2.4.4
Detergent-Induced Conformational Alterations
In the case of cytochrome b 6 f, exposure to detergents results in the progressive disaggregation of the
complex. Alterations can be more subtle. With the multiplication of MP structures obtained by X-ray
diffraction, NMR, and electron microscopy (EM) under a variety of conditions, such as in a bilayer
environment vs. in complex with various detergents, evidence has accumulated that, whereas some MP
structures are strictly identical in a lipid vs. a detergent environment (see e.g. Screpanti et al. 2006),
environmental differences may affect not only the dynamics but also the average conformation of MPs:
structures observed in detergent solution do not necessarily reflect faithfully the native structure(s)
adopted in the membrane (for reviews and discussions, see e.g. Cross et al. 2011; Zhou and Cross
2013; Zoonens et al. 2013).
Zhou and Cross have carried out a systematic examination of the TM region of MP structures
deposited in the Protein Data Bank. They note deviations from reasonable expectations, as well as
divergences between structures obtained in different environments (natural membranes, artificial ones,
detergent solutions, crystals) and using different techniques (solution-state and solid-state NMR, EM,
X-ray crystallographic analysis of crystals obtained either in detergent solutions or in lipid cubic
phases) (Zhou and Cross 2013). Some of these observations are relevant to the present discussion,
inasmuch as they suggest ways in which the transfer from a natural membrane environment to a
detergent one may affect MP structures, and plausible mechanisms underlying them. Whereas many
structures are consistent from one system to another and appear reasonable from a structural point of
view, suggesting that they do represent the native state (see e.g. Lanyi and Schobert 2007; Liao et al.
2012), others are more problematical. In some of them, TM α-helices appear splayed out, leaving large
interhelical gaps, and/or they are misoriented, with polar residues and residues expected to be involved
in helix-helix interactions facing the exterior of the TM helix bundle, so that, in situ, they would be
exposed to the membrane core. In other structures, some helices are too short to span a membrane or
out of register with nearby helices. In many cases, it remains a matter of opinion whether the structure
is actually aberrant or not: different structures for the same protein may well correspond to different
functional states, particularly for transporters (see the example of the calcium pump SERCA1a in
Chap. 1, § 1.6.3, Figs. 1.34 and 1.36). In some cases, however, it seems extremely likely that the native
fold has indeed been disturbed. Possible causes invoked to explain these distortions include the loss of
lateral pressure, attraction of TM hydrophilic side chains that normally face the interior of the TM helix
bundle by the hydrophilic polar head region of detergent micelles, or the intercalation of detergent
molecules between helices. It seems intuitive that, in the case or SERCA1a, for instance, the loss of the
PE molecule that lies between helices M2 and M4 in the calcium-free E2 state (see Chap. 1, Fig. 1.37B)
and its replacement by detergent molecules might well perturb the overall TM structure and, as a
consequence, the arrangement of the extramembrane domains and the functionality of the whole
protein. Detergent molecules intruding between TM segments are seen in many crystallographic
structures (see Zhou and Cross 2013), as well as in some MD simulations (see e.g. Khelashvili et al.
2013; Lee et al. 2016; Fig. 2.14).
MD simulations do suggest that the transfer from a lipid to a detergent environment may bring
about significant structural changes (see e.g. Cuthbertson et al. 2006; Choutko et al. 2011; RodríguezRopero and Fioroni 2012). In a recent study combining MD simulations and functional measurements,
it was concluded that solubilization with DPC (see Fig. 2.1) of the mitochondrial uncoupler UCP2
entails an artifactual widening of its TM region, turning it into a porin-like aqueous channel and
abolishing its proton carrier activity (Zoonens et al. 2013).
It may seem paradoxical that, on the one hand, MPs are so often inactivated by detergents and
can be so hard to handle in detergent solutions, and on the other hand, a large number of them (~100 as
80
2 Extracting Membrane Proteins from Their Native Environment
Detergent-Induced Conformational Alterations
In the case of cytochrome b 6 f, exposure to detergents results in the progressive disaggregation of the
complex. Alterations can be more subtle. With the multiplication of MP structures obtained by X-ray
diffraction, NMR, and electron microscopy (EM) under a variety of conditions, such as in a bilayer
environment vs. in complex with various detergents, evidence has accumulated that, whereas some MP
structures are strictly identical in a lipid vs. a detergent environment (see e.g. Screpanti et al. 2006),
environmental differences may affect not only the dynamics but also the average conformation of MPs:
structures observed in detergent solution do not necessarily reflect faithfully the native structure(s)
adopted in the membrane (for reviews and discussions, see e.g. Cross et al. 2011; Zhou and Cross
2013; Zoonens et al. 2013).
Zhou and Cross have carried out a systematic examination of the TM region of MP structures
deposited in the Protein Data Bank. They note deviations from reasonable expectations, as well as
divergences between structures obtained in different environments (natural membranes, artificial ones,
detergent solutions, crystals) and using different techniques (solution-state and solid-state NMR, EM,
X-ray crystallographic analysis of crystals obtained either in detergent solutions or in lipid cubic
phases) (Zhou and Cross 2013). Some of these observations are relevant to the present discussion,
inasmuch as they suggest ways in which the transfer from a natural membrane environment to a
detergent one may affect MP structures, and plausible mechanisms underlying them. Whereas many
structures are consistent from one system to another and appear reasonable from a structural point of
view, suggesting that they do represent the native state (see e.g. Lanyi and Schobert 2007; Liao et al.
2012), others are more problematical. In some of them, TM α-helices appear splayed out, leaving large
interhelical gaps, and/or they are misoriented, with polar residues and residues expected to be involved
in helix-helix interactions facing the exterior of the TM helix bundle, so that, in situ, they would be
exposed to the membrane core. In other structures, some helices are too short to span a membrane or
out of register with nearby helices. In many cases, it remains a matter of opinion whether the structure
is actually aberrant or not: different structures for the same protein may well correspond to different
functional states, particularly for transporters (see the example of the calcium pump SERCA1a in
Chap. 1, § 1.6.3, Figs. 1.34 and 1.36). In some cases, however, it seems extremely likely that the native
fold has indeed been disturbed. Possible causes invoked to explain these distortions include the loss of
lateral pressure, attraction of TM hydrophilic side chains that normally face the interior of the TM helix
bundle by the hydrophilic polar head region of detergent micelles, or the intercalation of detergent
molecules between helices. It seems intuitive that, in the case or SERCA1a, for instance, the loss of the
PE molecule that lies between helices M2 and M4 in the calcium-free E2 state (see Chap. 1, Fig. 1.37B)
and its replacement by detergent molecules might well perturb the overall TM structure and, as a
consequence, the arrangement of the extramembrane domains and the functionality of the whole
protein. Detergent molecules intruding between TM segments are seen in many crystallographic
structures (see Zhou and Cross 2013), as well as in some MD simulations (see e.g. Khelashvili et al.
2013; Lee et al. 2016; Fig. 2.14).
MD simulations do suggest that the transfer from a lipid to a detergent environment may bring
about significant structural changes (see e.g. Cuthbertson et al. 2006; Choutko et al. 2011; RodríguezRopero and Fioroni 2012). In a recent study combining MD simulations and functional measurements,
it was concluded that solubilization with DPC (see Fig. 2.1) of the mitochondrial uncoupler UCP2
entails an artifactual widening of its TM region, turning it into a porin-like aqueous channel and
abolishing its proton carrier activity (Zoonens et al. 2013).
It may seem paradoxical that, on the one hand, MPs are so often inactivated by detergents and
can be so hard to handle in detergent solutions, and on the other hand, a large number of them (~100 as
80
2 Extracting Membrane Proteins from Their Native Environment