551 New Flashcards
(175 cards)
1
Q
What wavelength is vis light?
A
400-700nm
2
Q
Why need Xrays and mostly soft ones to see proteins?
A
CC bonds are about 1.5 A, need short wavelength to see atoms.
3
Q
How do xrays diffract?
A
They interact with the electrons of atoms and diffract.
4
Q
What is a unit cell?
A
It is a repeating unit of the crystal that goes on for forever.
5
Q
What are nodes? Lattice?
A
Nodes are at the 8 corners of unit cells. Each node interacts with 8 unit cells. These nodes when connected form a lattice.
6
Q
What is translational symmetry?
A
It is that there is repetition between unit cells in the x,y, and z planes.
7
Q
What are atomic coordinates?
A
A node is picked as the origin 0,0,0. Then each atom has a x,y,z coordinate. If there is symmetry-related atoms in the same unit cell, only need 1 molecule and this symmetry to describe the entire crystal.
8
Q
What is contouring?
A
From electron density, any density below certain threshold not included. Anything above is included. From this shape can develop the molecule from it. This model is fitted into the electron density.
9
Q
Why is centre of Xray diffraction pattern blank?
A
Because have beam stop to stop undiffracted x rays from the direct beam hitting the detector.
10
Q
What is the correlation between array of spheres and diffraction pattern?
A
The lattice spacing is inversed in diffraction pattern (shorter distances in real space are longer in pattern).
11
Q
Benzene has 6 fold symmetry in real space, does its diffraction pattern have 6 fold symmetry?
A
No pattern has 2 fold symmetry.
12
Q
How do you see unrecorded reflections, how give coordinates?
A
Get diffraction pattern, but some beams not in state to be constructive interference on the film, so are unrecorded. To see these, need to rotate the crystal. These unrecorded reflections in reciprocal space have h,k,l coordinates. The center of the film is usually the origin.
13
Q
What does a FT do?
A
Fourier transform combines all structure factors from pattern. You get the electron density at x,y,z.
14
Q
Each diffracted xray is _________ of the contributions from all atoms in the unit cell.
A
the sum
15
Q
How are structure factors written?
A
They are vectors, you put Fa, Fb etc. and do it for any symmetrically related atoms in the same unit cell with a’, b’ etc.
16
Q
What is a structure factor?
A
It is a wave (fourier sum) created from adding all the waves from individual atoms.
17
Q
What do you get from XRD pattern? What don’t you get?
A
Get intensities and positions of points, do not get the phases.
18
Q
What holds protein crystals together?
A
There is water present in crystal, there are weak intermolecular forces so the crystal is fragile.
19
Q
Is it possible for 2 different crystal types to arise from the same growth conditions and mother liquor?
A
Yes
20
Q
What is crystal mosaicity?
A
It is that protein crystals are made of smaller blocks that are roughly aligned. Because of this, x rays will diffract slightly differently depending on orientation of smaller blocks. This is the mosaic spread of a crystal.
21
Q
Do protein crystals contain water, do they need it?
A
They need it, if they are dry, they do not diffract. The water molecules are ordered.
22
Q
How do we know that proteins in solution are the same as in a crystal?
A
They still have function while in crystal. Crosslinking also supports crystal structures, can do NMR and overlap structure with one from PXRD, they are generally similar.
23
Q
Describe hanging and sitting drop crystallization.
A
Hanging has some protein and precipitant in the drop. If [precipitant] is higher in the well below drop, then water leaves the drop so protein and precipitant concentrations go up, then protein conc. goes down as precipitant stays same as crystals form.
Sitting is good as detergents if needed for MB protein in drop reduce surface tension, so sitting more practical. These can be modified during experiment (like adding more precipitant to well etc.).
24
Q
When crystallization is too fast, what do you get?
A
Amorphous precipitation or you get really small crystals.
25
What happens if protein and precipitant concentrations stay high and it stays in nucleation part of phase diagram?
This is bad as all protein in sample going towards making new crystals and not growing existing ones.
26
What are the 2 types of seeding, explain the process.
Macroseeding is using tiny crystal, put in solution with protein and precipitant conc. so in growth conditions. Usually need to etch the surface of crystal as surface is poisoned, need to give surface for crystal to grow on.
Microseeding uses a crushed crystal that is diluted to get a few small nuclei for seeding.
27
How is microdialysis done?
Have a 'button' with a small well with the protein solution. It is kept in the well with MB. This button is placed in precipitant solution, the precipitant goes into the well (diffusion). Easy to make changes to precipitant solutions.
28
How is a microbatch crystallization done? Why work and what are the advantages (3)?
It is a small well containing the precipitant and protein already mixed together. This aq solution is covered by a layer of oil. The oil ensures that water from the sample leaves very slowly, and it also protects the sample from oxidation. Advantages are that isolated from air, very small volumes, and can be automated with robots due to this.
29
What are the 6 main factors that can be altered to optimize crystallization?
1. Concentration of protein, if too high, it might not be soluble when try to crystallize it.
2. Conc. and type of precipitant.
3. pH.
4. Temp, for kinetics or because protein more stable.
5. Additives = ligands, diff buffers.
6. Depending on method, the conditions needed to grow crystals is different (not always same conc. of proteins and precipitant).
30
How does PEG work as a precipitant?
It draws water towards it and away from the protein.
31
Why additives such as diff ligands used for crystallization?
When protein binds to ligands it may adopt a more stable conformation.
32
What is covariance in proteins?
It is if an AA is mutated, then the other(s) that interact with it will also be different.
33
What 2 main features of AlphaFold help with the predicting and determining model accuracy?
1. You can add ligands and post-translational mods to the data so it accounts for those when uses the sequence to simulate a protein structure.
2. plDDT (predicted local difference test) uses dark blue, light blue, yellow, and orange for each AA to help show what regions are expected to be more accurate, less accurate, are likely disordered.
34
What info do you gain from AlphaFold Predicted Aligned Error (PAE) graph?
It looks at the confidence of the relative position of 2 residues in the predicted model. Have y-axis, top is zero. X axis left is zero. So compare each AA to all the other AA's in the sequence. Diagonal line from top left to bottom right as this is the AA lining up with itself so 0% expected position error.
A square represents a domain. If going down Y axis see a horizontal white line, then it means that those AA's are far away in space from the others in the protein.
35
What data from PAE graph helps us with crystallization?
We can see which AA's are far apart from others, and with the colour representing expected accuracy can see what are more flexible regions. We can cut the sequence between domains at diff spots or cut off flexible regions to get better chances of crystallisation.
36
What are some limitations of AlphaFold (4)?
1. Only the protein is modelled, not the environment around it and how protein interacts with it unless you add ligands, etc.
2. Models are biased based on prior models, so an earlier model of a certain conformation of a protein will skew your model to resemble that one, even if yours has a diff conformation in vivo actually.
3. Can't see disordered regions.
4. Static structures, but proteins are dynamic.
37
Do you need to cleave all protein tags?
No, only the long ones usually. Short ones don't need a cleavage site after them.
38
What are and the orientation of the vectors and angles of the unit cell relative to xyz?
Vectors are right-handed, not always orthogonal (90 degrees). For xyz it is abc.
For angles, it is alpha between BC, B between AC, y between AB.
39
What are is the reasoning for what plane designation is?
It is abc, looking at how many times this particular plane cuts across each plane. So a plane that cuts across a once, b twice, and c 3 times is plane 123.
40
What are the miller indices for any planes parallel to a,b,c?
For planes parallel to a, it is 0kl, then h0l for B and hk0 for c.
41
Can you have negative miller indices?
Yes, you go perpendicular to the planes and if points towards +,- this is towards x pos, y neg. This means that for a 210 plane, the 1 is neg.
42
What is Bragg's law?
2dsin(theta)=n x lambda
d is distance between hkl planes. lambda is wavelength of x-rays. n is an integer (1,2,3,....) Theta is the angle that xrays hit the planes.
43
How are reciprocal unit cells drawn?
Draw line perpendicular to that particular plane and the length of this line is 1/interplanar spacing for that set of planes.
44
What is sphere of reflection, what do for us?
Radius is 1/wavelength of xrays used. Any point on sphere meets reciprocal lattice, get diffraction conditions. Crystal in centre of sphere. We derived Bragg's Law from it.
Rotate crystal means we rotate the lattice and get other nodes from lattice to touch edge of sphere.
45
Instead of rotating crystal to get other reciprocal space lattice nodes to be in diffraction conditions, what else could you do?
Change the wavelength, so like sphere of reflection, you change radius (1/wavelength), so it touches different points of lattice.
46
Can you have symmetry mirror and rotation plane at once with proteins?
No you can't, only one at once. If you have both, then you would need to have all the AA's to be R-handed not L-handed to give you the same protein molecule.
47
What is the convention for symmetry-related atoms in the same unit cell for atom position?
As each axis is 0-1, then you could have 1-x, y+1/2 etc. You essentially add or subtract multiples of 1 from any coordinate to get the other atom that is symmetrically related but the same atom for that molecule.
48
What does it mean that something has 3 fold symmetry? Where would the axis be for that for a cube?
It means that something when rotated once, during that 360 degrees it appears exactly the same 3 times. This is from corner to corner for a cube, so if you look at it facing one corner, you would also see 3 sides.
49
What are the following crystal systems: Tetragonal, orthorhombic, rhombohedral, hexagonal, monoclinic?
Tetragonal: It is a rectangle, lengths of 2 sides are equal.
Orthorhombic: Rectangle where none of the axis lengths are equal to each other.
Rhombohedral: All the sides are equal to each other like a cube, all of the angles are the same but not 90 degrees like a cube.
Hexagonal: like hexagonal prism: but only a segment of it.
Monoclinic is 2 angles same, no lengths are the same.
Triclinic: None of the angles or lengths are the same.
50
What are primitive vs centered unit cells?
Primitive have nodes at each of the corners. Centered have a lattice points at the corners too but also one in the center.
51
What are the following types of centering: Primitive, body-centered, face-centered? How many nodes do they have per unit cell?
Primitive is a node at each of the 4 corners. It has 1 node as it has 8 nodes shared with 8 unit cells.
Body centered has 1 at each corner but 1 other node in the center of it. It then has 2 nodes as 1 to itself and 1 from having 1/8 of each of the 8 corners.
Face centred has a node on each corner, and one on each face. From each corner, get a total of 1 node. From each face, it is shared by another unit cell so 6 nodes/2 cells means it gets 3 nodes total from that. So 3+1=4 nodes total.
52
What kind of symmetry in diffraction pattern? Is it the same or diff from crystal symmetry?
It is point group symmetry. Crystal has symmetry so DP will show symmetry. They are diff as DP symmetry will have 1 point that stays in place, no translational symmetry in DP.
53
What are the 2 categories of X-ray sources?
X-ray generators and synchrotrons.
54
How do X-ray generators work, what are the 2 types?
Have filament that gives off electrons, then voltage creates electric field that pulls electrons towards the anode. X-ray tube will have anode with water to cool it, or can have rotating anode so surface has chance to cool down before hit again. These X rays pass through Beryllium window as it is solid and not very electron dense. Need solid window so can keep inside of generator under vacuum.
55
What are the 3 energy level relaxations given off to produce X-rays. What is the mechanism for this?
Electron will kick out low level electron, so another from higher level will drop to fill the spot, producing x-ray. Have Ka1 and Ka2, similar energy levels in 2P (L level). Then have KB from the 3P (M level).
56
What filter do we use to block what type of x-ray emission before X-rays hit crystal?
Use nickel filter to absorb KB which is slightly shorter wavelength, doesn't work as well as Kalpha emission. Our anode is copper that is hit by electrons, so we use nickel (beside it on periodic table) to block out the emission from it. We do this otherwise we get double the points in DP.
57
How synchrotron produce X-rays?
Bend electrons to give off X-rays.
58
What is a goniometer?
It positions the crystal at diff angles as X-rays pass through it.
59
What would be a good element for beam stopper for x-rays?
Lead, it is electron dense.
60
Capillary vs cryoloop mounting of protein crystal?
Capillary will have tube with special glass filled with mother liquor, then from hanging drop put the drop on top of capillary, the crystal will sink to bottom of tube. Then remove most liquid and break off top of tube and seal it.
Cryoloop will scoop crystal out of solution, hold it as X-rays pass through it. Cold N2 blown over it to help proteins stay stable, but not freezing as ice will diffract xrays and damage crystal.
61
What is autoindexing?
It is when you take an image at 0 and 90 degrees of phi and take this image over +/- 1 degree. Then collect DP from each, and software will compare them and figure out possible unit cell shapes and orientation the unit cell is.
62
What is autoindexing?
It is when you take an image at 0 and 90 degrees of phi and take this image over +/- 1 degree. Then collect DP from each, and software will compare them and figure out possible unit cell shapes and orientation the unit cell is currently in.
63
What is the penalty in autoindexing, does the lowest penalty mean the best unit cell option?
Not always best option. You get the abc lengths and alpha, beta, gamma angles. If it comes back with angles not 90 degrees and 2 side lengths that are sort of close but not really, then the penalty will be higher.
64
How accurate is autoindexing for symmetry purposes?
It is geometric info, not measuring intensities of XRD pattern, so not symmetry info.
65
What is the number of measurable reflections?
It is when you have a limiting sphere with radius of 2/wavelength. Inside it you have sphere of reflection with radius of 1/wavelength.
66
What does mosaicity do to peak collected for the intensity of a diffraction pattern point?
More mosaicity is slight variations in the unit cell angles, they aren't all perfectly aligned, so get wider distribution of angles, so wider peak.
67
What is electron density a function of for FT?
It is a function of x,y,z, and does integrals for the structure factor (vector) of h,k,l.
68
What do we measure for each hkl value?
We get the intensities directly, then the square root of intensity is proportional to the abs value of the amplitude of the Fhkl structure factor term for that DP point.
69
How is the frequency of Fhkl related to the interplanar spacing?
It is inverse, so Fhkl=1/dhkl.
70
On XRD, how is larger angle of reflection correlated to frequency of structure factor and interplanar spacing?
A higher frequency fourier term will give higher angle of reflection so smaller interplanar spacing.
71
What would happen if you add F00 term to summation of structure factors to give electron density?
F00 had intensity of 1, so everything shifted by 1, so would look like everything is electron dense.
72
What is fj?
fj is the atomic scattering factor for atom j. For atom j, it would have fractional coordinates xj,yj, and zj.
73
What are rings on XRD pattern?
They are from ice. Ice has smaller unit cells, so higher angle of reflection.
74
As sin theta/wavelength increases on x axis, what happens to atomic scattering factors on y axis?
Atoms have electrons centered around them but are spread out, not in a certain point. So as resolution increases scattering from electrons in same atom cancel out, so get smaller scattering factor.
75
What are temperature factors?
They are because atoms not stationary, they move around a point.
76
When you have Ewald's sphere, when you rotate the lattice, how do you get concentric circles?
Diff points of lattice on the sphere, so you get a full circle concentric.
77
What are units of resolution in atomic scattering plot?
Y axis is atomic scattering factor, then x axis is resolution (sin theta/lambda).
78
What does a Wilson plot do?
It estimates the B-factor.
79
What are systematic absences?
They are when you get a pattern of seeing a point, then not in the same line.
80
What are Fridel Pairs?
They are hkl vectors that have same intensity if no anomalous scattering.
81
Do phases or amplitudes matter more to get image?
Phases.
82
Can you use phases from a similar structure to get a new structure?
Yes, if you have structure, can find phases used to get it, then apply those phases to your new protein.
83
What is single isomorphous replacement?
It is when add heavy atom to protein crystals.
84
How to locate atoms using amorphous replacement?
You do the DP of the PH, then minus the P, so get just H.
85
What does Patterson Function do?
It is based on electron density as get peaks of intensity at atom locations.
86
How to locate heavy atoms from Patterson Function?
You get the plot, then draw a vector between any 2 points.
87
How does space group symmetry help us find heavy atoms?
If you know what the equivalent positions at other coordinates other than x,y,z.
88
What is the Harker plane?
It is when interatomic vector v=1/2, it is between symmetry related atoms.
89
What is anomalous scattering?
It is when heavy atoms absorb x-rays of certain wavelengths.
90
What do you get from minimum and maximum of the peaks f' and f'' respectively?
f' minimum peak will give same xray energy value as inflection point of f'', and inflection point will give same value as max peak of f''.
91
What is used to replace Sulfur in AA's, which AA is more problematic?
Use selenomethionine, it is more easily oxidized than S.
92
What is the workflow for an experiment using selenomethionine?
Use media with SeMet instead of Methionine. Purify protein, use reducing agents, use MS to determine SeMet incorporation.
93
What 3 wavelengths do crystallographers select for data collection using anomalous scattering, which is the most important?
Peak, inflection, and remote. Peak is most important.
94
What does multiwavelength Anomalous Diffraction plot look like?
Both start at origin, draw structure factor with protein+heavy atom for both the wavelength where no anomalous scattering.
95
How do we correct for model bias using difference Fourier Techniques?
We draw 2 Fourier maps.
96
What is nonisomorphous phasing models?
It is if we add domain to protein, similar structure but diff protein packing.
97
What is the orientation search and rotation function?
It is giving correlation between Patterson maps with rotation function.
98
What is translation search?
It is when you look at correlation between measured structure factor amplitudes and model expected examples.
99
Why don't we get 100% of completeness of crystal structure?
Because points at edge of sphere don't diffract.
100
What is multiplicity of crystal structure data?
It is same as redundancy, how many times on average was reflection measured per point.
101
What is Rmerge value?
It is each measurement compared to the average intensity.
102
What is cc1/2 value?
It is correlation coefficient.
103
For the signal to noise (intensity/std deviation), what is the min value you can go to before it is too much noise?
You can go down to 2.
104
If you had high signal to noise ratio, and high cc1/2 value, and high redundancy, what is a reason you might not shrink the high resolution shell of your measurements?
If the completeness is low.
105
What is noncrystallographic symmetry?
It is when you have regions of local symmetry, but it is not present throughout the entire crystal.
106
How is resolution refinement monitored?
It is when select 10% of data randomly.
107
Why would AA that is pre-proline be restricted in movement?
Because 5 membered ring of proline will affect side chain placement of other AA.
108
Why would glycine have a wide rotational and torsional bond angle variation in peptide?
No side chain, less sterically hindered.
109
Refined model should have rms deviations of about ____ and ___ for bond lengths and bond angles respectively?
0.02 Angstroms for bond lengths, 2 degrees for bond angles.
110
What is the R-factor for refinement? What is the R free value?
It is structure factor (obs-calc)/obs.
111
What is anisotropic B value?
Isotropic is same in all directions, so anisotropic is not.
112
What does Luzzati plot show?
It is R-factor on y axis, then 1/distance representing resolution on x axis.
113
What does zero occupancy atom mean?
It means that we know atom has to be there, but we don't know what it is.
114
What is the flowchart for macromolecular crystallography?
1. Obtain crystals 2. Collect data 3. Obtain phases by molecular replacement.
115
What is microcrystal electron diffraction (microED)?
It is when we use really small crystals.
116
For electron source for microED, which one makes electrons with a greater energy spread?
Have thermionic emission, has greater energy spread than field emission gun.
117
How is Ewald's sphere different for microED?
As we use very small wavelength, we have sphere with 1/wavelength for radius.
118
What do we use to solve phase for MicroED?
Use molecular replacement.
119
What forces scatter electrons for MicroED?
Electrons scatter due to interactions with both electrons and nucleus of atoms.
120
How is Ewald's sphere different for microED?
As we use very small wavelength (0.025 A), we have sphere with 1/wavelength for radius, so LARGE radius. This causes the sphere to seem flat. Still rotate crystal to get lattice points on sphere to get recorded diffraction.
121
What do we use to solve phase for MicroED?
Use molecular replacement. Trial and error. First find orientation of molecule, then search for position.
122
What forces scatter electrons for MicroED?
Coulomb forces. Maps are therefore coulomb potential not electron density.
123
How are scattering factors different for MicroED compared to XRD?
XRD we have at resolution of basically zero, we have the amplitude of scattering at the atomic number of the atom. Not for MicroED, we have negatively charged atoms at different points than other atoms, not based on atomic number.
124
Why might you see fewer diffraction points for microED depending on the sample?
If have peptide instead of protein, then you have smaller unit cells, less in these unit cells, have weaker diffraction.
125
If have 2 crystals of same space group and angles, how could you tell if they are isomorphous?
If they have same unit cell lengths.
126
Is Rmerge, Rwork, Rfree higher or lower for electron beam?
All are generally higher than XRD.
127
What is rule of thumb for Rwork in XRD, but not in MicroED?
It is that the resolution should be 10 x higher than the value of R work.
128
When to use cocrystallization vs soaking?
For large ligand, and tight binding, use 1:1 concentration of ligand to protein, cocrystallization or small ligand, use lots of it as many will bind per 1 protein, this is more like soaking. Avoid precipitants that will displace ligands.
129
How would you crystallize for lead generation in drug design?
Have crystal, then soak them with different compounds. If it doesn't work, then do cocrystallization.
130
When deciding if your molecule can fit in region with lower electron density, what are temperature factor restraints?
Atoms bonded to each other, expected to move in similar fashion.
131
What 5 big things can use to judge quality of XRD model?
R-factor, Rfree, rms deviations from ideal bond lengths and angles, Ramachandran plot, Temperature factor plot, Luzzati Plot (Has R-factor for observed and calculated model, higher factor means more similar. Then on x axis have 1 over resolution).
132
Instead of electron density, Cryo-EM map is called a ______ ____ map.
Electric potential map.
133
What resolution is needed to see C-C bonds? What about protein secondary structure?
Need 1.5 A for C-C, and 3.5 A for beta sheets (a helic is 5.4 ish so 3.5 is good too).
134
What is used instead of physical lens to focus electron beam?
Electromagnets.
135
Why want a thin sample (usually) for cryo-EM? Why would we want a slightly thicker sample?
Electrons don't penetrate well, so need to blot excess liquid from EM grid so have thin sample. Thicker sample needed if protein needs lots of hydration around it.
136
What is BIG diff between Cryo-EM and XRD data collection?
Cryo-EM does not lose phases when go into Fourier Space, but XRD loses phases.
137
What is the image that we get from e beam hitting protein sample?
We get 2D projection of protein, hopefully protein is in different orientations so we can sum all them to get image.
138
Why can't we use cryoprotectants for Cryo-EM samples? What do we do instead?
They make it harder to blot away excess liquid, so sample is thicker and get larger noise:signal. We use liquid ethane to vitrify the sample instead.
139
Proteins sit at air-ice interface, and may have preferential orientation. What can we add to limit this?
Detergent, it disrupts this interface.
140
If use membrane mimetic for MB protein for Cryo-EM, is it going to be accurate to in vivo?
Not always.
141
In Cryo-EM, what is shots per hole?
In Cryo-EM grid, have many holes. It takes a while to get microscope to focus on new hole, so try to take as many images of the hole as possible before moving on. Would move beam a bit each image as electron beam damages sample.
142
Ideally, proteins would be orientated by a small angle, then if sum all these images we get good high res image. What is this ideal angle?
0.1 degrees.
143
What are Cryo-EM Charging effects?
It is when damage caused to sample by electron beam.
144
What is the problem with 3D reconstruction for cryo-EM?
It is that Cryo-EM gives 2D images, particles should hopefully be at random orientations.
145
What is the projection slice theorem, how does it help us with 3D reconstruction problem?
It is that each 2D image can be used to get 2D fourier transform, then combining these fourier transforms makes 3D fourier transform, then converted to real space.
146
We can computationally separate conformational states or species as ________.
Classes. You have class 1, 2 etc that may be open channel, then calcium bound etc.
147
What is Fourier Shell correlation? What is it used for in Cryo-EM?
It is a method to measure similarity between two 3D volumes over different spatial frequencies. It is used to estimate resolution of 3D reconstructions by dividing dataset into 2 random halves, then make 2 3D volumes. These two are compared.
148
What does RyR channel do?
It is in Smooth ER, when have increasing Ca in cytosol during action potential, it activates and releases more Ca into cytosol, get skeletal muscle contraction.
149
What effect do we see of imperatoxin on the RyR channel?
We see that channel stays in a subconductor state, some Ca leaks through.
150
What features of the imperatoxin allow it to bind to RyR receptor?
It is positively charged on one side, see that some amino acids need to be positive for toxin to work. The receptor has neg AA-s that form salt bridges with + toxin.
151
If Imperatoxin is phosphorylated on Thr, does it completely block RyR channel?
Yes it does.
152
What is a tomogram?
It is a movie of images.
153
How does computed tomography (CT) scan work?
It has X-ray source, see what shadow it casts. Do this at many angles, get images.
154
For XRC vs CT scanner, what do they have similarities? Differences?
Both use X-rays, collect data from different angles, add images to get final result. XRC uses diffraction by crystal, CT uses transmission through sample. Also XRC gives 3D map, CT scanner gives 2D cross sections, can be joined to give 3D image.
155
What is the general workflow for cryo-tomography?
1. Plunge freezing in liquid ethane 2. Use focused ion beam to get sample thin enough so electrons can be transmitted through (milling) 3. Tilt-series collection 4. Reconstruction.
156
Do proteins, viruses, bacteria, cells, and tissue samples need thinning, what type?
Viruses and proteins do not need thinning. Bacteria may, cells do use focused ion beam. Tissue samples are thick, so cut out and lift out a sample.
157
Why do we go from -60 to +60 degrees on Cryo-ET, why not more angles?
It is because beam is going through too much sample, too thick for transmission. Also, more electrons interacting with sample gives more damage.
158
What are data collection limits from cryo-ET?
We don't know if top or bottom of sample, it is arbitrary. We also have higher radiation damage at low angles, and less electron transmission near 60 degrees. So missing info.
159
How do we correct for missing wedges of data at low and high angles (near 0 and 60) for Cryo-ET?
We do a dual tilt. So rotate sample 90 degrees on other axis, then get another tilt series. Solves missing wedge.
160
What is ideal electron dose for Cryo-ET? Why do we have limited resolution in Cryo-ET?
1-10 electrons per angstrom squared. The damage destroys high resolution features, so get lower signal:noise.
161
What does dual-axis tomography do?
We do 2 tilt series at a 90 degree in-plane rotation. Reduces the missing wedge of info, and mitigates anisotropy (deals with unevenness in signal:noise ratio).
162
What is dose fractionation?
It is when total electron dose is split over 2 tilt series. Each series gets half the dose. Kind of get missing pyramid shape, better than traditional tilt series with 1 axis.
163
What is sub-tomogram averaging?
It is when put sample in rod holder, so it can be rotated 180 degrees. If we have many identical copies of protein/particle, then hopefully have them in random orientations.
164
What is segmentation in Cryo-ET?
It is when you determine what is the mitochondria etc from the picture.
165
In context of Cryo-ET, what is high structural density, visual overload, and lower signal:noise and denoising?
High structural density is as there are overlapping organelles and molecules, makes interpretation challenging. Visual overload is when small regions contain dense clusters of objects, hard to disentangle structures. Low Signal:noise is that as biological samples are fragile, using lower electron doses causes less damage, but get lower signal. We can use denoising filters to enhance signal.
166
What are pros/cons of manual and automated method of segmentation?
Manual can have high accuracy, but very time-consuming, and diff interpretations between people, not very scalable. Automated can reduce time to do, but need training data, good quality of it, and generally computationally expensive, but can be done.
167
What is sub-tomogram averaging?
It is when have multiple copies of structures, so we get them independently, then align them to increase signal:noise, get higher resolution. We are merging 3D images.
168
Sub-tomogram averaging is similar to single particle analysis, but what is a complication with dual alignment?
We need to align the raw tilt images. Also cells have lower number of particles to use than in purified sample. We therefore get lower signal.
169
Why might we not see lysines, glutamate, aspartates missing from tomography?
Lysines too flexible, and electrons damage glutamate and aspartate by decarboxylating them.
170
How use Cryo-ET to get order of protein complex assembly?
We knockout subunits 1 at a time, see what we get. Can get order, then get the individual segments and assemble them together to get full structure.
171
Spike protein needs to be _____ to allow fusion. This lets remainder of spike remain ____ state for a short time, before it closes again as open state is unstable.
Cleaved by host proteins, open state.
172
Why does imposing symmetry improve resolution? How applied to Covid-2?
It is that glycosylation in this case will then assign particle orientation for the trimeric spike protein. We can then essentially triple our data set, each image is duplicated so 3 copies total. This is because we can apply each image for each fold axis of orientation. This will increase resolution.
173
Why can't we see glycosylation for Cryo-ET?
It is because different ones at same sites, also they are flexible.
174
What difference in spike mainly do we see from Cov-1 to Covid-2?
Spike protein is more flexible, could increase infection.
175
How does COVID virus remain hidden from immune system?
Host proteins will add endogenous sugars to glycosylation sites = glycan shield.
