Lec 23 Flashcards
Why is cryo EM better than x ray crytallography
In cryo em you don’t need to make a crystal which takes a long time
What is the difference in vitreous ice and regular ice
Regular ice : expands and form lattices, crystalline ice, this is bad because for a protein crystal because the expansion of ice would shear and damage it
Vitreous ice: protects the crystal, you freeze the sample so fast that the water is frozen in place which is the vitreous ice
How can we make the sample freeze faster to form the vitreous ice in cryo em
Why can’t we just use cryoprotectants like in x ray crystallography
Liquid ethane
Because it will make the sample too thick
What are the ways you can treat the sample before doing cryo EM
For membrane proteins
Use a sugar gradient with a cross linker. This crosslinker stabilized the complexes between the protein and the gradient lets us isolate the complex
Need to optimize the buffer
Need to remove the sugars
For membrane proteins Need to add detergents, amphipol or nano disc bc they mimic the membrane which is the environment that the membrane protien is in
After preparing the sample for cryo em where you do put it
Apply it to a metal grid with squares, inside the squares are hole with the sample inside
Blot away excess liquid to make sample thin
Cool it in liquid ethane to preserve the sample
In x ray crystallography do the sample thinkenss matter
No that’s why we can use cryoprotectants
In cryo em what is the thickness of the ice in the metal grid
Where do we image and why
In the square where the sample is, that’s the thickest then as it moves out we get thinner
We image around the thick ice because that’s where the contrast is best
What needs to be optimized in cryo em to get best images and resolution
The distribution of particle and the ice thickness (want particles to be separate and ice to be thin)
The problems of the microscope optics
The preferential orientation
The grid type
How many times you shoot a hole with the electron beam , if shoot mutiple times you speed up the throughput of the images
But if you shoot too many times there are negative effects like charging effects and drift
What is preferential orientation in cryo EM
Al lot of the proteins like to orient themselves on one specific way in the ice and stay in that orientation
This is a problem because we need all different angles of a random distributed orientation to get the 3D image.
Why do we need many projection from diff angles for cryo em
If only one orientation we get the incorrect information
What is a CT scanner
It’s a machine that measures the scattering
Explain the reason why we need multiple angles using the cat scanner example
If we scan at only one angle (0 degrees) you can’t interpret it
If you add the scattering from a bunch of different angles at fewer slices of angles and a bunch of images you get better resolution
What are the similarities in cat scanner and x ray crystallography
They both use x rays
Both use data collected from many different angles (for crystallography we rotate the crystal)
Both add images to get the final result
What are the differences in x ray crystallography and CT scanner
The CT scanner uses transmission through the sample, x ray crystallography relies on diffraction by the crystal
Cat scanner gives 2D cross sections that come together to give 3D image
But crystallography gives a 3D map after Fourier transformation
How are images made in cryo EM
So you have the beam hitting the sample which has the protein in different orientations in the ice
Each orientation shows a 2D image
if you find the the angles of each image you can add them all up and make a 3D image
How can you find the angles from a 2D image from cryo em
Start with a purposely very low resolution image
Then do projection matching: take the low resolution image and match its angle to what the 2D image looks like
Whatever angle in the low resolution model simulation matches the raw data that’s the angle of the protein in that raw data
What is the biggest problem with the low resolution simulation matching the images angle
The angles aren’t accurate, youre limited to the angles of the model
If you can find the orientation (angles) of the images can you make a 3D structure
No
What is an example of how the reference based method can lead to bias
If you use a high resolution image as a reference, and average the noise from it
You get a lower resolution of the reference image back which has nothing to do with the data
This is why we don’t supply high resolution reference to get data. Need low resolution reference
Other than the reference model method to get angles what else is there
A method without using a reference to get the angles
Leads to less bias
To actually get the 3D image what has to happen
We need to go back and forth between Fourier transform and no Fourier transform (real space)
What is similar in cat scanner and cryo em
They both get a 3D image from many 2D projections
Do the angles need to be accurate in order to get the accurate 3D image
Yes
What is the average ing problem of cryo EM
If the protein has different domains that are flexible and not static
The parts that we get a 3D structure from are the ones that don’t move around too much
Hard to capture moving protein domains
If there is a high resolution 2D image at specific angles what does this mean
We summed multiple images of the particle at that one angle
What is the major advantage of cryo em
Tells all the different conformational states of the protein since imaging all the particles in the ice in different states
Once you get a 3D structure of the protein in cryo em what can you do
Can computationally separate each conformational state of that protein as classes
For example one class would be the open state and the other would be the closed state
If a protein in the closed state has many isoleucines in the inside what does this mean
If in open state it’s asparatate and glu what does this mean
The closed state is hydrophobic and has strip hinder which makes it so that the charged ions can’t go through
They are negatively charged residues that interact with the positive charge calcium and make it go through the channel
In cryo em do you get one single resolution of the 3D structure
No you get a range of resolution across the whole structure
So if resolution of 3.8 this is not the entire structure resolution , this is the overall resolution
Low resolution region of a 3D structure are due to
Flexibility of the protien in the region
