电子断层成像术
7. Tomography and Subtomogram Averaging
Tanmay Bharat
EM course 2014
What is tomography?
Steven and Aebi, Trends Cell Biol, 2003
What is tomography? tilt series is used to obtain a tomogram
Tomographic reconstruction: weighted back projection
The back projections must be appropriately weighted and filtered for tomographic reconstructions. Weighting cannot correct for the distortions caused by the missing wedge, but a priori knowledge such as size of the reconstructed object etc can minimize these effects. Smith, Digital Signal Processing 2011
Applications: Cells and cellular components
Sporulating bacteria
Nuclear pores
Plasmodium infected erythrocytes
Beck et al, Science 2004 Cyrklaff et al, Science 2011 Tocheva et al, Cell 2011
Lucic and Baumeister, JCB, 2013
What is tomography? Overall workflow
OK, this is great! To what can we apply tomography?
Lucic and Baumeister, JCB, 2013
Of course, in tomography we have very good prior information of orientations!!
*repeated from lecture 6 by Sjors Scheres
What is tomography? Overall workflow
Applications: tissues and embryoes
Human skin tissue (CEMOVIS) Zebrafish embryo (plastic embedding)
Al Amoudi et al, Nature 2007 Nixon et al, Traffic 2008
Tomographic data collection: tilt schemes
Biological samples can sustain limited dose, and this must be distributed over the tilt series. Linear tilt scheme ( tilt angle α is a constant) Saxton tilt scheme (αn+1 = αn + sin-1(sin(α0)*cos(α0)))
Applications: tissues and embryoes
So how can we make a sample for tomography?
Human skin tissue (CEMOVIS) Zebrafish embryo (plastic embedding)
Al Amoudi et al, Nature 2007 Nixon et al, Traffic 2008
Tomographic data collection: Tracking images
Track
Record
Tracking images are needed to keep the same area in the field of view, and focus is needed because the tilt axis of the grid does not necessarily pass through the area being considered. Track, tilt, track, focus, record is iterated.
2. Freeze substitution, low-temperature embedding
1. High pressure freezing
3. Sample sectioning at room temperature
Tokuyasu method for immuno-labelling
Tomographic reconstruction
*repeated from lecture 6 by Sjors Scheres
Tomographic reconstruction: weighted back projection
Let us suppose we take projections of our image along different directions Smith, Digital Signal Processing 2011
CEMOVIS: Cryo Electron Microscopy Of Vitrified Sec8ons
Vitrified samples can be sec1oned at at cryo temperatures – no stain, no embedding.
OK, now we have a sample. How should we get the data?
Tomographic data collection: Tracking images
Record
Track
Tilt axis
*low dose acquisition explained in lecture 4 by Lori Passmore
Tomographic data collection: the missing wedge
Tilt series ±30° Ice thickness increases as we tilt. Cannot tilt to 90°
There is missing information along the beam direction leading to artifacts in the resulting reconstruction Palmer and Löwe, Ultramicroscopy, 2013
*repeated from lecture 3 by Chris Russo
Tomographic data collection: energy filtration
OK, now we have our data. How should we make tomograms?
*repeated from lecture 3 by Chris Russo
Sample preparation: in vitro systems and viruses
Dubochet and colleagues 1984 EMBL
*repeated from lecture 1 by Tony Crowther *see also lecture 4 by Lori Passmore
Tilt series
Tomogram
The principle of 3D reconstruction analogous to single particle analysis
What is tomography? tilt series is used to obtain a tomogram
Sectioning of cryo-protected frozen (sucrose) samples, followed by immunolabelling, contrasting and imaging at room temperature.
Sample preparation: thick samples (CEMOVIS)
Cylindrical specimen holder to circumvent the missing wedge
Saxton et al, Ultramicroscopy, 1984 Palmer and Löwe, Ultramicroscopy, 2013
Tomographic data collection: energy filtration
Sample preparation: thick samples (CEMOVIS)
CEMOVIS: Cryo Electron Microscopy Of Vitrified Sec8ons
Vitrified samples can be sec1oned at at cryo temperatures – no stain, no embedding.
Tomographic reconstruction: weighted be projections, and smear them along the direction where they came from (tilt angle) and simply sum the smeared images. This is a simple back projection. Smith, Digital Signal Processing 2011