encounter (for general reviews on MP structure, see e.g. White and Wimley 1999; Popot and Engelman
2000; Vinothkumar and Henderson 2010; Buchanan et al. 2012; White et al. 2018, and references
therein). We will start with listing a few rules that are generally obeyed and then give some examples of
more complex situations. In the next two sections, we will consider membrane protein/lipid
interactions (§ 1.5) and the dynamics of TM protein regions (§ 1.6), both of which are essential to
understanding the problems of instability most MPs face when extracted from their native membrane
environment and how nonconventional surfactants can help mitigating them.
As more or less general rules, we can consider the following:
1. Protein surfaces exposed to the hydrophobic core of the membrane are highly hydrophobic.
This is achieved by three mechanisms: (i) membrane-exposed segments span completely the
hydrophobic core, leaving no loops exposed to it; (ii) they satisfy most of the hydrogen bonds
that can be formed between the >N—H and O¼C< groups of the main chain peptide bonds;
and (iii) most exposed side chains are nonpolar (cf. Figs. 1.6 and 1.7). This is a natural
consequence of the high cost of burying polar groups in the membrane core (Table 1.2). An
exposed hydrogen-bonding side chain like that of serine residues will tend to either bond back
with the TM surface of the protein or interact with complementary residues at the surface of
other TM regions. This is why, for instance, an Ala ! Glu mutation in the single TM α-helix
of the FGFR3 receptor of fibroblasts drives its dimerization, mimicking physiological
activation and resulting in cancer (Li et al. 2006). Kinks and other irregularities in TM helices
do exist, however, and often have important functional roles (see e.g. Popot and Engelman
2000 and examples shown below and in § 1.6).
Membrane protein structure revealed
# 2018 by Francis Haraux
ä
Fig. 1.6 (continued) from Paracoccus denitrificans (X-ray diffraction, 2.7-Å resolution; 1AR1;
Ostermeier et al. 1997), bc1 cytochrome bc 1 complex from beef heart (TM subunits only; X-ray diffraction, 3.7-Å resolution; 3BCC; Xia et al. 1997), PSI Photosystem I reaction center from Synechococcus
elongatus (α carbons only, side chains having not yet been identified at the time; X-ray diffraction, 4-Å
resolution; 2PPS; Krauss et al. 1996), MscL mechanosensitive ion channel from Mycobacterium tuberculosis (X-ray diffraction, 3.5-Å resolution; 1MSL; Chang et al. 1998) (Figure from Popot and Engelman
2000).
1.4 Membrane Protein Structure
17
2000; Vinothkumar and Henderson 2010; Buchanan et al. 2012; White et al. 2018, and references
therein). We will start with listing a few rules that are generally obeyed and then give some examples of
more complex situations. In the next two sections, we will consider membrane protein/lipid
interactions (§ 1.5) and the dynamics of TM protein regions (§ 1.6), both of which are essential to
understanding the problems of instability most MPs face when extracted from their native membrane
environment and how nonconventional surfactants can help mitigating them.
As more or less general rules, we can consider the following:
1. Protein surfaces exposed to the hydrophobic core of the membrane are highly hydrophobic.
This is achieved by three mechanisms: (i) membrane-exposed segments span completely the
hydrophobic core, leaving no loops exposed to it; (ii) they satisfy most of the hydrogen bonds
that can be formed between the >N—H and O¼C< groups of the main chain peptide bonds;
and (iii) most exposed side chains are nonpolar (cf. Figs. 1.6 and 1.7). This is a natural
consequence of the high cost of burying polar groups in the membrane core (Table 1.2). An
exposed hydrogen-bonding side chain like that of serine residues will tend to either bond back
with the TM surface of the protein or interact with complementary residues at the surface of
other TM regions. This is why, for instance, an Ala ! Glu mutation in the single TM α-helix
of the FGFR3 receptor of fibroblasts drives its dimerization, mimicking physiological
activation and resulting in cancer (Li et al. 2006). Kinks and other irregularities in TM helices
do exist, however, and often have important functional roles (see e.g. Popot and Engelman
2000 and examples shown below and in § 1.6).
Membrane protein structure revealed
# 2018 by Francis Haraux
ä
Fig. 1.6 (continued) from Paracoccus denitrificans (X-ray diffraction, 2.7-Å resolution; 1AR1;
Ostermeier et al. 1997), bc1 cytochrome bc 1 complex from beef heart (TM subunits only; X-ray diffraction, 3.7-Å resolution; 3BCC; Xia et al. 1997), PSI Photosystem I reaction center from Synechococcus
elongatus (α carbons only, side chains having not yet been identified at the time; X-ray diffraction, 4-Å
resolution; 2PPS; Krauss et al. 1996), MscL mechanosensitive ion channel from Mycobacterium tuberculosis (X-ray diffraction, 3.5-Å resolution; 1MSL; Chang et al. 1998) (Figure from Popot and Engelman
2000).
1.4 Membrane Protein Structure
17