Federico Forneris, PhD
The Armenise-Harvard Laboratory of Structural Biology
http://fornerislab.unipv.it
Cryo-EM
The light spectrum
D≈ =
λn
(NA)2
D≈(min) =
λ
2
Electron Microscopes were developed due to limitations
of Light Microscopes.
Scientific desire to see the fine details of the interior
structures of organic cells.
10,000x plus magnification is not possible using Light
Microscopes.
History of EM
Decrease
wavelength range
Increase
resolution limit
λ = 0.00251 nm (200 keV) R 0.2–0.3 nm
~
EM is microscopy with X-rays
Light Microscope Electron Microscope
Light Source Visible Light Electrons
Lens Type Glass Electromagnets
Magnificatio
n Method
Lens Movement
Current changes through
lens coil
Sample
viewing
Eyepiece
Camera
Fluorescent screen
X-ray camera
Vacuum No Yes
Electron vs Light Microscopy
Physics of electron microscopy: the electron source
Electron sources:
• W
• LaB6
• Field Emission Gun
Physics of electron microscopy: the need for vacuum
- Enable the electron beam to travel in straight lines (10-5 mbar)
- Tungsten filaments burn out in air
- Columns must be kept dust free
2-fold pumping
- mechanical pump, stage-1
- membrane diffusion pump (or turbo pump), stage-2
electrons are charged, and are therefore
deflected when they cross a magnetic field
Physics of electron microscopy: the electron lenses
electron beam
soft iron pole piece
electrical coil
The Electron Microscope
Cryo-EM in Pavia
ThermoFisher Glacios (FEG 200 kV, Cryo-TEM)
Typical image Pixel size 0.1 Å
JEOL JEM-1200-EX (LaB6 120 kV)
Typical image Pixel size 7 Å
Electron microscopes: TEM vs SEM
Detection of
electrons scattered by
the specimen
Detections of back-
scattered and
generated electrons
TEM Resolution is 10X higher than SEM
SEM examples
EM images are in contrast scale ONLY
Colors are a prerogative of visible light
Structural Biology with CryoEM
O’Reilly et al., Science (2020)
Patel et al., Science (2021) Ke et al., Nature (2020)
Abdella et al., Science (2021)
Wagner et al., Nature (2020)
TEM requires multiple projections for correct reconstitution
Techniques for image reconstitution
• Single-particle reconstruction
• Slow
• Simpler setup
• Higher resolution
• Requires multiple objects
• Tomography
• Fast
• Complex Setup
• Lower Resolution
• All on a single object
TEM requires multiple orientations of the same object
to obtain complete structural reconstruction
Sample preparation for EM analysis
• Vacuum
• Dehydration
• Physical damage
• Accelerated electrons
• Radiation damage
• Contrast!
• High contrast image
• No special temperature control
• Essentially no radiation damage
• Particle distorted
• Image = stain “shell” around the
particle
• Low resolution: 20-15 Å
• Great choice for initial sample
screening
• Low contrast image
• Sample maintained at cryogenic
temperature (85 K)
• High radiation damage
• Particle undistorted
• Image is of the actual particle
• Higher resolution obtained: 15-4 Å
• Best choice for reconstruction
Negative stain vs. Cryo-EM
Sample support for EM analysis
EM grids
Cryo-EM sample prep
Carbon is hydrophobic, protein solutions are hydrophilic
Negative Stain EM
Adattato da Science Magazine
http://www.sciencemag.org/news/2017/10/cold-clear-view-life-wins-chemistry-nobel
http://fornerislab.unipv.it
Cryo-EM sample prep
deposit blot plunge
• 3-4 µL of highly pure sample in solution
• Decide humidity, blot force, blot time
• Manipulation after plunge-freezing to prepare
sample holder cassette
Adattato da MRC LMB Cambridge
http://www2.mrc-lmb.cam.ac.uk/research/scientific-facilities-and-support-services/electron-microscopy/
http://fornerislab.unipv.it
The single-particle EM Experiment
5 degrees of freedom
to determine
Sample preparation for EM analysis
The lack of contrast in EM
• Collect image set (20-100 images, vary focus)
• Correct motion, astigmatism, defocus
• Perform contrast-transfer-function (CTF) correction for
each image
• Pick Particles (4000-100,000)
• Center, align, classify, make “class averages”
The single-particle EM Experiment
Fast readout, better quality
Software-based motion correction can (partically) account for molecular drift
Direct electron detectors are game-changers
Current frame rate (Gatan K3: 1500 frames/sec)
The single-particle EM Experiment
Adattato da MRC LMB Cambridge
http://www2.mrc-lmb.cam.ac.uk/research/scientific-facilities-and-support-services/electron-microscopy/
http://fornerislab.unipv.it
Averaging
Class Averaging
Projection matching and angular refinement
Density map from 2D projections
The single-particle EM Experiment
Collect Data 2-3 days
From days to
months
Pick Particles
Make Model
Refine
Average
Classify
Reconstitute
TEM requires multiple projections for correct reconstitution
(Cryo)Electron Microscopy
http://fornerislab.unipv.it
Adapted from A.J. Jakobi
(Cryo)Electron Microscopy
http://fornerislab.unipv.it
Adapted from A.J. Jakobi
(Cryo)Electron Microscopy
Adapted from A.J. Jakobi
http://fornerislab.unipv.it
EM Artefacts
Model Bias affecting noise
Averaging 1000 images of PURE WHITE NOISE using a model
intentionally to introduce bias
Shatsky M, J. Struct Biol, 2009

20231122(Cryo-EM).pdf

  • 1.
    Federico Forneris, PhD TheArmenise-Harvard Laboratory of Structural Biology http://fornerislab.unipv.it Cryo-EM
  • 2.
    The light spectrum D≈= λn (NA)2 D≈(min) = λ 2
  • 3.
    Electron Microscopes weredeveloped due to limitations of Light Microscopes. Scientific desire to see the fine details of the interior structures of organic cells. 10,000x plus magnification is not possible using Light Microscopes. History of EM Decrease wavelength range Increase resolution limit λ = 0.00251 nm (200 keV) R 0.2–0.3 nm ~
  • 4.
    EM is microscopywith X-rays
  • 5.
    Light Microscope ElectronMicroscope Light Source Visible Light Electrons Lens Type Glass Electromagnets Magnificatio n Method Lens Movement Current changes through lens coil Sample viewing Eyepiece Camera Fluorescent screen X-ray camera Vacuum No Yes Electron vs Light Microscopy
  • 6.
    Physics of electronmicroscopy: the electron source Electron sources: • W • LaB6 • Field Emission Gun
  • 7.
    Physics of electronmicroscopy: the need for vacuum - Enable the electron beam to travel in straight lines (10-5 mbar) - Tungsten filaments burn out in air - Columns must be kept dust free 2-fold pumping - mechanical pump, stage-1 - membrane diffusion pump (or turbo pump), stage-2
  • 8.
    electrons are charged,and are therefore deflected when they cross a magnetic field Physics of electron microscopy: the electron lenses electron beam soft iron pole piece electrical coil
  • 9.
  • 10.
    Cryo-EM in Pavia ThermoFisherGlacios (FEG 200 kV, Cryo-TEM) Typical image Pixel size 0.1 Å JEOL JEM-1200-EX (LaB6 120 kV) Typical image Pixel size 7 Å
  • 11.
    Electron microscopes: TEMvs SEM Detection of electrons scattered by the specimen Detections of back- scattered and generated electrons TEM Resolution is 10X higher than SEM
  • 12.
  • 13.
    EM images arein contrast scale ONLY Colors are a prerogative of visible light
  • 14.
    Structural Biology withCryoEM O’Reilly et al., Science (2020) Patel et al., Science (2021) Ke et al., Nature (2020) Abdella et al., Science (2021) Wagner et al., Nature (2020)
  • 15.
    TEM requires multipleprojections for correct reconstitution
  • 16.
    Techniques for imagereconstitution • Single-particle reconstruction • Slow • Simpler setup • Higher resolution • Requires multiple objects • Tomography • Fast • Complex Setup • Lower Resolution • All on a single object TEM requires multiple orientations of the same object to obtain complete structural reconstruction
  • 17.
    Sample preparation forEM analysis • Vacuum • Dehydration • Physical damage • Accelerated electrons • Radiation damage • Contrast!
  • 18.
    • High contrastimage • No special temperature control • Essentially no radiation damage • Particle distorted • Image = stain “shell” around the particle • Low resolution: 20-15 Å • Great choice for initial sample screening • Low contrast image • Sample maintained at cryogenic temperature (85 K) • High radiation damage • Particle undistorted • Image is of the actual particle • Higher resolution obtained: 15-4 Å • Best choice for reconstruction Negative stain vs. Cryo-EM
  • 19.
    Sample support forEM analysis EM grids
  • 20.
    Cryo-EM sample prep Carbonis hydrophobic, protein solutions are hydrophilic
  • 21.
  • 22.
    Adattato da ScienceMagazine http://www.sciencemag.org/news/2017/10/cold-clear-view-life-wins-chemistry-nobel http://fornerislab.unipv.it
  • 23.
    Cryo-EM sample prep depositblot plunge • 3-4 µL of highly pure sample in solution • Decide humidity, blot force, blot time • Manipulation after plunge-freezing to prepare sample holder cassette
  • 24.
    Adattato da MRCLMB Cambridge http://www2.mrc-lmb.cam.ac.uk/research/scientific-facilities-and-support-services/electron-microscopy/ http://fornerislab.unipv.it
  • 25.
    The single-particle EMExperiment 5 degrees of freedom to determine
  • 26.
  • 27.
    The lack ofcontrast in EM
  • 28.
    • Collect imageset (20-100 images, vary focus) • Correct motion, astigmatism, defocus • Perform contrast-transfer-function (CTF) correction for each image • Pick Particles (4000-100,000) • Center, align, classify, make “class averages” The single-particle EM Experiment
  • 29.
    Fast readout, betterquality Software-based motion correction can (partically) account for molecular drift Direct electron detectors are game-changers Current frame rate (Gatan K3: 1500 frames/sec)
  • 30.
  • 31.
    Adattato da MRCLMB Cambridge http://www2.mrc-lmb.cam.ac.uk/research/scientific-facilities-and-support-services/electron-microscopy/ http://fornerislab.unipv.it
  • 32.
  • 33.
  • 34.
    Projection matching andangular refinement
  • 35.
    Density map from2D projections
  • 36.
    The single-particle EMExperiment Collect Data 2-3 days From days to months Pick Particles Make Model Refine Average Classify Reconstitute
  • 37.
    TEM requires multipleprojections for correct reconstitution
  • 38.
  • 39.
  • 40.
    (Cryo)Electron Microscopy Adapted fromA.J. Jakobi http://fornerislab.unipv.it
  • 41.
  • 42.
    Model Bias affectingnoise Averaging 1000 images of PURE WHITE NOISE using a model intentionally to introduce bias Shatsky M, J. Struct Biol, 2009