signals, the structure of a hydrogenated ligand bound to the deuterated receptor trapped in DAPol. The
other two studies are more preliminary: Study 10.8 examines the feasibility of using A8-35 to study
two more GPCRs, the melanocortin-2 and -4 receptors, Study 10.11 that of using sNMR to gather
information on the extramembrane loops of C. trachomatis MOMP, an outer MP used as an immunogen in the formulation of vaccines (see Chap. 15).
10.3.3.1 Dynamics of the Transmembrane b-Barrel of OmpX
Time-averaged atomic structures of proteins often leave it uncertain how they achieve some of their
functions. OmpX from E. coli is a case in point. While electrophysiological studies suggest that OmpX
can translocate ions (Dupont et al. 2004; Arnold et al. 2007), structural studies describe this MP as a
rigid TM β-barrel, whose lumen is too densely packed with amino acid side chains to admit any solute
(Vogt and Schulz 1999; Fernández et al. 2001a, b, 2004). However, low-energy, ground-state
conformations do not preclude the existence of time-dependent conformational fluctuations that may
allow the transient formation of a channel.
In Study 10.3 (Table 10.1; Catoire et al. 2010b), the solvent accessibility of amide protons (
1 H
N )
of OmpX kept water-soluble either by A8-35 or by diC 6 PC was investigated by measuring the extent
of hydrogen/deuterium (H/D) exchange after extensive equilibration, providing insights into the
dynamics of the β-barrel. The resolution was improved by adding EDTA to the buffers, presumably
preventing traces of Ca
2+ ions from bridging some MP/A8-35 complexes. The accessibility of
1 H
N
protons involved in the β-barrel hydrogen bond network was assessed by comparing
15 N1 H TROSY
spectra collected after extensive equilibration (8 weeks at 4
C) of OmpX/A8-35 complexes in either
5 or 100% D 2 O (Fig. 10.21). Unsurprisingly, the most protected
1 H
N protons (in red in Fig. 10.22) are
found in the TM region of the barrel, with a few residues at the ends of each strand showing higher
accessibility. The number of protected
1 H
N protons differs from one β-strand to the other, the first two
strands, in particular, displaying only a few partially shielded
1 H
N protons in the central part of the TM
region.
H/D exchange measurements carried out in diC 6 PC at pH 8.0 yielded essentially the same
results. Most of those
1 H
N protons that do exchange do it to comparable degrees in the two
environments, suggesting that the overall dynamics of the barrel is similar. However, in diC 6 PC,
residues belonging to the end of the barrel that faces the periplasmic space tend to exchange somewhat
more than in A8-35. This is reminiscent of the observations reported by Gerhard Wagner and
colleagues on BR (Etzkorn et al. 2013; see § 10.3.1) and by Melanie J. Cocco and colleagues on
C. trachomatis MOMP (Feinstein et al. 2014; see § 10.3.3.3).
Overall, the observations on OmpX show that the barrel does not behave as a solid block, some
of its strands appearing more mobile (and/or accessible) than others. While it cannot be excluded that
the water molecules involved in the exchange originate from within the surfactant layer, they may also
come from the bulk solution via the lumen of the barrel. H/D exchange, in this hypothesis, would
provide insights into conformational changes that affect the packing of side chains in the lumen. One
interesting working hypothesis could be that ion flux is associated with small relative movements of
strands 1 and 2 that would disrupt the intraluminal network of hydrogen bonds (Böckmann and
Caflisch 2005), letting a transient channel form. A similar channel-gating mechanism has been
proposed for OmpA (Hong et al. 2006). Alternatively, the dynamics of the β-strands suggested by
H/D exchange measurements could reflect a transport mechanism distinct from ion flux, of which the
latter would only be a side effect (for a discussion, see Catoire et al. 2010b).
10.3.3.2 Conformation of LTB 4 Bound to the BLT2 Leukotriene Receptor
G protein-coupled receptors (GPCRs) mediate a wide range of physiological responses. Producing
them under a functional form and determining their 3D structure, their transconformations, and the
10.3 Solution NMR Studies of Amphipol-Trapped Membrane Proteins
483
other two studies are more preliminary: Study 10.8 examines the feasibility of using A8-35 to study
two more GPCRs, the melanocortin-2 and -4 receptors, Study 10.11 that of using sNMR to gather
information on the extramembrane loops of C. trachomatis MOMP, an outer MP used as an immunogen in the formulation of vaccines (see Chap. 15).
10.3.3.1 Dynamics of the Transmembrane b-Barrel of OmpX
Time-averaged atomic structures of proteins often leave it uncertain how they achieve some of their
functions. OmpX from E. coli is a case in point. While electrophysiological studies suggest that OmpX
can translocate ions (Dupont et al. 2004; Arnold et al. 2007), structural studies describe this MP as a
rigid TM β-barrel, whose lumen is too densely packed with amino acid side chains to admit any solute
(Vogt and Schulz 1999; Fernández et al. 2001a, b, 2004). However, low-energy, ground-state
conformations do not preclude the existence of time-dependent conformational fluctuations that may
allow the transient formation of a channel.
In Study 10.3 (Table 10.1; Catoire et al. 2010b), the solvent accessibility of amide protons (
1 H
N )
of OmpX kept water-soluble either by A8-35 or by diC 6 PC was investigated by measuring the extent
of hydrogen/deuterium (H/D) exchange after extensive equilibration, providing insights into the
dynamics of the β-barrel. The resolution was improved by adding EDTA to the buffers, presumably
preventing traces of Ca
2+ ions from bridging some MP/A8-35 complexes. The accessibility of
1 H
N
protons involved in the β-barrel hydrogen bond network was assessed by comparing
15 N1 H TROSY
spectra collected after extensive equilibration (8 weeks at 4
C) of OmpX/A8-35 complexes in either
5 or 100% D 2 O (Fig. 10.21). Unsurprisingly, the most protected
1 H
N protons (in red in Fig. 10.22) are
found in the TM region of the barrel, with a few residues at the ends of each strand showing higher
accessibility. The number of protected
1 H
N protons differs from one β-strand to the other, the first two
strands, in particular, displaying only a few partially shielded
1 H
N protons in the central part of the TM
region.
H/D exchange measurements carried out in diC 6 PC at pH 8.0 yielded essentially the same
results. Most of those
1 H
N protons that do exchange do it to comparable degrees in the two
environments, suggesting that the overall dynamics of the barrel is similar. However, in diC 6 PC,
residues belonging to the end of the barrel that faces the periplasmic space tend to exchange somewhat
more than in A8-35. This is reminiscent of the observations reported by Gerhard Wagner and
colleagues on BR (Etzkorn et al. 2013; see § 10.3.1) and by Melanie J. Cocco and colleagues on
C. trachomatis MOMP (Feinstein et al. 2014; see § 10.3.3.3).
Overall, the observations on OmpX show that the barrel does not behave as a solid block, some
of its strands appearing more mobile (and/or accessible) than others. While it cannot be excluded that
the water molecules involved in the exchange originate from within the surfactant layer, they may also
come from the bulk solution via the lumen of the barrel. H/D exchange, in this hypothesis, would
provide insights into conformational changes that affect the packing of side chains in the lumen. One
interesting working hypothesis could be that ion flux is associated with small relative movements of
strands 1 and 2 that would disrupt the intraluminal network of hydrogen bonds (Böckmann and
Caflisch 2005), letting a transient channel form. A similar channel-gating mechanism has been
proposed for OmpA (Hong et al. 2006). Alternatively, the dynamics of the β-strands suggested by
H/D exchange measurements could reflect a transport mechanism distinct from ion flux, of which the
latter would only be a side effect (for a discussion, see Catoire et al. 2010b).
10.3.3.2 Conformation of LTB 4 Bound to the BLT2 Leukotriene Receptor
G protein-coupled receptors (GPCRs) mediate a wide range of physiological responses. Producing
them under a functional form and determining their 3D structure, their transconformations, and the
10.3 Solution NMR Studies of Amphipol-Trapped Membrane Proteins
483