[gmx-users] Adding a substrate into protein-lipid bilayer system

[Justin Lemkul]

On 6/1/15 12:13 PM, MPI wrote:
>>>
>>>>> I wonder when it is appropriate to add NH3 into the system and how to
>>>>> setup the control of temperature coupling in thermostats.
>>>>> For example,
>>>>>
>>>>> 1.  add NH3 first with the protein and  then embed protein-NH3  into the lipid
>>>>> or
>>>>> 2.  embed the protein into  lipid with an equilibration,  followed by
>>>>> adding NH3 into the system
>>>
>>>
>>>> If you want NH3 to pass through the channel, I'd say (2), more or less, but
>>>> really you should add NH3 as a cosolvent as part of the equilibration.  Add the
>>>> NH3 in the box, add water, and equilibrate.
>>>
>>>        If choosing (2), what is the concern ?
>>> Why adding NH3 as part of water_and_ions after protein-lipid  is
>>> equilibrated already ?
>>> Choosing (2), I have to go through equilibration twice.
>>>
>> No you don't.  You add NH3 into the box, then solvate with water.  See
>> http://www.gromacs.org/Documentation/How-tos/Mixed_Solvents
>> I'm saying that (2) is in general more logical for your purpose (as I understand
>> it) but you don't do two separate equilibrations.  Moreover, it is unlikely that
>> you'll be able to find space in an equilibrated layer of water to insert NH3
>> molecules.
>
> Hi Justin,
>
>    Thank you very much for the details.
>
>
>>> Also,  if the substrate is a very small ligand, is the protocol the
>>> same ?  Namely, adding a new substrate AFTER protein-lipid is
>>> equilibrated.  Then perform another equilibration for the substrate +
>>> protein-lipid.
>>>
>> This is context-dependent.  If you want to simulate a protein-ligand complex in
>> a bilayer, you add the ligand first into the protein (crystal structure,
>> docking, etc).  If you want to observe partitioning from bulk solvent of the
>> ligand into the protein's binding site, then it's the same.  Put the required
>> number of ligands into the box where the solvent will be, then add the solvent
>> (water).
>
> If I understand this correctly, you actually suggested adding a ligand
> BEFORE adding the solvent and BEFORE the protein is embedded into
> lipid bilayer. Is this correct ?
>
> I am concerned about the sequential  order of adding a substrate and
> embedding a protein into a bilayer.
>
> Which one goes first ?
>

1. Protein into bilayer
2. Add ligands/cosolvent
3. Add water
4. Add ions

>
>
>>>
>>>
>>>>> Also,  if the substrate has a charge, which order is preferred ?
>>>
>>>
>>>> Charge shouldn't affect the protocol.
>>>
>>> Did you mean  when adding a new substrate, eg, NH4+ or Ca2+  into a
>>> protein-lipid system,  one has to remove all ions  and water in the
>>> equilibrated protein-lipid system,  put back ions and water again and
>>> finally equilibrate the new system right after ?
>>>
>> Or just build it in sequence from scratch without having to remove water and
>> ions.  That's what I'm suggesting above.  The cosolvent/ligand/whatever is
>> always added before the water.  It's easier and faster in practical terms,
>
> Before the water, it means "you add ligand at the very beginning
> BEFORE protein is embedded into bilayer, correct ? "
>

No, see above.

> If so, what about the temperature coupling ?
>
>
> (1) tc-grps         = Protein DPPC  water_and_ions_and_LIG
>
> or
>
> (2) tc-grps         = Protein_LIG  DPPC  water_and_ions
>

This is tricky.  You have a freely diffusing species that will pass into/through 
the protein.  So at some times it may be part of the solvent, sometimes it might 
be bound to the protein.  I can't predict what the differences might be here. 
Strictly speaking, multiple thermostats aren't correct, but are common, 
especially in such heterogeneous systems.

>
> It seems to me if LIG = NH3/NH4,  then (1) is preferred  but if LIG is
> a small drug "MOVING" along the channel,  dose (2) make sense ?
> or  (3) tc-grps         = Protein DPPC LIG  water_and_ions ?
>
>
>> because if you have to remove solvent, then what was the point of the initial
>> equilibration?
>
> The initial equilibration is for inserting protein into the bilayer.
> One can add any substrate afterwards without  worrying about the
> process of inserting a protein into bilayer. Dose it make sense to you
> ?
>

I suppose you can do this if you need to.  Depends on how you're inserting the 
protein, I suppose, and how much equilibration you intend to do before 
introducing your ligands/cosolvents.

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==================================================