Using Structural Models and Other Data

Overview

The purpose of these experiments is to familiarize you with the techniques of modeling, to build a familiarity with the process, and an understanding of its strengths and limitations. The goal of these experiments is for you to be able to critically examine a modeling study, and to be able to initiate a modeling study on a problem of your own. While the program AMMP will be used in the experiments, learning to use this program is not a major point of the lessons. You will not be tested on AMMP. There are many programs available for modeling, but despite the advertising copy they are very similar in choice of algorithms and quality of results.

The focus of this lesson is to examine how to use multiple structures to extract motifs. We will look at two related problems where structures and databases can be used together to extract information which is not obvious from either on its own.

The experiments

In order to perform these experiments you will need a workstation running Linux and supporting MESA or OGL or an INTEL PC running Windows 95, 98 or NT.

The first experiment is to examine how multiple ligand structures can be used to elucidate the important features of a binding site. While this will be done in a simple manner, somewhat more sophisticated techniques using correlation analysis form a basis for rational design (or at least rational choice) for developing novel lead compounds.

The second experiment is the converse of the first experiment. Here differences in protein sequence will be correlated with differences in inhibition.

Experiment 1, Trypsin inhibitors

The structure and inhibition constants for a series of mono-amine trypsin inhibitors was determined as part of a model study for drug design. Can we use this structural data to understand the relative effectiveness of the inhibitors? Retreive the files and find out.
Compound Ki File
Phenyl Ethylamine11.0mMpea.ammp
Amino-methyl cyclohexane1.17mMamc.ammp
Fluoro-benzylamine0.43mMfbs.ammp
Phenyl Propylamine32.5mMppa.ammp
Phenyl butylamine20.0mMpba.ammp
Tranylcypromine13.3mMpta.ammp

The trypsin core structure is 3ptn.ammp. You will need the scripts minimize.ammp and analyze.ammp.

  1. CONFMA analysis Start AMMP and read in the inhibitors. ( use either "read <file>;" (i.e. read amc.ammp;) or the menu File|input ammp file). The inhibitors are aligned as they bind to trypsin. Certain parts of the compounds will tend to be aligned. While we could be sophisticated and use a property-based alignment, visual analysis will suggest that certain groups are important. Which ones? (write them down because you'll need them for the next steps). Do the molecules cluster with respect to the observed Ki values? Analysis like this is applied to a structural database of compounds with known activity to suggest what are the important functional groups. This is the first step in a rational choice of a new compound to make.
  2. In the Protein For each inhibitor start AMMP and load trypsin. and the inhibitor. (It is easier to load the inhibitor first, turn off the autocenter feature (in the draw window controls|toggle autocenter) and then load trypsin. That way you're already centered on the inhibitor. Otherwise center on residue 195 (the catalytic serine)) (you don't have to do all the inhibitors at once). Using the groups you selected as being important in step 1 look at the distances to different residues in the protein. How well do the inhibitors fit? (You can read pairs of inhibitors in to compare them) Is there any correlation between distances and Ki?
  3. Energetic Analysis Energy calculations can be used to aid in understanding how molecular recognition works. For several of the inhibitors (time permitting), use the scripts minimize.ammp and analyze.ammp to get an estimate of the interaction energy. Only read one inhibitor and trypsin together; it will not work if you have several inhibitors read in at the same time. Read analyze.ammp, note the interaction energy (The interaction energy is given by the number following Vnonbon total external), then read analyze.ammp again. Pay attention to how much the atoms move during minimization ( the show tethers command will do this). Does either estimate of interaction energy correlate with the Ki?

The data for this experiment are presented in:
Harrison R.W. and Kurinov I.V. (1994) "Prediction of Novel Serine Protease Inhibitors" Nature Structural Biology 1(10) 735-743.

Experiment 2, Drug resistant mutations in HIV protease

HIV protease is an essential enzyme in the lifecycle of the virus and many effective drugs have been developed to inhibit it. However this provides a selective pressure for the development of resistant mutations. In this experiment we examine resistant mutations (v32i,v82a,v82f) and sequinavir. (These mutations are not those which arise in vivo but have measureable differences in Ki).
V82F model structure
The mutations were modeled with the homology techniques discribed in the previous lessons and are in the files: v32isaq.ammp, v82asaq.ammp, and v82fsaq.ammp. The model for the wild type is found in hivsaq.ammp.

Load each model and look at the mutation. It is important to look at the mutation in both monomers (e.g. residue 82 and 182). Remember the kinds of differences seen with the trypsin inhibitors. Are there similar structural differences here? Based on the model can you derive a reasonable hypothesis for its mode of resistance? How would you construct an experiment to test the hypothesis?


The data for this experiment can be found in:
Weber, I.T. and Harrison, R.W.(1999) "Molecular Mechanics Analysis of Drug Resistant Mutants of HIV Protease" Protein Engineering., in press.