Questions: Cryo-ET

Question 1

What is the key difference between single-particle cryo-EM and cryo-electron tomography?

• A. Single-particle cryo-EM uses electrons; cryo-ET uses X-rays
• B. Single-particle cryo-EM images many identical copies of purified molecules and averages them for a high-resolution structure; cryo-ET images individual, unique biological objects (cells, organelles) by tilting to reconstruct their 3D volume, sacrificing resolution for the ability to see structures in their native context
• C. Cryo-ET requires crystals; single-particle cryo-EM does not
• D. There is no difference; they are the same technique

Answer: (B)

Single-particle cryo-EM achieves high resolution by averaging many images of identical (purified) molecules — the information deficit from low-dose imaging of each particle is overcome by having millions of copies. Cryo-ET images unique objects (a specific region of a specific cell) and achieves 3D information by tilting rather than by having multiple copies. This means cryo-ET typically achieves lower resolution (~20-40 Angstroms for a single tomogram), but it reveals the native cellular context — where complexes are located, what they interact with, and how they are organized in the cell. Subtomogram averaging of repeated structures can improve resolution to ~5-10 Angstroms.

Question 2

Cryo-ET can image the interior of a whole mammalian cell at high resolution without any sample preparation beyond vitrification.

• T. True
• F. False

Answer: False

Mammalian cells are too thick (5-10 micrometers) for electrons to penetrate — the electron beam can only pass through specimens ~200-500 nm thick. Intact mammalian cells must be thinned by focused ion beam (FIB) milling, which uses a gallium or xenon ion beam to ablate frozen cellular material, leaving a thin lamella (~100-200 nm) that is electron-transparent. FIB milling is performed on the frozen, vitrified specimen and preserves the native cellular ultrastructure within the lamella. Small cells (bacteria, thin cellular extensions) can be imaged directly. FIB-milling has been a transformative advance for cryo-ET, making the interior of any cell type accessible to tomographic imaging.

Question 3

How does subtomogram averaging improve the resolution achievable by cryo-ET?

Model answer: When the same macromolecular complex (e.g., a ribosome, a nuclear pore complex, or a coat protein on a vesicle) appears multiple times within a tomogram or across multiple tomograms, each instance can be extracted as a 'subtomogram' — a small 3D volume. These subtomograms are then aligned and averaged, just as single-particle cryo-EM averages 2D particle images. The averaging improves signal-to-noise and resolution because the noise in each subtomogram is random while the signal (the common structure) reinforces. With enough copies (~1,000-100,000), subtomogram averaging can achieve sub-nanometer resolution — approaching that of single-particle cryo-EM, but with the advantage that the structures were imaged in their native cellular environment.

Subtomogram averaging bridges the gap between cellular imaging (low resolution, native context) and structural biology (high resolution, purified sample). It has been used to determine in-situ structures of ribosomes on the endoplasmic reticulum, nuclear pore complexes in the nuclear envelope, and coat proteins on transport vesicles — revealing how these machines function in their actual cellular setting.