The pathogenic organism Staphylococcus aureus makes a pigment called staphyloxanthin. The pigment imparts a golden color (hence aureus), but more importantly, contributes to virulence by protecting the bacteria from being killed by the host immune system. The S. aureus enzyme CrtM may be a good drug target because it catalyzes a key step in staphyloxanthin synthesis:

A cholesterol biosynthesis inhibitor blocks Staphylococcus aureus virulence. Liu CI, Liu GY, Song Y, Yin F, Hensler ME, Jeng WY, Nizet V, Wang AH, Oldfield E. Science. 2008 Mar 7;319(5868):1391-4.

We will view and compare different structures of this enzyme.

1. Launch ChimeraX
2. Show the Side View by clicking in the Graphics tab of the toolbar, or using the menu (Tools... General... Side View), or using a command:

   Command: tool show "Side View"

3. Fetch the structure from the Protein Data Bank:

   Command: open 3w7f

   The protein is shown as ribbons, with ligands and nearby residues as sticks.
4. There are two copies of the enzyme, chains A and B. Delete chain B:

   Command: delete /b

The enzyme combines two 15-carbon molecules of farnesyl pyrophosphate to form a 30-carbon lipid. This structure contains farnesyl thiopyrophosphate, which differs from the substrate by having a sulfur (yellow) in the place of one oxygen.

5. Focus the view on the ligand residues and label them:

   Command: view ligand
   Command: label ligand

   Label height is 0.7Å by default, but can be changed, for example:

   Command: label height 1.0

In this structure, the farnesyl thiopyrophosphate molecules are named FPS. Three magnesium ions help to offset the negative charges on the phosphates. The ions are shown as greenish spheres; clicking into the ChimeraX window and hovering the mouse cursor over each shows information in a pop-up balloon. Metal coordination bonds from FPS, water, and the protein are shown with dashed purple lines. Lines drawn to indicate interactions other than covalent bonds are called pseudobonds. Hovering the cursor over a pseudobond or bond shows balloon information about its end atoms and length.

6. Although water could be included in the various analyses,
   remove it to simplify the tutorial:

   Command: delete solvent

7. Label the residues with sidechains shown:

   Command: label @@display

   In this command, @@display refers to an atom-level attribute named display that indicates whether the atom is displayed. Essentially, the command is saying: “label all residues with displayed atoms.”

It looks like several sidechains could be donating hydrogen bonds to phosphate oxygens. (Although the structure does not include hydrogens, we know they are there!)

One of the displayed residues is Ser 21. To measure a distance:

1. Ctrl-click to pick the sidechain oxygen of Ser 21
2. Shift-Ctrl-doubleclick on the nearest phosphate oxygen
3. click Distance in the resulting context menu

Similarly, measure the distance between the sidechain oxygen of Tyr 248 and the same phosphate oxygen. Equivalent commands would be:

Command: distance :21@OG :302@O1B
Command: distance :248@OH :302@O1B

Distance measurements are shown as yellow dashed pseudobonds. The distances seem consistent with hydrogen bonds. However, rather than measuring many distances and trying to remember the appropriate hydrogen-bonding distances for different types of atoms, you can just use the hbonds command or H-Bonds tool.

1. To limit the H-bond search, first select the FPS residues:

   Command: select :FPS

   (or menu: Select... Residues... FPS)
2. Find H-bonds with one end selected, and report the results in the Log:

   Command: hbonds sel restrict cross reveal true log true

   The same thing could be done using the H-Bonds tool (menu: Tools... Structure Analysis... H-Bonds) with options checked to Reveal atoms of H-bonding residues, Limit by selection: with exactly one end selected, and Log, other settings default.

The H-bonds are shown as light blue dashed pseudobonds. The two measured distances were found to be consistent with hydrogen bonding. Details can be viewed in the Log.

Remove the “hydrogen bond” pseudobonds:

Command: ~hbonds

Remove the distance measurements:

Command: ~distance

Clashes/Contacts is similar to H-Bonds but can also identify nonpolar interactions:

• clashes – unfavorable interactions where atoms are too close together; close contacts
• contacts – all kinds of direct interactions: polar and nonpolar, favorable and unfavorable, including clashes

We will identify contacts of the FPS residues. Make sure the FPS residues are still selected, then execute the contacts command:

Command:  contacts sel restrict cross reveal t log t select t

(This could also be done with the Contacts tool, in the menu under Tools... Structure Analysis.)

In the Log, the atom-atom contacts are listed in order of decreasing VDW overlap: positive where the atomic VDW spheres are intersecting, zero if just touching, negative if separated by space. By default, slightly separated spheres are still considered contacts. Distances between the atomic centers are also given.

We will use the atoms selected by the command to get a list of the contacting residues.

The Log lists the details of each atomic contact, but one might want a simpler list of interacting residues instead. First, deselect everything that is not protein:

Command: ~sel  ~protein

Now list the selected residues in the Log using the info residues command:

Command: info residues sel

You can see Ser 21, Tyr 248, and several others in the list. To clean up for the next section, remove the contact pseudobonds and clear the selection:

Command: ~contacts
Command: select clear

Focus the view on Tyr 248:

Command: view :248

Ctrl-click to select any atom in Tyr 248, then press the up arrow key on the keyboard to promote the selection to the whole residue (if the initial selection is a bond instead of an atom, you will need to press the up arrow twice). Alternatively, you can select the whole residue with the following command:

Command: select :248

We will set up clash checking and rotate the sidechain interactively. Start the Clashes tool from the menu under Tools... Structure Analysis, or use the command:

Command: tool show Clashes

In the Clashes dialog, keep most options as defaults, except:

• Turn on the option to Limit by selection, set to: with exactly one atom selected
• Turn on the option to Reveal atoms...
• Set the Frequency of checking to continuously (until dialog closed)
  (...and do not Close the dialog yet, although it can be moved aside)

Alternatively, this setup could be done with the command:

Command: clashes sel restrict cross reveal t continuous t

As reported in the Log, no clashes with Tyr 248 are found in its initial position.

Measure the sidechain chi1 and chi2 angles of Tyr 248:

Command: torsion :248@n,ca,cb,cg
Command: torsion :248@ca,cb,cg,cd1

The Log reports chi1 and chi2 as –106.801° and –142.054°, respectively.

Set the function of the right mouse button (= trackpad + Alt on Windows, trackpad + on Mac) to bond rotation by clicking in the Right Mouse tab of the toolbar or using the command:

Command: mouse right bond

With the assigned button, click on the Tyr 248 CA-CB bond (closest to the ribbon) and drag to change chi1 interactively, or on the CB-CG bond (the next one out, adjacent to the ring) to change chi2. Any clashes are shown as bright purple pseudobonds and updated automatically as the bond is rotated.

Close the Clashes dialog. If clash pseudobonds are still shown, remove them:

Command: ~clashes

Restore the original chi1 and chi2 angles of Tyr248:

Command: torsion :248@n,ca,cb,cg -106.801
Command: torsion :248@ca,cb,cg,cd1 -142.054

Next, we will compare the conformation of Tyr 248 in the structure to tyrosine rotamers from a library. With Tyr 248 still selected (if not, Ctrl-click it), start the Rotamers tool:

Command: tool show Rotamers

In the resulting dialog, click OK to show TYR rotamers from the Dunbrack library.

The dialog lists rotamer sidechain (chi) angles and probabilities from the library. The rotamers are displayed, and you can see that none matches the original conformation of Tyr 248 (chi1 –107°, chi2 –142°). Specific rotamers can be shown by clicking/dragging to choose rows in the list.

The probabilities are based only on the residue's backbone angles (given in the Log: phi –74° and psi –32.2°), NOT on any interactions with the rest of the structure. Use the rotamer dialog to calculate H-bonds and Clashes with default settings. The new columns show that each rotamer forms several clashes but no H-bonds. Earlier in the tutorial, we found that Tyr 248 H-bonds with ligand and avoids clashes, which may compensate for its nonrotameric (presumably strained) conformation.

A sidechain can be replaced with a rotamer of the same amino acid type or “mutated” to a different type. With a single rotamer chosen in the dialog, either click Use Chosen Rotamer(s) to replace the sidechain with that rotamer or Cancel to close the dialog without changing the structure.

Unlike Tyr 248, Tyr 41 resembles the two highest-probability tyrosine rotamers given its backbone angles. If you like, focus the view on Tyr 41, Ctrl-click to select one of its atoms or bonds, and use Rotamers to show and evaluate rotamers of tyrosine or some other type of amino acid at that position.

When finished with Rotamers, click Cancel on the rotamer dialog. Clean up for the next section and zoom back out:

Command: select clear
Command: label delete
Command: ~hbonds; ~clashes
Command: view

FPS and the natural substrate both have a highly polar/charged “head” and a long hydrocarbon “tail.” The enzyme pocket is very large and deep, as needed to accommodate two of these molecules.

A surface representation is best for showing the shape of the pocket. ChimeraX provides a command that calculates the molecular lipophilicity potential (MLP) map for proteins. Either click on the “color by hydrophobicity” icon in the Molecule Display tab of the toolbar or use the mlp command:

Command: mlp

This shows a molecular surface and colors it from dark cyan for most hydrophilic (least hydrophobic) areas, through white, to dark goldenrod for the most lipophilic (most hydrophobic).

As expected, the mouth of the pocket near the phosphates and ions is mostly polar, while the rest of the pocket is largely hydrophobic. However, the pocket is so deep that it is hard to see when the whole surface is shown. Showing just the pocket region allows viewing it from the outside.

Limit protein surface display to atoms within 6.3 Å of ligand and show only the largest patch:

Command: surface ligand @<6.3 visiblePatches 1

Hide the ribbon and the protein atoms:

Command: hide ribbon
Command: hide protein

Amino acid hydrophobicity is one example of an attribute. Another attribute commonly used in structure analysis is atomic B-factor. B-factor values indicate which parts of the structure are more or less flexible or disordered.

Display all atoms by choosing Presets... Sticks from the menu or by using the command:

Command: preset sticks

Now the atoms are shown as sticks. Show the ligands as spheres:

Command: style ligand sphere

The style command sets atom/bond display styles.

The color command provides a number of options for coloring structures, including coloring by attribute. Color the structure by B-factor:

Command: color byattribute bfactor

The lowest B-factors (blue) are in the protein core, the highest (red) in a loop over the active site and the C-terminus on the opposite side. Different color mappings could be applied, for example:

Command: color byattribute bfactor palette rainbow range 25,75

B-factor is an atom attribute, but residue and model attributes can also be used for coloring.

Comparing different structures of a protein is another way to evaluate flexibility. We have been viewing a structure bound to substrate analogs, 3w7f. A structure of the same enzyme without ligands is also available. Fetch the “empty” structure and apply the default (original look) preset:

Command: open 2zco
Command: preset original

Now the first structure is tan and the new structure is sky blue, as shown in the Model Panel (Tools... General... Model Panel).

The structures are in completely different positions, so the next step is to superimpose them. This is easily accomplished with the matchmaker command (same as mmaker).

Command: mm #2 to #1 showAlignment true

Matchmaker details

MatchMaker first generates a sequence alignment using both residue types and secondary structure (tries to align helix with helix and strand with strand), then fits the sequence-aligned residues in 3D by pairing their CA atoms. The showAlignment true option displays the sequence alignment. By default, the fit is iterated so that far-apart pairs are not included in the final match. This allows the most similar parts to superimpose more tightly and the dissimilar parts to stand out. Final match statistics are reported in the Log, and the pairs used in the final fit are shown with light orange boxes on the sequence alignment.

The structures are highly similar except for the loop at approximately residues 52-56; you can see residue numbers by clicking into the sequence alignment window and hovering the cursor over the corresponding one-letter codes. In that window, click any part of the light orange boxes to select all the residues used in the final fit; these form the common core, the most conformationally similar parts of the structures. The selection can be inverted if you want to do something to the less similar parts, for example:

Command: select  ~sel
Command: disp sel & protein
Command: select clear

The command select ~sel inverts the selection, and disp sel & protein displays the selected atoms that are also classified as protein. The net effect is to show the side chains of the conformationally dissimilar loop.

We want to morph between two states, the bound state (3w7f) and the empty state (2zco), using the morph command. Morphing involves calculating a series of intermediate, interpolated structures between the original input structures. The series of structures is treated as a trajectory that can be replayed, saved to a coordinate file, or saved as a movie.

Command:  morph #1,2 wrap true frames 60

The wrap option means to wrap around from the last model specified to the first, in this case, to morph from the bound to empty state and then back to the bound state. The command plays the morph trajectory once and shows a slider interface that may be used to replay the trajectory or to save it as an MP4 movie. Note that you will need to move the slider to the beginning manually to record the whole trajectory as a movie file.

Although not all of the atoms in the morph are displayed at first, all atoms that are present in both input structures are also present in the morph. Their display can be controlled just as in any other structure, for example:

Command: show #3:lys
Command: hide #3:lys

The ligand FPS was only present in one of the structures, 3w7f, so it is not included in the morph trajectory. However, atoms from the original structures can be displayed along with the trajectory. Display of 3w7f (model #1) can be re-enabled by checking the (shown) box in the Model Panel or with the command:

Command: show #1 models

Then, show only the ligand residues and ions from that model:

Command: hide #1 surfaces,ribbons,atoms
Command: show #1:fps,mg