An Activation Switch in the Rhodopsin Family of G Protein-coupled Receptors

We aimed at understanding molecular events involved in the activation of a member of the G protein-coupled receptor family, the thyrotropin receptor. We have focused on the transmembrane region and in particular on a network of polar interactions between highly conserved residues. Using molecular dynamics simulations and site-directed mutagenesis techniques we have identified residue Asn-7.49, of the NPxxY motif of TM 7, as a molecular switch in the mechanism of thyrotropin receptor (TSHr) activation. Asn-7.49 appears to adopt two different conformations in the inactive and active states. These two states are characterized by specific interactions between this Asn and polar residues in the transmembrane domain. The inactive gauche+ conformation is maintained by interactions with residues Thr-6.43 and Asp-6.44. Mutation of these residues into Ala increases the constitutive activity of the receptor by factors of ∼14 and ∼10 relative to wild type TSHr, respectively. Upon receptor activation Asn-7.49 adopts the trans conformation to interact with Asp-2.50 and a putatively charged residue that remains to be identified. In addition, the conserved Leu-2.46 of the (N/S)LxxxD motif also plays a significant role in restraining the receptor in the inactive state because the L2.46A mutation increases constitutive activity by a factor of ∼13 relative to wild type TSHr. As residues Leu-2.46, Asp-2.50, and Asn-7.49 are strongly conserved, this molecular mechanism of TSHr activation can be extended to other members of the rhodopsin-like family of G protein-coupled receptors.

We aimed at understanding molecular events involved in the activation of a member of the G protein-coupled receptor family, the thyrotropin receptor. We have focused on the transmembrane region and in particular on a network of polar interactions between highly conserved residues. Using molecular dynamics simulations and site-directed mutagenesis techniques we have identified residue Asn-7.49, of the NPxxY motif of TM 7, as a molecular switch in the mechanism of thyrotropin receptor (TSHr) activation. Asn-7.49 appears to adopt two different conformations in the inactive and active states. These two states are characterized by specific interactions between this Asn and polar residues in the transmembrane domain. The inactive gauche؉ conformation is maintained by interactions with residues Thr-6.43 and Asp-6.44. Mutation of these residues into Ala increases the constitutive activity of the receptor by factors of ϳ14 and ϳ10 relative to wild type TSHr, respectively. Upon receptor activation Asn-7.49 adopts the trans conformation to interact with Asp-2.50 and a putatively charged residue that remains to be identified. In addition, the conserved Leu-2.46 of the (N/S)LxxxD motif also plays a significant role in restraining the receptor in the inactive state because the L2.46A mutation increases constitutive activity by a factor of ϳ13 relative to wild type TSHr. As residues Leu-2.46, Asp-2.50, and Asn-7.49 are strongly conserved, this molecular mechanism of TSHr activation can be extended to other members of the rhodopsin-like family of G protein-coupled receptors.
Genome sequencing projects have identified the G proteincoupled receptor (GPCR) 1 family as one of the largest class of proteins with more than 800 human sequences. These sequences have been classified into five main families, glutamate, rhodopsin, adhesion, frizzled/taste2, and secretin (1). GPCRs are involved in passing chemical signals across the cell membrane. The incoming signal may arrive in the form of neurotransmitters, peptides, divalent cations, proteases, hormones, and sensory stimuli such as photons and gustatory or odorant molecules (2,3). Atomic level details of a three-dimensional structure of a GPCR are only known for the inactive form of rhodopsin, the light photoreceptor protein of rod cells (4,5). Rhodopsin is formed by an extracellular N terminus, seven ␣-helices, which cross the cellular membrane (TM 1-TM 7) connected by hydrophilic loops, and a cytoplasmic C terminus containing an ␣-helix (HX 8) parallel to the cell membrane. The overall structure of the helical bundle seems common in the rhodopsin-like GPCR family because of the large number of conserved sequence patterns in these 7 TM segments (6): GN in TM 1, (N/S)LxxxD in TM 2, (D/E)RY in TM 3, CWxP in TM 6, and NPxxY(x) 5,6 F in TM 7 and HX 8, among others.
Activation of GPCRs is commonly discussed in terms of the extended ternary complex model, which proposes an equilibrium between inactive and active states (7). Recent advances in the ␤ 2 -adrenergic receptor have shown that agonist binding, leading to the active form of the receptor, is a multistep process (8 -10). Agonist binding or constitutively active mutations disrupt a series of constraining intramolecular interactions that keep the receptor in the inactive conformation. Recently identified constraints within the transmembrane domain are i) the conformational transition of Trp of the CWxP motif in TM 6 (11,12), ii) the interaction of Asn of the NPxxY motif in TM 7 with either residues located in TM 6 (13) or the backbone of TM 6 through a water molecule (14), iii) the ionic lock between Arg of the highly conserved (D/E)RY motif in TM 3 and a partly conserved Asp/Glu residue in the cytoplasmic end of TM 6 (15), and iv) the aromatic-aromatic interaction between Tyr of the NPxxY motif in TM 7 and an aromatic residue located in HX 8 (16,17). Release of these constraints induce rigid body motions of several if not all transmembrane helices (18,19).
The thyrotropin receptor (TSHr), together with the luteinizing hormone/chorionic gonadotropin and follicle-stimulating hormone receptors, is a member of the glycoprotein hormone receptor family of GPCRs. These receptors are characterized by a large N-terminal ectodomain, responsible for high affinity binding of the hormone and by a rhodopsin-like helical bundle, responsible for signal transduction (20,21). An interesting peculiarity of the TSHr is the wide spectrum of mutations capable of increasing its constitutive activity (22), offering a valuable tool to study molecular mechanisms of GPCR activation. We have shown that in the glycoprotein hormone receptor family an Asp residue in TM 6, maintains the highly conserved Asn (in the NPxxY motif in TM 7) in the inactive state through a complex hydrogen bond network (13). Similar constraining interactions have been reported for the luteinizing hormone/ chorionic gonadotropin (23,24). Moreover, this Asn residue is crucial in the activation mechanism by both the natural agonist (13) and activating mutations, including those affecting the (E/D)RY motif that release the ionic lock between TM 3 and 6 (25). Interestingly, mutation of this Asn to Ala in the C5a receptor abrogates receptor signaling (26). Thus, Asn of the NPxxY motif seems to be a key residue in the stabilization of both the inactive and active states of other receptors as well. We aim in this work to study the interactions of the Asn side chain in the inactive and active state of the TSHr by a combination of molecular modeling, functional assays, and site-directed mutagenesis.

Molecular Modeling and Molecular Dynamics Simulations
A model of the transmembrane domain of the TSHr plus helix 8 that expands parallel to the membrane was constructed by homology modeling using the crystal structure of bovine rhodopsin (PDB code 1L9H) as a template (14). The molecular models for the mutant receptors containing the single D2.50N (standardized nomenclature 2 (27)) or double D2.50N/N7.49D substitutions were built from the derived model of WT TSHr, by changing the atoms implicated in the amino acid substitutions by interactive computer graphics. These structures were placed in a rectangular box (ϳ60 ϫ 72 ϫ 70 Å in size) containing a lipid bilayer (60 molecules of palmitoyloleoylphosphatidylcholine and ϳ7650 molecules of water in addition to the receptor structure), resulting in a final density of ϳ1.0 g cm Ϫ3 . The receptor-lipid bilayer systems were subjected to 500 iterations of energy minimization and then heated to 300 K in 15 ps. This was followed by an equilibration period (15-250 ps) and a production run (250 -500 ps) at constant pressure with anisotropic scaling, using the particle mesh Ewald method to evaluate electrostatic interaction. During the processes of minimization, heating, and equilibration a positional restraint of 10 kcal mole Ϫ1 Å Ϫ2 was applied to the C ␣ atoms of the receptor structure. This simulation protocol seems adequate to adapt the rhodopsin template to the structural requirements of the wild type and mutants receptor side chains, with the aim of understanding the local helix-helix interactions that keep the receptor in the inactive state. Structures were collected for analysis every 10 ps during the production run (25 structures/simulation). The molecular dynamics (MD) simulations were run with the Sander module of AM-BER 8 (28), the ff99 force field (29), SHAKE bond constraints in all bonds, a 2 fs integration time step, and constant temperature of 300 K coupled to a heat bath. The molecular electrostatic potentials were calculated with GRASP (30).

Reagents
Plasmid pBluescript SK ϩ was from Stratagene, plasmid pSVL was from Amersham Biosciences, restriction enzymes were from Invitrogen and New England Biolabs (Beverly, MA), and Pfu Turbo polymerase from Stratagene. Monoclonal antibody (mAb) BA8, obtained by genetic immunization (31), was directed against a conformational epitope on the TSHr ectodomain. Bovine TSH (bTSH) was from Sigma.

Construction of TSHr Mutants
Mutations were introduced in the hTSHr by site-directed mutagenesis as described previously (32). The primers are available upon request. Amplified fragments of SK ϩ hTSHr containing the appropriate mutations were subcloned in pSVL hTSHr cDNA (between XhoI and XbaI restriction sites). All constructs were amplified in DH5␣FЈ-electrocompetent cells, and recombinant DNA from selected clones was purified and sequenced for confirmation of PCR-generated fragments.

Transfection Experiments
COS-7 cells were used for all transient expression experiments, which were performed according to two different protocols (20). Briefly, 300,000 or 2 million cells were seeded in 3.5-or 10-cm Petri dishes, respectively, at day 1 and transfected based on the DEAE-dextran transfection method (32) at day 2. For the 3.5 dishes protocol, 2 days later cells were used for flow immunofluorometry and cAMP determinations; whereas for the 10-cm dishes at day 3, cells were detached by trypsinization, centrifuged, resuspended in 16 ml of culture medium and seeded in 24-well plates (1 ml/well). At day 5, cells were used for flow immunofluorometry and cAMP determinations. Duplicate dishes were used for each assay. Cells transfected with pSVL alone were always run as negative controls.

Cell Surface Expression
Expression of wt and diverse hTSHr mutants was quantified by flow immunofluorometry with the mAb BA8 exactly as described preciously (31). The fluorescence of 10,000 cells/tube was assayed by FACScan FLOW cytofluorometer (BD Biosciences). Cells transfected with pSVL alone and pSVL hTSHr were always run as negative and positive controls, respectively.

Determination of cAMP Production
For cAMP determinations, culture medium was removed 48 h after transfection and replaced by Krebs-Ringer-HEPES buffer for 30 min. Thereafter, cells were incubated for 60 min in fresh Krebs-Ringer-HEPES buffer supplemented with 25 M phosphodiesterase inhibitor Rolipram (Laboratoire Logeais, Paris, France) in the absence or presence of various concentrations of bTSH (Sigma). The medium was discarded, replaced with 0.1 M HCl, and the extracts were dried under vacuum, resuspended in water, and diluted appropriately for cAMP measurements. Duplicate samples were assayed in all experiments; results are expressed in pmol/ml. Because of the linear relationship existing between basal cAMP accumulation and receptor cell surface expression (32), basal cAMP was normalized to cell surface expression for each of the constructs. To this end, specific cAMP accumulation (equal to cAMP of receptor-transfected cells Ϫ cAMP of the pSVLtransfected cells) is divided by the specific FACS value (equal to fluorescence of receptor-transfected cells Ϫ fluorescence of pSVL-transfected cells). Specific basal activity was defined as ϭ (cAMP(receptor) Ϫ cAMP(pSVL))/(FACS(receptor) Ϫ FACS(pSVL)). The values were then normalized to the basal activity of the WT TSHr, arbitrarily set to 1. Concentration-effect curves were fitted with Prism v3.03 (GraphPad Software, Inc., San Diego, CA). Fig. 1a shows a detailed view of the Asp-2.50/Asn-7.49/Asp-6.44 environment of the molecular model of the TSHr in the inactive conformation obtained by unrestrained MD simulations (see "Experimental Procedures"). Importantly, the Wat 1 set of water molecules found in the most recent structure of rhodopsin (1L9H) are included in the model (14). These structural waters mediate a number of interhelical interactions through the highly conserved Asp-2.50 and Asn-7.49 amino acids. The lack of coordination of the polar Asp-2.50 in both rhodopsin and the TSHr leads us to suggest that this set of water molecules is also present in the TSHr and is putatively involved in maintaining the inactive form of the receptor. Wat 1a links TM 2, TM 3, and TM 7, acting as a hydrogen bond donor in the interaction between the side chain of Asp-2.50 and the backbone carbonyl of Ala-3.35 and as a hydrogen bond acceptor in the interaction with the side chain of Asn-7.45. Wat 1b is located between TM 2 and TM 7, linking the backbone carbonyl of Asn-7.45, the backbone N-H amide of Asn-7.49, the side chain of Asp-2.50, and the side chain of Asn-7.49. As shown in Fig. 1a, a network of hydrogen bond interactions restrains the side chain of Asn-7.49 in the 1 ϭ gaucheϩ rotamer con-formation, pointing toward TM 6. In particular, both Thr-6.43 and Asp-6.44 interact with Asn-7.49 through a water molecule (Wat 1c ). Wat 1c is also bridging the side chain of Asn-7.49 and the backbone carbonyl of Leu-6.40. It is important to note that Wat 1c is very mobile during the simulation and can also achieve, in other snapshots, interactions with other residues like Thr-6.43 or Ser-3.39. These proposed networks of interactions partly diverge from the previously described model (13) because of the presence of these functional water molecules. Finally, Asn-7.45 is interacting with Asp-6.44, Cys-6.47, and Wat 1a , whereas Tyr-7.42 is hydrogen bonding, via the OH group, one of the O ␦ atoms of Asp-6.44 and the O ␥ H group of Ser-3.36 (not depicted for clarity). A detailed description of the role of these side chains in maintaining the inactive conformation of the homologous follicle-stimulating hormone receptor has recently been published (33).

Molecular Modeling and Molecular Dynamics Simulations
We have previously suggested that upon receptor activation the side chain of Asn-7.49 undergoes a conformational transition from gaucheϩ to trans (13). Fig. 1b shows the computer model of the TSHr where the side chain conformation of Asn-7.49 has been initially modified from its inactive gaucheϩ conformation to a trans conformation. The recently obtained structure of metarhodopsin I has revealed that early stages of activation does not involve large rigid body movements of helices but local side chain relocations (12). Thus, this modeling exercise only aims at exploring the possible interactions of Asn-7.49 in the active state of the receptor, which could correspond to early stages of the activation process when side chain relocations have not yet been translated into major structural changes. Regardless of these limitations, it seems clear that the conformational transition of Asn-7.49 allows for a direct interaction with Asp-2.50. As a consequence the polar O ␦ atom of Asn-7.49 reorients toward the intracellular side and becomes accessible. During the simulations the positively charged side chain of Arg-6.36, highly conserved within the glycoprotein hormone receptor family, hydrogen bonds the O ␦ atom of Asn-7.49 (not shown). Fig. 1, a-d, also shows the position of the highly conserved side chain of Leu-2.46.

Experimental Results
The role of the residues theoretically predicted to interact with Asn-7.49 in its "inactive" or "active" rotamer conformation was experimentally tested by site-directed mutagenesis and functional assays (Table I, Figs. 2 and 3). The following mutants were engineered.
Thr-6.43 and Asp-6. 44 Mutants-The molecular model depicted in Fig. 1a predicts that both Thr-6.43 and Asp-6.44 are involved in the interaction with Asn-7.49. Thus, we engineered the D6.44A, D6.44N, T6.43A, and the double T6.43A/D6.44N mutant receptors (Figs. 2 and 3). These mutants are expressed at the 26 -96% range (Table I and Figs. 2A and 3A). D6.44N functions as WT TSHr in terms of constitutive activity, whereas D6.44A increases basal activity by a factor of ϳ14 (Table I) as was previously shown (13). Notably, T6.43A also increases basal activity by a factor of ϳ10 relative to WT receptor that is comparable in magnitude to the D6.44A mutant receptor (Table I). Strikingly, with a specific constitutive activity (or SCA) of ϳ3 (see Table I) the double T6.43A/D6.44N mutant is less constitutive than T6.43A (SCA of ϳ10) but higher than the WT-like D6.44N mutant (SCA of 1.1). This suggests that Thr-6.43 has a more significant role in constraining the side chain conformation of Asn-7.49 in the presence of Asp-6.44 than Asn-6.44. All the mutants are able to respond to bTSH.
Asp-2.50 and Asn-7.49 Mutants-The process of TSHr activation appears to require the release of Asn-7.49, from the interactions with the residues in TM 6, toward Asp-2.50 (13). Thus, single D2.50A, D2.50N, and N7.49D, and double D2.50N/ N7.49D mutant receptors were engineered (Table I and Figs. 2 and 3) to explore the hypothesis of a direct Asp-2.50**Asn-7.49 interaction in the active state of the TSHr. These mutants are expressed at the 24 -115% range (Table I and Fig. 2A). D2.50A and D2.50N show a loss in basal activity (Table I and Fig. 2C). Furthermore, the typical concentration-response curve illustrated in Fig. 3C for the D2.50N construct shows higher EC 50 and lower E max values than WT receptor. Thus, Asp-2.50 seems crucial to stabilize the active state of the receptor. On the other side N7.49D shows a significant increase in constitutive activity, reaching more than 11 times that of wild type TSHr (Table  I and Fig. 2C) as was previously shown (13). Notably, addition of the D2.50N mutation on the N7.49D background, making the D2.50N/N7.49D double mutant, reverses completely the phenotype to levels of constitutive activity even below wild type-like behavior (Table I and Fig. 2C). Moreover, these mutants efficiently respond to bTSH (Fig. 2B).
The Arg-6.36 Mutant-Arg-6.36 was replaced by Ala to explore the possibility of a direct interaction between the Asp-2.50**Asn-7.49 pair and Arg-6.36 in the active state of the receptor as suggested during MD simulations. The R6.36A mutant exhibits reduced constitutive activity relative to wild type TSHr (Fig. 2C), but the receptor can be activated by the extracellular ligand (Fig. 2B). This data disregards the possibility of Arg-6.36 as the counterpart to the Asp-2.50**Asn-7.49 pair in the active state of the receptor.
Leu-2.46 Mutants-We also performed the L2.46A, L2.46I, and L2.46W mutations to investigate the role of this highly conserved Leu residue, which belongs to a conserved hydrophobic layer (34), and it is located in the proximity of Asn-7.49 in the TSHr molecular model. The cell surface expression of the L2.46A mutant is only 12% of WT receptor (Table I and Fig.  3A). L2.46A shows a significant increase in basal activity when normalized to surface expression (Fig. 3B). The concentrationresponse curve illustrated in Fig. 3D shows that L2.46A is activated efficiently by TSH with similar EC 50 to WT receptor ( Table I). Introduction of a ␤-branched Ile residue at this 2.46 position only shows a decrease in receptor expression at the cell surface (Table I). Basal activity and ligand-induced cAMP accumulation seem to be unaffected (Fig. 3). In contrast, L2.46W decreases basal activity, and the mutant receptor is almost unable to be activated by the hormone (Fig. 3). DISCUSSION Previous work (13,25) showed the central role of Asn-7.49 in the activation of the TSHr, acting as an on/off switch by adopting two different conformations in the inactive and active states. In the present work we aimed at identifying the partners of Asn-7.49 regulating this conformational transition.
The TD6. 44 (Fig. 1a). Accordingly, both D6.44A and T6.43A single mutants show a significant increase in constitutive activity relative to WT TSHr, by factors in the order of ϳ14 and ϳ10 (Table I and Fig. 2), respectively.
As the TD6.44 motif is specific to the glycoprotein hormone receptor family, the question arises of whether the inactivation of Asn-7.49 through interactions with partners in TM 6, can be extrapolated to the other members of the GPCR family. It has recently been shown that an internal water molecule in the crystal structure of rhodopsin (14) bridges the side chain of Asn-7.49 (Asn-302 in bovine rhodopsin) and the backbone carbonyl of Gly-6.40 (Gly-257), keeping the Asn side chain toward TM 6. This suggests a conserved mechanism where Asn-7.49 of the NPxxY motif is restrained toward TM 6 in the inactive state of all rhodopsin-like GPCRs by intramolecular interactions that diverge among GPCR subfamilies.
The Asp-2.50**Asn-7.49 Interaction in the Active State of the TSHr-The modeling and experimental data presented above, together with previously reported results (13), suggest that the release of the Asn-7.49 side chain, from interactions with the Thr-6.43/Asp-6.44 residues, is a necessary step in TSHr activation. Disruption of these interactions has a large energetic cost that must be compensated by the formation of new stabilizing interactions. We propose Asp-2.50 as the most likely partner of Asn-7.49 in the active state of the TSHr (Fig. 1b). In this mode of interaction the polar O ␦ atom of Asn-7.49 is free and points toward the intracellular side (see below). This carboxylate oxygen has been suggested as necessary in TSHr activation (35).
Some members of the GPCR family receptors exchange the NPxxY motif by DPxxY in TM 7 (6). Several works have suggested a Asp-2.50/Asn-7.49 direct interaction in the active state of GPCRs (36,37). These two residues are found interchanged in the gonadotropin-releasing hormone receptor and engineering such a swap leads to functional receptors in a number of cases (36,37). Notably, this D2.50N/N7.49D swapping in the TSHr leads to a mutant receptor with lower constitutive activity than WT TSHr (Fig. 2). This suggests that, despite the presence of a negative charge at both 6.44 and 7.49 positions, there is a hydrogen bond network constraining Asp-7.49 toward TM 6. Fig. 1c shows the molecular model of the D2.50N/ N7.49D double mutant in the inactive conformation (see "Experimental Procedures"). The side chain of Asp-7.49 is interacting with Asn-2.50 and Thr-6.43 (through Wat 1c ), whereas Asp-6.44 is coordinated by Ser-3.39, Tyr-7.42, and Asn-7.45 (not all depicted in Fig. 1d for clarity). This finding reinforces the hypothesis of the presence of Wat 1c in the Asn-7.49 environment. We suggest that activation of the D2.50N/  (Fig. 1, c and d). In this active conformation the N ␦2 H group of Asn-2.50 would act as a donor in the hydrogen bond interaction with the O ␦1 atom of Asp-7.49. Similarly to WT TSHr, the polar O ␦2 atom of Asp-7.49 would be accessible and pointing toward the intracellular side (Fig. 1, b and d).
The Counterpart of Asp-2.50**Asn-7.49 in the Active State of the TSHr-Our various simulations (Fig. 1) indicate that upon formation of the 2.50**7.49 interaction in WT TSHr or D2.50N and D2.50N/N7.49D mutants the O ␦ atom of Asn-7.49 or Asp-7.49 points toward the intracellular side. Fig. 1, e-g, shows the resulting molecular electrostatic potential of the receptor structure at the accessible surfaces of Asn-7.49 or Asp-7.49 (see "Experimental Procedures"). Clearly, because of the presence of the negatively charged Asp-2.50, the O ␦ atom of Asn-7.49 is more polar than a regular carbonyl of neutral Asn (compare Fig. 1, e with f) and resembles the polarity of the negatively charged O ␦ atom of Asp-7.49 (Fig. 1, compare f with g). Furthermore, although D2.50A is still able to induce cAMP production in the presence of the ligand (Fig. 2), the N7.49A mutant cannot be activated by TSH (13). Thus, it seems reasonable to suggest that Asn-7.49 is a key residue in receptor activation, whereas the role of Asp-2.50 is to increase the reactivity of the O ␦ atom of Asn-7.49 toward an intracellular counterpart (i.e. a positive charge, see below).
Although our MD simulations suggest that the nearby Arg-6.36 (specific to the glycoprotein hormone receptor family) could interact with the O ␦ atom of Asn-7.49, the R6.36A mutant can still be activated by TSH ruling out this arginine as a key counterpart to the Asp-2.50**Asn-7.49 pair in the active state of the receptor. Still, we expect positively charged residue(s) to interacts with the acidic pocket described above.
It had been suggested that the highly conserved Arg-3.50 of the (D/E)RY motif at the bottom of TM 3 also performs a conformational change during the process of receptor activation: the ionic counterpart of Arg-3.50 in the inactive state would be the contiguous D/Glu-3.49, replaced by Asp-2.50 in the active state (38). This appears currently unlikely, as the D2.50A mutant is still activable. However, the charged side chain of Arg-3.50 could be the ionic counterpart of the Asp-2.50**Asn-7.49 pair, interacting with the closer and polar O ␦ atom of Asn-7.49. This hypothesis is also supported by the fact that addition of the N7.49A mutation to the constitutively active E3.49A or E3.49Q mutant receptors, dramatically lowers their activity to levels of wild type TSHr (25). Thus, the release of the Arg-3.50 side chain by either E3.49A or E3.49Q mutations is only stable in the presence of Asn-7.49, which could indicate a direct Arg-3.50**Asn-7.49 interaction. Moreover, a conformational link between Asp-2.50, Arg-3.50, and Tyr-5.58 has been shown to be critical for muscarinic M 3 receptor activation (39). However, it is important to note that large confor-mational changes of TM 3, TM 6, and TM 7 are required to bring the side chains of Arg-3.50 and Asn-7.49 into proximity. Importantly, cross-linking studies of rhodopsin indicate that such a large conformational change of TM 3 toward TM 7 occurs upon activation (40).
Alternatively, the putative positive charge supposed to be the counterpart of Asn-7.49 in the active state could also be located in the G protein. This would be in agreement with previous data suggesting that  in the cholecystokinin B receptor is specifically involved in G q protein activation (41). In this receptor, the N7.49A mutant keeps similar expression levels and high affinity cholecystokinin binding compared with WT receptor. It can form stable complexes with G q upon ligand binding but lacks the ability to activate phospholipase C and protein kinase C.
The Role of the Highly Conserved Leu-2.46 in the Process of TSHr Activation-Leu-2.46 is part of the (N/S)LxxxD motif in TM 2, highly conserved among all the members of the rhodopsin-like family of GPCR. Until now, only one group has reported mutagenesis data on Leu-2.46 mutants in rhodopsin (42). Similar to their findings, the L2.46A single mutant in the TSHr shows a significant increase in constitutive activity by a factor of ϳ13, relative to wild type TSHr (Table I and Fig. 3). The removal of the side chain (mutation into Ala) implies the loss of an interaction important for keeping the inactive state. Leu-2.46 belongs to a cluster of hydrophobic residues (43). It was suggested that to enable the movement of TM 3 and 6, necessary for receptor activation, hydrophobic contacts in this cluster might undergo a rearrangement (34). Madabushi et al. (42) extended this idea by suggesting that the signal originating from ligand binding would be transduced through this core toward the G protein-coupling site.
Considering the effect of the L2.46A mutation in constitutive activity of the TSHr, comparable in magnitude to the T6.43A, D6.44A, or N7.49D mutants, it is tempting to propose an alternative role for Leu-2.46 in the activation mechanism. The allowed conformations of Leu in ␣-helices are restricted to gaucheϩ and trans rotamers (44). In the gaucheϩ rotamer (observed in rhodopsin), the side chain of Leu-2.46 points toward TM 3 and away from the Asn-7.49**Asp-2.50 pair. In trans conformation the side chain of Leu-2.46 would be located halfway between Asp-2.50 and Asn-7.49 and, thus, interfering with the conformational transition of Asn-7.49. Therefore, the bulky and ␥-branched Leu-2.46 could also constrain the conformation of Asn-7.49 toward TM 6 by occupying the space between Asn-7.49 and Asp-2.50 in the TSHr. However, additional knowledge of the residues participating in the network of interactions that stabilize the active conformation of the receptor is indispensable for further understanding the active role of Leu-2.46.
Conclusions-The rhodopsin family of GPCRs is characterized by a number of highly conserved charged and polar residues located within the TM region. Mutagenesis studies indicate that most if not all of these amino acids are involved in the activation mechanism. Our work aimed at understanding the molecular roles and relationships of these residues and their conformational regulation during receptor activation. The highly conserved Asn-7.49 appears to have a well defined inactive conformation, locked by interactions that are specific to each subfamily. Upon activation, it must reorient toward another conserved residue, Asp-2.50. This combination of polar and charged side chains leads to a particular electrostatic landscape, which, in our view, is at the center of receptor activation. The highly negative potential produced in the TM 2-TM 3 region could force reorientation of another highly conserved charge, Arg-3.50, and drive the important conformational reorganization of the TM bundle. Alternatively, negative electrostatics could be directly involved in the interaction with the G protein. As these key residues are common to all rhodopsin-like GPCRs, we believe that these molecular steps form a conserved mechanism.