Chloroethylclonidine and 2-Aminoethyl Methanethiosulfonate Recognize Two Different Conformations of the Human α2A-Adrenergic Receptor*

The substituted cysteine-accessibility method and two sulfhydryl-specific reagents, the methane-thiosulfonate derivative 2-aminoethyl methanethiosulfonate (MTSEA) and the α2-adrenergic receptor (α2-AR) agonist chloroethylclonidine (CEC), were used to determine the relative accessibility of engineered cysteines in the fifth transmembrane domain of the human α2A-AR (Hα2A). The second-order rate constants for the reaction of the receptor with MTSEA and CEC were determined with the wild type Hα2A (cysteine at position 201) and receptor mutants containing accessible cysteines at other positions within the binding-site crevice (positions 197, 200, and 204). The rate of reaction of CEC was similar to that of MTSEA at residues Cys-197, Cys-201, and Cys-204. The rate of reaction of CEC with Cys-200, however, was more than 5 times that of MTSEA, suggesting that these compounds may interact with two different receptor conformations. MTSEA, having no recognition specificity for the receptor, likely reacts with the predominant inactive receptor conformation (R), whereas the agonist CEC may stabilize and react preferentially with the active receptor conformation (R*). This hypothesis was consistent with three-dimensional receptor-ligand models, which further suggest that α2A-AR activation may involve the clockwise rotation of transmembrane domain 5.

more) conformations or allosteric states, R and R* (5). In the absence of ligand, the inactive state R predominates. Agonist binding stabilizes the receptor protein in its active state R*, promoting its coupling with G-proteins. The resulting G-protein activation initiates a cascade of intracellular events leading to physiological responses.
Recently, a technique utilizing charged methanethiosulfonate derivatives to probe the accessibility of substituted cysteine residues has been used to study the structure of the dopamine D2 and the ␤ 2 -adrenergic receptors (4, 6 -8). This method identifies residues that form the surface of the binding-site crevice by replacing consecutive amino acid residues in the membranespanning segments with cysteine, one at a time. The sulfhydryl side chain of a cysteine can face the water-accessible bindingsite crevice, the interior of the protein, or the lipid bilayer. Sulfhydryls facing the binding-site crevice react rapidly with sulfhydryl-specific methanethiosulfonate derivatives such as 2-aminoethyl methanethiosulfonate (MTSEA), with formation of a covalent bond with the free sulfhydryl group of an accessible cysteine residue (4,8). MTSEA has no preference for receptor conformations, and it reacts, therefore, with the predominant (inactive) form of the receptor (R). In this method, agonists cannot be used to stabilize and thereby to study the structure of the active receptor species, because the presence of an agonist in the binding pocket would interfere with the access of MTSEA to the engineered cysteines. The activated state of the receptor (R*) can, however, be detected with MTSEA by using a constitutively active receptor mutant (7). This approach, combined with molecular modeling, has suggested rotation and/or tilting of the TM6 segment of the ␤ 2 -AR as a consequence of receptor activation (7).
We recently used the agonist chloroethylclonidine (CEC) to map the structure of TM5 of the human ␣ 2A -AR with a modified substituted cysteine-accessibility method (9). The reactive aziridinium ion derivative of CEC forms a covalent bond with the sulfhydryl side chain of a cysteine exposed in the bindingsite crevice and accessible to CEC. This method has the advantage of introducing recognition specificity by using a thiolreactive group incorporated into an affinity ligand of the target receptor. In functional studies, CEC irreversibly activates ␣ 2 -AR, both prejunctionally in the rat vas deferens and postjunctionally in the dog saphenous vein (10,11), suggesting that the compound is covalently tethered to an endogenous cysteine near the binding site.
In our previous study, TM5 of H␣2A was shown to be consistent with an ␣-helix with residues Val-197, Cys-201, and Ser-204, pointing into the binding-site crevice, residues Ile-198, Ser-199, Ile-202, and Gly-203, facing the lipid bilayer, and Ser-200, pointing partly toward TM4 (9). Surprisingly, the pattern of accessibility to MTSEA in TM5 of the dopamine D2 receptor was not consistent with a fixed ␣-helical structure for TM5 (8). A possible explanation for these results is that this region changes conformation. Because this region is known to be directly involved in agonist binding and because the intracellular end of TM5 interacts with G-proteins, conformational changes in TM5 may be critical to the activation mechanisms of these receptors.
To explore such a conformational change in the H␣2A receptor, we have compared the rates of reaction of the agonist CEC and of MTSEA with accessible cysteines at positions 197, 200, 201, and 204 in TM5 of H␣2A. Nonaccessible cysteines at positions 202 and 203 and a H␣2ASer201 mutant (containing no substituted cysteines) were used as controls. MTSEA reacted with accessible cysteines with the following rank order of reactivity: Cys-200 Ͻ Cys-201(wt) Ͻ Cys-197 ϭ Cys-204. In contrast, for CEC, the cysteine at position 200 was found to be the most reactive (rank order of the rate constants: Cys-204 ϭ Cys-201(wt) ϭ Cys-197 Ͻ Cys-200). Our findings suggest that the agonist CEC and MTSEA may recognize two different receptor conformations, with MTSEA reacting with the predominant inactive form and CEC reacting with the agonist-dependent active conformation of TM5 of H␣2A. Such a hypothesis is consistent with three-dimensional models of the receptor-MTSEA and receptor-CEC complexes.
Mutagenesis and Expression Vectors-Site-directed mutagenesis was performed utilizing the Altered Sites II in vitro mutagenesis system (Promega, Madison, WI) as described previously (9). The wild type H␣2A and the mutated receptor cDNAs were subcloned into the KpnI/ BamHI sites of the expression vector pREP4 (In Vitrogen, NV Leek, The Netherlands).
Cell Culture and Transfections-Adherent CHO cells (American Type Culture Collection, Manassas, VA) were cultured as reported previously (9). The pREP4-based expression constructs were transfected into cells using the Lipofectin reagent kit (Life Technologies, Inc.) (9). Hygromycin B (Roche Molecular Biochemicals)-resistant (550 g/ ml) cell cultures were examined for their ability to bind the ␣ 2 -AR antagonist [ 3 H]RX821002. The transfected cells chosen for further experiments were subsequently maintained in 200 g/ml hygromycin B.
Reactions with CEC and MTSEA-Six large culture flasks (175 cm 2 ) of CHO cells were harvested into chilled phosphate-buffered saline, pelleted, washed, suspended in 2.5 ml of ice-cold 50 mM K ϩ -phosphate buffer (pH 7.4 at 21°C), and homogenized with an Ultra-Turrax homogenizer (model T25, Janke & Kunkel, Staufen, Germany; setting 9500 rpm, twice for 10 s). Aliquots (45 l) of cell homogenate were incubated for 10 min at room temperature. Five l of freshly prepared CEC or MTSEA dilutions were added to each vial to reach the appropriate final concentration. Reactions took place at room temperature for 15 min (CEC) or 2 min (MTSEA); the reaction mixtures were then diluted 20-fold with 50 mM K ϩ -phosphate buffer (pH 7.4 at 21°C). Residual ␣ 2 -AR binding was assessed by incubating the homogenate (30 -60 g of protein/assay tube) with 2.5 nM [ 3 H]RX821002. Nonspecific binding was determined by including 10 M phentolamine in parallel assays.
The second-order rate constant (k) for the reaction of CEC or MTSEA with wild type H␣2A and each susceptible mutant was estimated by determining the extent of the reaction after a fixed time, 15 min (CEC) or 2 min (MTSEA), with seven concentrations of CEC (0.5, 1, 5, 10, 50, 100, and 500 M) or MTSEA (5,20,50,100,150,200, and 250 M) as previously reported (8). The fraction remaining of the initial total binding, Y, was fit to (span ϫ e -kct ) ϩ plateau, where k is the second-order rate constant (M Ϫ1 s Ϫ1 ), c is the concentration of CEC or MTSEA (M), t is the reaction time (s), and plateau is the fraction of binding remaining at saturating concentrations of reagent. For Cys-197, Cys-200, and Cys-204, the highest concentration of MTSEA and of CEC produced nearly complete inhibition of ligand binding (see below). Thus, for all of these fits, the plateau was set to a constant value of 0, and the span was correspondingly fixed to 1 because binding in the absence of added reagent is by definition 100%. The plateau was not fixed for the other CEC fits, because the data did not show complete inhibition at the highest concentration of CEC tested. For these fits the span was set equal to 1 Ϫ plateau. The results were analyzed with GraphPad Prism (GraphPad Software, San Diego, CA).
Measurement of Change in Extracellular Acidification Rate-Extracellular acidification rates were measured using the Cytosensor® Microphysiometer (Molecular Devices Corp., Menlo Park, CA) (13). CHO cells were seeded into 12-mm capsule cups at 3 ϫ 10 5 cells/cup and incubated in 5% CO 2 at 37°C for 20 h. The capsule cups were loaded into the sensor chambers and perfused with running medium (bicarbonate-free ␣-minimum essential medium supplemented with 2 mM glutamine, 26 mM NaCl, 50 units/ml penicillin, and 50 g/ml streptomycin) at a flow rate of 100 l/min. CEC and UK 14,304 were diluted with the running medium. Following base-line stabilization, cells were exposed to agonists for 15 min, and a 30-min wash period was employed between successive agonist exposures. The alterations in the extracellular acidification rate were calculated as the difference between the maximum effect after agonist exposure and the average of three measurements taken immediately before agonist addition.
Molecular Modeling-The H␣2A receptor model contains seven transmembrane helices, and it was built using the computer program MODELLER 4.0 (14). The model is based on a template (15) generated using electron density of the frog rhodopsin (16) and the analysis of ϳ500 sequences of G-protein-coupled receptors (15). Receptor models for each mutant were generated by replacing corresponding residues at positions 197, 200, or 204, one at a time, with a cysteine, and by replacing Cys-201 with a serine.
CEC was docked manually in the binding-site crevice into a position that allows hydrogen bonding between the protonated nitrogen atom in the imidazoline ring of CEC and Asp-113 in TM3 of the receptor. Simultaneously, a covalent bond was created between CEC and the cysteine residue in TM5. Each receptor-CEC complex was then energyminimized using CHARMm (17) (Molecular Simulations Inc., San Diego, CA) using harmonic constraints for backbone atoms to maintain the helical structures of the TMs. Energy minimizations were also performed without constraints to completely relax the structure of the complex. MTSEA was docked into the same region of the binding-site crevice as CEC, and a disulfide bond between the sulfur atom of MTSEA and the cysteine residue of the receptor was created.

Site-directed Mutagenesis and Transfections-
The introduced mutations were confirmed, and the absence of secondary mutations was verified by dideoxy sequencing of doublestranded DNA. Mutated and wild type receptors were expressed in CHO cells. Hygromycin B-resistant cell cultures were examined for their ability to bind the ␣ 2 -AR antagonist radioligand [ 3 H]RX821002. Three clones from each transfection expressing the expected receptor were isolated for the preliminary experiments, and one cell line from each transfection was expanded for further experiments and subsequently maintained in 200 g/ml hygromycin B.
Ligand  (9) and is also shown in Table I. In all investigated cell lines expressing wt and mutant receptors, the addition of CEC to competition binding assays inhibited specific binding of 2.5 nM [ 3 H]RX821002 with steep monophasic competition curves (9). The concentration of CEC that inhibited specific [ 3 H]RX821002 binding by 50% (IC 50 ) was used to calculate apparent K i values (inhibition constant) according to the Cheng-Prusoff equation (19). Competition binding analysis of the control mutants H␣2ASer201, H␣2ASer201Cys202, and H␣2ASer201Cys203 containing no exposed cysteines in the binding cavity in our receptor model, indicated that the lack of irreversible inhibition of binding by CEC in these receptors was not due to a lack of binding affinity but rather to the absence of accessible cysteine residues on the surface of the binding-site crevice (apparent K i ϭ 260 Ϯ 32, 345 Ϯ 38, and 808 Ϯ 37 nM, respectively) ( Table I).
Reactions with CEC- Fig. 1A illustrates the inhibition of antagonist binding produced by CEC. Binding to the mutants Cys-197, Cys-200, and Cys-204 is very sensitive to CEC, whereas the mutants Ser-201, Cys-202, and Cys-203 are relatively resistant to CEC, and the wild type receptor Cys-201 shows intermediate sensitivity. The rate and extent of receptor inactivation by CEC with cysteine substitutions for Ile-202 and Gly-203 was comparable with that of H␣2ASer201, which contained no substituted cysteines, suggesting that residues 202 and 203 face the lipid bilayer or are closely packed against other membrane-spanning segments.
To quantitate the susceptibility of the engineered cysteines to CEC, the second-order rate constants for reaction with these compounds were determined as described under "Experimental Procedures" (Table II). It is apparent from these data ( Fig. 1) that H␣2ASer201 is not resistant to the effects of CEC and that it contains, therefore, one or more relatively slowly reacting cysteines. For all of the mutants except H␣2Awt(Cys201), however, the simple single exponential decay function gave a reasonable fit of the CEC data, especially at low concentrations of CEC (Fig. 1A). This is likely the case because Cys-197, Cys-200, and Cys-204 are so reactive and because reaction at these positions produces such a large inhibition of binding that the relatively slowly reacting endogenous cysteines are nearly irrelevant to the resulting pattern of inhibition. For the H␣2wt(Cys201) receptor, however, the plateau determined by this fit was substantially greater than the experimentally determined value, even though the fit is consistent with the inhibition seen at low concentrations of CEC. If the plateau was fixed to 0, however, a single exponential fit provided a gross underestimate of the rate of reaction (data not shown), a conclusion based on the significant inhibition of binding to this receptor observed at even very low concentrations of CEC. The data for the wild type receptor are better fit by a 2-site exponential decay function, but the variation in the rates obtained for the very reactive component for this receptor (and for the other mutants) was too great to make this a reliable method for analyzing the data (data not shown). Thus the value shown in Table II is an approximation of the rapid rate of reaction with Cys-201, and the remaining inhibition is likely because of reaction with the other more slowly reacting cysteines in H␣2ASer201.
Reactions with MTSEA-MTSEA inhibited binding to all of the mutants (Fig. 1B), with Cys-202 and Cys-203 being the most resistant, consistent with the relative inaccessibility of these cysteines. The results for the Ser-201 mutant were not accurately described by an exponential decay function; this receptor was relatively resistant to MTSEA up to a concentration of 125 M but showed quite significant inhibition at higher  (19) (although this calculated apparent K i results from a combination of reversible and irreversible inhibition and is not a true K i as typically determined for reversible binding, it nevertheless supports the argument that CEC has substantial affinity for all the mutants tested). The characterization of some of these cell lines has been reported elsewhere (9  MTSEA concentrations. Consequently, no estimate for the second-order rate constant of inactivation of this receptor with MTSEA could be computed. The reasons for this unexpected inhibitory pattern remain unclear but may reflect the reactivity of other cysteines present in the Ser-201 background. In Fig.  1B the MTSEA results for the Ser-201, Cys-202, and Cys-203 mutants are presented without curve-fitting. Reasonable single exponential decay fits were obtained for Cys-197, Cys-201, and Cys-204. For Cys-200, the data also show a "lag" at low concentrations, similar but less pronounced than that seen with Ser-201 (see above), and thus the fit is less satisfactory than that seen with the other three reactive residues. This results in an overestimation of the rate of reaction of Cys-200 with MTSEA, although it is impossible to determine the exact rate of reaction of this substituted cysteine in the presence of the Ser-201 background.
Receptor-mediated Changes in Extracellular Acidification Rate-CEC increased the extracellular acidification rate in CHO cells expressing H␣2Awt in a concentration-dependent and irreversible manner (Fig. 2). After CEC treatment followed by extensive washing, the reversible ␣ 2A -AR agonist UK 14,304 was unable to further increase the acidification rate, consistent with the complete reaction of the entire receptor population with the covalently tethered agonist CEC and persistent activation. Exposure of CHO cells expressing H␣2Awt and H␣2ASer201Cys200 receptors to UK 14,304 resulted in reversible increases in the rate of extracellular acidification (Fig. 3). These results confirmed the efficacy of the tethered agonist and the functionality of the mutant receptor. The functionality of this mutant receptor was also shown in a [ 35 S]GTP␥S binding assay. 2 Receptor Modeling-Our current receptor model is based on an ␣-carbon template of the transmembrane helices in the rhodopsin family of G-protein-coupled receptors (15). Mutant receptor models were built by replacing the corresponding residues. Because only two residues at a time were replaced, and the mutations were quite conservative, such as Cys-201 to Ser and Ser-200 or Ser-204 to Cys or Val-197 to Cys and because radioligand antagonist binding to the mutants was similar to that of wt receptor, it is reasonable to assume that these mutations did not have major effects on the receptor structure; therefore, these residues were simply replaced in the wt model.
CEC was docked manually into the binding-site crevice of the wild type receptor and the mutants in a position that allows hydrogen bonding between Asp-113 in TM3 and the protonated nitrogen atom of the imidazoline ring in CEC. At the same time, the other end of the CEC molecule was able to form a covalent bond between a reactive carbon atom and the sulfhydryl group of the cysteine residue in TM5 of the receptor. The models were then energy-minimized (Fig. 4B). In the H␣2Awt receptor, in the H␣2ASer201Cys197 mutant, and in the H␣2ASer201Cys204 mutant, cysteine residues are readily accessible in the binding-site crevice, and an ideal covalent bond can be formed. In the mutant H␣2ASer201Cys200, the cysteine residue is less accessible, and the predicted covalent bonding distance is longer than in the other cases, suggesting that a conformational change is necessary to shorten this distance and thereby to boost the rate of reaction between CEC and this cysteine. When the minimization is made without constraints, the cysteine residue can move. This allows the orientation of the TM5 helix to change, and makes the cysteine at position 200 more accessible for covalent bonding with CEC (Fig. 4C).
The starting position of MTSEA was similar to that of CEC. Fig. 4A shows the energy-minimized models of the receptor-MTSEA complexes. MTSEA forms ideal disulfide bonds with sulfhydryl residues in the H␣2Awt receptor, the mutant H␣2ASer201Cys197, and the mutant H␣2ASer201Cys204. However, the disulfide bond in the mutant H␣2ASer201Cys200 was slightly longer, and the cysteine residue was less accessible in the binding-site crevice than it is in the H␣2ASer201Cys197, H␣2ASer201Cys204, or H␣2Awt receptor models. As MTSEA has no other points of attachment in the binding-site crevice, energy minimization of the H␣2ASer201Cys200 mutant without constraints did not make the Cys-200 residue more accessible, as was the case with CEC. DISCUSSION Using the substituted cysteine-accessibility method and irreversible binding of the agonist CEC, we have previously mapped the residues accessible in the binding-site crevice in TM5 of human ␣ 2A -AR (9). Surprisingly, this pattern of accessibility was quite different from that observed in TM5 of the homologous dopamine D2 receptor (8). We postulated that the differences in accessibility might relate to the use of the agonist CEC. CEC has been used to discriminate between ␣ 1 -AR subtypes in functional assays (20,21), and it has also been shown to irreversibly bind to H␣2A and H␣2C but not H␣2B (9,22). In functional studies with tissue preparations, CEC has been shown to behave as an ␣ 2 -AR agonist, both at prejunctional and postjunctional ␣ 2 -ARs (10,11,23,24). However, the agonist action of CEC has previously not been shown in recombinant cell lines expressing only one defined receptor subtype. In this work, the agonist effects of CEC in recombinant CHO cells expressing H␣2A were confirmed with Cytosensor® microphysiometry. These results suggest that CEC is covalently tethered to the side chain of Cys-201 in such a manner that it produces irreversible receptor activation.
In the current study, the rate of reaction of engineered cysteines in TM5 of human ␣ 2A -AR was determined using two different sulfhydryl-specific agents, CEC and MTSEA. The rates of reaction of CEC and MTSEA were similar at the wt receptor (Cys-201). The rates of reaction of the two reagents were likewise similar, with cysteine substituted for residues 197 and 204. In contrast, the rate of reaction with a cysteine at 2 J. M. Peltonen, unpublished observations.

TABLE II
Second-order rate constants (k) of inactivation of the wild type and mutated human ␣ 2A -ARs by MTSEA and CEC The data represent the mean Ϯ S.E. of three to six separate experiments performed in triplicate. k(mut)/k(wt) and k(mut)/k(H␣2ASer201) were obtained by dividing each k value by the k determined for the wild type or H␣2ASer201 receptor. The ratio CEC/MTSEA represents the relative reactivity of CEC and MTSEA at each receptor. ND, not reliably determined. position 200 was more than 5-fold faster for CEC than for MTSEA. The true difference in rate is even greater than 5-fold, because the fit obtained for the inhibition of Cys-200 by MTSEA, as discussed above, significantly overestimates the rate of reaction of this residue with MTSEA. At particular positions including the reaction of Cys-200 with MTSEA, a simple single-exponential decay function is clearly inadequate to deal with the complex molecular interactions underlying the experimental results. There must be other sites in the receptor protein in addition to the investigated cysteines that are susceptible to inactivation by CEC and MTSEA, and these sites may also interact with each other. In addition, residual CEC and MTSEA present in the ligand binding assays may further complicate the situation, although experiments in which the reagents were removed by extensive washing gave similar results (data not shown). Nonetheless, both the qualitative results and the derived second-order rate constants support a similar reactivity of Cys-197, Cys-201, and Cys-204 to both CEC and MTSEA, and this contrasts with the substantially greater rate of reaction of Cys-200 with CEC than with MTSEA. The concept that G-protein-coupled receptors may possess some degree of spontaneous or constitutive activity and thus activate their cognate G-proteins in the absence of receptor agonist has recently become widely accepted (25,26). G-protein-coupled receptors are thought to exist in an equilibrium between two (or more) interconvertible allosteric states, active (R*) and inactive (R) conformations. Binding of a receptor agonist causes an increase in the ratio of R* to R (5,25). It has also been suggested that the R state can isomerize, not to a single active conformation, but to multiple different ligandspecific active conformations (27,28). This multistate model suggests that the biological efficacy of an agonist may be a consequence of the qualitative and quantitative preference of the ligand for conformational variant forms of the receptor rather than just affecting the equilibrium between only two alternative states (28). MTSEA has no significant affinity for the receptor and, thus, reacts with cysteines accessible in the predominant resting receptor form (R). The activated state of a receptor can only be investigated with this reagent by using constitutively active mutant receptors (7). When the agonist CEC binds to TM5 of H␣2A, however, it likely stabilizes one or several of the active conformations (R*) and reacts with engineered cysteines accessible in the active conformation. Thus, the difference in the reactivity of CEC and MTSEA toward the substituted cysteine in position 200 in the H␣2ASer201Cys200 mutant may result from a difference in the accessibility of this residue in the active and the inactive forms of the receptor.
There are several other factors that might also contribute to a difference in the rate of reaction of a cysteine with CEC and with MTSEA. Local steric factors might favor the reaction of one reagent. Although this is difficult to rule out, MTSEA is somewhat smaller than CEC, and such a selective interference with the reaction of MTSEA would be surprising. Because the methanethiosulfonate derivatives react vastly faster with the thiolate than with the thiol form of their substrates (29), changes in the ionization state of the sulfhydryl would have a significant impact on the rate of reaction of MTSEA. It is significant to note, however, that the relative difference in rates results in significant part from the very rapid reaction of CEC at the Cys-200 position. The simplest explanation for this rapid reaction is the proximity of Cys-200 to the reactive moiety of bound CEC.
We have previously used a H␣2A receptor model based on the electron diffraction structure of bacteriorhodopsin (30). The model was consistent with the ␣-helical structure and orientation of TM5 in H␣2A-AR (9), and the observed helical structure is also consistent with our current revised receptor model based on the ␣-carbon template for the TM helices in the rhodopsin family of G-protein-coupled receptors (15). When CEC was docked into the receptor model based on the bacteriorhodopsin structure, residues 200, 201, and 204 were equally accessible to form the covalent carbon-sulfur bond with CEC or with MTSEA. In our current receptor model, however, the TM3 helix is tilted into the binding-site crevice, thereby limiting the size of the binding site and the access of ligands to the residue at position 200. Energy minimizations without constraints indicated that the presence of CEC may influence the orientation of position 200, thereby optimizing the interactions between the receptor and CEC, i.e. both hydrogen bonding of CEC with Asp-113 in TM3 and covalent bonding between CEC and Cys-200 in TM5 are simultaneously favored. Our results with MTSEA support the prediction that Cys-200 in the H␣2ASer201-Cys200 mutant is not quite as readily accessible as Cys-201 in H␣2Awt, Cys-197 in the H␣2ASer201Cys197 mutant, and Cys-204 in the H␣2ASer201Cys204 mutant. Removing constraints during the energy minimization does not cause any conformational changes in TM5 in the H␣2ASer201Cys200-MTSEA complex.
These results suggest that CEC and MTSEA may recognize different conformations of the receptor, with CEC accessing the cysteine at position 200 more readily than MTSEA. This may be because of agonist-dependent clockwise rotation of TM5 of the activated human ␣ 2A -AR. The current receptor model thus appears to more accurately predict the three-dimensional receptor structure than the previous bacteriorhodopsin-based receptor model.