1
Q
Benefits of CryoEM
A
- allows molecules to the studied in native environment
- biochemically functional buffers
- visualise conformations/membrane proteins
- low concentrationss
2
Q
Principle of EM
A
- high energy electrons are waves
- smaller wavelength than interatomic distance so high resolution
- diffraction limit not a problem
- electrons focussed by magnetic lenses
- images more powerful than diffraction patterns
- no need to solve problem
3
Q
Limitations of EM
A
- high vacuum to avoid unwanted scattering of electrons
- high energy electrons are also ionizating radiation
- forced to limit exposure of sample to avoid damage
- CryoEM samples move when irradiated
4
Q
Dubochet’s Vitrification Method
A
- freezing in glass like ice
- makes buffer invisible but maintains sample hydration
- plunged into liquid ethane
5
Q
Frank’s Image Analysis
A
- imaging 2D projections of original object
- randomly oriented proteins are hit by electrons beams leaving a trace on the image
- computer discriminates between traces and fuzzy background placing similar ones in the same group
- uses similar traces to generate high resolution image
- computer generates high resolution structure in 3D
6
Q
Henderson’s Vision for High Res EM
A
- created high resolution structure using EM
- 2D crystals of membrane proteins
- low dose imaging to mitigate impact of ionizing radiation
- direct electron detection
7
Q
Resolution Revolution
A
- advances in image processing
- increased complexity and smaller objects visualised
- major advances in technology software: direct detectors gives higher signal:noise, higher precision in images
- can take images of different states/conformations
- higher ratio gives precision in orientation assignments
- revolution in information content extracted from images
8
Q
Wave Particle Duality
A
- electrons can be both (move as wave but hits in one place like a particle)
- electron guided by wave but has defined position
- electrons in vacuum behave as light with wavelength and can be influenced by electric and magnetic field
- electrons either hit object to scan surface or goes through to show inner structure
9
Q
Electron Microscope Component
A
- aperture: highly scattered electrons blocked
- magnetic lenses: bend electrons
- vacuum conditions prevent scattering of air
- influenced/refocused by lenses
- objective lenses amplify
10
Q
Electron Scattering by Atom
A
- 80% transmitted and 20% scattered
- gives low signal:noise
- elastic electron conserved energy and changes direction (gives signal)
- inelastic electron released as ionizing radiation (gives noise)
11
Q
Electron Microscope Structure
A
- source
- condenser
- sample
- objective lens/aperture
- projector system
- imaging
12
Q
Electron Source
A
- electrons extracted by heating source
- broad electron source
- field emission gun: much more focused beam
13
Q
Electron Wave Focusing
A
- electromagnetic lens producing magnetic field to bend electron path
- changes way electron travels through it
14
Q
3 Lenses of Microscope
A
- condenser lens: controls intensity/convergence of beam
- objective lens: generates contrast in image
- projector lens: magnify image
15
Q
Aperture
A
- removes highly scattered electrons
- reduces spot size and abnormalities
- size of hole controls diameter of electron beam
- aperture after each lens system
16
Q
ID Sensor
A
- higher throughput compared to film
- lower sensitivity
- conversion of information from electron to light back to electron that is detected on the camera
- delocalised signal in fiber optic coupling lowers signal:noise ratio
17
Q
DD Sensor
A
- fast readout and movie like frames
- electron to electron conversion and high sensitivity
- uses less extreme electron doses to reduce damage to material
18
Q
Beam Induced Motion
A
- unknown reason
- can computationally correct for this by aligning individual snapshots and frames of snapshots
- sharper image
19
Q
Fourier Transforms in CryoEM
A
- way of splitting something into component sine waves
- sine waves created for 3D sinusoids
- contrast representing amplitudes
20
Q
Amplitude Contrast
A
- electrons as particle
- enhanced by objective aperture
- detecting amplitude difference in scattered and unscattered electrons
21
Q
Phase Contrast
A
- electrons as wave
- interaction with sample causes scattered electrons to have a phase shift
- objective lens focuses scattered beam and directs it back to the detector
- different path length to detector gives different phase shifts
- scattering to higher angles gives higher resolution
22
Q
Oscillating Contrast
A
- in image you detect oscillations of positive and negative contrast as a function of scattering angle of the wave
- oscillation as a function of the angle carries information at different resolution ranges
- low angles have lower phase shifts so contribute minimally to the image
- oscillations of positive/negative contrast varying as a function of resolution range
23
Q
Contrast Transfer Function
A
- The contrast transfer function (CTF) mathematically describes how aberrations in a transmission electron microscope (TEM) modify the image of a sample.[1][2][3][4] This contrast transfer function (CTF) sets the resolution of high-resolution transmission electron microscopy (HRTEM), also known as phase contrast TEM
- The contrast in HRTEM comes from interference in the image plane between the phases of scattered electron waves with the phase of the transmitted electron wave. When an electron wave passes through a sample in the TEM, complex interactions occur. Above the sample, the electron wave can be approximated as a plane wave. As the electron wave, or wavefunction, passes through the sample, both the phase and the amplitude of the electron beam is altered. The resultant scattered and transmitted electron beam is then focused by an objective lens, and imaged by a detector in the image plane.
24
Q
Phase Shift in Focusing
A
- when beam is focussed you bent electrons more strongly and shorten their path
- control amount of defocus and control amount of phase contrast applied to lens
- phase shift controlled through defocusing of objective lens and this generates contrast
- additional phase shifts affect image contrast at different resolution ranges
- image can be corrupted by CTF
25
Effects of CTF
- inversion of contrast at different resolution ranges
- causes deconvolution of signal
- correct for this function to get rid of image distortion
26
Phase Plates
- another way to generate contrast via phase contrast by lens defocusing
- use phase plates as mechanical way of generating 90 degree optimal phase shift
-
27
Sample Requirements
- size: need enough atoms to scatter enough to reliably detect
- quantity: don't need much
- environment: examine membrane proteins/native-like environment
28
Sample Heterogeneity
- need homogeneity for higher resolution
- purification: SEC or affinity
- quality assessment: SDS Page, Activity assay
- negative stain: key advantage using stain to assess quality of proceedings as you go
- can also have conformational heterogeneity/flexibility
29
High Resolution Challenges
- vacuum samples removes water
- ionizing radiation produced destroys sample
- inelastic electrons contribute significantly to noise levels
30
Negative Stain
- heavy metals scatter electrons due to dense nucleus
- contributes to amplitude contrast
- stain forms cast around object and sample is destroyed by vacuum
- image negative cast of protein revealing shape and solvent excluded surface
- sample applied on charged carbon film and stained
31
Negative Stain Pros/Cons
- high speed: info about purity and conformations
- high contrast: high atomic number of strains/lots of A contrast
- radiation hard: sample undamaged by radiation
- resolution limited: grain size of stain
- protein distortion: if sample not fully embedded cast doesn't fully form
32
Freezing Particles
- ideally you get particles in a thin super cooled vitrous later with surface tension holding meniscus of particles between holes in carbon film
- in reality there are forces at the air water interface causing denaturation
33
Minimizing Radiation Damage
- destroys long range order and high resolution shells suffer
- low dose image reduces amount of electrons exposed to radiation
- identify where we take images and underfocus to generate contrast/find focus away from material
- take lower dose images: possible due to direct detectors that still allow for good signals
- collecting movies with fast readout using different parts/frames to get highest resolution images
- damage mitigated but not stopped
34
Sources of Resolution Limits
- properties of EM data
- sample preparation
- alignment accuracy (small globular molecules)
- biochemical properties: conformational and compositional heterogeneity
35
Process of CryoEM
1. biochemistry and sample purification (negative stain for quality control)
2. plunge freezing
3. low dose sample navigation
4. low dose imaging
5. processing and data analysis
6. 3D reconstruction
36
Single Particle analysis
- take 2D images and reconstruct original object
- need to know the orientation of the object in ice
- need many views of sample to get idea of structure
- 2D images are projections of 3D objects
37
Projection Theorum
- fill empty box of 3D object with all different views and angles of original object
- 2D projection is a central slice through the 3D fourier transform of the object
- FT the 2D slice and set slice into Fourier space
- inverse FT the fourier space to get 3d reconstruction
38
Applying Projection Theorum
- filled 3D FT -> inverse FT will reconstruct object
1. electron source
2. orientation of electron source (a,b)
3. molecule in laboratory frame
4. projection of experimental image
5. 2D ft is a central slice of the 3D FT of the original object
- lots of projections from lots of orientations
39
Orientation Parameters
- need to determine orientations in images to appropriately orient all central slices together
- orientation defined as in plane rotations and translations
- noise and low contrast leads to inaccuracy in alignment and degrades image quality
- heterogeneity gives a less accurate picture
- put particles in same orientation to take an average and increase signal:noise
40
Single Particle Averaging
- biological machines are all moving and have intermediates so averaging is difficult
- low contrast and noise can make it hard to distinguish states
- single particle analysis averages over individual particles
- if you pick 2 different particles you create something that doesn't exist
- computer can help separate and classify particle states
41
Signs of Averaging over Mixed States
- not full
- fragmented density
- shows averaging between something occupied and absent
- use purification techniques to separate states and get a better final image
42
Imaging Process
- protein purification
- negative stain, inital model, initial model reprojection used as orientation refinement
- cryoEM purified sample, subframe collection, particle picking and aligned and averaged frames, defocus determination and CTF correction, particle alignment and classification, final structure
- CTF correction artefacts on image caused by defocusing
43
Motion Correction
- fast frame rate moves can correct for beam induced motion in ice
44
Defocus Estimation
- taking images at different defocus ranges shifts origin to fill in space
- when you image at certain defocus the phase contrast difference between positive and negative contrast depends on path length of the wave after interacting with lens to get to focal plane
- CTF impact on contrast and resolution range changes
- deconvolutes signal from images
- estimate defocus parameters and CTF to apply correction
45
Particle Classification
- select with computer
- template matching
- not perfect
- averaging improve signal:noise
- much higher resolution
- classify images together representing same view and orientation
- identify heterogeneity of sample
46
Projection Matching
- determine orientation
- initial model is a 3D object rotated at fixed angles
- each particle finds best match to orientation in computer
- projections of particle undergo 2D FT and are oriented in fourier space by projection mapping with model based projections
- inverse FT to obtain object
47
Map Validation
- CCD detectors
- low signal: noise
- can't determine particle orientation
- inaccuracies in alignment
- incorrect structures
eg. HIV envelope spike structures
- used template based on selection methods and picked particles that didn't contain particles
48
Model Bias
- be skeptical of own work
- need to be confident you can actually see particle
- need high enough dose and defocus to see clear phase contrasts
- real particles will show high resolution features not present in raw data
- do features in map match reported resolution
- can we use multiple references to initiate refinement and obtain same model
