How to interpret electron density maps and build a structure?

Question 1

What are the steps for electron density map interpretation

Answer 1

1. X-rays are scattered by electrons, so this is what we see
2. Calculate phases and observed amplitudes
3. Improve phases (by solvent-flattening or averaging)
4. Interpret map by building model (autotrace or hand building)
5. Improve model by refinement and rebuilding (iterative process)

Question 2

Why is resolution important

Answer 2

1. Information about the finest details comes from the highest- angle part of the diffraction pattern.
2. Fine details can only be observed if the crystals are well-ordered, as any difference between molecules, or motion during the experiment, blurs the image.
3. Low resolution images are less precise and have more gross errors due to misinterpretation.

Question 3

What are all the steps to build a final model

Answer 3

1. The inferred phase information and observed diffraction amplitudes are used to calculate an electron density map
2. The map plots the probability of finding electrons at each point in the unit cell
3. Density is interpreted as a series of atoms that follow the known atomic connections (protein amino acid sequence)
4. Building a model into the electron density involves interpretation and prior knowledge

Question 4

What prior knowledge can be used

Answer 4

1. Protein/solvent regions,
2. C-alpha trace,
3. main chain,
4. peptide direction,
5. sequence assignment,
6. side chain conformations,
7. disulphides,
8. metal ions,
9. glycosylation etc.

Question 5

What is solvent flattening

Answer 5

1. Put a mesh on to define which part is protein and which parts are solvent

Question 6

How is the modelled map built

Answer 6

1. The building of the model is done by a crystallographer using interactive graphics programs
2. Alternatively, computer programs can be used to automatically interpret the structure

Question 7

How do you build the model

Answer 7

1. Generally you will only build what you can see
2. If resolution is good enough one can thread protein sequence into electron density map
3. Automated building programs can thread high resolution data
4. Often times one builds a partial model and then iteratively improves the model
5. With the help of new maps to build loops and other mobile elements
6. This initial model is then subjected to a variety of refinement procedures

Question 8

How do you deal with water molecules, ions and ligands

Answer 8

1. The scattering of X-rays by electrons is insensitive to what kind of molecule the atom is doing the scattering
2. As long as an atom is well ordered (found in the same position in each unit cell) it will show up in the electron density map and can be modeled
3. X-ray structure can reveal any bound ligands, ions and well ordered water molecules with no special effort
4. Since water molecules are small, you need reasonable resolution (~2.5 Å) to be sure that it’s not just noise

Question 9

What is a Trace protein background

Answer 9

1. In reality electron density map is usually very “messy” and hard to interpret
2. Computer-based automatic tracing can help but manual tracing is still most important
3. In skeletonization, the map is reduced to a series of lines that run between the peaks of density, through regions of reasonably high density
4. With this representation, a lot of information can be presented quite simply

Question 10

What is the skeleton of protein main chain

Answer 10

1. The skeleton serves as starting point of building backbone
2. This is aided by the known amino acid sequence
3. Much “human” efforts are needed at this stage

Question 11

How do you fit the first model

Answer 11

1. The problem of fitting an initial model is to see the forest for the trees
2. To get an overview of a density map, it needs to be simplified in some way
3. This is generally done through skeletonization, or by making a “bones” representation of the map
4. Since this is a high quality map which shows a clear b-structure, the tracing is relatively straightforward
5. A 4-stranded b-sheet is easily built from the electron density map with the aid of graphics program

Question 12

How do you fit side chains

Answer 12

1. Once main chain is more or less traced, attention turns to the side chains
2. Look at what side chains fit the density

Question 13

Describe refinement and refitting of model

Answer 13

1. Usually a partial model is built
2. Computer refinement can help to optimize the fitting between model and map
3. Manual rebuilding for further optimization
4. The interactive process goes on until satisfactory

Question 14

Describe refinement and restraints

Answer 14

1. Building the model by hand (or by computer) gives a reasonable model but not one that is necessarily the best possible fit to all the available information
2. To improve our built model, we “refine” it
3. Refinement seeks to find the model that best predicts our original observations while simultaneously satisfying what we know about the chemical structure of proteins (e.g. phenyl rings are flat)
4. Restraints (secondary structure, hydrogen bonding, multiple copies of molecules in the unit cell) can be employed to prevent protein from flying apart or reduce noise
5. One needs to judiciously choose which restraints to employ
6. Within the refinement programs are physics & chemistry libraries based on known properties of proteins (bond angles, lengths, etc.) to further restrain atom sampling

Question 15

What does refinement do

Answer 15

1. In essence, refinement will move positions of protein atoms to minimize the following equation
2. R = E(sum)(|Fo-Fc|)/E(sum)(Fo)
3. F0 is observed amplitude (from the X-ray experiment) and Fc is calculated amplitude from the model
4. Data fit term- the R-factor (reliability)
5. Measures how closely the diffraction amplitudes predicted by our model matches the diffraction amplitudes we actually observed in the experiment

Question 16

Why are different maps used

Answer 16

1. Used throughout the refinement process after a model has been built
2. A way to remove potential phase bias and errors

Question 17

What are the different types of maps

Answer 17

1. F0-Fc map
2. 2F0-Fc map
3. Usually one examines a F0-Fc map to identify errors in a model and a 2F0-Fc map to guide the construction of the new model
4. Omit map

Question 18

Describe use of F0-Fc map

Answer 18

1. F0 (native/observed) map differs from Fc (calculated) map where there are missing or wrongly placed atoms
2. Produces positive/negative peaks in areas where F0 differs from Fc
3. The refinement process ends when the F0-Fc map is essentially featureless

Question 19

What is 2F0-Fc map

Answer 19

1. F0 plus a difference map F0-Fc

2. Map should look like the corrected model

Question 20

What is omit map

Answer 20

1. Used to reduce the bias introduced by phases calculated from the model
2. If the structure is refined with the atoms from the questionable region omitted from the model, the model bias can be reduced and more details of the correct chain trace can be observed
3. To examine troublesome regions:
4. If interpretation of part of an electron density map is somewhat doubtful
5. Surface loop cannot be traced satisfactorily
6. Electron density map for only part of the structure that is known correctly