Cryo-Electron Tomography (Cryo-ET): Is Your Sample Ready for In Situ Structural Biology?

The most promising applications of cryo-electron tomography are not only, “What does this protein look like?” but also, “What does this structure look like where it actually functions?”

That distinction matters. The purified complex can also offer a perfect level of structural detail whilst eliminating membranes, adjacent complexes, cell structure or viral structure important for the biological question. Cryo-ET tackles this challenge in a different way, by taking many projections of a vitrined sample to reconstruct a 3D volume.

At present, Longlight offers Cryo-ET as an in situ structural approach to biological assemblies, membrane associated complexes and viral systems as part of its Cryo-electron microscopy service.

The practical question, however, for the data collection is,

Is the biological question really one that does require a tomogram?

Take the Biological Question before the Microscope

Sometimes, scientists liken Cryo-ET to single-particle analysis, suggesting that one had to be superior to the other.

It's typically not an apples-to-apples comparison.

Single-particle cryo-EM might be a more direct path to a high-resolution averaged structure if a purified, relatively homogeneous macromolecular complex can answer the biological question. However, Cryo-ET becomes very attractive if the position/orientation, molecular neighbourhood, membrane environment or cellular organisation of the target is relevant.

Longlight's guide to near-atomic-resolution Cryo-EM offers some background to the single-particle path; Cryo-ET poses another question: what can be seen when biological context is maintained?

First Physical Limit—Electron Accessibility

If a sample is biologically interesting, it does not necessarily mean that it is an electron-accessible sample.

Direct cryogenic tomography may be possible for viruses or other isolated organelles, small cells and thin peripheral areas of cells. Some in situ workflows include specimen thinning, which is necessary for useful transmission imaging of larger eosinucleic cells that are too thick to image.

One well established method of making thin lamellae of cells is by cryo focused ion beam (FIB) milling. Cryo-FIB-milled eukaryotic cells have been used for in situ studies to make internal structures accessible for tomography and subsequent structural analysis.

Sample planning should not start with the voltage of the microscope! Should include 'from where' and 'how', the electron beam will reach the target.

A Tilt Series is NOT a pile of photos, it's a Dose Budget.

Cryo-ET does not get the information about the third dimension from a single exposure.

This specimen is tilted many times and a series of 2D projections are captured. This is due to the radiation sensitivity of frozen biological specimens and the projections must be given a limited total electron-dose budget. The effective thickness that the electrons pass through is also increased at larger tilt angles and this means that less signal may be usable.

In conventional tomography a complete ±90° acquisition is also not possible due to mechanical geometry and the thickness of the specimen. The unsampled information creates the well known missing wedge in Fourier space.

But more projections do not necessarily equal better. The interplay between tilt range, dose distribution, quality of alignment, stability of the specimen and the visibility of the target must be considered.

Determine if you need a Tomogram, a repeated structure.

Not all Cryo-ET projects need to culminate in an almost atomic molecular model!

Often the tomography alone provides an answer to the question: Where is a structure located? What happens to a membrane when it's distorted? What is the organization of the viral components? Which complexes are adjacent to each other?

A different goal will be displayed when there are multiple copies of the same macromolecular complex. They repeated targets can be extracted as subtomograms, aligned, classified and averaged to enhance the structural signal.

In this article, Longlight already provides a detailed explanation of this route. The key issue for a new project is thus made up front: Is this a unique 3D architecture or are there sufficient similar copies for averaging?

The Hidden Bottleneck Can Become Target Localization.

Choosing the best cellular area may be as important as obtaining a technically good tilt series.

Suppose that the target is only a small part of a large cell. Even if you have a pretty tomogram from the wrong area, it is still the wrong data set.

In order to be collected, they should, therefore, take into account the abundance of the target, the approximate localization of the cells, its recognizable morphology, the information available on the labeling, and whether another localization strategy is necessary.

This biological map can significantly enhance the logic of the acquisition plan for membrane associated targets, viral assembly sites, organelles or spatially restricted complexes.

Pre-session planning for Data Path:

When a tilt-series has been acquired, a Cryo-ET project is not complete.

The images must still then be aligned, reconstructed into tomograms and examined, searched for targets and segmented, and if suitable, transferred to particle extraction, classification and subtomogram averaging.

That's why the desired deliverable should be established in advance of collection.

In Longlight's current Cryo-EM service, we focus on the delivery of final density maps, but also of raw data and intermediate processing files. This is useful when a researcher wants to check processing decisions, conduct subsequent analysis within the system or return to the data set at a later date.

Match Platform to Stage of the Project

The quality of the instrument should not be used to remedy an indeterminate sample question.

Initial screening can indicate if specimen preparation, particle distribution, ice quality, and/or target visibility should be improved prior to the investment of expensive collection time.

Longlight today announced a staged cryo-EM platform consisting of Talos L120C G2 for screening, Glacios 2 at 200 kV for routine cryo-EM/tomography and Titan Krios G4 at 300 kV for demanding high-end acquisition. It is also linked to upstream protein workflows through its broader تطبيق البروتين portfolio that pairs structural analysis with its portfolio.

The research rule of thumb is:

Sample is ready, go to the next instrument, not because it is available.

This is the Final Cryo-ET Project Check.

With regard to committing a sample to Cryo-electron tomography, there are five questions to consider:

Is there a need for native spatial context? Is the area of interest in the molecule accessible to the electron? Is there a reliable sight of the target? Ew, one tomogram or repeated particles? Do you know what type of data products will be needed postacquisition?

Once those answers are decided, equipment selection and the processing strategy becomes much easier.

The power of cryo-ET is that it preserves information that is lost during purification. However, this is only going to be useful, if the specimen preparation, the imaging strategy and the biological question are co-designed as one project.

Researchers can consult with Longlight on an in situ structural project after describing to them the type of sample they will be using, the target location, the type of structure they are looking to achieve, and the downstream analysis they will be using.

الأسئلة المتداولة

Q1. Cryo-electron tomography (Cryo-ET)?

Cryo-ET involves taking several images of a biological sample in the vitrines from different angles and then reconstructing this into a 3D tomogram. One major benefit is that it maintains the structural information in a near-native biological environment.

Q2. What are the proper conditions for using Cryo-ET over single particle cryo-EM?

Consider cryo-ET when the position of the cell, membrane organization, molecular neighbors or native architecture is a key factor to the research question. When the purified particle/averaged molecular structure is the focus, SPA is often more direct.

Q3. Should cryo-FIB milling be used for all samples for Cryo-ET?

No. It is possible to image naturally thinned specimens directly. In thicker regions of the cell, thinning may be required to create an electron transparent part of the cell to acquire the tomography.

Q4. What is missing in Cryo-ET?

Lacks 3D information from the limited angular range that can be acquired during tilt series. May cause anisotropy, reconstruction artifacts, which need to be taken into account in analysis.

Q5. Does the subtomogram averaging need to be performed for every project carried out in Cryo-ET?

No. A single tomogram can suffice in answering questions related to morphology or spatial organization. The value of subtomogram averaging lies in the possibility of alignment and averaging of several similar copies of a target when they are available.