[gmx-users] Difficulty building a topology for a synthetic branched PEG-peptide molecule

[Pablo Englebienne]

Hi all,

I am trying to build a topology for a synthetic molecule that consists 
of peptides connected by oligoethyleneglycol (I'll call it PEG) linkers 
terminated with an amine and a carboxylic acid:

-NH2-CH2-[CH2-O-CH2]n-CH2-C(=O)-

The system looks like this:

N-(PEG)-C-N-(peptide2)-Lys(C-ter)-NZ-C-(PEG)-N

So:
* the C-terminus of a PEG linker is attached to the N terminus of the 
peptide
* the terminal Lys on the peptide is attached to the C-terminus of a PEG 
linker

I was able to successfully build a topology for this molecule by:
* defining appropriate residues (for the PEG chains and the Lys with a 
PEG attached on the NZ) in a local copy of the forcefield file, adding 
the residues' topologies to aminoacids.rtp
* using the specbond.dat file to define the bond between the NZ in the 
Lys and the PEG linker
* adding the residues to residuetypes.dat with a "Protein" type
* calling pdb2gmx with the -ter option to assign the protonation states

Now, I need to extend the topology to a molecule like this one:

N-(peptide1)-C-N-(PEG)-C-N-(peptide2)-Lys(C-ter)-NZ-C-(PEG)-N-C-(peptide3)-N

The difficulty with this molecule is that it has 2 N-termini and a 
single C-terminus (in the Lys with the PEG attached to the NZ 
sidechain). pdb2gmx recognizes the whole molecule as a peptide, but 
treats the last residue as a C-terminus, when it actually is an N-terminus.

I found in the description for specbond.dat 
(http://www.gromacs.org/Documentation/File_Formats/specbond.dat) that 
for a branched peptide the "-chainsep" option of pdb2gmx can be used, so 
I started to work on that. I split the molecules in 2 chains like this:

N-(peptide1)-C-N-(PEG)-C-N-(peptide2)-Lys(C-ter)-NZ- || 
-C-(PEG)-N-C-(peptide3)-N

I reversed the order of the residues in the second chain, so that the 
residues are in N-to-C order. With this, pdb2gmx recognizes the proper 
termini when called as:

$ pdb2gmx -f substrate_edited-reversed.pdb -ter -chainsep interactive

I tried setting the protonation of the termini as charged for the "real" 
termini and None for the artificial one (the one that should be handled 
by specbond.dat), but I get an error message:

---[pdb2gmx output]---
[...]
Splitting PDB chains based on TER records or changing chain id.
Merge chain ending with residue LYSS27 (chain id ' ', atom 71 NZ) with
chain starting with residue GLU28 (chain id 'p', atom 308 OE2)? [n/y]
n

Merged 1 chains into one molecule definition

There are 2 chains and 0 blocks of water and 47 residues with 303 atoms

   chain  #res #atoms
   1 ' '    27    198
   2 'p'    17    105
[...]
Identified residue ARG1 as a starting terminus.
Identified residue LYSS27 as a ending terminus.
9 out of 9 lines of specbond.dat converted successfully
Special Atom Distance matrix:
                   PEA118
                   CG1125
   LYSS27   NZ198   2.762
Select start terminus type for ARG-1
  0: NH3+
  1: NH2
  2: None
0
Start terminus ARG-1: NH3+
Select end terminus type for LYSS-27
  0: COO-
  1: COOH
  2: None
0
End terminus LYSS-27: COO-
[...]
Identified residue GLU28 as a starting terminus.
Identified residue PEA47 as a ending terminus.
9 out of 9 lines of specbond.dat converted successfully
Select start terminus type for GLU-28
  0: NH3+
  1: NH2
  2: None
0
Start terminus GLU-28: NH3+
Select end terminus type for PEA-47
  0: COO-
  1: COOH
  2: None
2
End terminus PEA-47: None
[...]
-------------------------------------------------------
Program pdb2gmx, VERSION 4.5.4
Source code file: pdb2top.c, line: 1035

Fatal error:
There is a dangling bond at at least one of the terminal ends. Select a 
proper terminal entry.
For more information and tips for troubleshooting, please check the GROMACS
website at http://www.gromacs.org/Documentation/Errors
-------------------------------------------------------
---[pdb2gmx output]---

I think that the problem might stem from the fact that the C-terminus in 
the second chain is not a real peptide; I changed residuetypes.dat to 
have the PEG residues as "Other", which causes pdb2gmx to recognize the 
last aminoacid as a C-terminus, but treating it as "None" yields the 
same error. Unfortunately, there is nothing about this error in 
http://www.gromacs.org/Documentation/Errors .

Any suggestions on how to make this work will be greatly appreciated!

-- 

Pablo Englebienne, PhD
Dept. of Biomedical Engineering
Dept. of Chemistry and Chemical Engineering
Institute for Complex Molecular Systems (ICMS)
Eindhoven University of Technology, TU/e
PO Box 513, HG -1.26
5600 MB Eindhoven, The Netherlands
Tel +31 40 247 5349