Protein X-ray crystallography

Question 1

What is required to produce a crystal structure?

Answer 1

A single crystal of the protein of interest, that is of sufficiently large volume to produce measurable diffracted intensities
As well as some expensive equipment

Question 2

General overview of X-ray crystallography

Answer 2

An X-ray beam is diffracted by the protein crystal, which generates a patten of spots whose intensities are measured by a detector
From this, a 3D map of the electron densities responsible for the diffraction is derived
A 3D atomic structure can then be obtained by fitting a model to the electron density

Question 3

How can proteins be promoted to form crystals?

Answer 3

When the solution in which they are dissolved reaches supersaturation
Under these conditions, the protein molecules can pack together in a repeating array, held together by non-covalent interactions
The resulting crystals can then be used in X-ray crystallography

Question 4

Why is protein crystallisation generally carried out in water?

Answer 4

Because proteins are biological molecules that function in aqueous environments

Question 5

Drawback of protein crystallisation being carried out in water

Answer 5

The protein crystals that form are typically up to 70 % water - the protein is swamped by the solvent
The loose packing of the protein molecules within the crystal lattice gives lower diffraction quality

Question 6

Advantage of ‘soaking’ crystals in water

Answer 6

Allows small molecule ligands (e.g. drugs) to diffuse through the solvent channels and bind to the active site

Question 7

Why is crystallisation of a particular protein unpredictable?

Answer 7

Because proteins vary greatly in their physicochemical properties
Determining appropriate crystallisation conditions is a slow, trial-and-error-based process
Proteins are also sensitive to temperature, pH and the chemical composition of the solution

Question 8

Intensity of the scattered waves is proportional to…

Answer 8

…the square of the amplitude

Question 9

What is the “phase problem” of X-ray crystallography?

Answer 9

The detector can only measure the intensities of the scattered waves, not their phases
It is impossible to reconstruct the electron density without phase information

Question 10

Methods for solving the phase problem

Answer 10

1. Molecular Replacement method - can be used if the coordinates of a similar protein are available. The “old” data can be combined with the “new” data to derive the electron density map. This solves the phase problem by providing estimates of the phases of the new structure from a previously known structure
2. Isomorphous Replacement method - involves introducing a heavy (electron-dense) atom into the crystal without perturbing the shape of the crystal. The differences in the intensities of the diffraction spots for the original and the heavy-atom crystals can be used to derive estimates of the phase. Historically the most common approach to solving the phase problem. Usually requires several heavy-atom derivatives – “multiple isomorphous replacement”

Question 11

Isomorphous crystals

Answer 11

Are closely similar in shape

Belong to the same space group

Question 12

Once phase estimates are available…

Answer 12

…an electron density map can be calculated (via Fourier transform)

Question 13

Model building

Answer 13

The process by which the electron density map is interpreted in terms of a set of atoms and atomic coordinates

Question 14

From electron density to 3D structure

Answer 14

A preliminary model is determined (model building)
This can then be refined against the experimental data - the electron density can be calculated for the model structure, which is then compared to the observed density
The 3D structure is then modified iteratively (‘crystal refinement’) until a best-fit is obtained

Question 15

Crystal refinement is…

Answer 15

…a matter of interpretation of the electron density

It involves computer modelling and judgement by the crystallographer

Question 16

Resolution

Answer 16

The minimum distance at which 2 objects can be observed as separate entities
Smaller distance = higher resolution

Question 17

Resolution required for structure-based drug design

Answer 17

Resolution > 2.5 A

i.e. structure resolved with detail > 2.5 A

Question 18

Reasons why the positions of atoms within the crystal structure can be ill-defined

Answer 18

1. The same atoms in different molecules in the crystal may adopt different positions/conformations - i.e. the same side chain/residue can have multiple conformations - X-ray crystallography takes an an average ‘snapshot’ of all these so the image would be blurry if there are multiple conformations
2. Atoms are not static - they vibrate around their equilibrium positions, which also leads to blurring

Question 19

Why can X-ray crystallography not resolve H atoms in most protein crystals?

Answer 19

Their low electron density leads to weak scattering

Usually have to be added by computer modelling

Question 20

Common assumptions made when protein crystal structures are used in structure-based drug design

Answer 20

1. The protein structure is correct (does not contain errors and is complete)
2. The structure of the ligand and its interactions with the protein are correct
3. The protein-ligand structure is relevant for drug design

Question 21

Why can it not be assumed that the protein structure from X-ray crystallography is always correct?

Answer 21

An X-ray crystal structure is one crystallographer’s subjective interpretation of an experimental electron density map expressed in terms of an atomic model
The interpretation may contain errors, may be incomplete or may be open to alternative interpretation

Question 22

Most important factors determining the quality of the final model

Answer 22

Resolution of the study (low resolution increases probability of errors and incomplete modelling of the data)
Experience of the crystallographer

Question 23

Most important experimental technique for studying protein structure

Answer 23

X-ray crystallography

Question 24

Factors to be aware of regarding protein structure and computational modelling

Answer 24

1. The molecular models of protein structures from crystallography are the product of considerable human intervention (e.g. in the model building/refinement)
2. The models represent an average over all the molecules in the crystal over the whole time-course of the experiment - some groups in many protein crystal structures will adopt alternative conformations. Amino acid side chains often have 2 or more well-ordered conformations
3. Some parts of the structure may not be resolved by crystallography, especially if they are very mobile and have no well-defined conformation or position in the large number of molecules in the crystal. For example, surface loops/terminal regions of the protein
4. Crystal structures of macromolecules such as proteins are not the equivalent of small molecule crystal structures, which have a restricted conformation. Protein crystal structures are the best fit to the available experimental data, which contains static and dynamic disorder, as well as experimental errors
5. Molecular mechanics energy minimisation of a protein crystal structure typically reduces the energy of a protein crystal structure by a large amount (e.g. by relaxing large numbers of close interatomic contacts), which can change the structure in subtle but important ways. This doesn’t mean the crystal structure is wrong - it is just that MM methods aim to give a good structure of a single protein molecule, whereas a crystallographic structure is an average
6. Proteins undergo a wide range fo complex internal motions - a crystal structure contains the effects of averaging the many different protein conformations produced by these motions, and the effects of the motions themselves during the experiment
7. For high-resolution structures, a combination of 2 or more different structural models may provide a better fit to the experimental data than a single structure