You are here

How to define a theozyme from the .pdb crystal structure without ligand?

2 posts / 0 new
Last post
How to define a theozyme from the .pdb crystal structure without ligand?

I  want to graft a coenzyme binding sites to a scaffold,  so the first thing is to define a theozyme. However, the exsiting .PDB structures  binding the coenzyme are not accessible, what I can download is only protein crystal structure without ligand. I that way, what should i do? Should i DOCK the ligand to the protein at first? could anyone explain me the procedure? Thanks a lot!

Post Situation: 
Mon, 2019-05-20 21:18

The theozyme defines the ideal geometry you want to use in your protein-small molecule interaction.

This geometry can come from multiple locations. In the initial concept (and where the term "theozyme" comes from), the geometry is derived from QM calculations of a transition state and what the ideal geometry of interacting residues would be to promote that particular geometry. (See papers by Kendall Houk for more details on how this works.)

What's been found, though, is that you don't necessarily need QM calculations to come up with a decent theozyme for Rosetta, particularly if you're looking for just binding and not catalysis. You can mock this up from other sources. As you indicate, one of the prime sources is to look at an existing protein which binds to this small molecule, and copy those interactions.

But you don't necessarily need a full binding interaction. You can make a hypothetical binding, by looking at the structure of the ligand and saying things like "A good binder will probably have a hydrogen bond to there, and there's a pi-pi interaction there, and another hydrogen bond there ..." You can then take these idealized interactions, put them together with what the typical geometries are for hydrogen bonds of this type and pi-pi interactions of that type, and create a theozyme based on your desired interaction geometry. (Even if such a geometry doesn't already exist in a native protein.)

Keep in mind, though, that not all geometries can be supported by all backbone scaffolds. Often times it's helpful to come up with a set of potential interaction geometries, and then try each of them in turn. Often you might want to only impose a subset of interactions from your full set of desired interactions. (That is, if your ideal interaction would have six interactions, try theozymes with only 5 of those interactions, or only 4 or 3. While the full set of six might not be realizable simultaneously, some significant subset might be accessible.)

Tue, 2019-07-09 12:31