general structure, it is similar to BR, but there is no trace of an evolutionary relationship between them,
and their functions are completely different. Among GPCRs, rhodopsin is also peculiar in having a
very low level of basal activity (see below), which endows our eyes with the possibility to detect even
very weak visual signals. This feature turned out to be important in permitting its crystallization,
because it reduces conformational variability. Other than that, rhodopsin is an archetypical GPCR. It
was the first one whose 3D crystallographic structure was ever established, in 2000, by Krzysztof
Palczewski and colleagues (Palczewski et al. 2000).
The signaling functions of GPCRs are illustrated in Fig. 2.16 in the case of the β 2 adrenergic
receptor, whose natural ligands are adrenaline and noradrenaline. Similar in this respect to the
pentameric ligand-gated ion channels (pLGICs), which we have discussed in Chap. 1 (§ 1.6.2),
GPCRs are in equilibrium between two or more conformational states: binding of an extracellular
ligand shifts the equilibrium toward the state(s) with the highest affinity for the ligand. However, at
variance with pLGICs, conformational transitions do not control the opening and closing of a TM
channel. Rather, the reorganization of the protein around the ligand-binding site is coupled, through the
TM region, to the opening of a cleft in the cytoplasmic face of the receptor. Cytosolic proteins bind
to it and propagate and amplify the signal. Two main routes are the activation of stimulatory or
inhibitory heterotrimeric G proteins, on the one hand, and coupling to arrestin, on the other. Both of
them are schematized in Fig. 2.16, along with some of the downstream cellular events. An important
feature of GPCRs is that most of them are delicately poised between the inactive and active states, in
which case they exhibit a basal level of activity (constitutive activity). As symbolized in the inset to
Fig. 2.16, they can therefore be regulated both up and down. “Agonists,” which have much more
affinity for the active state, can turn them on completely. “Partial agonists,” which have only a
moderate preferential affinity for the active state, achieve only partial activation even at saturation.
“Inverse agonists” favor the inactive state and decrease the activity below the constitutive level.
Finally, “neutral agonists” block the binding site – and, thereby, prevent regulation by other ligands
– without themselves affecting the conformational equilibrium. In the case of rhodopsin, the cofactor
in its resting conformation, 11-cis-retinal, is a very potent inverse agonist, whereas the light-induced
isomer, all-trans-retinal, is an agonist.
The view just presented of regulation being achieved simply by modulating the equilibrium
between two conformational states is, however, widely oversimplified, GPCRs being able “to sample a
continuum of conformations with relatively closely spaced energies” (see Kobilka and Deupi 2007,
Rosenbaum et al. 2009, and references therein).
The conformational flexibility of GPCRs has been a nightmare for crystallographers, as (i) it
considerably complicates obtaining crystals of receptors in a well-defined conformation and (ii) it
probably contributes to their high sensitivity to detergents. Crystallizing GPCRs has required heroic
efforts at stabilization, which typically have involved (i) crystallizing in the presence of a high-affinity
ligand, in order to try to shift the equilibrium as completely as possible toward a single form, while
stabilizing it against conformational excursions and (ii) various attempts at further stabilization,
including binding of antibody F ab fragments, or camelid heavy chain antibodies (nanobodies), fusion
of phage T4 lysozyme in the large cytosolic loop that links TM helices 5 and 6, or “conformational
thermostabilization,” which aims at accumulating mutations that improve the homogeneity and
stability of the target receptor in detergent solution (reviewed in Rosenbaum et al. 2009; Tate 2012).
Mutants can be selected either by directed evolution (Sarkar et al. 2008; Dodevski and Plückthun
2011) or by systematically scanning the sequence for stabilizing mutations and then combining enough
of them to achieve the desired stability. We will take as an illustration the latter approach, whose
principle and main results are reviewed with great clarity by one of its main promoters, Christopher
G. Tate, in Tate (2012), on which the following summary is based (for an update, see Magnani et al.
2016).
86
2 Extracting Membrane Proteins from Their Native Environment
and their functions are completely different. Among GPCRs, rhodopsin is also peculiar in having a
very low level of basal activity (see below), which endows our eyes with the possibility to detect even
very weak visual signals. This feature turned out to be important in permitting its crystallization,
because it reduces conformational variability. Other than that, rhodopsin is an archetypical GPCR. It
was the first one whose 3D crystallographic structure was ever established, in 2000, by Krzysztof
Palczewski and colleagues (Palczewski et al. 2000).
The signaling functions of GPCRs are illustrated in Fig. 2.16 in the case of the β 2 adrenergic
receptor, whose natural ligands are adrenaline and noradrenaline. Similar in this respect to the
pentameric ligand-gated ion channels (pLGICs), which we have discussed in Chap. 1 (§ 1.6.2),
GPCRs are in equilibrium between two or more conformational states: binding of an extracellular
ligand shifts the equilibrium toward the state(s) with the highest affinity for the ligand. However, at
variance with pLGICs, conformational transitions do not control the opening and closing of a TM
channel. Rather, the reorganization of the protein around the ligand-binding site is coupled, through the
TM region, to the opening of a cleft in the cytoplasmic face of the receptor. Cytosolic proteins bind
to it and propagate and amplify the signal. Two main routes are the activation of stimulatory or
inhibitory heterotrimeric G proteins, on the one hand, and coupling to arrestin, on the other. Both of
them are schematized in Fig. 2.16, along with some of the downstream cellular events. An important
feature of GPCRs is that most of them are delicately poised between the inactive and active states, in
which case they exhibit a basal level of activity (constitutive activity). As symbolized in the inset to
Fig. 2.16, they can therefore be regulated both up and down. “Agonists,” which have much more
affinity for the active state, can turn them on completely. “Partial agonists,” which have only a
moderate preferential affinity for the active state, achieve only partial activation even at saturation.
“Inverse agonists” favor the inactive state and decrease the activity below the constitutive level.
Finally, “neutral agonists” block the binding site – and, thereby, prevent regulation by other ligands
– without themselves affecting the conformational equilibrium. In the case of rhodopsin, the cofactor
in its resting conformation, 11-cis-retinal, is a very potent inverse agonist, whereas the light-induced
isomer, all-trans-retinal, is an agonist.
The view just presented of regulation being achieved simply by modulating the equilibrium
between two conformational states is, however, widely oversimplified, GPCRs being able “to sample a
continuum of conformations with relatively closely spaced energies” (see Kobilka and Deupi 2007,
Rosenbaum et al. 2009, and references therein).
The conformational flexibility of GPCRs has been a nightmare for crystallographers, as (i) it
considerably complicates obtaining crystals of receptors in a well-defined conformation and (ii) it
probably contributes to their high sensitivity to detergents. Crystallizing GPCRs has required heroic
efforts at stabilization, which typically have involved (i) crystallizing in the presence of a high-affinity
ligand, in order to try to shift the equilibrium as completely as possible toward a single form, while
stabilizing it against conformational excursions and (ii) various attempts at further stabilization,
including binding of antibody F ab fragments, or camelid heavy chain antibodies (nanobodies), fusion
of phage T4 lysozyme in the large cytosolic loop that links TM helices 5 and 6, or “conformational
thermostabilization,” which aims at accumulating mutations that improve the homogeneity and
stability of the target receptor in detergent solution (reviewed in Rosenbaum et al. 2009; Tate 2012).
Mutants can be selected either by directed evolution (Sarkar et al. 2008; Dodevski and Plückthun
2011) or by systematically scanning the sequence for stabilizing mutations and then combining enough
of them to achieve the desired stability. We will take as an illustration the latter approach, whose
principle and main results are reviewed with great clarity by one of its main promoters, Christopher
G. Tate, in Tate (2012), on which the following summary is based (for an update, see Magnani et al.
2016).
86
2 Extracting Membrane Proteins from Their Native Environment