Most biogenic amine G protein-coupled receptors contain a conserved aspartic acid residue positioned near the intracellular side of the second transmembrane-spanning (TMS) domain that is the primary site of allosteric modulation by sodium ions and pH. Recently, zinc ions and amiloride derivatives were found to allosterically modulate antagonist binding to dopamine receptors. In the current study, the wild-type D4 dopamine receptor showed an 8-fold decrease in zinc affinity in the presence of 120 mM NaCl, but the binding of zinc to the neutral TMS2 D4-D77N mutant was completely sodium-insensitive. In contrast to zinc, methylisobutylamiloride (MIA) binding to the wild-type D4receptor was virtually unaffected by sodium. In addition, the binding affinity for MIA was essentially unchanged in the presence of an IC50 concentration of zinc and vice versa. Furthermore, MIA binding affinity was decreased 4-fold for the D4-D77N mutant and increased 30-fold for the TMS3 mutant D4-M107V, even though the binding affinity for zinc was similar to the wild-type D4 background for both mutants. These findings demonstrate for the first time the existence of three distinct sites of allosteric modulation within a G protein-coupled receptor.
Allosteric modulation by sodium ions is a hallmark of many G protein-coupled receptors (GPCRs), including most of the aminergic GPCRs. The molecular explanation for this is that the sodium binding site is formed by a strictly conserved aspartate residue situated approximately three helical turns from the intracellular side of the second transmembrane-spanning domain (TMS2) (Horstman et al., 1990; Neve et al., 1991a; Donnelly et al., 1994). Sodium generally decreases the agonist binding affinity while increasing the binding for some classes of antagonists (Motulsky and Insel, 1983; Neve, 1991b). In contrast, antagonist binding to some GPCRs has been shown to be allosterically displaced by 5-amino-substituted amiloride derivatives, and in the case of the α2A-adrenergic receptor and the D2 dopamine receptor, it has been postulated that sodium and substituted amiloride derivatives bind to distinct sites (Wilson et al., 1990; Strange, 1997). Notably, the 5-amino-substituted amiloride derivatives such as methylisobutylamiloride (MIA) allosterically decrease antagonist binding to D2dopamine receptors (Hoare and Strange, 1996) in a manner similar to zinc (Schetz and Sibley, 1997; Schetz et al., 1999). However, zinc but not MIA binding affinity for the D2 dopamine receptor is sodium-dependent (Schetz et al., 1999; S. R. J. Hoare, personal communication). Importantly, all three of these allosteric modulators have been shown to act directly on the GPCR and, although each of them is chemically distinct, their molecular mechanisms appear to be related. A major question, however, is whether their shared or interrelated effects are due to interaction at a shared site or at three physically distinct sites. The D4 dopamine receptor was chosen as the model system for this investigation to simplify the interpretation of allosteric-allosteric interaction experiments, since zinc inhibition of antagonist binding to the D4 dopamine receptor subtype is noncompetitive (Schetz et al., 1999). In addition, [3H]methylspiperone was chosen as the primary radioligand with which to study allosteric interactions, because it is an antagonist and its binding to dopamine receptors is relatively sodium-insensitive. In this report, we demonstrate that sodium, zinc, and MIA occupy three physically distinct allosteric binding sites on the D4 dopamine receptor protein.
Materials and Methods
All drugs were purchased from Research Biochemicals International (Natick, MA). Analytical grade binding and wash buffer reagents were from Sigma Chemical Co. (St. Louis, MO) and Fluka Chemical Corp. (Ronkonkoma, NY) and cell culture supplies were purchased from Life Technologies (Gaithersburg, MD). Ultrapure zinc chloride was purchased from Aldrich Chemical (Milwaukee, WI). The [3H]methylspiperone (NET856, 85.5 Ci/mmol) was purchased from DuPont-New England Nuclear (Boston, MA).
Mutations in the rat D4 dopamine receptor were created using QwikChange, a DpnI-based site-directed mutagenesis kit (Stratagene, La Jolla, CA), and verified by 33P dideoxy-nucleotide sequencing using Sequenase (Amersham, Piscataway, NJ). The naming convention for the mutant receptors begins with the name of the wild-type receptor background followed by the single-lettered code for the amino acid to be mutated and its position, and then ending with the corresponding amino acid substitution. For example, the D4-D77N mutant has a D4 background that has been mutated from an aspartate at position 77 to an asparagine.
Transfection of Wild-Type and Mutant DNA.
The pcDNA3 plasmid constructs containing either the wild-type or a mutant dopamine receptor were transfected into COS-7 or CHO cells using CaPO4 precipitation (Invitrogen, Carlsbad, CA). Specifically, 20 μg of plasmid DNA was mixed with a final volume of 1 ml of CaPO4/HEPES solution and the resulting precipitate was added dropwise to 20 to 30% confluent cells on a 150-cm2 plate in a total media volume of 20 ml. The following day, the media were removed by aspiration and replaced with fresh media. COS-7 cells were then grown to confluence and harvested (transient transfection system), whereas CHO cell were first selected with geneticin (Life Technologies) and single colonies were expanded before harvesting (stable transfection system).
Preparation of Membranes and Radioligand Binding.
Membranes from transient or stable cell lines expressing wild-type or mutant D4 dopamine receptors were isolated and used in radioligand binding assays as described previously (Schetz et al., 1999). In the case of saturation isotherm binding, membranes were equilibrated with increasing concentrations of [3H]methylspiperone (0.03–3 nM). For competition experiments, membranes were equilibrated with a fixed concentration of [3H]methylspiperone (ca. 500 pM) and increasing concentrations of the competing ligand. Nonspecific binding was defined by 5 μM (+)-butaclamol.
Calculations and Data Analysis.
All experiments were performed in triplicate and repeated three to four times. All averaged values are reported as a geometric mean with a standard deviation. The error bars in all the figures are standard error bars. The equilibrium dissociation constant (K D) of the primary radioligand was measured by saturation isotherm analysis. Inhibition constant (K i) values for dopamine and MIA were calculated from IC50 values using the Cheng-Prusoff equation K i = IC50/(1 + [ligand]/K D). This form of the equation assumes a purely competitive interaction and a pseudo Hill slope of 1. Even though MIA is an allosteric modulator of [3H]methylspiperone binding to D4 dopamine receptors, itsK i affinity value derived from the competitive form of the equation is a good approximation because its binding to the D4 dopamine receptor is highly cooperative (Hoare et al., 2000). In the case of dopamine, the best-fit curve has a pseudo Hill slope significantly less than unity, and consequently, the K i affinity values are K 0.5 value approximations. Since zinc is a noncompetitive allosteric inhibitor of methylspiperone binding to D4 receptors, itsK i binding affinity value was taken to be approximately equal to its IC50 value, i.e., the noncompetitive form of the Cheng-Prusoff equation (Cheng and Prusoff, 1973). A 95% confidence interval was used for all curve-fitting procedures and for comparing different curve-fitting models using GraphPad Prism, version 2.0. The statistical measures of fit were the F test, the run test, and a correlation coefficient.
Sodium ions modulate a variety of different G protein-coupled receptors via electrostatic interaction with a highly conserved aspartic acid located approximately three helical turns from the intracellular side of the second transmembrane-spanning domain. Consequently, our strategy for determining whether the allosteric modulators sodium, zinc, and MIA bind to separate or related sites on the D4 dopamine receptor was to first create a sodium-insensitive D4 mutant by neutralizing the sodium binding site. This neutralization was achieved by mutating the corresponding negatively charged aspartic acid (D) at position 77 in TMS2 of the D4 dopamine receptor to a neutral asparagine (N). For comparison, another mutant D4dopamine receptor was constructed by replacing the D4 subtype-specific methionine in the first turn of TMS3 with the corresponding valine residue, which is present in all four other dopamine receptor subtypes. To ensure that the mutant receptors were properly folded and expressed in COS cells, the resulting D4-D77N and D4-M107V mutant dopamine receptors were assayed for their ability to specifically bind with high affinity to the D2/D3/D4-selective radioligand antagonist [3H]methylspiperone. Direct determination of [3H]methylspiperone equilibrium dissociation constants (K D) by saturation isotherm analysis yielded a K D = 294 ± 30, 84 ± 29, and 138 ± 22 pM for the wild-type D4, the D4-D77N mutant, and the D4-M107V mutant dopamine receptors, respectively (n = 3). Since the D4-D77N and D4-M107V mutants bind [3H]methylspiperone with high affinity and [3H]methylspiperone binding to wild-type dopamine receptors is displaced by zinc and MIA but not by sodium ions, [3H]methylspiperone was selected as a suitable primary radioligand for measuring sodium, zinc, and MIA interactions by competition binding.
Zinc ions inhibit [3H]methylspiperone binding to the wild-type D4 dopamine receptor expressed in COS cells with an 8-fold lower affinity in the presence of 120 mM sodium than in the absence of sodium (Fig.1A). This sodium-sensitive decrease in zinc affinity observed for the wild-type D4receptor was abolished in the D4-D77N mutant, even though the D4-D77N mutant retains its D4 wild-type binding affinity for zinc (Fig. 1A). Like zinc, dopamine displayed sodium-sensitive binding to the wild-type D4 receptor, which was lost in the D4-D77N mutant receptor (Fig. 1B). However, dopamine binding affinity for the D4-D77N mutant also increased slightly (i.e., 2.5-fold after applying the Cheng-Prusoff equation for a competitive inhibitor) (Fig. 1B). In contrast to dopamine and zinc, MIA binding affinity for the wild-type D4 dopamine receptor was not significantly affected by sodium ions (Fig. 1C). In further contrast to dopamine and zinc, MIA binds to the D4-D77N mutant dopamine receptor with a 4-fold lower apparent affinity than it does to the wild-type D4 dopamine receptor (Fig. 1C).
Having established the interaction of sodium ions with the conserved aspartic acid in TMS2, we next sought to determine whether the apparent similarity in the molecular mechanisms for zinc and MIA modulation of antagonist binding is a result of coupling between their binding sites. We reasoned that if zinc and MIA bind to wild-type D4 dopamine receptors at the same site (competitive) or structurally linked (allosterically coupled) sites, then the affinity of either of these modulators should be influenced by the presence of the other modulator. Remarkably, the apparent affinity values for MIA-[3H]methylspiperone inhibition curves performed in the absence of zinc were not significantly different from those done in the presence of a concentration of zinc that produces approximately a half-maximal effect (Fig.2A). However, in the reverse experiment with zinc-[3H]methylspiperone inhibition curves, the apparent affinity for zinc is significantly increased in the presence of a concentration of MIA that produces about a half-maximal effect compared with in the absence of MIA, although the effect was relatively small, i.e., about 3-fold (Fig. 2B). The combined zinc-MIA experiments allowed us to rule out any competitive allosteric interactions between zinc and MIA binding at two separate sites. A purely competitive binding of zinc and MIA for an identical site could also be ruled out, since the apparent affinity for zinc measured in the presence of an IC50 concentration of a perfectly competitive ligand would be significantly decreased.
The apparent distinction between zinc and MIA binding sites was investigated further by comparing zinc and MIA binding properties for a different mutant D4 dopamine receptor than the sodium-insensitive D4-D77N mutant. In contrast to the D4-D77N mutant, which binds MIA with a 4-fold lower affinity (Fig. 3A), the TMS3 mutant D4-M107V dopamine receptor had a 30-fold higher affinity for MIA (Fig. 3A). The pronounced and opposing effects of these two point mutants on MIA binding were not mimicked by zinc. In fact, zinc binding to both mutant D4 receptors was similar to the wild-type D4 dopamine receptor background from which they were derived (Fig. 3B). Furthermore, the molecular mechanisms of allosteric modulation for zinc and MIA at the wild-type D4 dopamine receptor are distinct: the primary effect on [3H]methylspiperone binding for zinc is a change in the maximum number of binding sites (Schetz et al., 1999), whereas for MIA it is a change in affinity (Hoare et al., 2000). Thus, the combined data from experiments using wild-type and sodium-sensitive, and sodium-insensitive mutant D4 dopamine receptors demonstrates at several levels that sodium, zinc, and MIA each bind a distinct site on D4 dopamine receptors.
A growing appreciation for the role of receptor conformation in defining GPCR receptor pharmacology has resulted in a renewed interest in allosteric modulation of GPCR proteins. The term allosteric modulation is frequently synonymous with conformational change, because allosteric modulators are defined as occupying a different site than the “primary” site of ligand binding. Consequently, the conformational effect of the allosteric site on the primary site must somehow be propagated through the protein structure, when the allosteric site is occupied by other proteins (Kramer and Karpen, 1998), drugs (Hoare and Strange, 1996), or ions (Neve, 1991b; Schetz and Sibley, 1997). Allosteric modulators also offer a potentially important means for modulating therapeutic responses to drugs (Ehlert, 1986; Birdsall et al., 1995; Lazareno and Birdsall, 1995; Strange, 1997; Schetz et al., 1999). Moreover, some endogenous modulators appear to have a physiological role in modulating ligand-induced receptor function. For example, millimolar concentrations of sodium ions acting from the intracellular side accelerate the dissociation of agonist from the GPCR. The same sodium-induced changes in the receptor that result in a decrease in agonist affinity also increase the binding for some classes of antagonists (Motulsky and Insel, 1983; Neve, 1991b). These results suggest that the binding of sodium ions to an allosteric site on the receptor changes the shape of the receptor's ligand binding pocket such that the favored conformation is the one associated with a less active state of the receptor. Sodium has been shown to have this effect on dopamine receptors (Neve, 1991b), and more recently, it was found that zinc and derivatives of amiloride allosterically decrease antagonist binding to dopamine receptors (Hoare and Strange, 1996;Schetz and Sibley, 1997). The question that arises is whether the interrelated pharmacological properties for sodium, zinc, and MIA are a result of differential allosteric interaction of these three modulators at the same site or at two or three distinct sites on the D4 dopamine receptor protein.
Since all three allosteric modulators of D4dopamine receptors bind with relatively low affinity (10−5–10−2 M), it is not possible to use radiolabeled derivatives (e.g.,65Zn2+) to directly measure their binding using rapid filtration techniques. Instead, the binding of these allosteric modulators to D4 dopamine receptors had to be measured indirectly by measuring their ability to inhibit the high-affinity binding of the D2-like antagonist [3H]methylspiperone. Although some information concerning their respective binding sites can be obtained by indirectly measuring their binding properties in the presence of one another, the allosteric nature of their interactions coupled with the indirect method of measuring their binding precludes the definitive determination of whether sodium, zinc, and MIA bind to three distinct sites. For these reasons, we adopted a classical approach to studying physical chemistry phenomena, which is to perturb the system and then study the relative changes. The “perturbation” was achieved by first mutating the wild-type D4 dopamine receptor, and then, relative changes in binding properties for the three allosteric modulators were measured either singly or in combination. By analogy with other catecholamine GPCRs (Donnelly et al., 1994), the site of sodium modulation corresponded to a conserved aspartic acid residue located approximately three helical turns from the intracellular side of the second transmembrane-spanning domain or aspartate at position 77 of the D4 dopamine receptor. Mutation of this negatively charged aspartate 77 to the neutrally charged but similarly sized asparagine resulted in a total loss of sodium sensitivity of the D4 dopamine receptors as judged by its loss of sodium-sensitive binding for dopamine and for zinc ions. The D4-D77N mutant also bound MIA with approximately 4-fold lower affinity than the wild-type D4 receptor, but had essentially no effect on zinc binding affinity. A different mutant located in the first helical loop of the extracellular side of TMS3 had an approximate 30-fold increase in MIA binding affinity, again with essentially no effect on zinc binding. Thus, in perturbed or mutant D4 dopamine receptors the effect on the binding of MIA and zinc is clearly different.
Overall, several lines of evidence indicate that the allosteric modulators sodium, zinc, and MIA bind to three distinct sites on the D4 dopamine receptor protein. First, the allosteric modulation of [3H]methylspiperone binding to wild-type D4 dopamine receptors by zinc ions is sodium-sensitive, whereas allosteric modulation by MIA is insensitive to sodium ions. Second, the sodium sensitivity of zinc binding was eliminated in the D4-D77N mutant, without affecting the apparent binding affinity for zinc. Thus, zinc does not act at the sodium binding site (aspartate 77) and sodium does not act at the zinc binding site, rather sodium is allosterically modulating the binding of zinc. Third, the allosteric effects produced by zinc and MIA appear to be mutually exclusive at the wild-type D4 dopamine receptor. For example, the apparent affinity for either allosteric modulator is the same or slightly higher when the other allosteric modulator is present at its IC50 concentration. However, this would not be the case if zinc and MIA bound to the same site, i.e., a purely competitive binding. Consequently, zinc and MIA do not appear to be exerting their macroscopically similar allosteric effects on [3H]methylspiperone binding by binding to a common site. Fourth, mutant D4 dopamine receptors were identified that either increase or decrease MIA binding affinity but in either case zinc binding is essentially unaffected. Fifth, the molecular mechanisms of allosteric modulation for zinc (Schetz et al., 1999) and MIA (Hoare et al., 2000) are distinct. These last three results, in particular, suggest that zinc and MIA do not share a common binding site on the D4 dopamine receptor and demonstrate that similar allosteric outcomes can result from binding to physically distinct allosteric sites.
To our knowledge, this report is the first demonstration of three distinct sites of allosteric coupling to the “binding site crevice” on a GPCR. This novel finding has several important implications with respect to conformational pharmacology. First, the presence of multiple sites of allosteric modulation should allow for the development of higher affinity modulators with increased selectivity using polymer-linked ligand dimers as has been recently shown for GTPases and cyclic nucleotide-gated channels (Kramer and Karpen, 1998). Second, these findings suggest that two different allosteric sites on the protein may induce similar conformational effects on the binding site crevice, and consequently, produce a similar pharmacological outcome. Third, as the number of allosteric sites of interaction on a given protein increases so will our understanding of which are the critical intra protein-protein interactions that produce a desired protein conformation. In other words, more tools are available for dynamically probing the conformational pharmacology of membrane-bound protein receptors.
We thank Julio Caesar Rodriguez for independently confirming the relative changes in zinc affinity at the wild-type and D4-D77N mutant dopamine receptors as a function of sodium ions.
- Received July 19, 2000.
- Accepted November 1, 2000.
Send reprint requests to: Dr. John A. Schetz, Department of Pharmacology, University of Mississippi, P.O. Box 1848, University, MS 38677-1848. E-mail:
This work was supported by National Institutes of Health. Some of this work was presented as a poster at the American Society for Pharmacology and Experimental Therapeutics meeting, 2000.
- G protein-coupled receptor
- transmembrane-spanning domain
- Chinese hamster ovary
- aspartic acid
- The American Society for Pharmacology and Experimental Therapeutics