Biology Reference
In-Depth Information
first equilibrate the membrane and solvent around the protein prior to the equilibration
of the whole system. The standard procedure is based on the application of positional
restraints of 10 kcal mol
1
˚
2
to all heavy atoms of the receptor to progressively re-
duce such restraints to zero until the complete release of the protein. Such simulation
conditions allow the closure of small holes at the lipid-protein interface, frequently
generated at this step while the system is adopting an adequate box size. Nowadays,
the available computational resources enable to equilibrate GPCR-membrane systems
during 20-100 ns, mostly depending on the complexity of the membrane environment.
The use of coupled pressure baths in NPT simulations puts a limit to the speed of
the simulation as the integration time step cannot be higher than 2 fs in such
simulations. Nevertheless, once the system is equilibrated, we can release the pressure
coupling and switch to the canonical ensemble (NVT, number of particles (
N
), volume
(
V
), and pressure (
T
) remain constant) where larger integration time steps (i.e., 4 fs)
can be used and, as a result, the system can be simulated for longer times.
Studying the dynamic behavior of GPCRs in such native-like membrane systems
can give valuable and meaningful insights into GPCR functioning. However, as for
any other protein-membrane system, the statistical significance of a particular result
or event obtained through MD simulations of GPCRs is based on the number of times
one is able to reproduce it. In this line, production runs need to be replicated as many
times as possible to support the relevance of the observed dynamic events. Since
most of the systems are coupled to temperature bath (i.e., NVT and NPT simula-
tions), we can take advantage of the randomness introduced by the “friction” term
of certain thermostats and simply restart the production to yield different replicates.
Ideally, any property studied in the reference system (i.e., the first system simulated)
should be calculated for all the replicated production runs.
4.2.3
Studying GPCR-lipid interactions from simulations:
quantification of membrane remodeling and residual mismatch
The lipid environment plays an important role in the membrane GPCR interactions.
The membrane composition modulates both function and conformation of mem-
brane proteins (
Engelman, 2005; Phillips, Ursell, Wiggins, & Sens, 2009
). Despite
the relevance of membrane-mediated GPCR modulation, many aspects remain still
unclear at a molecular level. Specific lipid-receptor interactions and unspecific in-
teractions that are mediated via bulk properties of the membrane (e.g., membrane
thickness and elasticity) have been reported to modulate receptor functioning.
One particular phenomenon of an unspecific membrane effect is the hydrophobic
mismatch, where the shape of the membrane changes by getting deformed according
to, among other factors, the type and conformation of the protein embedded in this
membrane (
Huang, 1986; Nielsen, Goulian, & Andersen, 1998
). Such adjustment
between membrane and proteins can be explained in terms of the energy cost due
to the aforementioned hydrophobic mismatch (
Lundbaek, Collingwood,
Ing´ lfsson, Kapoor, & Andersen, 2010; Mondal et al., 2011
). Thus, two different
contributions affect this mismatch: on the one hand, the remodeling of the membrane
