MSCI815 Module 5.
3D viewers for protein/nucleic acid structure
"September 20, 1999: Version 1.0.
It took a year of work to achieve this version 1.0, yet it is far from finished. I have many plans for the Protein Explorer and will continue to develop it as fast as I am able. At present, it comprises over 10,000 lines of HTML, over 5,000 lines of javascript, and over 500 lines of RasMol/Chime script. "
This module is a survey of 3D viewers for looking at macromolecules. There are several ways to do this. The most expensive option is to buy a dedicated computer system like a Silicon Graphics Oxygen 2 system. These are found at X-ray crystallography labs and they require some learning to operate them effectively. Most researchers will not use this option, unless thay have a friendly crystallographer that will help them to use it. Another more practical method is to use 3D browser software. We will talk about two types in this session. The first is Cn3D from the NCBI. This is free software you can download and install. This has been done on the class computers so we can use the software today. If you want to get the software for your own computer go to this link.
The other option is a browser plugin called Chime (current version 2.6 SP3). Chime is a requirement for software products written by MDL Information Systems, Inc. (formerly Molecular Design Limited). Chime and the software that uses it is targeted at chemical and pharmaceutical companies that have a need to display molecules. The rest of us in Academia can also benefit by being able to see the 3D interactive graphics posted at more than 120 websites. For a list see this list. You may have to register before you can view the list, but that is free. These sites include the PBS program Nova, which has a graphic of the structure of diamond. Once you have the Chime plugin, you can use Protein Explorer. This is the successor to RasMol. Protein Explorer was written by Eric Martz and it is web based so you do not have to have it on your own computer, but that is an option if you want to make your own interactive graphics of protein structures for your own web pages. I am trying to put up some P450 3D images on my site, but I am having some problems getting it to work.
Protein Explorer began in Oct. 1998 as experimental software. The first version 1.0 was released a year later.
Taken from Eric Martz' history of Protein Explorer
The program suite has continued to undergo revisions and is now at version 1.901.
We will start with Cn3D from the NCBI website and finish with Protein Explorer.
From the NCBI homepage, scroll down the left frame to Molecular databases Sequences, structures and taxonomy. In the map click on 3D structures. Now click on Cn3D on the left frame. Cn3D is a helper application for your browser to help you view over 10,000 molecules in the NCBI's structure database called MMDB (Molecular Modeling database). These 10,000 structures have been linked to the Entrez browser and the Blast search tool so that PubMed entries, sequence entries in GenBank and blast hits are linked to their respective structures. If your protein does not have a structure, you can click on Related Sequences or Protein Neighbors to find one that does have a structure link.
The Cn3D Tutorial
The following is a abbreviated version of the Cn3D Tutorial at NCBI. For more detail please go to that site. This tutorial uses the PTEN protein for an example. The first part of the tutorial shows how to find the structure at NCBI. Before we do this please start the Cn3D program on your computers. Once that is running jump to the NCBI homepage. Drag the pulldown menu from nucleotide to structure, then type in PTEN and return. This will call up one link to the protein 1D5R. If you click on this you will be taken to the MMDB structure summary page for PTEN. There is a button marked view/save structure. Click this button. That will pull up the structure in the Cn3D window for structure. There is a second window for the sequence. These two windows are linked, so you can highlight regions of the sequence and they will light up in the structure. This is very useful for locating mutations, helices, active sites, phosphorylation sites etc.
Now you are thinking, can I open any PDB file (Brookhaven Protein Database structural coordinate file) in this Cn3D viewer? The answer is not directly. The MMDB structures were composed from the PDB files, but they have been automatically parsed and checked for errors and the structures you are seeing in Cn3D are not just renderings of the PDB file. However, all the proteins in PDB that are real and not theoretical models are in MMDB. You should be able to find a known structure and see it. Just for fun let's try CYP2C5, the only mammalian cytochrome P450 structure known. It is 1DT6 If you get this structure, this is a view of the general P450 structure you have been working on in this class. The I-helix runs from 262-295, then there is a prolin that usually kinks helices and may signal a turn. The residues 297-309 are the J-helix. Select the long helix protein sequence and see the two helices in the protein structure light up yellow. Put your cursor on the structure window and hold down the mouse button then drag the mouse. This will rotate and twist the structure. Do this until you get a good view of the two yellow helices. Under the VIEW pull down menu you can ZOOM IN. Try that about twice to get a closer look.
A note on structure codes: The PDB has four character codes that all start with a number. 1JME is an other P450 CYP102. The MMDB has kept the same codes. If you know the PDB code you can search for it in the structure section of NCBI.
Another way to find a structure is to do a PubMed search. The example used is again PTEN. Search for PTEN structure. The 29th hit on the second page has
Related Articles, Nucleotide, OMIM, Protein, Structure on the right side. Click on structure. This will take you to the MMDB entry. To find a structure by blast searching, select the PDB as the database (instead of nr or ests). Use a protein against protein search (not tblastn). The results will give you PDB codes and if you click on one of the hits you can get to the structure from the structure link on the top right part of the page.
Lets go back to the P450 image. The File menu lets you save the image as an asn.1 format file. These can be openned by Cn3D for later use. If you want to make a graphic for publication or web pages, the image can be saved as a .PNG (Portable Network Graphics) format file. PNG is an image format like .GIF, but it is better than .GIF.
The VIEW menu has the zoom controls and a reset command if you mess everything up too badly. The drawing settings menu is also here. This is a complex set of tabs and check boxes that controls the appearance of the structure. The STYLE and COLOR pull down menus on the tool bar beside VIEW are subsets of the drawing settings. Let's look at STYLE first. These options control the display of the protein backbone. The default is secondary structure that shows helix and sheets superimposed on a worm-like representation of the backbone. Neighbor replaces the worm with a rigid thicker tube and does not have the helix and sheets. Note the heme in the center of the protein. Wireframe shows all sidechains in a very thin line representation. Zoom in and find the heme iron (a transparent sphere with Fe in the middle. Note the yellow line going to the center of the iron. This is the heme thiolate ligand S(-) from cysteine in the heme signature sequence. It is one of the three amino acids that never varies in P450s.
Tubular is a very pleasing view that is a thicker version of wireframe. Heirarchy makes the sidechains wireframe and the backbone tubular. It is easier to see into the structure this way. Spacefill shows the atoms as overlapping spheres. Ball and stick makes these spheres smaller so they are connected by rods.
The color menu controls what color the protein parts are displayed in. The default is secondary structure where helices are green, sheets ar orange and coil is blue. You can change these colors under the options menu. Cycle object will color two different aligned structures different colors so they are easier to see. Cycle domain will color different proteins domains different colors. Cycle chain will color different subunits different colors. Hydrophobicity will color residues according to how hydrophobic they are. Blue is hydrophilic and red is hydrophobic. This is best used with spacefill and zoomed out to see the protein surface and look for colored patches of blue or red. Temperature gives the thermal motion factor of each atom. Red is more mobile (disordered in the crystal).
From the DRAWINGS SETTINGS MENU there are tabs marked label. This can turn on or off labels of the N-terminal, C-terminal and spacing of labels along the sequence. The font can be selected. The SHOW/HIDE tab controls what is seen. If there are mutiple subunits or multiple domains these can be turned off to be able to see other parts of the molecule better.
To label your own specific residues and color them, or change the appearance of them. You can select the residues you want by selecting them in the sequence window or double clicking on them in the drawing window. Then, from the annotate tab, you can create a name and description for the selected residues, like I-helix, then you can change the style and color and click add, then apply, to show this modified region in the display.
The NCBI has a database called VAST that has alignments of related structures. All the proteins in PDB have been parsed into domains and all the domains (about 18,000) have been aligned structurally. These structural alignments can be called up in the Cn3D viewer to see how two related sequences compare in their 3D fold. These structural alignments can be from two whole proteins or just the individual domains of proteins. Go to VAST homepage
The Cn3D viewer can show these structural alignments. Typing in the ID like 1DT6 in the box labeled Structure Summary via PDB/MMDB Code: can bring up a list of sequence neighbors and structure neighbors. This page shows structural neighbors A, A1 and A2. These refer to whole protein neighbors and individual domain neigbors. If you click on A1, bacterial P450 BM3 is listed as a neighbor and the structural alignment can be viewed by checking the first check box and clicking on view alignment. P450 alignment The two proteins are shown in different colors to aid in interpretation.
Another feature in Cn3D is mapping of related sequences onto a structure to show conservation seen in a sequence alignment. This is different than the structural alignment shown above. In this case related sequences are imported and alignmed with the master sequence. A color scheme is then used to show the most conserved residues as red and the least conserved as blue. We will not go through how to do this in detail, but it is covered in the alignment section of the Cn3D tutorial. An example of a structure showing the conservation of residues is shown here
Protein Explorer
Go to the Protein Explorer Protein Explorer. front door page and click on the Quick Start Protein Explorer link. We will be following the 1 hour quick tour.
You should have a window with Gal4 bound to DNA spinning slowly. This has a lot of red dots (water molecules) shown. You can toggle them off with the hide show water button. You can stop the rotation with the toggle spinning button. Grab the image with your cursor. Hold down the mouse and drag. The molecule will be twisted and turned to give you a view you want. Try this.
You can see two chains in different colors. These are the different subunits of the protein. The DNA is blue. Get a view down the center of the double helix of DNA. The red spheres are cadmium. Clicking on any vertex will give the residue identity in the window on the bottom left.
If you press the molecule information button and then SEQUENCES, the chains are given below. This page has 4 chains, two DNA and two protein. The two protein chains are identical so only one is shown. Note that the first 7 amino acids are not in the crystal structure. They are disordered. This is indicated by 7 dots.
On the molecule information page, one link on this page is to help/index/glossary. This can save you if you do not know how to do something, or you do not know what a term means. Show counts gives a list of the items in the structure, including disulfide bonds, number of chains, number of atoms etc.
Click on Explore More. This takes you to Quick Views. From this point on follow the Quick Tour from number 9. You may want to open it in a second window so you can flip back and forth.
