Despite the lipid-mediated contacts between BR trimers, purple membrane crystals are
extremely well ordered and diffract electrons to high resolution, which has permitted to establish the
3D structure of BR first at medium resolution (7 Å in the membrane plane), revealing the presence of
seven TM helices (Henderson and Unwin 1975), and later to high resolution (3.5 Å in the membrane
plane), permitting an atomic model to be built into the electron density map (Henderson et al. 1990;
Grigorieff et al. 1996). BR however proved extremely reluctant to form well-ordered 3D crystals from
detergent solutions, and it is only in lipid cubic phases (see Chap. 11, § 11.2.2.2) that such crystals
could finally be obtained, from which the X-ray structure was solved to, initially, 2.5-Å resolution
(Landau and Rosenbusch 1996; Pebay-Peyroula et al. 1997; Rummel et al. 1998). As of 2015,
84 X-ray diffraction, 6 electron diffraction, and 3 NMR structures of BR had been deposited in the
Protein Data Bank, 21 X-ray structures reporting light-induced structural changes and changes induced
by mutations, changes in pH, thermal annealing, or X-ray-induced photoreduction (reviewed in
Wickstrand et al. 2015). The best of these structures reach <2-Å resolution, which permits, in
particular, to visualize water molecules and protein-bound lipids.
An abundant, often contentious literature deals with the structural changes undergone by BR at
each step of the photocycle, as well as with the structural effects of mutations created to either mimic
one or the other of these states or to slow down the cycle so as to favor their observation, the ensemble
of which is easily bewildering to the nonspecialized reader. A remarkable piece of meta-analysis has
been carried out by Richard Neutze and colleagues to compare and classify all of the structures
reported as of 2014, sort out reproducible observations from possible artifacts (such as those due to
radiative damage, mixed states, crystal twining, crystallization conditions, etc.), and try to come up
with a unified model (Fig. 1.25; Wickstrand et al. 2015; for earlier discussions, see e.g. Kühlbrandt
2000; Neutze et al. 2002; Hirai et al. 2009). According to this model, at the core of the photocycle is, as
a consequence of the isomerization of retinal, a rearrangement of the Schiff base and its environment
(including a key water molecule), the result of which is to (i) shift close to one another the pK of the
Schiff base and that of Asp85, which are widely different in the resting state of BR, and (ii) move these
two groups closer in space. These two changes allow the Schiff base proton to be transferred to Asp85
(Step 1 in Fig. 1.24B), generating State M 1 ; after which, once the electrostatic attraction that existed
between the positively charged Schiff base and the negatively charged carboxylate of Asp85 vanishes,
the two groups separate (Fig. 1.25A, B). In the second part of the cycle (transition from M 1 to M 2 ),
helix F moves outward (Fig. 1.25A), leading water molecules to transiently order in the cytoplasmic
half of the channel, thereby facilitating the reprotonation of the Schiff base from Asp96 (Step 3 in
Fig. 1.24B), while other structural changes help control proton release from Asp85 and return to the
ground state. Throughout the cycle, helix movements are limited to relatively modest displacements of
helix F and part of helix C.
Also of relevance to some of the observations to be discussed later are two further pieces of data.
First, Giuseppe Zaccai and his colleagues have shown, using neutron scattering measurements, that, in
order to function, BR needs a “soft” environment. Stiffening either by lowering the temperature or as a
result of dehydration will stop the photocycle (Ferrand et al. 1993; Lehnert et al. 1998; Zaccai 2000,
2004). Second, the photocycle of BR is sensitive to the chemical nature of its environment and, in
particular, to the nature of the lipids surrounding it, among which squalene and phosphatidyl
glycerophosphate (Dracheva et al. 1996; Joshi et al. 1998; Hendler and Dracheva 2001; Lee et al.
2015).
1.6 Dynamics of Transmembrane Regions and the Function of Membrane Proteins
35
extremely well ordered and diffract electrons to high resolution, which has permitted to establish the
3D structure of BR first at medium resolution (7 Å in the membrane plane), revealing the presence of
seven TM helices (Henderson and Unwin 1975), and later to high resolution (3.5 Å in the membrane
plane), permitting an atomic model to be built into the electron density map (Henderson et al. 1990;
Grigorieff et al. 1996). BR however proved extremely reluctant to form well-ordered 3D crystals from
detergent solutions, and it is only in lipid cubic phases (see Chap. 11, § 11.2.2.2) that such crystals
could finally be obtained, from which the X-ray structure was solved to, initially, 2.5-Å resolution
(Landau and Rosenbusch 1996; Pebay-Peyroula et al. 1997; Rummel et al. 1998). As of 2015,
84 X-ray diffraction, 6 electron diffraction, and 3 NMR structures of BR had been deposited in the
Protein Data Bank, 21 X-ray structures reporting light-induced structural changes and changes induced
by mutations, changes in pH, thermal annealing, or X-ray-induced photoreduction (reviewed in
Wickstrand et al. 2015). The best of these structures reach <2-Å resolution, which permits, in
particular, to visualize water molecules and protein-bound lipids.
An abundant, often contentious literature deals with the structural changes undergone by BR at
each step of the photocycle, as well as with the structural effects of mutations created to either mimic
one or the other of these states or to slow down the cycle so as to favor their observation, the ensemble
of which is easily bewildering to the nonspecialized reader. A remarkable piece of meta-analysis has
been carried out by Richard Neutze and colleagues to compare and classify all of the structures
reported as of 2014, sort out reproducible observations from possible artifacts (such as those due to
radiative damage, mixed states, crystal twining, crystallization conditions, etc.), and try to come up
with a unified model (Fig. 1.25; Wickstrand et al. 2015; for earlier discussions, see e.g. Kühlbrandt
2000; Neutze et al. 2002; Hirai et al. 2009). According to this model, at the core of the photocycle is, as
a consequence of the isomerization of retinal, a rearrangement of the Schiff base and its environment
(including a key water molecule), the result of which is to (i) shift close to one another the pK of the
Schiff base and that of Asp85, which are widely different in the resting state of BR, and (ii) move these
two groups closer in space. These two changes allow the Schiff base proton to be transferred to Asp85
(Step 1 in Fig. 1.24B), generating State M 1 ; after which, once the electrostatic attraction that existed
between the positively charged Schiff base and the negatively charged carboxylate of Asp85 vanishes,
the two groups separate (Fig. 1.25A, B). In the second part of the cycle (transition from M 1 to M 2 ),
helix F moves outward (Fig. 1.25A), leading water molecules to transiently order in the cytoplasmic
half of the channel, thereby facilitating the reprotonation of the Schiff base from Asp96 (Step 3 in
Fig. 1.24B), while other structural changes help control proton release from Asp85 and return to the
ground state. Throughout the cycle, helix movements are limited to relatively modest displacements of
helix F and part of helix C.
Also of relevance to some of the observations to be discussed later are two further pieces of data.
First, Giuseppe Zaccai and his colleagues have shown, using neutron scattering measurements, that, in
order to function, BR needs a “soft” environment. Stiffening either by lowering the temperature or as a
result of dehydration will stop the photocycle (Ferrand et al. 1993; Lehnert et al. 1998; Zaccai 2000,
2004). Second, the photocycle of BR is sensitive to the chemical nature of its environment and, in
particular, to the nature of the lipids surrounding it, among which squalene and phosphatidyl
glycerophosphate (Dracheva et al. 1996; Joshi et al. 1998; Hendler and Dracheva 2001; Lee et al.
2015).
1.6 Dynamics of Transmembrane Regions and the Function of Membrane Proteins
35