Protein-Ligand Visualization with VMD

Key Terms:  molecular surface, sequence/structural alignment, RMSD, rendering; induced fit, structural complementarity, binding recognition; buried hydrophobic surface area, electrostatic specificity.

Homework:  Explain Discussion Questions 1-7 in complete paragraphs. Incorporate your rendered images into your report with figure captions explaining what is being shown. Summarize the publications associated with each PDB structure examined: Files: day5papers.zip

Focus Question:  What are the determinants of drug binding and design? How do they relate to the driving forces of protein folding we learned in tutorial 4?

Highly Recommended:   Bring an external mouse.

Download the VMD installer file from the VMD Downloads page. The website will ask for us to pick a username, password, and create an account.

Macs:

• Macs 2020 and later:  download VMD 1.9.4 ARM64 (M1 and later processor chips).

• Troubleshooting: If you download the wrong code, bad CPU error will show in Terminal when you try to open VMD. Identify your processor (M series or Intel) by clicking on the apple in the upper-left of your Desktop: Apple> About This Mac. (Normally we wouldn't use an alpha test release because of potential new bugs but 1.9.4 alpha is the preferred version.)

• OS 10.14 (2018) or earlier:  choose VMD version 1.9.3 (32-bit), which supports fast GPU rendering (OpenGL).

• Program Installation:  On a Mac it will download a .dmg disk image. Double click the dmg file to mount the virtual drive: it should open a folder with the VMD program (colored icon), if you don't see the disk image you will need to navigate to it in your Finder like you would an external USB drive. Now in a separate Finder window, open your Applications folder. To actually "install" the VMD software on your cmputer, you need to drag the VMD icon from the mounted disk image window and drop VMD inside your /Applications folder (you should see it actually take a few seconds to copy the program files). When the copying bar reaches full completion, you can unmount the disk image like you would a USB drive (click the eject button). Once unmounted, you can delete the .dmg installer file because the program files are now installed in your Applications/ folder, where they have full access permissions needed to run. Finally, if Mac won't let you double-click VMD.app, it should run correctly if you right-click on the icon inside your Applications folder and select open.

 PC:  Choose the official release for Windows (1.9.3, without CUDA).

• Open the installer file and it will install VMD under Program Files> University of Illinois.

• Open your Program Files folder and right-click the vmd executable file to see the option for creating a short-cut icon on your Desktop.

Part 1.  Membrane Protein Receptor

Download the 3RZE structure file from the Protein Data Bank:  rcsb.org

  1.  Search for the entry with the following 4-digit PDB ID:  3RZE.

  2.  On the upper-right of the PDB page for 3RZE:

• Click 'Download Files'  and on the drop-down menu select PDB Format (plain text).

  You may need to right-click on PDB Format and select 'Save Link As..'

• Save the 3rze.pdb file by selecting your Desktop and clicking on 'New Folder', which you can name  day5.

It is a bit of a hassle to open pdb files in VMD using the mouse. If you know the path a program was to installed to, you can often use an alias to call it from a Unix terminal..

 Command line (Mac):

  alias vmd '/Applications/VMD\ 1.9.3.app/Contents/MacOS/startup.command'

  alias safari '/Applications/Safari.app/Contents/MacOS/Safari &'

  vmd ~/Desktop/day5/3rze.pdb

To manually open the pdb file:  Double click VMD.app to open a visualization window. Behind this primary visualization window it will also open a terminal where it explains what the program is currently doing. Verify from the startup message in the terminal that it detected your graphics card using the OpenGL renderer. Otherwise the graphics calculation/rendering will be performed on your CPU instead. CPUs are best at serial calculations and will not be as responsive when rotating the molecule. 3D graphics calculations are inherently highly parallel.

# If you couldn't get the alias to work in Terminal.. in the 'VMD Main' window:

1.  Select File > New Molecule...   this will open the window 'Molecule File Browser'.

2.  Next to 'Filename', click  Browse..  The 'Choose a molecule file' window opens. It may be behind your open windows (ctrl-Down arrow for Mission Control on Mac). Double click to enter your Desktop folder, then double click on day5, select 3rze.pdb and click 'Open'.

3. Back in the 'Molecule File Browser' window, 'Determine file type:' should list 'PDB' -- if not, select it from the drop down.

  Under [Browse], click [Load] and a structure should appear in the VMD Display window.

Setting up the Display:  

•  Drag the lower-right corner of the VMD Display window to enlarge. If the window goes blank, left-click and drag in the center of the window to rotate the molecule.

•  Under 'VMD Main' > Display, change perspective view to orthographic and uncheck the Depth Cueing box (atoms in the distance will be too dark). In 'Display Settings', set Near Clip = 0.01. Atoms that are closer to the viewer will be hidden from view (in front of the clipping plane).

•  Under VMD Main > Graphics > Representations, change 'Drawing Method' to NewCartoon and 'Coloring Method' to Secondary Structure.

•  You may also wish to change the VMD Main > Graphics> Color> Display> Background> to White.

How to View the Molecule:

1. Scroll up with your mouse wheel or touchpad function to enlarge the molecule to fill your window. This can also be accomplished by tapping s (scale mode reveals mouse <=> arrows) and click-dragging the mouse left/right.

2. Tap t and double-click and drag your mouse to translate the protein into the center of the frame.

3. Tap c and click exactly on the center of an atom to center the protein to rotate about that point.

4. Tap r to enter rotate mode (reveals pointer) again and at any time. Left-click drag in a circle to get a feel for the shape of your molecule. 

# Note:  if your display window goes blank, your molecules have translated off the screen: re-center the system by selecting 'Display > Reset view', or tap the = key.

3RZE is the structure of histamine H1 receptor (H1R). H1R is ubiquitously expressed and in involved in allergy and inflammation. It is a Gq-coupled GPCR, consisting of seven transmembrane helices that insert in the plasma membrane. By convention, membrane proteins are oriented with the z-axis perpendicular to the plane of the lipid bilayer. Rotate the molecule to see that the large 7-alpha helix bundle is oriented along the z-axis. Below the helix bundle is a small globular domain that resides in the cytosol, which we are not interested in. Note how by default the molecular system rotates about the center point of all the atom coordinates.

If you look back at the terminal output, your will see that the VMD program had to determine the bond structure using a distance search. This is because the coordinate PDB file only contains information on the types and locations of all the atoms, not how the amino acids or functional groups are connected (which is the molecular topology).

Q1:  Describe the protein you are looking at. If you scroll up and read the terminal output when it loaded, how many Atoms, Bonds, and Residues has VMD found?  

Residues are the number of amino acids encoded by the gene sequence.

We are done loading the molecule, so click the red button to exit/close the small 'Molecule File Browser' and 'New Molecule' windows. These small windows seem to get hidden or lock up over time, and we will be opening another molecule later and will want fresh windows. VMD is free scientific software and is not what we would call "user-friendly".

Customize Graphical Representations:

The header comments in 3rze.pdb explain the structure:

    TITLE     STRUCTURE OF THE HUMAN HISTAMINE H1 RECEPTOR IN COMPLEX WITH DOXEPIN

The structures we'll be looking at were solved by a technique known as X-ray crystallography. H1R was crystallized bound to a drug ligand, doxepin. Doxepin is an antidepressant that antagonizes various GPCRs. At low doses it serves as an antihistamine and first-generation antagonist for the endogenous hormone histamine. Histamine, derived from decarboxylation of the amino acid histidine, is responsible for mediating the allergic response.

Bring up the VMD Main > Graphics > Representations window.

• You can create multiple representations to display different parts of the same pdb structure uniquely. Click the 'Create Rep' box. Create Rep duplicates the current Rep or whatever Rep you have selected.

• Click on the second entry in the large white box to highlight it yellow or green. In the white box under Selected Atoms, the default all  and type:  not protein enter  (this selects the doxepin ligand). The mouse cursor needs to be over the white text box while you type. Now the NewCartoon representation only works for macromolecules, so we'll change it next..

• Change 'Drawing Method' to VDW (van der Waals spheres). Change the ligand's Coloring Method to:  Name. This space-filling representation shows the oxygen atoms red, nitrogens blue, and carbon atoms cyan.

• Finally, change the ligand's 'Drawing Method' to Licorice, so we can see its functional groups.

  3RZE.pdb - Graphical Representations:

DrawStyle   Color    Selected Atoms

NewCartoon SecondaryStruct  all

Licorice    Name       not protein

• Licorice is a bond stick representation. Notice how there are no white hydrogen atoms in the structure--that is because the protons did not have enough electron density to show up in the X-ray diffraction.

• Cartoon ribbons are a standard notation for tracing the polypeptide backbone from the N- to the C-terminus. Alpha helices are purple and beta sheets are yellow.

Q2:  a) Is the beta sheet parallel or antiparallel?  b) Is the bound doxepin exposed more to extracellular water or phospholipids in the membrane bilayer? 

Rotate the view to find the ligand inside the top of the transmembrane bundle. It has a phosphate anion sitting directly above it. Second-generation antagonists bear a negative carboxyl group that displaces this phosphate, binding nearby basic lysine residues with higher specificity than doxepin. Notice how this pdb file actually overlays two copies of the molecule doxepin. Each copy is a binding pose: the molecule can adopt two different orientations within the binding pocket. Compare the two poses by using the selections 'resid 1200' or 'resid 1201'.

Let's visualize the pharmacophore binding interactions. In VMD Main > Graphics set the following:

 3RZE.pdb - Graphical Representations:

DrawStyle   Color   Selected Atoms

NewCartoon  ResType  all

Licorice    Name     resid 1201 1202

Licorice    Name     resname LYS

• ResType colors the backbone by the hydrophobicity of the amino acid residue. Gray segments mark hydrophobic or aliphatic sidechains, green = hydrophilic residues, blue = basic groups (Lys, Arg, His), and red = acidic carboxyl sidechains (Asp/Glu).

• If you forget what a Rep does, double-click its line in the white box to hide or show it (entry becomes red).

Let's get ready to save an image of the ligand in its binding pocket.

• In VMD Main > Display, change Axes to Off.

• Rotate and scale the view so that the transmembrane helices fill the frame and are pointing up with doxepin on top (by convention the extracellular space is on top). Rotate around the cylinder until you find a good view of the ligand binding interacdtions that are not obscured by a helix.

Render a 2D image of the current view..

• Select VMD Main > File > Render... > Tachyon (internal, in-memory rendering), Filename = gpcr.tga > Start Rendering. Alternatively you can use Tachyon, which lets you increase the pixel resolution with -res x y (File Controls lets you Browse.. and pick a folder in the separate window.)

• Rendering generates a 2D bitmap image from our 3D model. The render creates a temporary .dat file and then a lossless, uncompressed .tga image that can be opened by double-clicking on its icon in the folder. Once you are satisfied with a render that is crisp and fills the window, select File > Export... and save the .tga file as a compressed .png or .jpg (lossy) image.

• Delete the .dat and .tga files. You will drag-and-drop the compressed image file into Word when preparing your lab report.

• If you have trouble rendering, you can take a screen shot, but the image may not be as crisp (Mac: shift-command-4 and crop the area).

Q3:  Which two lysine residues bind the phosphate anion in the ligand binding pocket?  Use the label atoms command..  VMD Main > Mouse > Label: Atoms. Then click on a nitrogen atom and record the residue number (detailed info is printed in the Terminal).

Once you're happy with this visualization and your render(s), click on the 3rze.pdb entry in VMD Main window (line turns green or yellow) and select Molecule > Delete Molecule.

Part 2.  Nucleosome Core Complex

Download the 1ZLA.pdb structure file from the Protein Data Bank:  rcsb.org

In your VMD application...  VMD Main > File > New Molecule > Browse... 1zla.pdb (file type: PDB) > Load. Close 'Molecule File Browser'.

The header comments in 1ZLA.pdb explain the structure:

   HEADER    PROTEIN BINDING/VIRUS/DNA               05-MAY-05   1ZLA

   TITLE     X-RAY STRUCTURE OF A KAPOSI'S SARCOMA HERPESVIRUS LANA                

   TITLE    2 PEPTIDE BOUND TO THE NUCLEOSOMAL CORE   

The nucleosome is the first level of organization in chromatin structure and the basic structural unit of DNA packaging in eukaryotes. It contains 146 DNA base pairs wrapped in 1.7 left-handed super-helical turns around a histone protein core. 

Bring up the VMD Main > Graphics > Representations.

• Make sure 1ZLA.pdb is the 'Selected Molecule'. It may have auto-loaded all the chains. Delete all but the first representation.

• In your remaining Rep (atom selection = all), select the entry and change 'Coloring Method' to Chain. Change 'Drawing Method' to NewCartoon. You should see eight different colored histone monomers with double-helical DNA wrapped twice around the octameric protein complex.

Zoom in so the nucleosome particle fills the Display window (s mode to scale; r mode to rotate).

Q4:  Describe the geometry and symmetry of the protein and nucleic acid constituents in the nucleosome core particle. 

Bound to the histones in white/pink you will see a small peptide hairpin in thin cyan. Let's visualize the binding complementarity.

1.  Highlight the NewCartoon rep and change Selected Atoms to:  not chain K and protein.

2.  Click Create Rep and change its Selected Atoms to:  chain G or chain H. Change Drawing Method to  Surf (near the bottom of the list) and change Coloring Method to: ResType.

3.  Create another representation and select  chain K. Drawing Method = Licorice. Coloring Method = ResType. The peptide is hard to see: increase Bond Radius to 0.4 and Bond Resolution to 32.

4.  Create a final representation selecting  'chain I J'  drawn as NewRibbons. Coloring Method = Chain.

5.  Change the Material for the ligand and DNA (last two representations) from Opaque to BrushedMetal, converting RGB colors to CMYK for better printing reproduction.

 1ZLA.pdb - Graphical Representations:

DrawStyle   Color   Selected Atoms

NewCartoon  Chain    not chain K and protein

Surf        ResType  chain G or chain H

Licorice    ResType  chain K   Material=BrushedMetal 0.4

NewRibbons  Chain    chain I J  Material=BrushedMetal

• Surf draws a surface around the protein as a whole. The bumpy molecular surface is the solvent-accessible surface area, which is traced by running a 1.4 angstrom sphere (water molecule) around all the van der Waals radii of the protein atoms. For large proteins, Surf will be slower to load and rotate.

The small bound peptide is latency-associated nuclear antigen (LANA), a viral encoded protein in Kaposi's sarcoma-associated herpesvirus (KSHV). LANA modulates viral and cell gene expression. When KSHV infects chromosomal DNA in AIDS patients, it commonly leads to Kaposi’s sarcoma, a type of cancer. LANA is one of the few encoded proteins that is highly expressed in all latently infected tumor cells.

Coloring by residue type shows the electrostatic molecular surface of a protein.

• Basic residue sidechains are positive (blue) and can form an attractive interaction, or salt bridge, with acidic functional groups which are negative (red). Hydrophobic sidechains are nonpolar (grey). Polar sidechains (green) are hydrophilic but are also uncharged.

Q5:  Describe the ligand. Is the molecule acidic or basic? Does the histone binding pocket interact with LANA via acidic or basic surface sidechains?

Render images of the nucleosome and the binding pocket..

• Zoom out and frame the whole chromatin particle showing its spooled DNA. Select VMD Main > File > Render... > Tachyon (internal, in-memory rendering), Filename = nucleosome.tga > Start Rendering.

• Zoom in tight on the binding pocket for the LANA peptide on the histone surface (don't get so close that you clip any ligand atoms from view). Create another render virus.tga  showing the electrostatic complementarity.

Once you're happy with this visualization and your render(s), click on the 3zla.pdb entry in VMD Main window (line turns green or yellow) and select Molecule > Delete Molecule, or just close VMD..

Over time VMD can become unresponsive. For the next section, I recommend closing your current VMD session by selecting File > Quit.

Proceed to:    Part 3