Print

Print
RE: Refining two different covalent species at the same position.
Stephen,
 
If the first molecule is given an altcode (e.g. 'A'), then it will (well, it *should*) interact normally with the rest of your protein. Set its occupancy as you belive it should be set.
 
Then give the second molecule altcode 'B'; it will NOT interact w/ A, but will interact w/ the protein. Also give it an occupancy (presumably = 1 - occ-A).
 
You *may* need to fake out the refinement program (e.g., refmac) by also giving a *dummy* first molecule in your coordinates (xyz can be identical), but with altcode 'B', occ = 0.0. The same holds true for 2nd molecule (but w/ altcode 'A', occ = 0.0).
 
Then refine. This should work, I think, even given your somewhat more complicated covalent/noncovalent situtation.
 
If it doesn't, try SHELX (e.g., I used it to refine this structure, http://www.rcsb.org/pdb/explore.do?structureId=1KMV, where DMSO was present at low occupancy instead of [and overlapping] NADPH ).
 
Dave

David Borhani, Ph.D.
D. E. Shaw Research, LLC
120 West Forty-Fifth Street, 39th Floor
New York, NY 10036
[log in to unmask]
212-478-0698
http://www.deshawresearch.com


From: CCP4 bulletin board [mailto:[log in to unmask]] On Behalf Of Stephen Connelly
Sent: Tuesday, January 13, 2009 4:44 PM
To: [log in to unmask]
Subject: Re: [ccp4bb] Refining two different covalent species at the same position.

Calling all Crystallographers, Please Help!

RE: Refining two different covalent species at the same position.

I have a protein that is a homo-tetramer ABCD with two symmetrical active sites sitting on a two-fold axis that runs through the AB/CD interface (space group P21212). We developed an inhibitor that would covalently bind to a Lysine residue at the opening of each of the 2 active sites. The thing is with this symmetrical active site there are two possible binding options in each active site to bind to i.e., bind to either Lys15 (A) or Lys15' (C) sitting opposite from each other across this two-fold axis.

I collected a dataset at 1.35A and have refined everything except the ligand to create an omit map (using aniostropic B-factors and using H) in Refmac5, it looks great. The issue is now is I have density at Lys15 and Lys15' for both a covalently bound Lys residue and the native unbound Lysine residue which arises from the two possible binding orientations in the symmetrical active site.

Does anyone know how to build and refine two different species in the same position? for example in mine I would be refining at position 15, both a modified ligand/Lysine and just the normal LYS residue from the Refmac libraries?

I have tried in Coot to build in an alternative conformation at this position and it generates another ligand/Lysine residue. I then edited the PDB file to remove all the extra atoms to the terminal NZ to creat a Lysine residue and name this LYS (as per the refmac dictionaries) setting the occupancy to 0.5 for each. I then refined this using the ligand/Lysine cif library file made in CCP4 for the covalent species hoping Refmac will recognize the the unbound conformation labeled LYS as a 'LYS' residue from the library files. I still have issues in Coot with regularizing it or in real space refinement where it will not allow it to move or refine? Its as if it is now not recognized as a 'LYS' residue? All sorts of funny things happen after a round of refinement including the changing of occupancies, but infuriatingly it does not seem place this into the density?

Is it even possible to refine two different species in the same position? I guess that with most covalent inhibitors they appear at 100% occupancy and this is not a problem using an imported dictionary cif file to refine this. The problem is due to the two possible binding conformations in a symmetrical active site I am left with an effective 50% occupancy between the bound/unbound states. 

If you think you could help please let me know, Steve at [log in to unmask] <mailto:[log in to unmask]>

Many thanks

Steve