Single-particle electron microscopy Flashcards

1
Q

What does X-ray crystallography require

A
  1. Requires crystals and x-rays- solid (crystalline state)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What does Cryo-electron microscopy and NMR spectroscopy

A
  1. Biomolecules in solution state
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What is overview of single-particle electron microscopy (EM)

A
  1. Single-particle EM images are projections
  2. Sample preparation
  3. Computational reconstruction methods
  4. 2D image analysis
  5. 3D reconstruction
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What are the steps in 2D image analysis

A
  1. Image preprocessing
  2. Particle picking
  3. Image clustering and class averaging
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What are steps in 3D reconstruction

A
  1. Reconstruction with known view angles
  2. Structure refinement with unknown view angles
  3. Calculating an initial structure
  4. Fitting atomic-resolution models to lower-resolution EM structures
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What is the basic idea of EM

A
  1. We want the structure of a “particle”: a molecule (e.g., protein) or a well-defined complex composed of many molecules (e.g., a ribosome)
  2. We spread identical particles out on a film, and image them using an electron microscope
  3. The images are two-dimensional (2D), and each particle is positioned with a different, unknown orientation.
  4. Given enough 2D images of particles, we can computationally reconstruct the 3D shape of the particle
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What are some dramatic recent improvements to Single-particle EM

A
  1. Invention of better cameras
  2. Until recently, electrons were detected either by photographic film, or by scintillator-based digital cameras which converted electrons to photons for detection
  3. New “direct-electron detectors” can detect electrons directly, substantially improving image resolution and quality
  4. Better computational reconstruction techniques
  5. Single-particle EM is thus coming into much wider use, and may challenge crystallography as the dominant way to determine experimental structures
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What is single-particle EM’s major advantage over crystallography

A
  1. It does not require formation of a crystal
  2. Particularly advantageous for large complexes, which are usually difficult to crystallize
  3. Also avoids structural artifacts due to packing in a crystal lattice.
  4. In EM, particles are in a more natural environment
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

What is disadvantage of single-particle EM

A
  1. Single-particle EM’s resolution is (typically) generally lower than that of crystallography. However recent advances have shown that high resolution data can be obtained.
  2. Reconstructing structures of very small proteins from EM images may be difficult, because images from different orientations look similar (i.e., “a blob”)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Why is single-particle EM is particularly advantageous for large complexes

A
  1. large complexes tend to be harder to crystallize
  2. The computational reconstruction problem in single-particle EM is usually easier to solve for large particles than for small ones
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What do particles need to be prepared for

A
  1. To survive in the electron microscope (in a vacuum, under electron bombardment), the particles are usually prepared in one of two ways
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

What are the two methods to prepare the particles

A
  1. Negative staining
  2. Vitrification
  3. Usually you’ll perform negative staining to check your protein sample, make sure everything is okay. Then you’ll move to vitreous ice to take high resolution images.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Describe what negative staining is

A
  1. Coat particles with heavy metal salt crystals
  2. This increases contrast (particles are easy to pick out from background)
  3. It limits resolution to ~20 Å and can introduce artifact
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What is vitrification

A
  1. Particles are embedded in ice (vitreous ice: flash frozen, not crystalline)
  2. This gives less contrast, but enables much higher resolution (beyond 4 Å)
  3. High-resolution single-particle EM relies on vitrification and is thus referred to as cryo-electron microscopy (cryo-EM)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

What are particle picking methods

A
  1. Particle picking can be difficult, because the images are low- contrast and noisy
  2. Images may also have contaminants that should be ignored
  3. A variety of automated and semi-automated methods have been developed
  4. Often this is still done manually, at least to seed automated methods with suitable templates
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

How can you reduce noise

A
  1. Averaging similar images reduces noise
  2. The images in each row represent the same ideal image but with different corrupting noise
  3. If we average the images in each row (in the sense of averaging corresponding pixels), we end up with a less noisy image, because the noise in the different images tends to cancel out
  4. In practice, the particles under the microscope will be in different orientations, and these orientations aren’t told to you in advance, so you need a way to group particles with a similar orientation.
17
Q

What is clustering

A
  1. Group together images with similar view angles
  2. Then align them to one another and average them together to reduce noise
  3. To do this, divide images into several classes (with each class representing a set of similar view angles)
  4. We need to determine both what the classes are and which images should be assigned to each class
  5. This is a clustering problem
  6. Group images such that the images within a group are similar, but images in different groups are different
  7. In machine learning terminology, this is “unsupervised learning”
18
Q

How do you refine a structure

A
  1. If we’re not given the view angles for each particle, but we have a decent initial 3D model, then iterate the following steps to improve the model:
  2. For each projection (i.e., each class average), find the view angle that best matches the 3D model
  3. Given the newly estimated view angles, reconstruct a better 3D model (e.g., using filtered back-projection)
  4. This is called 3D projection matching
19
Q

What is a high-resolution single-particle EM structure

A
  1. you can see alpha helixes
  2. And you can see electron densities corresponding to each side-chain.
  3. The best EM structures are still not as high-resolution as the best crystal structures, but it’s remarkable that you can achieve atomic-resolution structures with this approach.
20
Q

How can you obtain atomic-resolution models from lower-resolution EM

A
  1. Often we have high-resolution x-ray crystallography structures of each individual protein in a complex whose low-resolution structure was determined by single-particle EM.
  2. We can fit the high-resolution structures into the EM density
21
Q

Advantages of single particle analysis summary

A
  1. Does not require crystals
  2. Samples can be partially inhomogeneous
  3. Physiological conditions possible
  4. Requires small amount of sample
  5. Rapid - many steps automated
  6. May eventually even be possible in vivo
  7. Investigating biomolecules at atomic resolutions allows us to explore the very architecture of life.
22
Q

What is cryo-electron microscopy single particle analysis (Cryo-EM) facilitating the study of

A
  1. Dynamic biological processes
  2. Protein structures
  3. Protein complexes, aggregates
  4. Large virus assemblies
23
Q

What are the limitations- summary

A
  1. Radiation damage
  2. Precision of image alignment
  3. Numbers of particles averaged
  4. Conformational heterogeneity
  5. Orientational preferences