[gmx-users] E. coli's Inner Membrane Lipid Selection

Thomas Piggot t.piggot at soton.ac.uk
Sat Feb 18 23:34:44 CET 2017


For 1), perhaps but it depends upon what you are doing and what you are 
wanting to look at. Force field choice is (IMHO) likely to be more 
important (and CHARMM36 should be a good choice as long as you get all 
your mdp settings, etc., correct).

For 2), the composition looks fairly reasonable bar a couple of points. 
I'm not aware of PS normally being a lipid found in E. coli, so I 
wouldn't have this. I'm not completely positive what the Y tail is in 
your/the CHARMM-GUI acronym (I think it is probably palmitoleoyl) but I 
believe cardiolipin should have a similar tail composition to PG, as in 
E. coli cardiolipin is synthesised directly from a combination of two 
PG's (unlike in mitochondria, where it is made de novo IIRC). So 
something more like PYPY for the four cardiolipin tails would likely 
give a better representation of the real system. To be honest, how 
complex you go with the membrane will likely depend upon what you want 
to do. A simple 75% POPE, 25% POPG is a fairly accurate representation 
of the membrane. I know there isn't a huge amount of oleoyl (18:1 
delta9) tails in all the available E. coli experimental data, but oleoyl 
is a nice mix of the two most commonly found sn-2 chains: palmitoleoyl 
(16:1 delta 9) and cis-vaccenyl (18:1 delta 11). That said, if you want 
as accurate as possible, you can have all sorts of different types of 
lipid in there, with different tail types. At some stages in the cell 
cycle there are high amounts of cyclic rings in the tails instead of 
double-bonds, for example. Plus the ratio of PG:cardiolipin changes 
during the cell cycle too, as cardiolipin is synthesised and broken down 
(IIRC). Does, this matter? Would this impact upon your simulations (as 
per your question 1)? These really are questions for you to address. In 
my view, probably not a huge amount, but you can't really tell for sure 
in your case without extensive testing (e.g lots of different repeat 
simulations in each type of membrane, long enough to ensure convergence 
and a statistical comparison of the results you get for what you 
consider to be important properties that you want to compare). A lot of 
work, that probably isn't worth it.

So, I guess my answer for both of these, is that it really depends upon 
what you are wanting to study. For 1), it is hard to tell without 
extensive testing for your example. For 2), how complex you go with your 
composition is also up to you and what you are wanting to look at. At 
the one extreme, a simple PC membrane could well be good enough, for 
example, if the protein has been shown to experimentally behave fine in 
this type of membrane. If you're not really sure, I would personally 
probably go with the simple 3:1 POPE:POPG mixture as a simple yet 
reasonable representation of the membrane. One step further up in terms 
of complexity would be to include cardiolipin and have some more 
realistic tails (1-palmitoyl 2-palmitoleoly and 1-palmitoyl 
2-cis-vaccenyl plus perhaps some cyclic ringed tails in there too 
depending if there is a certain point in the cell-cycle you are wanting 
to study).



On 18/02/17 21:12, Jonathan Saboury wrote:
> Hello all,
> I'm trying to perform MD on a membrane protein AcrB. I am using charmm-gui
> to build the membrane but having difficulty deciding on membrane
> constituents.
> AcrB is a protein in E. coli's inner membrane: http://www.nature.com/nature/
> journal/v419/n6907/images/nature01050-f5.2.jpg
> Based on the below literature, I have chosen the following membrane: PYPE
> 70%, PYPG 14%, DPPS 8%, TYCL2 8%.
> Characterization of the Escherichia coli membrane structure and function
> during fedbatch cultivation: https://www.ncbi.nlm.nih.gov/
> pmc/articles/PMC514524/pdf/1475-2859-3-9.pdf
> My two main questions are:
> 1.) Does lipid choice affect MD simulation much? I'm using charmm36
> 2.) Does my lipid selection make sense? If not, why and what would you use?
> Thank you for your time!! :-)

Dr Thomas Piggot
Visiting Fellow
University of Southampton, UK.

More information about the gromacs.org_gmx-users mailing list