Modulation of Agonist Binding to Human Dopamine Receptor Subtypes by l-Prolyl-l-leucyl-glycinamide and a Peptidomimetic Analog

The tripeptide l-prolyl-l-leucyl-glycinamide (PLG), also known as melanocyte-stimulating hormone release-inhibiting factor, has been demonstrated to possess a variety of pharmacological activities in the central nervous system (Srivastava et al., 1988; Drucker et al., 1994; Reed et al., 1994). A series of previous clinical studies showed that this tripeptide possessed substantial therapeutic activity in Parkinson's disease, 3,4-dihydroxy-l-phenylalanine-induced dyskinesia, antipsychotic drug-induced tardive dyskinesia, and depression (Ehrensing, 1974; Barbeau et al., 1978; Ehresing et al., 1994). However, the therapeutic activity of PLG still needs much improvement in terms of its efficacy and potency. For example, Barbeau et al. (1978) have previously shown that PLG improved the signs of Parkinson's disease in some patients yet failed to detect significant improvement in other patients. However, these authors did report consistent clinical trends, suggesting that further work on this tripeptide would be beneficial.

Currently, little is known about the mechanistic action of PLG; thus, in view of reports that PLG has demonstrated potential therapeutic effects in many central nervous system disorders (Mishra et al., 1983, 1986), it would be of interest to study the mechanistic manner in which PLG modulates central DA receptors to produce an effective and potent drug for people suffering from such disorders. PLG has also been shown to enhance the binding of various agonists such as apomorphine, 2-amino-6,7-dihydroxy-1,2,3,4,-tetrahydronaphthalene, and N-propylnorapomorphine (NPA) to striatal DA receptors, whereas antagonist binding (haloperidol and spiperone) was unaffected (Bhargava, 1983; Srivastava et al., 1988). Moreover, PLG and its peptidomimetic analogs have been shown to modulate the affinity states of DA receptors possibly by enhancing its interaction with guanine nucleotide-binding proteins (G protein) (Srivastava et al., 1988; Costain et al., 1997).

Previous studies using bovine brain synaptosomal membrane fractions showed that PLG and its peptidomimetic analogs primarily modulate agonist binding to DA D₂ receptors (Srivastava et al., 1988). However, brain tissue contains a heterogeneous population of various DA receptor subtypes; therefore, it has not been possible to study the interaction of PLG with individual DA receptors using striatal membranes. Therefore, the present study was undertaken to establish specific DA receptor modulation by PLG and its potent peptidomimetic analog 3(R)-[(2(S)-pyrrolidinylcarbonyl)amino]-2-oxo-1-pyrrolidineacetamide (PAOPA) (Fig. 1) and to further determine whether the coupling of DA receptors to the G protein is required for such a modulatory effect. To execute this study, human neuroblastoma SH-SY5Y cells were transfected with the cDNA of specific DA receptor subtypes, thus generating stably expressing cell lines. This cell line was used for the following reasons: 1) it is a neuronal cell line that has been shown to display a high rate of DA receptor subtype expression (picomoles per milligram of protein) upon transfection (Nair et al., 1996; Hillion et al., 2002); 2) these cells synthesize and release DA as well as housing tyrosine hydroxylase and other related enzymes; and 3) various G proteins as well as adenylyl cyclase, phospholipase, GTPase activity, protein kinase A, and protein kinase C are present in this cell line (Kazmi and Mishra 1989; Lambert and Nahorski, 1990). Last, although this cell line does not express DA receptors, it does express α₂-adrenergic, opiate (Kazmi and Mishra, 1987, 1989), N-methyl-d-aspartate (Nair et al., 1996), adenosine A₂ (Salim et al., 2000), and muscarinic receptors (Jope and Song, 1997). The expression of these other receptors in this cell line offers a distinct advantage over other cell lines, because it allowed us to compare the effects of PLG on other related G protein-coupled receptors without performing multiple transfections. Overall, in comparison with other cell lines, this cell line is somewhat similar to endogenous brain tissue.

We also investigated whether PLG requires the receptor/G protein complex to induce its modulatory effect through the use of suramin. Suramin behaves like a direct antagonist of heterotrimeric G proteins (such as G_o/G_i) since it competes with the receptor for the docking site on the α-subunit of the G protein and can uncouple receptor/G protein complexes, thus preventing GDP dissociation that is essential for Gα subunit activation (Beindl et al., 1996; Waldhoer et al., 1998; Chung and Kermode, 2004).

Collectively, our results indicate PLG and PAOPA display modulation of agonist binding to the D_2S, D_2L, and D₄ receptors, whereas agonist binding to the D₁ and D₃ receptors remains unaffected. Additionally, agonist binding to the α₂-adrenergic receptor (G_i-coupled) remained unaffected by PLG or PAOPA. Moreover, conversion of the higher affinity agonist binding state of the D_2L receptor by guanosine 5′-(β,γ-imido)triphosphate [Gpp(NH)p] was attenuated by PLG and PAOPA. Furthermore, the results suggest that PLG requires the G protein to be coupled to the D_2L receptor to modulate agonist binding.

Materials and Methods

Materials. SH-SY5Y cells (five passages) were obtained from Dr. A. Blume's laboratory (Roche Institute of Molecular Biology, Nutley, NJ). [³H]Spiperone (50 Ci/mmol), [³H]NPA (25 Ci/mmol), [³H]quinpirole (40 Ci/mmol), [³H]rauwolscine (82 Ci/mmol), and [³H]SCH23390 (52 Ci/mmol) were purchased from PerkinElmer Life and Analytical Sciences (Boston, MA). [³H]7-OH-DPAT (160 Ci/mmol) was purchased from GE Healthcare (Little Chalfont, Buckinghamshire, UK). PLG, DA, and other routine chemicals, including unlabeled spiperone, (+)-butaclamol, and quinpirole, were purchased from Sigma-Aldrich (St. Louis, MO). CHO cells expressing human DA D₃ receptors were also obtained from Sigma-Aldrich. PAOPA was synthesized as described previously (Yu et al., 1988). Suramin was obtained from Tocris-Cookson Inc. (Ellisville, MO), whereas ascorbic acid was purchased from BDH Inc. (Toronto, ON, Canada).

Cell Culture and Gene Transfection. Human DA D_2L receptor cDNA was subcloned into the mammalian expression vector pRC/RSV (Invitrogen, Carlsbad, CA), whereas human D₄ (cDNA-genomic hybrid; Van Tol et al., 1992), D_2S, and D₁ receptor cDNAs were cloned into pCD-PS expression vectors (Invitrogen). DNA was prepared using the Flexiprep plasmid isolation kit (Pfizer, Inc., New York, NY). Human neuroblastoma SH-SY5Y cells were grown at 37°C under 5% CO₂, 95% air in RPMI 1640 medium supplemented with 10% fetal calf serum, 1 mM glutamine, 50 U ml^–1 penicillin, and 50 U ml^–1 streptomycin. Transfection of cells with respective cDNA was performed with 10 μg of DNA to 3 × 10⁷ cells as described previously by Nair et al. (1996). Cells were selected with geneticin (2 μg/ml) (Invitrogen), and single colonies were expanded by limiting dilution to generate a stable transfection system.

Membrane Preparation and Receptor Binding Assays. The SH-SY5Y cells, stably expressing cloned human DA receptors, were grown to confluence and membranes were prepared as described previously (Nair et al., 1996). The DA receptor agonist binding assays were carried out as described previously (Srivastava et al., 1988; Nair et al., 1996). The assay buffer contained 50 mM Tris-HCl, 5 mM KCl, 4 mM MgCl₂, 1 mM EDTA, and 120 mM NaCl, pH 7.6, with 100 μg of membrane protein in a total volume of 1 ml. Protease inhibitors were added to the assays as a precaution against proteolysis of the membrane protein as well as the tripeptide PLG and its analog PAOPA. The concentration of [³H]NPA used for the D_2S, D_2L, and D₄ receptor assays was 0.25 nM, whereas that for the D₁ receptor was 0.5 nM. For the D₃ receptor binding assay, CHO cell membranes were suspended in the above-described buffer with the addition of 0.05% ascorbic acid to prevent the degradation of [³H]7-OH-DPAT (1 nM). Specific binding was defined as the difference between the radioactivity bound in absence and presence of 1 μM (+)-butaclamol (D₁, D₂, and D₄ receptors) and 1 μM DA (D₃ receptor). The [³H]quinpirole (2 nM) binding assay conditions were the same as for [³H]NPA, with the exception that the assay tubes also contained 120 mM glucosamine. The binding of [³H]rauwolscine to α₂-adrenergic receptors and [³H]SCH23390 to D₁ receptors was carried out as described previously (Srivastava et al., 1988). The contents of the tubes (triplicate) were incubated for 1 h at 25°C. At the end of the incubation period, assays were terminated by rapid filtration through a Brandel cell harvester (Brandel Inc., Gaithersburg, MD), and radioactivity bound on filters was counted in a Beckman liquid scintillation counter (model LS5 KTA).

For the [³H]spiperone/NPA competition assays, 0.2 nM [³H]spiperone, and various indicated concentrations of NPA were added to the assay mixture. The competition assays were carried out both in the absence and presence of either 100 μM Gpp(NH)p, 1 μM PLG, or Gpp(NH)p and PLG together.

Cell Membrane Pretreatment with Suramin and DA. Cell membrane protein (200 μg) was incubated alone, with 10 mM suramin, and with 10 μM suramin + 100 μM DA for 15 min at 25°C. The DA was dissolved in 0.1% ascorbic acid. The membranes were then washed three times to remove any unbound DA and suramin, and the resulting pellet was resuspended and used for the receptor binding assays. The assay buffer contained 50 mM Tris, 5 mM MgCl₂, 1 mM EDTA, 0.1 mM dl-dithiothreitol, 0.1 mM phenylmethanesulfonyl fluoride, 100 μg/ml bacitracin, and 5 μg/ml soybean trypsin inhibitor, pH 7.4, with 40 μg of membrane protein in a total volume of 1 ml, in each assay tube. The concentration of [³H]NPA used was 1 nM, whereas the concentration of PLG and PAOPA used was 10 μM and 1 nM, respectively. The contents of the tubes (triplicate) were incubated for 1 h at 25°C. At the end of the incubation period, assays were terminated by rapid filtration in the same manner described above for other binding assays.

Data Analysis. The data were analyzed using GraphPad Prism software, version 4.1 (GraphPad Software Inc., San Diego, CA). The receptor binding data were analyzed by weighted nonlinear curve fitting for agonist versus antagonist competition curves. The data were analyzed for either one- or two-site populations of binding sites together with statistical analysis comparing the goodness of fit between one-or two-affinity state models. A two-site model was selected only if a statistically significant improvement of the fit of the data were obtained over a one-site model. The IC₅₀ values (concentration of competing ligands that inhibits the binding of the labeled ligand by 50%) obtained from the competition curves were converted to K_i values (inhibitory constant) using the Cheng and Prusoff equation (Cheng and Prusoff, 1973). The data for Scatchard analysis were analyzed by nonlinear regression using GraphPad Prism, and Student's t tests were used to compare differences among values between the experiments. Dose-response curves were analyzed by one-way repeated measures of analysis of variance (ANOVA) using GraphPad Prism software followed by Newman-Keuls post hoc test. The multiple comparison procedure used is indicated in the appropriate figure legend.

Results

PLG and the Peptidomimetic Analog PAOPA Increase [³H]NPA Binding to the DA D_2L, D_2S, and D₄ Receptor Subtypes. To determine whether PLG and PAOPA increase agonist binding to specific DA receptor subtypes, receptor binding assays were carried out with the agonist [³H]NPA for various DA receptor subtypes. The results outlined in Fig. 2 demonstrate that both PLG and PAOPA increase agonist binding to the D_2L, D_2S, and D₄ receptors, whereas agonist binding to the D₁ ([³H]NPA) and D₃ ([³H]7-OH-DPAT) receptors was not statistically significantly affected. The stimulatory effect is concentration-dependent and displays a bell-shaped curve. The bell-shaped curve is consistent with the previously reported findings for in vivo animal experiments (Mishra et al., 1997; Costain et al., 1999) and human clinical studies using PLG (Drucker et al., 1994; Ehrensing et al., 1994). In comparison with PLG, PAOPA displayed similar efficacy and ranged from being 10 to 1000 times more potent.

PLG and PAOPA Increase [³H]Quinpirole Binding to the DA D_2L, D_2S, and D₄ Receptors. To determine whether PLG or its potent analogs also increase the binding of another D₂ receptor family agonist, the binding of [³H]quinpirole was also examined in these cultures. As shown in Fig. 3, A and B, both PLG and PAOPA significantly increased the binding of [³H]quinpirole to the D_2L, D_2s, and D₄ receptors. Although PAOPA displayed a similar efficacy to PLG, it was found to be 1000 times more potent for the D_2L receptor and 100 times more potent for the D_2S receptor, and it displayed the same potency for the D₄ receptor.

PLG Decreases the Equilibrium Dissociation Constant (K_d) for [³H]NPA Binding to the D_2L, D_2S, and D₄ Receptors. To determine whether the increase in agonist binding was due to an increase in B_max or a decrease in K_d, Scatchard analysis of [³H]NPA binding in the absence and presence of 1 μM PLG was performed on membranes collected from cultures transfected with the D_2L, D_2S, and D₄ receptor. The data shown in Fig. 4 reveal a single class of binding sites with a K_d of 0.20 ± 0.015 nM (mean ± S.E.M.) and a total number of binding sites of 1259 ± 104 fmol/mg protein. When D_2L-transfected membranes were treated with PLG (1 μM), Scatchard plot analysis illustrated a significant decrease (P < 0.01) in the K_d value (0.09 ± 0.004 nM) without any significant increase in B_max (1473 ± 109 fmol/mg protein). Similarly, both PLG and PAOPA affected the equilibrium dissociation constant of [³H]NPA for the D_2S and D₄ subtypes (Figs. 5 and 6). The Scatchard plot analysis shown in Figs. 5 and 6 revealed a K_d value of 0.25 ± 0.020 nM and a B_max of 470.0 ± 30.0 fmol/mg protein in the absence of PLG for the D_2S receptor. When the membranes were treated with PLG, there was a significant decrease (P < 0.01) in the K_d value (0.10 ± 0.009 nM), without significantly affecting the B_max (488.0 ± 40.0 fmol/mg protein). Similarly, the K_d value obtained for the D₄ receptor subtype in the absence of PLG was 0.35 ± 0.026 nM with a significant decrease (P < 0.05) in the presence of PLG (0.14 ± 0.013 nM). The B_max values, however, did not show a significant difference in the absence and presence of PLG, (1680 ± 170 fmol/mg protein without PLG and 1710 ± 195 fmol/mg protein with PLG).

PLG Does Not Affect Antagonist [³H]Spiperone Binding to the DA D_2L Receptors in Transfected Cells. To determine whether antagonist binding is affected by PLG in D₂-transfected cells, binding assays with [³H]spiperone were carried out. In contrast to agonist [³H]NPA binding to the D_2L, D_2S, and D₄ receptors, the binding of [³H]spiperone was not affected by the addition of PLG. Representative binding parameters obtained for the D_2L receptors demonstrated that the K_d and B_max values (Scatchard analysis) remained unchanged for both control (K_d of 0.09 ± 0.005 nM; B_max of 1214 ± 122 fmol/mg protein) and PLG-treated membranes (K_d of 0.07 ± 0.003 nM; B_max of 1256 ± 128 fmol/mg protein). These results are consistent with other reports in which no changes in [³H]spiperone binding in striatal membranes were observed (Bhargava, 1983; Chiu et al., 1983a; Srivastava et al., 1988; Mishra et al., 1990). Similar observations were made for the D_2S and D₄ receptors (data not shown), thereby extending these results to other D₂ receptor family members.

PLG and PAOPA Attenuate Gpp(NH)p-Induced Conversion of High-Affinity Agonist Binding State to Low-Affinity State in D_2L Receptors. To establish whether the high-affinity agonist state of the D₂ receptor is affected by PLG, [³H]spiperone/NPA competition experiments were carried out in the presence of PLG and Gpp(NH)p. The representative results from the [³H]spiperone/NPA competition for the D_2L receptor are shown in Fig. 7. Analysis of the competition curves revealed that NPA discriminates between two D_2L receptor affinity states that are labeled by [³H]spiperone. The two states, agonist high-affinity and agonist low-affinity, were present in approximately equal proportions (53 and 47%, respectively) in the control-transfected cell membranes; however, they significantly differed with respect to their K_i values (0.065 nM for the high-affinity state and 94 nM for the low-affinity state).

In the PLG-treated D_2L receptor membranes, there was a leftward shift in the competition curve with an almost 3-fold decrease in K_i for the high-affinity agonist binding site without any significant change in the K_i for the low-affinity site. The ratio of high- to low-affinity sites in the PLG-treated membranes significantly increased from 1.12 (control) to 2.70 (Table 1).

The nonhydrolyzable analog of GTP, Gpp(NH)p, caused a significant rightward shift in the competition curve for the D_2L receptor; this decreased the proportion of high-affinity binding sites to 18% with a K_i of 0.15 nM. However, the conversion of the high- to low-affinity state was significantly attenuated by PLG. For example, the percentage of receptors in the high-affinity state was 34% with a K_i of 0.040 nM in the PLG-treated membranes (Table 1). Yet, the affinity of the receptors remaining in the low-affinity state was unaffected (Table 1). PAOPA demonstrated similar results to PLG (Table 2), yet PAOPA was almost 1000 times more potent than PLG.

PLG Does Not Affect the Conversion of the High-Affinity State to the Low-Affinity State by Gpp(NH)p for the DA D₁ and α₂-Adrenergic Receptors. To establish the specificity of PLG with respect to modulating the D₂ receptor subtype, DA D₁ and α₂-adrenergic receptors were also used. The D₁ receptor antagonist/agonist competition curve using [³H]SCH23390 (antagonist) and SKF38393 (agonist) revealed a high- and low-affinity state of agonist binding, although the binding parameters were unaffected by the addition of PLG. The values obtained from controls were K_H, 3.00 ± 0.25 nM; K_L, 96.0 ± 7.0 nM; percent R_H, 30.0 ± 2.5; and percent R_L, 70.0 ± 9.0. Similar values were obtained from the PLG-treated group and were not significantly different (K_H, 2.60 ± 0.18 nM; K_L, 92.0 ± 8.0 nM; percent R_H, 34.0 ± 5.0; and percent R_L, 66.0 ± 9.0).

The α₂-adrenergic receptor antagonist/agonist competition curves, using [³H]rauwolscine as an antagonist and BHT-920 as the agonist, displayed a high-affinity and low-affinity state of agonist binding as reported previously in this cell line (Kazmi and Mishra, 1989). The binding parameters were also unaffected by PLG or even by its highly constrained analog PAOPA. The values obtained in the control membranes for PLG (K_H, 2.60 ± 1.1 nM; K_L, 102 ± 13 nM; percent R_H, 43.0 ± 3.2; and percent R_L, 57.0 ± 4.5) were not significantly different from those obtained in the PLG-treated membranes (K_H, 2.48 ± 0.051 nM; K_L, 102 ± 13.0 nM; percent R_H, 45.0 ± 4.0; and percent R_L, 55.0 ± 5.1). Even the highly potent, conformationally constrained analog PAOPA did not alter the affinity state of the α₂-adrenergic receptors, with the control values being K_H, 2.9 ± 0.21 nM; K_L, 107 ± 14 nM; percent R_H, 43.0 ± 3.0; and percent R_L, 57.0 ± 5.0 and the PAOPA-treated values being K_H, 2.7 ± 0.30 nM; K_L, 110 ± 11 nM; percent R_H, 46.0 ± 4.0; and percent R_L, 54.0 ± 4.7.

Effect of Suramin on PLG-Induced Increase in Agonist Binding. To establish whether PLG and PAOPA display modulatory effects on agonist binding to D₂ receptors in the absence of G protein coupling, we investigated agonist binding in the absence and presence of suramin. Suramin is a polysulfonated naphthylurea and is considered to uncouple the receptor from the G protein that results in decreased high-affinity agonist binding (Nakata, 2003; Nickolls and Strange, 2003; Chung and Kermode, 2004). Both PLG and PAOPA, as shown in Fig. 8, were unable to modulate [³H]NPA binding to D_2L receptors in the presence of suramin. PLG and PAOPA increased agonist binding in membranes pretreated with suramin and DA, but they had no effect on the membranes treated with suramin alone. Furthermore, to establish whether this lack of PLG modulation is due to a direct competition with suramin, a competition experiment was performed. As indicated in Fig. 9, the competition curve revealed that PLG does not compete with suramin for the receptor binding site.

Discussion

The main findings of the present study are first that both PLG and a conformationally constrained analog (Fig. 1) increase agonist binding to specific DA receptor (D_2S, D_2L, and D₄) subtypes in the D₂ receptor family (Figs. 2 and 3). Second, this increase in agonist binding is dependent upon D₂ receptor/G protein coupling (Fig. 8).

The results of this study provide evidence that PLG and PAOPA enhance agonist binding (NPA and quinpirole) to the DA D_2L, D_2S, and D₄ receptors in a manner that is selective for the agonist high-affinity state of the receptor. The ineffectiveness of PLG and PAOPA on spiperone (D₂ receptor antagonist), SCH23390 (D₁ receptor antagonist), and the α₂-adrenergic receptor confirms the specificity of both these peptides toward D₂ receptor agonist binding. Furthermore, results from the D₁, D₃, and α₂-adrenergic receptor binding ascertain that both agonist and antagonist binding to these receptors is unaffected by PLG and PAOPA.

The results from [³H]NPA and [³H]quinpirole binding to the DA D_2L, D_2S, and D₄ receptors reveal that the effect of PLG and PAOPA are dose-dependent with a maximal effect at 0.1 to 1.0 μM for PLG and 0.001 to 0.1 μM for PAOPA. This dose response corroborates the phenomenon of a “therapeutic window” in which the pharmacological response for PLG can be elicited in animal models (Drucker et al., 1994; Mishra et al., 1997; Costain et al., 1999; Sharma et al., 2003) and human clinical trials (for review, see Mishra et al., 1986).

The PLG- and PAOPA-induced increase in agonist binding for the D₂ receptor is selective for the high-affinity state; this is evident from the antagonist/agonist competition experiments carried out for the D_2L, D_2S, and D₄ receptors (Table 1), where a significant increase in the affinity and population of high-affinity binding sites was observed for [³H]NPA. The number of low-affinity sites for [³H]NPA decreased as a result of PLG and PAOPA treatment without a significant change in affinity. Thus, these results suggest a conversion from the D₂ receptor low-affinity form to the high-affinity form for agonists in the presence of PLG and PAOPA.

The increase in the affinity of the high-affinity agonist binding site is consistent with the Scatchard analysis of [³H]NPA binding for the D_2L, D_2S, and D₄ receptors in PLG-treated membranes, since the [³H]NPA concentrations used (up to 1 nM) have been reported to selectively label the high-affinity form of the DA D₂ receptor (Titeler and Seeman, 1979). High- and low-affinity states of the D₂ receptors are an indication of coupling and noncoupling, respectively, of the receptor with the G protein Gi_αβγ GDP/Go_αβγ GDP. Therefore, the observed increase in the proportion and affinity of the binding sites after PLG and PAOPA treatment may be due to an increase in the association of the D₂ receptor with Gi_αβγ GDP.

The enhancement of the affinity and proportion of high-affinity agonist binding sites by PLG and PAOPA and their attenuation on the effect of Gpp(NH)p (Fig. 1; Tables 1 and 2) demonstrate an increased interaction between DA D₂ receptors and G proteins. PLG was unable to increase agonist binding in the presence of suramin, which uncouples the receptor/G protein complex; this suggests that PLG will only interact with D₂ receptors when they are coupled to the G protein α-βγ subunits. However, the antagonistic effect of suramin on G proteins can be reversed by the addition of agonist (Beindl et al., 1996). In this case, the increased concentration of DA (100 μM) reversed the inhibitory effect of suramin because the agonist-liganded receptor competes with suramin for binding to the G protein. Increasing the number of active receptors in the membrane by increasing agonist occupancy can overcome the inhibitory effect of suramin on the receptor/G protein complex (Beindl et al., 1996). Additionally, the suggestion that PLG requires the receptor/G protein complex may be part of the reason for the lack of significant PLG modulation associated with the D₃ receptor subtype. In contrast to the D₂ and D₄ receptor subtypes, the receptor/G protein interaction associated with the D₃ receptor subtype is weaker because GTP does not modulate high-affinity agonist binding to the D₃ receptor (McAllister et al., 1993; Filteau et al., 1999). It is also possible that PLG cannot modulate the D₃ receptor for reasons similar to that of GTP; further research is required to explore this possibility.

Previous studies have used bovine striatal membranes, which contain a heterogeneous population of DA receptors, as a means of testing the action of PLG on the DA receptor. However, because the mechanism of the interaction between PLG and the DA receptor is still not clear, we have attempted to study this interaction in cultures stably transfected with individual DA receptor subtypes, which provided a model based analysis/interpretation of the modulatory effect of PLG. This model facilitated specific attention to binding parameters as well as experimental manipulation to identify the role of high versus low affinity and G protein coupling.

Additionally, the importance of the D_2L receptor/G protein complex is indicative of the possible site of interaction with PLG by which this endogenous neuropeptide may be modulating and increasing the number and affinity of the D_2L receptors in the high-affinity state. We suggest a mechanism that proposes that PLG causes a conformational change in the receptor, presumably involving a putative PLG binding site on the D₂ receptor (Chiu et al., 1983a,b); this would lead to an increased association of the receptor with the G protein (G_i) by adjusting the equilibrium between the receptor, G protein, and receptor/G protein complex (Costain et al., 1997). Such a mechanism would explain the enhancement of high-affinity agonist binding and the effect of PLG opposite to that of Gpp(NH)p.

If PLG is acting at a distinct site on the D₂ receptor, it is possible that PLG is acting as an allosteric ligand. The conformational change is frequently synonymous with the term allosteric modulation, because allosteric modulators are described as modulators that occupy sites other than the “primary” site of ligand binding, the orthosteric site. Both DA D₂ and D₄ receptors have been shown to be allosterically modulated by amilorides and zinc (for review, see May and Christopoulos, 2003). Furthermore, many receptors of the G protein-coupled receptor superfamily are allosterically modulated. Allosteric modulators modify receptor conformation to cause a change in the binding of the orthosteric ligand (May and Christopoulos 2003; Kenakin, 2004); this is what PLG does precisely—it increases the binding of an agonist (orthosteric ligand) to D₂ receptors, perhaps in an allosteric manner. However, unlike zinc or amilorides, PLG requires the D₂ receptor to be coupled to the G protein; therefore, the molecular mechanisms of allosteric modulation for PLG are more distinct than other modulators. The simplest mechanism that describes an allosteric interaction between two ligands binding at distinct sites on a receptor is referred to as the ternary complex model (Kenakin, 2004). In this model (Fig. 10), each ligand binds to the receptor with its own affinity constant with the symbol (α) acting as the cooperative factor that is used to quantify the magnitude of the change in affinity of one ligand that is caused by the binding of the second ligand. The ability of the allosteric modulator (PLG) to cause a change in affinity for the agonist (NPA or quinpirole) binding at the orthosteric site relates to the degree in which the modulator (PLG) induces a change in the receptor conformation. The ternary complex model has already successfully quantified the behavior of several other allosteric modulators of G protein-coupled receptors (Christopoulos, 2002).

The findings reported in this study show important implications with regard to the interaction of PLG and conformational pharmacology of DA D₂ receptors. First, the presence of an endogenous neuropeptide site of allosteric modulation should allow for the synthesis of higher affinity modulators, such as PAOPA shown here, with increased selectivity. Second, PLG analogs such as PAOPA provide additional tools for probing the conformational pharmacology of plasma membrane bound G protein-coupled DA D₂ receptors. Third, PLG displays no effect in the absence of orthosteric ligands; therefore, the normal spatial and temporal pattern of physiological signal transduction and termination is maintained, with the only effect of PLG being able to increase the pattern of signaling as suggested by Christopoulos and Kenakin (2002). Finally, there is very little possibility of toxic effects of modulators, such as PLG or PAOPA, because there is a ceiling to their effects after a maximal dose. Interestingly, this modulatory effect is reduced at higher concentrations.

In conclusion, we have unequivocally shown the selective modulation of the human D_2S, D_2L, and D₄ DA receptor subtypes by PLG and provided a theoretical mechanistic action of PLG; this can now be used to perform even more detailed studies on this tripeptide that may lead to the development of a more effective and potent drug for the treatment of neurological disorders, such as Parkinson's disease and antipsychotic drug-induced tardive dyskinesia.