Biophysical Techniques Flashcards
1
Q
Principles of diffraction
A
- Arises from elastic scattering of radiation in directions other than incident direction
- Diffracted waves interact with each other + produce interference patterns
- Diffraction pattern from a single molecule has info needed to generate a model, but need large no for practical reasons e.g. signal to noise ratio
- Crystal lattice = regularly repeated arrangement of molecules
- Needed as scattering from individual molecules is too weak to measure
- Crystal = 3D arrangement of atoms in real space w/ coordinates xyz
- Its diffraction pattern = 3D array in reciprocal space w/ coordinate system hkl
- The 2 are related by FT
2
Q
Diffraction
Bragg equation
A
- In a crystal, only some angles give constructive interference + finite intensity, these spots = reflections
- If x rays, incident at angle θ are scattered from adjacent crystal layers separated by a, there is a path-length difference btw scattering from adjacent layers that is 2p=2dsinθ
- To get constructive interference btw 2 scattered waves, this path difference needs to be an integral no. of wavelengths. Gives Bragg equation: 2dsinθ = nλ
3
Q
Diffraction
Sources of radiation
A
- Radiation must have a λ that is of the same order of magnitude as the crystal lattice
- Produced in the lab by striking a target of metal w/ e-s accelerated at a high voltage
- OR produced w/ synchotrons. High E e-s travel in a circular orbit in a vacuum
- Orbiting e-s emit intense synchotron radiation at points of curvature
- Crystal monochromator is used to select a wavelength appropriate for the experiment
4
Q
Diffraction
The sample
A
- Usual samples = crystals or fibres
- Crystallisation consists of 2 main steps:
1. Nucleation = molecules dispersed in solvent start to gather into clusters
2. Crystal growth = clusters need to reach a critical size before they become nucleus - Highly flexible + disordered regions are unfavourable
- As the concentration of protein ↑ the protein becomes ↓ soluble until protein comes out of solution
5
Q
Diffraction
Vapour diffusion
A
- 25ul of purified protein is mixed w/ equal amount of reservoir solution
- Precipitant, like ammonium sulphate, binds H20 molecules + ↓ available H20, mimics ↑ protein conc
- Solution is suspended on a droplet under a cover slip, which is put on top of reservoir
- Vapour diffusion → net transfer of H20 from protein solution in the drop to the reservoir until precipitant conc. in both are equal
- Drop shrinkage ↑ both the conc. of both protein + precipitant, moving conditions diagonally into nucleation
6
Q
Diffraction
Batch experiment
A
- Crystal crops are incubated in low-density paraffin oil
- Aqueous crystallisation drops are denser than the paraffin oil → stay beneath the oil where protected from evaporation + contamination
- Protein is brought directly into the nucleation zone + samples are mixed at their final conc. at the start so conditions are constant + crystals x usually dissolve
- x be used for crystallisation trials containing a small volatile organic molecules as they can dissolve into the oil e.g. phenol
7
Q
Diffraction
Other preparation for non-protein compounds in complex w/ crystal
A
- Enables investigation of the structure of proteins interacting w/ ligands, like cofactors or inhibitors + heavy metal ions needed to solve the phase problem
1. Co-crystallistaion: protein + crystal are crystallised together. Only method for producing crystals of protein in complex w/ large ligands like nucleic acids
2. Soaking protein crystals in mother liquor that contains the ligand. Preferred when want to compare structure of pure protein w/ protein-ligand complex as crystals ↑ likely to be same form as pure protein
8
Q
Diffraction
Detectors
A
- Area detectors are used
- E.g. = charged coupled device (CCD)). Give fast readout times + are ideal for synchrotron use where higher throughput is required
- Large phosphor screen is used to convert incident X-rays into optical photons which are transmitted directly to CCD w/ tapered optical fibres
- The distance between two adjacent spots in a row/column is inversely proportional to the unit cell dimensions
9
Q
Diffraction
The phase problem
A
- All detectors measure the intensity of the reflections
- We can calculate the amplitude through its proportionality to the square root of the intensity
- A wave is defined by amplitude, which we can infer, + phase, so there is a loss of information - the phase problem
10
Q
Diffraction
Isomorphous replacement
A
- Patterson function = FT of I(hkl) {I(hkl) → p(uvw)
- As I(hkl) = measurable quantity, P(uvw) can be obtained directly
- Direct methods involving solutions by the Patterson function can be applied to small molecules + used to locate heavy atoms
- For large molecules, P(uvw) gives v complicated patterns → need isomorphous replacement
- Use a heavy metal ion (strong scatterer)
- This changes intensities of the reflections from P(hkl) to Ph(hkl) P=protein, Ph = protein + heavy atom)
- Changes in intensity can be measured + a Patterson difference map obtained from Ft of /p(hkl)-/ph(hkl)
- If this map can be interpreted, we know phase + amplitude of Fh (hkl)
- We can work out structure factors for Fh so we can work out amplitude of native protein /Fp/ + heavy atom derivative /Fph/ → solution for phase of each structure factor
- Solving the structure like this gives 2 possible solutions for the phases, use at least 1 other heavy atom derivative
11
Q
Diffraction
Single + multiple isomorphous dispersion
A
- Similar to IR but heavy atoms x have to be as large (often selenomethionine replaces methionine)
- Takes advantage of capacity of heavy atom absorbing X-rays of specified lengths
- As a result, Friedel’s law x hold so reflections hkl x = hkl
- This inequality of symmetry-related reflections = anomalous scattering
- In MAD, a complex data set is generated w/ native crystals giving /Fp/ for each of the native reflections
- Data set using heavy atom derivative is collected at the same λ → /Fph/
- 3rd data set is generated, maximises anomalous scattering by the heavy atom at a different x-ray length. Non-equiv Friedel pairs used to establish phases of reflections in heavy atom data + used to identify native phase
12
Q
Molecular replacement
A
- Choice for determining the structure of a macromolecule for which is similar to another whose structure is already known
- Calculation involves solution of the rotation + translation functions (rotated in 3D, calculate structure factors + agreement btw, then place in each position in unit cell)- Fact that split into 2 ↑ efficiency
13
Q
Diffraction
Refinement
A
- Often, initial set of phases are determined + e- density map for the diffraction pattern is calculated
- e- density map is used to identify portions of the structure that can be used to stimulate a new set of phases w/ higher precision
- From this, a new e- density map can be created, more accurate atom positions can be derived → even better phase angle
- Models can be improved by varying position x,y + z and mobility B to minimise residual by least squares method
- New set of phases = a refinement
14
Q
Diffraction
Least squares refinement
A
- An estimate of how good the agreement btw observed data (Fobs) + calculated (Fcalc) is given by
R =Σ(Fobs-/Fcalc/)/ΣFobs - R of 0.2 is usually obtained for 2.5A resolution data. Lower res = lower R value
- Oftem several cycles of refinement are needed
15
Q
Diffraction
Rigid body refinement
A
- Proteins have domains w/ predictable structure, it’s possible to refine lengths of 2o structure, like a helices or B strands
- All distances btw atoms are fixed so only 6 variables for each block of structure, somewhat simplified
- Used for proteins that are difficult e.g. refinement of metabotropic glutamate receptor
16
Q
Diffraction
Validation
A
- Typical Rfactor = 15-25%
- Powerful check = Ramachandran plot (plot of the main-chain dihedral angles in a polypeptide chain)
- Phi and psi angles are not used as restraints in refinement programs and hence their quality are independent checks of the model quality.
17
Q
Diffraction
Temperature factors
Free R factor
A
- Describes how much an atom oscillates/vibrates around the position specified in the model
- All B factors higher than 50A^2 suggests observed atom is disordered
Free R-factor = important against incorrect model building
Here, 5-10% of reflections are quarantined from the refinement
Decreases = correct model building, increase = model is being built wrong
A decrease in the conventional R-factor at the expense of an increase in the free R-factor is diagnostic that the model is being overfitted,
18
Q
NMR
Overview
A
- Detects reorientation of nuclear spin in an applied magnetic field, Bo
- Advantages = ability to detect changes in chemical structure + environment and magnetic waves readily penetrate membranes + tissues w/o damaging
- Only technique to report protein dynamics at atomic resolution
- Disadvantage = fairly weak signals compared to other forms of spectroscopy due to small E separation btw E levels in nucleus
19
Q
NMR
Background
A
- Spin
- Nuclei in many isotopes have spin e.g. 13C
- Nucleus w/ spin has a magnetic moment, un, that can interact w/ an applied static magnetic field. Bo
- If nucleus has spin I, interaction w/ Bo means there are 2I + 1 possible E levels - Bo field
- In a typical NMR experiment, a strong Bo field is applied to the sample
- This Bo is generated by a current in a solenoidal coil made of ‘superconductor’ wire - B1 field
- In addition to static Bo, rotating B1 is needed to induce phase coherence in the x-y plane
- Usually generated by a resonator coil
- Coil is tuned to resonant frequency of the signal to be detected
20
Q
NMR
Spectral parameters
A
- An NMR spectrum is characterised by a series of lines or resonances
- There are 5 parameters that define such lines
- Intensity
- Absolute conc. x be obtained reliably, but relative conc. can be obtained accurately from measurement of resonance intensity - Chemical shift
- Nuclei are surrounded by e-s; Bo induces currents in the e- clouds that ↓ effective field experienced at the nucleus
- Reference compound, usually tetramethylsilane
- Intrinsic shift = characteristic of a particular 2o structure,
- Induced shift = arises from influence of environmental effects. Local environment shifts allow resonances of groups w/ same chemistry to be resolved
- References for a-helices and B sheets have also been assembled
- Forms basis of chemical shift index which compared to CD spectra = info at residue level - Spin-spin coupling + multiplet structure
- Spin-spin coupling = magnetic interactions btw neighbouring, non-equivalent NMR-active nuclei
- Measured by spin-spin coupling constant J
- 3 equivalent nuclei give rise to a quartet w/ intensity ratio 1:3:3:1, 2 equivalent = 1:2:1 - T2 relaxation time
- T2 = time constant of the decay of Mxy components
- Line width of an NMR signal is determined by T2 - short T2 = broader lines
- T2 relaxation = caused by transient magnetic fields at any frequency. So, T2 ↓ as molecular reorientation rates slow - T1 relaxation time
- T1 = time constant that describes recovery of M2 (magnetisation along Bo field direction) after an applied perturbation
- T2 < T1 as return magnetisation to z direction inherently causes loss of magnetisation in xy plane
- Short T1 = magnetisation recovers rapidly + a spectrum can be acquired in ↓ time
21
Q
NMR
Nuclear overhauser effect
A
- A dominant relaxation mechanism = dipolar interactions with other spin magnetic moments
- An important consequence of relaxation = nuclear overhauser effect, which can be used to determine intra-molecular distances
- NOE effect is the change in population of one nucleus when another magnetic nucleus close in space is perturbed by irradiation
- Short range through-space interaction
- Can divide inter-atomic distances which can help confirm the 3D structure of a molecule
22
Q
Types of NMR
1D
A
- Has 2 dimensions: X axis corresponds to the frequency axis (chemical shift in ppm) + Y axis = intensity
- From 1D NMR, can work out folded state easily from TMS region - if unfolded, no structured hydrophobic region
Can tell how close to a e- -ve atom through shift
23
Q
Types of NMR
2D
A
- Has 2 frequency axis
- Building blocks of preparation, evolution, mixing and detection
- Involves a series of 1D experiments in which a single decay has been altered in length
- Both evolution + detection are time periods, t1 + 2, during which chemical shift + scalar couplings evolve
[1H-1H] COSY Experiment
- Through-bond NMR
- Correlates nuclei via their scalar couplings
- Consists of a single RF pulse → specific evolution time → 2nd pulse → measurement period (t2)
- Protons that are more than 3 chemical bonds apart give no cross signal because the 14J coupling constant are close to 0
- Cross signal btw H^n + H^a are important - this phi torsion angle of the protein backbone can be derived from the 3J coupling constant btw them
- There are 2 types of peak:
1. diagonal peak (have same frequency coordinate on each axis + appear along the diagonal)
2. Cross peak (have different values for each frequency + indicate couplings btw pairs of nuclei
[1H-1H]-NOESY experiment
- Examples of through-space experiment
- Uses dipolar interaction of spins (NOE) for correlation of protons
- The intensity of NOE is proportional to 1/r^6
- Normally a signal is only observed if distance is smaller than 3A
- Correlates protons that are distant in aa sequence but close in space
Major advantage of 2D spectra = improved spectral resolution + easier identification of correlation btw peaks
24
Q
Types of NMR
3D
A
- Can easily be constructed from a 2D experiment by adding additional evolution time + 2nd mixing period btw 1st mixing period + direct data collection
Triple resonance
- Used for assignment of larger proteins
- Triple resonance as 3 different nuclei (1H,13C,15N) are correlated
- Only contain a few signals on each frequency, so issue with spectral overlap ↓
2nd type involves using 2 tandem NMR experiments e.g. 1H-NOESY-TCOSY spectrum where t1 is NOESY evolution time and t2 = TCOSY mixing time
25
Q
NMR
Refinement
A
- Calculation of structures is done by incorporating various observed NMR restrains into a molecular dynamics simulation protocol
- MD simulations calculate the positions + velocity of protein atoms in a series of small steps
- Several 100 structures are produced + structures of a subset that has the lowest E are superimposed
- Following computation of structures, likely a no. of experimental ambiguities + mis-assignments so cycle of iterative corrections
Many visualisation software packages that perform chemical shift assignments automatically e.g. SPARKY 3
26
Q
NMR
Size limit + solutions
A
- In order to detect through-bond correlations, need good linewidth compared to values of coupling constant
- In 1H spectra of small proteins, J is large enough for through bond correlations to be detected for atoms up to 3 bonds apart
- As molecules ↑ in Mw, they tumble more slowly in solution → faster T2 relaxation → ↑ linewidth
- Observations in through-bond experiments e.g. COSY are more difficult when linewidth is similar to coupling constant
27
Q
Solution to NMR issues
A
- Extend into 3D: only a few signals are contained, ↓ the issue of spectral overlap
- Deuteration: swap 1H for 2H. Simplifies the spectra + gives better relaxation properties due to the fact that 2H nucleus has a much smaller magnetic dipole than 1H
- TROSY: useful pulse sequence that can extend the accessible molecular weight range. When 1 doublet is broader than another due to chemical shift anisotropy, TROSY can be used to detect the narrower signal. Works best in high magnetic fields
28
Q
MRI
A
- An image of water protons in the specimen is obtained by collecting spectra while different field gradients are applied to the sample
- These field gradients ‘label’ the position of sample components in space, which allows images to be constructed
- Contrast MRI
- To generate a good image, it’s important to have intensity variation in signals from different parts of a sample MR I
- One way to get contrast = exploit different T1 + T2 values in a sample
- Functional MRI = 2 MRI images are acquired in quick succession whilst the subject forms specified brain tasks
- Difference signals are superimposed
