As imaging capabilities have advanced, so too have sample preparation instruments. In cryoEM sample preparation, the first step is the successful vitrification of the samples. Hydrated samples are frozen to form amorphous ice, thus avoiding the damage caused by the formation of ice crystals. The leading strategies are plunge-freezing in liquid ethane for molecules and small cells or high-pressure freezing (HPF) for larger samples up to a few hundred microns in thickness. Advances in plunge-freezing technology allow the better capture of molecules in all orientations by reducing the self-assembly along the air/water interface⁶. Additionally, advances in HPF technology allow for the light and electrical stimulation of samples moments before freezing⁷. Most recently, a new HPF technology has been developed that can interface with a fluorescence microscope, allowing for a short time delay (<2 s) between the fluorescence imaging of live cells and HPF freezing⁸.

The above advances have energized the field of cryoEM and inspired the development of new methods. There has also been a reinvigoration of established techniques. Here, we present a collection of methods focused on advancing cryoEM with the aim of disseminating these strategies to the rapidly expanding research field.

While the improvements in imaging capabilities and sample preparation instruments have been a major driving force behind the advancements in the field of cryoEM, a major hurdle in this research continues to be the hands-on preparation of samples. Truong et al.⁹ present a method that takes advantage of lipid monolayer formation at the air/water interface. A lipid layer forms on the TEM grid and acts as a scaffold for the membrane proteins of interest. For multilamellar or stacked membrane samples, Johnson et al.¹⁰ present a strategy to “peel” off the excess layers, leaving a single layer on the grid that is appropriate for TEM imaging. In another article, Chang et al.¹¹ recognize that some proteins are temperature dependent and that the relevant structure is observed at elevated temperatures. They describe a modified plunge-freezing method that allows samples to be applied to the grid at elevated temperatures while still achieving vitreous ice after plunge-freezing. To further control the process of cryo grid preparation, Kang et al.¹² present a high-throughput method to produce micropatterned chips for TEM grids in which the pores have a controllable depth. This method may provide another variable to control the thickness of amorphous ice during grid preparation.

The correlation of multiple imaging modalities is impactful. Bieber et al.¹³ describe a cryo workflow in which a frozen grid is imaged in a cryo fluorescence microscope to locate a fluorescent region of interest. The grid is then transferred to a cryoSEM/FIB, where fiducials are used to align the fluorescence image with the SEM image, thus allowing the fluorescent region to be targeted for FIB milling. The target region is thinned to a few hundred nanometers, and then the grid is transferred to a cryoTEM for tomography on the thinned target location. DiCecco et al.¹⁴ demonstrate a correlation between liquid TEM and cryoTEM, thus moving closer to the goal of coupling high-resolution cryoTEM data with the TEM data of molecules in dynamic motion.

The automated acquisition of large data sets is also an area of interest. Valuable instrument time needs to be maximized, and automation is a key step. Kumar et al.¹⁵ describe the click-by-click setup of an automated data acquisition software for cryoTEM imaging.

As the field of cryoEM continues to develop, further advances will be seen in sample preparation, and the impact of cryoEM will be felt broadly across more disciplines. Some advances will involve not new techniques but the strengthening of existing techniques. The techniques will be faster, easier, and more robust, with higher throughput. As automation increases the data output, key areas of development will be in image analysis and data processing. Furthermore, AI will play an important role. We should expect to see a reduction in instrumentation costs and footprint, which will allow this technology to become more broadly accessible, although these developments may be slower to materialize.

The authors have nothing to disclose.

We thank all the authors for their contributions to this collection and our colleagues for the continuing progress in this field. This work was supported in part by the Koch Institute Support (core) Grant P30-CA14051 from the National Cancer Institute. We thank the Koch Institute's Robert A. Swanson (1969) Biotechnology Center for technical support, specifically Peterson (1957) Nanotechnology Materials Core Facility.