Cryo-Electron Microscopy
Pavel Plevka
Transmission Electron Microscope
(TEM)
Electron source:
Thermal emission from
heated cathode
Focussing:
Electro-magnetic Lenses
Detection:
Phosphor screen or CCD
camera (former times:
negative)
Vacuum!
Pro & Con of cryo-EM
Pro
• Short wavelength => high resolution
• Strong interaction with materials => good contrast
• Electromagnetic lenses => standard optics (in contrast to X-ray
crystallography)
• High intensity is easy to produce
• Inner structure of biomolecules is accessible
Con
• High vacuum requires special treatment of sample
• Sample has to be thin to avoid 100% absorption
• Low contrast of biomolecules
• Electron beam damages sample => short measurements
Why electrons?
Visible Light:
λ = 400 – 600 nm
Electrons:
λ = 0.002 – 0.004 nm
Remember: wave-particle dualism
de Broglie: λ = h / m*v
E = ½ * m * v²
Sample Preparation - Staining
⇒ To increase contrast: heavy atoms interact with electrons stronger
than biomolecules (C, N, O, S, P)
• Positive Staining
treat sample with solution of salt like uranyl acetate, lead citrate,
osmium tetraoxide – object is black on light background
• Negative Staining
place sample on dried film of heavy metal salt – object is light spot
on black background
• Shadowing
spray thin layer of heavy metal on sample to produce a shadow
Disadvantage:
Size of stain reduces resolution to about 20-30 Å
Alternative: Cryo-EM
• to avoid harsh staining which may change
the structure of your sample
• stabilization of sample by rapid freezing of
sample in liquid ethane to form vitreous ice
• into electron microscope at low
temperatures to keep sample stable in
hydrated state in vacuum
• thickness of ice layer as small as possible!
Advantage:
• sample structure unchanged
• inner structure of molecule is accessible
Types of Samples
• Periodic arrangement
=> 2D electron crystallography
small or membrane proteins < 200 kDa
resolution up to 2.5 Å
• Random arrangement
=> single particle technique
macromolecular complexes > 200 kDa
up to atomic resolution
• Large Organelles (Golgi, ER), whole cells
=> tomography
resolution > 40 Å
Ribosome
Bacteriorhodopsin
Whittaker-Shannon sampling theorem
For a given sampling frequency, the maximum
frequency you can accurately represent without
aliasing is the Nyquist frequency. The Nyquist
frequency equals one-half the sampling frequency,
as shown by the following equation.
toobjective
Macromolecules in water / vitreous ice
are phase objects
Signal to noise ratio
Classification and averaging
(principal component analysis)
Overview of steps
Challenge:
Processing of
images to filter
noise from signal
Image Processing
Improve signal to noise by
Overlaying many pictures
Problem:
objects are arranged in
different orientations
X
Y
α β γ
Alignment & Classification
Rough alignment of pictures (up to 100,000)
Correspondence analysis groups images using
statistical methods
Challenging process requiring at least 5 parameters
Images with same orientation will be averaged
(class averages)
Raw Images
Class Averages
3D Reconstruction
When the angles between the different classes are known
(estimated), a 3D model can be calculated.
Iterative Process:
3D model is used to
generate 2D images
which are fed into
statistical analysis of
images (alignment and
classification).
And then?
Try to interpret 3D map,
e.g. try to fit known
crystal structures into
electron density map
Enterovirus replication cycle
Virus
replication
factory
Acidic pH induced genome release of
echovirus 18
Reference-free two-dimensional class
averages of genome-releasing particles
Asymmetric structures of open particles
Model of enterovirus genome release
Human cardiovirus Saffold virus 3
(2.5Å resolution)
Human Parechovirus 1 @ 3.1Å