Simulating G Protein-Coupled Receptors in Native-Like Membranes: From Monomers to Oligomers - Methods in Cell Biology - page 75

contacts between membrane and protein. Here, the authors assess the residual mis-
match energy by AG res :

N TM X

N TM
i¼ 1 s res SA res ;i
where N TM is the number of TM segments in the protein, SA res, i is the exposed surface
area in each TM segment, and

AG res ¼ AG res ;i

i¼ 1

s res is a constant value, namely, 0.028 kcal mol 1 ˚ 2
( Ben-Tal, Ben-Shaul, Nicholls, & Honig, 1996; Choe et al., 2008 ), describing the
surface transfer energy of peptide residues between hydrophobic and polar
environments.
Frequently, the assessment of the SA res, i involved in unfavorable hydrophobic-
hydrophilic contacts is performed by means of solvent-accessible surface area
(SASA) calculations. These calculations can be carried out by a variety of available
software such as the NACCESS software ( http://www.bioinf.manchester.ac.uk/
naccess/ ) used by Mondal et al. (2011) in their recent studies of GPCRs. Herein,
two types of unfavorable contacts are considered as contributors for the residual en-
ergy penalty, namely, hydrophilic residues facing the hydrophobic part of the bilayer
and hydrophobic residues facing the hydrophilic part of the bilayer or water. To iden-
tify these contacts, Mondal et al. compare the hydrophobic length of each helical
segment along the z -axis with that of the surrounding membrane. It is important
to note that SA res, i is calculated from the SASA value of the whole residue. However,
such residue-wise consideration is prone to bias the SASA calculation by including
favorable contacts such as nonpolar atoms of polar residues in contact with the hy-
drophobic part of the bilayer. Thus, an atom-wise consideration of residues could
generate a more sensitive filtering of the unfavorable hydrophobic-hydrophilic con-
tacts. In spite of these pitfalls, the herein described assessment of the SA res, i is still a
useful method to qualitatively study residual exposure in GPCR-membrane systems.
Due to the relatively recent release of the 3D-CTMD model, this powerful tool has
not been widely used. To our knowledge, only a few published works perform this type
of analysis ( Khelashvili et al., 2012;Mondal et al., 2011, 2013 ). The original workwhere
Mondal et al. describe thismethodology ( Mondal et al., 2011 ) is a nice example showing
its capabilities. In this study, the CTMD model is used to describe how the lipid com-
position affects membrane remodeling in GPCR-membrane systems. Thus, they stud-
ied theGPCR rhodopsin embedded inmembranes of different lipid composition. One of
the conclusions of this study is that thicker bilayers (namely, pure diC 20:1 PC and 7:7:6
C 18:0 -C 22:6 PC/C 16:0 -C 18:1 PC/cholesterol membranes) become thinner at the
membrane-protein boundary, whereas thinner bilayers (namely, pure diC 14:1 PC and
diC 16:1 PC) become thicker. After quantification of the residual mismatch energy for
these systems, the authors show how the TM segment 1 (TM1) in rhodopsin embedded
in pure diC 14:1 PC is most likely involved in a potential mismatch-driven oligomeriza-
tion interface. In addition, the authors describe a higher predominance of the TM4 seg-
mentwhen rhodopsin is embedded in thickermembranes, suggesting this segment could
also be involved in a potential oligomerization interface (i.e., 7:7:6 C 18:0 -C 22:6 PC/