[gmx-users] To David van der Spoel about " Modeling of enzyme-substrate complexes for the metalloproteases MMP-3, ADAM-9 and ADAM-10 "

[Иванов Миша]

Dear David vad der Spoel,

Over 10 years our laboratory has been exploring the prokaryotic Zn-dependent carboxypeptidase T (CPT) from Thermoactinomyces vulgaris homologous to the well known eukaryotic carboxypeptidases A (CPA) and B (CPB). At present, we are studying  The Principles of Dual Substrate Specificity of Carboxypeptidase T" mainly by protein engineering. However we want to enlarge our studies and to perform molecular dynamics study of peptide bound to CPA, CPB and CPT. We performed a short (1ns) modeling of CPA with tetrapeptide FFVF (the structure of a resembling complex is available in PDB-bank) on the Gromacs force field, in Gromacs 3.1.2 (under Windows). The mdp options were taken from your publication "Modeling of enzyme-substrate complexes for the metalloproteases MMP-3, ADAM-9 and ADAM-10". And we found, that the enzyme preserves its overall structure, however the position of bound substrate changed drastically. Several important ionic and hydrophobic contacts were broken. As we are n
 ovice in molecular modeling we don't deny the possibility, that we've made some faults in this modeling. 

We read your publication  Modeling of enzyme-substrate complexes for the metalloproteases MMP-3, ADAM-9 and ADAM-10" with a great pleasure. In your study Gromacs reproduced correctly the structure of the enzymes and their substrates. That is why we decided to post our questions to you.
1)	When you introduced Zn and Ca ions and used the charge +2 for them, did you changed the charges of their ligands to make the system Zn-ligands electro neutral? If the histidines were treated as HISA and their charges were derived from the force field Gromos 43a1, that you used?
2)	How did you choose the protonation state of Asp, Glu, Lys, Arg, and His? Were the protonation state for all these residues determined automatically by pdb2gmx, or you made it by an another approach?
3)	Positions of the two N-terminal Phe of the substrate in the crystal structures of CPA is stabilized by  edge-to-face" interactions with protein Tyrosines (interaction of the slightly positively charged edge of one aromatic ring with the slightly negatively charged face of a second aromatic ring), beyond the van der Waals interactions. We would like to know, whether a  simple" molecular dynamics can predict such interactions or we have to make several improvements?
4)	If it possible, we would like to ask you to send us the Gromacs files of modeling of MMP-3 to enable us to obtain a molecular dynamics trajectory and carefully analyze the substrate behavior in the enzyme binding cleft. This will also give us a possibility to verify our mdp-settings and distance restraints settings. Namely we would like to obtain the following files:
a)	md.mdp (mdp file of modeling without positional restraints)
b)	b4md.gro (gro file of the MMP-3 with substrate in water after molecular dynamics with positional restraints)
c)	mmp3.top (the topology file (files) of the protein with substrate, counter ions, and with Zn-ligand distance restraints.

Sincerely yours,
Andrew Grishin, PhD-student
V.M. Stepanov Laboratory of Protein Chemistry
Institute of Genetics and Selection of Industrial Microorganisms,
Moscow, Russian Federation.