Electron Microscopy 2 Flashcards by Misha Gandhi

Can em teach us more

Most current knowledge = based on 2d images - classical tem images are 2d projections
Cells and tissues are 3d - can we get 3d info in tem
Is everything observed to date real - resin embedding = harsh ans causes deformation - can we do better

How well did you know this?

Not at all

Perfectly

Describe our current knowledge of molecules

50 angstroms - in resin - resolution limited buys ample preparation not tem
Can em get better resolution - through sample prep and methods of image analysis

How well did you know this?

Not at all

Perfectly

Why is 2d not enough

2d projection can give v misleading info about 3d structure

How well did you know this?

Not at all

Perfectly

How to retrieve 3d from 2d

Serial sections
Tomography

How well did you know this?

Not at all

Perfectly

Describe serial sectioning

Er- complex morphology = need 3d image
Imaging of consecutive sections from ultramicrotomy for 3d view - each separate, keep every section in order = stack and get 3d
Serial section from sectioning allows 3d view of structure - resolution limited by section thickness
In z= depends on cutting, best = 30nm but v high

How well did you know this?

Not at all

Perfectly

How are serial sections done

Stack manually or automatically using tool
Need confocal lm - sections thinner, dept of field small = better resolution
New Vision of the er
Using novel ultra thin sectioning technique - in which sections are 30-40nm
Stacked er sheets connected by helicoidal membrane motifs

How well did you know this?

Not at all

Perfectly

What is tomography

Like ct scan
Resolution of serial sections limited bu thickness of sections - 50-100nm - thinner sections lead to better resolution
Tomography allows to get a better 3d info about structures in thickness of section -300nm

How well did you know this?

Not at all

Perfectly

Describe principle of tomography

Take pics from diff angles
3d object —> set of 2d projections —> 3d reconstruction
Know angle = then do back projection - based on degrees of pics and can reconstruct
-60-100degrres = 61 images then back project to 3d structure

How well did you know this?

Not at all

Perfectly

Describe tomography history

3D reconstructions from projection images:
– Cormack and Hounsfield: Nobel prize in 1979 for Xray computed tomography
3D reconstructions from
EM images:
- Klug: nobel prize in 1982 for applying tomography principle in TEM (DeRosier and Klug 1968)

Ct scna invented, computer happened - calculate and back project images
Tomography waasnt poplar until 2000s —> bc need to rotate em by hand = hard but now automatic and computer controlled

How well did you know this?

Not at all

Perfectly

Describe tilt series

Data collection = stack images from diff angles
= can get reconstructed tomography lots of detail, only 300nm section could do 5 micron cell —> serial section and tomography but hard

How well did you know this?

Not at all

Perfectly

Problems with conventional em sample prep - fixation

Slow= seconds to mins
Conformation changes of protein
Permeability changes of membranes
Osmotic effect leads to dimensional alteration
Loss of diffusibel ions and small molecules
Masking of antigens

How well did you know this?

Not at all

Perfectly

Problems with conventional em sample prep - Dehydration

Shrinkage
Conformation changes of proteins
Loss of lipid

How well did you know this?

Not at all

Perfectly

Problems with conventional em sample prep - Embedding

Mechanical effects = hardening
Shrinkage during polymerization
Loss of lipids

How well did you know this?

Not at all

Perfectly

Problems with conventional em sample prep - Thin sectioning

Compression
Knife marks - if not perfectly sharp

How well did you know this?

Not at all

Perfectly

Problems with conventional em sample prep - Staining

Staining artefacts - precipitation of heavy metals

How well did you know this?

Not at all

Perfectly

Problems with conventional em sample prep - Tem

Interpretation mistakes

How well did you know this?

Not at all

Perfectly

Is ervything observed to date real

Biological sample is 70% water. Classical processing: Totally remove H2O and we cook it!!! Does that really preserve the structure of cellular component?
Technological adavances = can look at samples in more native conditions
= cryo em - observation of sample frozen

How well did you know this?

Not at all

Perfectly

Describe what happens when biological sample frozen

Cannot just freeze or use liquid nitrogen = damage = forms crystallized ice

Describe vitrification

Transformation of substance into a glass — a non crystalline amorphous solid
Sample to be in solid state but non crystallized form
Liquid state but behaves as solid

Describe cryo em

Vitrification of biological samples in solution on em grid by plunging small vol of sample quickly into liquid ethane at liquid nitrogen temp
= freeze so quick no crystals formed

Suspension sprayed = tiny droplets —> through slit —> on to em grid - has lots of particles —> dip in liquid ethane

Why do cryo-EM improve ultrastructure preservation?

Instead of using chemicals to fix = freeze v fast = so everything in place
Ice crystals damaging
If freeze v fast - at least 10,000 degrees Celsius/second = ice crystals do not form but amorphous ice = vitrified water
Observe sample at cold temp = no ice crystals
Sample frozen hydrated and in its native state

Why liquid Ethane and not liquid nitrogen for cryo em

Liquid ethane = -72degrees
Liquid nitrogen = -195 degrees
Water at room temp = 20degrees
If use liquid nitrogen = leidenforst efefct = will happen and will crystallize
Leidenfrost effect = formation of gas barrier between hot surface and a boiling liquid if temp diff great enough

Compare cryo em to classical em

See more detail but distorted with classical
See evrurthing with cryo (could do sectioning with cryo em but super hard = have to cut vitrified ice, electrostatic so section sticks on knife)

How to see proteins directly in cell in cryo em

Thin area or cellular organelles
Cryo electron tomography

What does cryo em give us - image

3d tomography Can see thin structures But low contrast images compared to resin Structure biological sample and must limit e- exposure = radiation of sample

Can we see atomic details in cryo em

Higher res Need to understand how cells functions at molecular level and how proteins function = use single particle em

What do we see at high resolution in tem

Cannot see well in regular classical em = need high magnification or higher res images

Describe single particle analysis - gen

Applies to protein complexes Assign orientation of projections in relative to 3d structure and calculate the reconstructed 3d structure

Describe acquire protection images of single particle analysis

Purified proteins vilify and put on grid and take pics = Acquire projection images Of diff ribosomes at diff orientations

Describe classifying projections of single particle analysis

For similarities = then see that there are similar groups = can improve signal to noise = have 2 diff classes = Assume ribosomes have same structures = imaging single particle at diff orientations —> figure out orientation of projections and then do back projection = get 3d structure of ribosome with single particle assumption

What is requirement for single particle analysis

Purified protein complexes not organelles or cells

Describe cryo em workflow

1- purify = sample prep 2= vitrify then sample freezing and loading 3- automated image acquisition 4= data storage and handling 5= to get image processing = 3d reconstruction

Describe single particle analysis particle picking

Identify location of protein complexes - particles in the image Manual picking = used to be manual Vs Automatic picking = software

Describe single particle analysis classification

2d images need to be grouped based on their appearance - same orientation, same conformation and same composition Used to discard wrongly picked particles or damaged particles Then average them to improve signal to noise ratio

Describe single particle analysis 3d alignment and reconstruction

Projection of initial model = protein we looking at low res —> Model 3d projection, matching of experimental image with projection —> Reconstruction = assign kno angles of proteins = we have it to image = 3d structures = high res

Describe cryo em fro structure determination advantages

Easy sample prep Molecules in closer to native state Requires only small amount of sample - 0.1mg More forgiven on sample purity Give info on sample dynamics

Structure of 2019 cov spike in perfusion conformation

Spike glycoprotein = in receptor of cilia, can understand life cycle -> can find Therapy for it Ex of single partical cryo em Plasmid that expresses spike protein Vitrification and do single particle cryo em 3d variability analysis = see binding domains, first structures of spike proteins = help design vaccine Covid variant = location of mutation in spike proteins = can see this What makes omicron so transmissible = put in receptor too, mutated = weakened antibody binding but still strongly binds receptors

What can we see using singe particle cryo em

Full structure of ribosome Microtubules very clearly Actin filaments Tau proteins

What allows for Best preservation of samples

Very rapid freezing - cryo em

What happens when frozen hydrated samples are hit with electrons

Frozen hydrated samples v sensitive to electron beam Amount of electrons required to get 3d high resolution info from a frozen section will destroy sections much faster than images can be collected

How to see purified protein complexes

Using averaging methods or single particles population average = average many images of identical complexes = purified protein complexes

Do we stain with cryo em

Nawww Only contrast = density difference between proteins and water = v low Defocus to get more contrast then reform image (bc deforms) = more contrast - can also use special cameras

Describe single particle vs tomography

With spa = have to separate things Cannot overlap particles, cannot use for almost everything inside cells

What destroys high res info

Resin embedding