vs. bilayer systems (Dörr et al. 2016) is not necessarily warranted, at least certainly not in all cases. It
seems likely that, as is the case for bicelles (see Chap. 3, § 3.2), the size of the discs, and, therefore, the
protein/lipid ratio, is modulated by experimental conditions, in particular the SMA/lipid ratio in the
final preparation, as indeed observed (Li et al. 2015a; Zhang et al. 2015b; Grethen et al. 2017).
To which extent the composition of the lipids and their organization in SMALPs, in particular
their distribution between the two monolayers, faithfully mimics the situation experienced in vivo by
the MPs they host remains to be ascertained. A key observation, discussed in Chap. 4 (§ 4.3.2), is the
fact that the lipid content of SMALPs exchanges very rapidly (within seconds), probably according to a
bulk mechanism that involves lateral fusion and fission events (Cuevas Arenas et al. 2017; Grethen
et al. 2018). Given that SMA seems to have no preference for extracting specific lipids as long as they
are in the fluid phase (Dominguez Pardo et al. 2017), these observations suggest that the preferential
association of specific lipids with MPs extracted in SMALPs that has been observed in some
experiments (Prabudiansyah et al. 2015; Rehan et al. 2017) reflects an equilibrium situation, not a
kinetically frozen one. By the same token, one should expect that lipids that, in the original membrane,
were located either in the outer or in the inner monolayer will rapidly mix and that whatever may
remain of the original compositional asymmetry be due to specific binding to the protein.
5.3.2
Particle Size and Organization
Detailed structural studies have been carried out on BR/A8-35 complexes, using, in particular, SANS,
Eq-AUC, and SV-AUC (Study 5.16 in Table 5.4). An introduction to these techniques and their
application to MP/APol complexes is provided in Chap. 9. In both cases, advantage was taken of the
possibility to modulate the contrast of the APol with the solvent by deuterating its side chains, which
increases both its buoying density and its neutron scattering length density. The density of the solvent
was modulated thanks to the use of H 2 O, D 2 O, and D 2
18 O (Gohon et al. 2008). These studies will be
described in Chap. 9, § 9.3. They show that BR and its associated lipids occupy, as expected, the center
of the particles, whose overall mass is ~90 kDa (Table 5.5), the APol forming a belt adsorbed onto the
TM region of the protein/lipid complex. The thickness of the belt – its extension away from the surface
– can be estimated to ~1.7 nm (see Chap. 9, Fig. 9.18). For the most part, these measurements, obtained
by a wide range of approaches, provide consistent estimates of the composition, size, and mass of the
complexes. Combining composition and s-values yields R S % 3.6 nm, whereas Eq-AUC and SV-AUC
analyses give R S % 3.8 nm. SANS yields an R g
value (radius of gyration at infinite contrast) of ~3.0
nm, which, for spherical particles, would correspond to R S % 3.8 nm. By EM after negative staining
(NS-EM), one observes particle half-widths and half-lengths of ~3.2 and ~4 nm. The outlier is SEC,
which provides significantly larger estimates (R S % 5.0 Æ 0.15 nm). Other observations suggest that,
for reasons that are probably electrostatic in origin, since the same phenomenon is not observed with
NAPols (Sharma et al. 2012; see below), SEC tends to overestimate the hydrodynamic radius of
MP/A8-35 complexes as compared to MP/neutral detergent ones. Indeed, SEC (Zoonens et al. 2007)
yields much larger apparent R S differences between tOmpA/A8-35 complexes (R S % 4.3–3.7 nm) and
tOmpA/dihexanoylphosphatidylcholine (diC 6 PC) ones (R S % 2.6 nm) than is indicated by an NMR
study of their respective rotational correlation times, τ c (Zoonens et al. 2005). The NMR data suggest
that SEC overestimates the R S of tOmpA/A8-35 particles by as much as 30–50% (Zoonens et al. 2007),
a difference close to that (~30%) observed for BR/A8-35 complexes between SEC data, on the one
hand, and SANS and AUC ones, on the other (Gohon et al. 2008). Why such a phenomenon does not
affect the determination by SEC of the Stokes radius of pure A8-35 particles (Chap. 4, § 4.3.1.2.2)
5.3 Composition, Organization, Dynamics, and Solution Properties of Membrane. . .
279
seems likely that, as is the case for bicelles (see Chap. 3, § 3.2), the size of the discs, and, therefore, the
protein/lipid ratio, is modulated by experimental conditions, in particular the SMA/lipid ratio in the
final preparation, as indeed observed (Li et al. 2015a; Zhang et al. 2015b; Grethen et al. 2017).
To which extent the composition of the lipids and their organization in SMALPs, in particular
their distribution between the two monolayers, faithfully mimics the situation experienced in vivo by
the MPs they host remains to be ascertained. A key observation, discussed in Chap. 4 (§ 4.3.2), is the
fact that the lipid content of SMALPs exchanges very rapidly (within seconds), probably according to a
bulk mechanism that involves lateral fusion and fission events (Cuevas Arenas et al. 2017; Grethen
et al. 2018). Given that SMA seems to have no preference for extracting specific lipids as long as they
are in the fluid phase (Dominguez Pardo et al. 2017), these observations suggest that the preferential
association of specific lipids with MPs extracted in SMALPs that has been observed in some
experiments (Prabudiansyah et al. 2015; Rehan et al. 2017) reflects an equilibrium situation, not a
kinetically frozen one. By the same token, one should expect that lipids that, in the original membrane,
were located either in the outer or in the inner monolayer will rapidly mix and that whatever may
remain of the original compositional asymmetry be due to specific binding to the protein.
5.3.2
Particle Size and Organization
Detailed structural studies have been carried out on BR/A8-35 complexes, using, in particular, SANS,
Eq-AUC, and SV-AUC (Study 5.16 in Table 5.4). An introduction to these techniques and their
application to MP/APol complexes is provided in Chap. 9. In both cases, advantage was taken of the
possibility to modulate the contrast of the APol with the solvent by deuterating its side chains, which
increases both its buoying density and its neutron scattering length density. The density of the solvent
was modulated thanks to the use of H 2 O, D 2 O, and D 2
18 O (Gohon et al. 2008). These studies will be
described in Chap. 9, § 9.3. They show that BR and its associated lipids occupy, as expected, the center
of the particles, whose overall mass is ~90 kDa (Table 5.5), the APol forming a belt adsorbed onto the
TM region of the protein/lipid complex. The thickness of the belt – its extension away from the surface
– can be estimated to ~1.7 nm (see Chap. 9, Fig. 9.18). For the most part, these measurements, obtained
by a wide range of approaches, provide consistent estimates of the composition, size, and mass of the
complexes. Combining composition and s-values yields R S % 3.6 nm, whereas Eq-AUC and SV-AUC
analyses give R S % 3.8 nm. SANS yields an R g
value (radius of gyration at infinite contrast) of ~3.0
nm, which, for spherical particles, would correspond to R S % 3.8 nm. By EM after negative staining
(NS-EM), one observes particle half-widths and half-lengths of ~3.2 and ~4 nm. The outlier is SEC,
which provides significantly larger estimates (R S % 5.0 Æ 0.15 nm). Other observations suggest that,
for reasons that are probably electrostatic in origin, since the same phenomenon is not observed with
NAPols (Sharma et al. 2012; see below), SEC tends to overestimate the hydrodynamic radius of
MP/A8-35 complexes as compared to MP/neutral detergent ones. Indeed, SEC (Zoonens et al. 2007)
yields much larger apparent R S differences between tOmpA/A8-35 complexes (R S % 4.3–3.7 nm) and
tOmpA/dihexanoylphosphatidylcholine (diC 6 PC) ones (R S % 2.6 nm) than is indicated by an NMR
study of their respective rotational correlation times, τ c (Zoonens et al. 2005). The NMR data suggest
that SEC overestimates the R S of tOmpA/A8-35 particles by as much as 30–50% (Zoonens et al. 2007),
a difference close to that (~30%) observed for BR/A8-35 complexes between SEC data, on the one
hand, and SANS and AUC ones, on the other (Gohon et al. 2008). Why such a phenomenon does not
affect the determination by SEC of the Stokes radius of pure A8-35 particles (Chap. 4, § 4.3.1.2.2)
5.3 Composition, Organization, Dynamics, and Solution Properties of Membrane. . .
279