B. Comparing a structure with structures related by homology or by functionality

From DISI

B. Comparing a structure with structures related by homology or by functionality

i) Bring up the Protein Information page of structure 2nv6 as before, using Text Search. Navigate to the corresponding Ligand Information page. We will first carry out a similar binding sites search to find sites that are closely related by homology. Relibase+ uses the FASTA program to do this.

ii) Click on Similar Binding Sites Search at the top of the Ligand Information page. In the resulting Binding Site Analysis page type 2.5 in the Lowest Resolution box. Leave all else at default. Click on Submit.

iii) 21 structures should be retrieved. You can link through to the individual structures by clicking on the links pdb#### in the second column. All the structures are of the same enoyl [acyl-carrier protein] reductase from Mycobacterium tuberculosis. We can superimpose them in the same reference frame. Type 7.0 in the Radius of sphere around Ligand box. This sphere defines the atoms that will be visualised and also the backbone α-carbons that will be used for the superposition. Click on Submit. In the Superposition Analysis page that comes up, a lot of information on the differences between the reference structure (2nv6) and the other structures is presented in table form. Further information is found by following the links in the table. View all the structures in ReliView by clicking on Show in ReliView.

iv) We can look for consistent differences between sets of structures. Examine the region near the pyridylcarbonyl portion of the ligand. This ligand and related ones bring about a conformational change in the protein by pushing aside the Phe149 side chain. Where the ligand is the normal co-factor for this enzyme NADH, which lacks the pyridylcarbonyl fragment, no conformational change is seen. (Doing the following to may help to see this most clearly: Highlight 2NV6-A_1 in the Protein Explorer pane (top left of ReliView): Right-click within this pane and select colours…: Select your favourite football team’s colour: Now click on the Ligands tick box in the All Entries row to turn off all ligands; Finally click on the Ligands tick box of the 2NV6-A_1 row. You can repeat this process for other ligand protein complexes in the set)

v) Most of the ligands are closely related so we can look in other structures for the exocyclic carbamoyl group that we were considering in A ix). In 2nv6 we found this group made a very unusual NH2..NH close contact. Now if you compare all structures you can see that in many of the structures this carbamoyl group has been assigned as being flipped by 180º. You can investigate whether the inter- and intra-molecular contacts for this orientation are more reasonable than in the alternative orientation. If you like, use IsoStar to confirm your conclusions.

vi) Oxidoreductase enzymes are common. If we wish to target this particular enzyme we may also want to analyse the binding sites of other similar oxidoreductases, especially those in Homo sapiens. Return to the Binding Site Analysis page. This time set a homology maximum of 95 and a homology minimum of 25. We will also lower the resolution limit to 2.0Å. Thirty nine (39) sequences should be returned, for which the highest homology is 33.2%. Link through to some of these structures to see where they originate from. A number of organisms should be represented, including E. Coli, and Thermus thermophilus. A carbonyl reductase from Homo sapiens is represented by 2hrb. We can superimpose the active sites as before. Overall the shape of binding site and position of NADH co-factor is very consistent over all structures. Note however the superpositions aren’t always perfect as we are trying to superimpose structures with low homology. You can see this by comparing the superposition of 2nv6 and 2hrb in ReliView.

vii) We could have relaxed both the minimum homology threshold and the resolution threshold and collected many more related structures (you can try it to see). Clearly analysing the similarities and differences of these structures to 2nv6 by hand is not an easy task. What we really want to do is to focus on the area of the binding site that we want to target in a drug–design program, and identify those other binding sites which have a similar shape and binding properties in that same region, regardless of homology. We can do this in Relibase+ using the CavBase module.

viii) CavBase allows users to compare portions of the active site on function and shape, not homology. CavBase searches need to be carefully prepared and they can take a long time to run. For this workshop a CavBase search has been previously prepared and run. Click on Home at the top level and then select Cavity Similarity Results.

ix) Select search_pdb2nv6.1_17_3. This loads a search carried out on a portion of the 2nv6 active site. The top hundred solutions have been saved. The top scoring cavity is in 1cyd. Notice it only has 18% homology with 2nv6. Click on pdb1cyd.12 to view details of the two cavities. A viewer, Cavity Viewer, that displays an overlay of the two cavities, will also be brought up. The surfaces of the cavities, where they match, and the pseudocentres which have been matched up in both cavities, are displayed. You can view the region of the 2nv6 active site that the search has been based on by clicking on the Search Setup tab at the top of the Cavity Viewer pane.

x) 1cyd is a mouse protein. The sixth entry in the ranked similarity list is 2ag5 a human oxidoreductase (DHRS6). This appears to naturally have an open region beyond the NADH group, which is being identified by CavBase as being similar to 2nv6. Anyone considering designing drugs to target this portion of the tuberculosis bacterium protein might want to consider possible side effects brought about by inhibition of DHRS6 and related proteins.

This ends Part1B - Part2 - Back to Start