Finding the Structure of a Protein

Suppose that we are interested in the biosynthesis of aminocyclopropanes and would like to find structural information on important active site residues in any available aminocyclopropane synthases. To begin, we would go to the Structure main page and enter “aminocyclopropane synthase” in the Search box. Pressing Enter displays a short list of structures, one of which is 1B8G, 1-aminocyclopropane-1-carboxylate synthase. Perhaps we would like to know the species from which this protein was derived. Selecting the Taxonomy link to the right shows that this protein was derived from Malux x domestica, or the common apple tree. Going back to the Entrez results page and selecting the PDB code (1B8G) opens the Structure Summary page for this record. The species is again displayed on this page, along with a link to the Journal of Molecular Biology article describing how the structure was determined. We immediately see from this page that this protein appears as a dimer in the structure, with each chain having three 3D domains, as identified by VAST. In addition, CD-Search has identified an “aminotran_1_2” CD in each chain. Now we are ready to view the 3D structure.

Viewing the 3D Structure

Once we have found the Structure Summary page, viewing the 3D structure is straightforward. To view the structure in Cn3D, we simply select the View 3D Structure button. The default view is to show helices in green, strands in brown, and loops in blue. This color scheme is also reflected in the Sequence/Alignment Viewer.

Locating an Active Site

Upon inspecting the structure, we immediately notice that a small molecule is bound to the protein, likely at the active site of the enzyme. How do we find out what that molecule is? One easy way is to return to the Structure Summary page and select the link to the PDB code, which takes us to the PDB Structure Explorer page for 1B8G. Quickly, we see that pyridoxal-5′-phosphate (PLP) is a HET group, or heterogen, in the structure. Our interest piqued, we would now like to know more about the structural domain containing the active site. Returning to Cn3D, we manipulate the structure so that PLP is easily visible and then use the mouse to double-click on any PLP atom. The molecule becomes selected and turns yellow. Now from the Show/Hide menu, we choose Select by distance and Residues only and enter 5 Angstroms for a search radius. Scanning the Sequence/Alignment Viewer, we see that seven residues are now highlighted: 117-119, 230, 268, 270, and 279. Glancing at the 3D Domain display in the Structure Summary page, we note that all of these residues lie in domain 3. We now focus our attention on this domain.

Viewing Structure Neighbors of a 3D Domain

Given that this enzyme is a dimer, we arbitrarily choose domain 3 in chain A, the accession of which is thus 1B8GA3. By clicking on the 3D Domain bar at a point within domain 3, we are taken to the VAST Structure Neighbors page for this domain, where we find nearly 200 structure neighbors.

Restricting the Search by Taxonomy

Perhaps we would now like to identify some of the most evolutionarily distant structure neighbors of domain 1B8GA3 as a means of finding conserved residues that may be associated with its binding and/or catalytic function. One powerful way of doing this is to choose structure neighbors from phylogenetically distant organisms. We therefore need to combine our present search with a Taxonomy search. Given that 1B8G is derived from the superkingdom Eukaryota, we would like to find structure neighbors in other superkingdom taxa, such as Eubacteria and Archaea. Returning to the Structure Summary page, select the 3D Domains link in the graphic display to open the list of 3D Domains in Entrez. Finding 1B8GA3 in the list, selecting the Related 3D Domains link shows a list of all the structure neighbors of this domain. From this page, we select Preview/Index, which shows our recent queries. Suppose our set of related 3D Domains is #5. We then perform two searches:

    1. #5 AND “Archaea”[Organism]
    2. #5 AND “Eubacteria”[Organism]

Looking at the Archaea results, we find among them 1DJUA3, a domain from an aromatic aminotransferase from Pyrococcus horikoshii. Concerning the Eubacteria results, we find among the several hundred matching domains 3TATA2, a tyrosine aminotransferase from Escherichia coli.

Viewing a 3D Superposition of Active Sites

Returning to the VAST Structure Neighbors page for 1B8GA3, we want to select 1DJUA3 and 3TATA2 to display in a structural alignment. One way to do this is to enter these two Accession numbers in the Find box and press Find. We now see only these two neighbors, and we can select View 3D Structure to launch Cn3D.

Cn3D again displays the aligned residues in red, and we can highlight these further by selecting Show aligned residues from the Show/Hide menu. The excellent agreement between both the active site structures and the conformations of the bound ligands is readily apparent. Furthermore, by selecting Style/Coloring Shortcuts/Sequence Conservation/Variety, we can easily see that the most highly conserved residues are concentrated near the binding site (Figure 6).