Cryo-EM
A purified sample is applied to a holey grid, blotted to a thin film and plunged into liquid ethane. Thousands of movies are collected automatically, and hundreds of thousands of particle images are aligned and averaged into a 3D map.
Our ruleSimulations prepare you for the bench but don't replace hands-on practice.
Single-particle analysis
A purified sample is applied to a holey grid, blotted to a thin film and plunged into liquid ethane. Thousands of movies are collected automatically, and hundreds of thousands of particle images are aligned and averaged into a 3D map.
How it differs from the standard method
- Screen grids for ice thickness and particle distribution before a long collection.
- Target 0.8–2.5 µm defocus and 40–60 e⁻/Ų fractionated into 40–50 frames.
- Watch for preferred orientation; tilted collection or detergent additives can help.
1Imaging molecules in glass
Proteins are made of light atoms that barely deflect electrons, and the electron dose needed to see one molecule clearly would destroy it. Cryo-EM gets around both problems.
Single-particle analysis
Purified complexes in random orientations in thin ice. Each particle is a 2D projection; finding their orientations reconstructs the 3D map.
Tomography
One region of a cell or lamella is imaged at many tilt angles and reconstructed as a 3D volume. Repeated objects in it can be averaged.
Electron diffraction
Nanocrystals too small for X-rays are rotated in the electron beam and their diffraction is processed like crystallographic data.
2Where contrast comes from
A thin specimen changes the phase of the electron wave rather than absorbing it, and phase is invisible to a detector at perfect focus. Imaging slightly out of focus converts phase shifts into intensity, but in a pattern that alternates with spatial frequency: the contrast transfer function (CTF).
3Why resolution depends on ice and particles
The resolution of an average grows only with the logarithm of the number of particles: each ten-fold increase buys the same step in 1/d². How steep that step is depends on an overall B-factor that collects everything that blurs the images, including thick ice, beam-induced motion, poorly aligned particles and a heterogeneous sample.
4How it fails
- Empty holes or broken particles: the ice is thinner than the particle, or the concentration too low. Particles that touch the air–water interface often denature or adopt one orientation.
- Grey, low-contrast images that stall at medium resolution: the ice is too thick.
- Crystalline ice: the grid cooled too slowly, for example in liquid nitrogen instead of liquid ethane.
- Blurred particles: frames summed without motion correction and dose weighting.
- A map that only shows one view, stretched in one direction: preferred orientation at the air–water interface.
- Resolution at the Nyquist limit and a noisy map: the half maps were not refined independently, so the FSC is meaningless.
Key terms
Practise these as flashcards →Amorphous, glass-like frozen water formed by cooling at around 10⁵ K/s or faster.
Contrast transfer function: how the microscope's defocus and aberrations transfer each spatial frequency into image contrast.
Electrons per square ångström delivered to the specimen; 40–60 for a single-particle movie.
Aligning the frames of each movie to undo beam-induced motion, and down-weighting high frequencies in later, more damaged frames.
Sorting particles by appearance to discard junk and separate conformations.
Fourier shell correlation between two independently refined half maps; resolution is quoted where it falls to 0.143.
The region of Fourier space a tomogram never samples because the stage cannot tilt to ±90°; it elongates features along the beam.
Aligning and averaging many copies of a structure extracted from tomograms.
When to use it
- Large complexes, membrane proteins in detergent or nanodiscs, and flexible assemblies that resist crystallisation.
- Samples with several conformations that you want to separate rather than average.
- Structures in their cellular context (tomography), or nanocrystals too small for X-ray diffraction (MicroED).
Compared with alternatives
Usually higher resolution and faster per structure for small, crystallisable proteins and for ligand series once a crystal form exists.
Size and overall shape in solution in an afternoon, useful to check a sample before committing to grids and microscope time.
Room-temperature negative stain or resin sections: quick checks of particle quality or cell ultrastructure, but not near-atomic structure.
Sub-techniques
Choosing one here switches the whole page, including the protocol and the lab, to it.
A purified sample is applied to a holey grid, blotted to a thin film and plunged into liquid ethane. Thousands of movies are collected automatically, and hundreds of thousands of particle images are aligned and averaged into a 3D map.
- Screen grids for ice thickness and particle distribution before a long collection.
- Target 0.8–2.5 µm defocus and 40–60 e⁻/Ų fractionated into 40–50 frames.
- Watch for preferred orientation; tilted collection or detergent additives can help.