Abstract
The aim of this study was to use pharmacological inhibition of protein kinase A and mutation of potential protein kinase A phosphorylation sites to determine the role of protein kinase A-catalyzed phosphorylation of the dopamine D1 receptor in agonist-stimulated desensitization and internalization of the receptor. To facilitate purification and imaging of the D1 receptor, we attached a polyhistidine tag to the amino terminus and enhanced green fluorescent protein to the carboxyl terminus of the receptor (D1-EGFP). D1-EGFP was similar to the untagged D1 receptor in terms of affinity for agonist and antagonist ligands, coupling to G proteins, and stimulation of cyclic AMP accumulation. D1-EGFP and two mutants in which either Thr268 or Ser380 was replaced with Ala were stably expressed in NS20Y neuroblastoma cells. Pretreatment with the protein kinase A inhibitor H-89 (N-[2-(p-bromocinnamylamino)ethyl]-5-isoquinolinesulfonamide) or substitution of Ala for Thr268 reduced agonist-stimulated phosphorylation of the receptor and resulted in diminished trafficking of the receptor to the perinuclear region of the cell. Substitution of Ala for Thr268 had no effect, however, on agonist-induced receptor sequestration or desensitization of cyclic AMP accumulation. Substitution of Ala for Ser380 had no effect on D1 receptor phosphorylation, sequestration, desensitization, or trafficking to the perinuclear region. We conclude that protein kinase A-dependent phosphorylation of the D1 receptor on Thr268 regulates a late step in the sorting of the receptor to the perinuclear region of the cell, but that phosphorylation of Thr268 is not required for receptor sequestration or maximal desensitization of cyclic AMP accumulation.
The dopamine D1 receptor belongs to the superfamily of heptahelical receptors that modulate the activity of effectors such as adenylate cyclase by activation of specific heterotrimeric GTP-binding proteins (G proteins). For many G protein-coupled receptors (GPCRs), phosphorylation by protein kinases such as protein kinase A (PKA) or G protein-coupled receptor kinases (GRKs) is an early step in agonist-induced desensitization, the diminished responsiveness that occurs after continuous or repeated exposure of the receptors to agonist (Hausdorff et al., 1990). The mechanisms of desensitization and resensitization have been perhaps most thoroughly characterized for the β-adrenergic receptor, a Gαs-coupled receptor that is structurally and functionally homologous to the D1 receptor. Occupation of the β-adrenergic receptor by agonist stabilizes a conformation of the receptor that increases its phosphorylation by GRK, which in turn increases the affinity of the receptor for β-arrestin (Krupnick and Benovic, 1998). The coupling of the receptor to heterotrimeric G proteins is inhibited by the binding of β-arrestin, which is likely to be the immediate cause of rapid desensitization. β-Arrestin also acts as an adaptor protein to facilitate receptor endocytosis (Krupnick and Benovic, 1998). Once internalized the receptor is either resensitized by dephosphorylation and then recycled back to the cell surface (Pippig et al., 1995), or directed to an intracellular compartment for degradation (Tsao et al., 2001).
Although this model closely describes the regulation of some GPCRs, the regulation of other receptors differs significantly in ways that may be both cell- and receptor-dependent (Koenig and Edwardson, 1996). For example, GPCR desensitization and internalization can occur despite greatly reduced or undetectable agonist-induced phosphorylation of the receptor (Malecz et al., 1998; Sadeghi et al., 1998; Olivares-Reyes et al., 2001), and dephosphorylation/resensitization can occur without internalization (Gardner et al., 2001).
Desensitization of the dopamine D1 receptor has been demonstrated in a variety of tissue preparations, including brain slices and retina, primary neuronal culture, neuroblastoma, astrocytoma, or kidney cell lines expressing an endogenous D1 receptor, and cell lines expressing a recombinant D1 receptor (Memo et al., 1982;Barton and Sibley, 1990; Zhou et al., 1991; Ng et al., 1994; Sibley and Neve, 1997; Jiang and Sibley, 1999; Ventura and Sibley, 2000). The responsiveness of the D1 receptor is probably regulated by PKA- and GRK-catalyzed phosphorylation of the receptor, because dopamine-induced desensitization is temporally associated with or preceded by receptor phosphorylation (Ng et al., 1994; Gardner et al., 2001), enhanced by activation of PKA or overexpression of GRK (Zhou et al., 1991; Tiberi et al., 1996), and reduced by inhibitors of PKA or GRK (Zhou et al., 1991). Furthermore, desensitization of the D1 receptor is blunted in cells deficient in PKA (Ventura and Sibley, 2000), and mutation of a potential site of PKA phosphorylation of the D1 receptor, Thr268, reduces the rate of agonist-induced desensitization (Jiang and Sibley, 1999). Ser380 has also been proposed to be a site of phosphorylation by PKA, because a peptide comprised of D1 receptor amino acid residues 372 to 442 is phosphorylated by PKA on Ser380 (Zamanillo et al., 1995).
In this study, we assessed the role of two potential sites of phosphorylation by PKA, Thr268 and Ser380, in agonist-induced phosphorylation, desensitization, and internalization of the dopamine D1 receptor. We created a D1 receptor with enhanced green fluorescent protein (D1-EGFP) at the carboxyl terminus and a hexa-histidine tag at the amino terminus. Wild-type D1-EGFP and mutants in which an alanine residue was substituted for Thr268 (T268A) or Ser380 (S380A) were stably expressed in NS20Y neuroblastoma cells. We now report that mutation of Thr268, but not Ser380, prevented dopamine-induced phosphorylation of D1-EGFP and redistribution of the receptor to the perinuclear region of NS20Y cells, without altering maximal desensitization of cyclic AMP accumulation or sequestration of the receptor away from the extracellular surface of the membrane.
Materials and Methods
Construction of Tagged and Mutant Receptors.
We attached a polyhistidine tag to the N terminus of the rhesus macaque D1 receptor (Machida et al., 1992) by using the polymerase chain reaction (PCR). The entire amplified portion of the gene was sequenced to confirm the absence of random mutations and cloned into pcDNA3. To place the EGFP to the C terminus, we used the QuikChange mutagenesis kit (Stratagene, La Jolla, CA) to introduce aBamHI site at Pro445 of the histidine-tagged D1 receptor, which resulted in the loss of the terminal Thr residue (position 446), and the substitution of Asn for His444. A HindIII-BamHI fragment containing the histidine-tagged D1 receptor was cloned into pEGFP-N1 (CLONTECH, Palo Alto, CA), creating a histidine-tagged D1 receptor with EGFP tethered by a six-residue linker at the C terminus to form the construct D1-EGFP (Fig. 1). Mutants of the D1-EGFP construct were made in which two potential sites of phosphorylation by PKA, Thr268 and Ser380, were changed to alanine, yielding the mutants T268A and S380A (Fig. 1). Untagged alanine substitution mutants of Thr268 and Ser380 for expression in C6 glioma cells were constructed separately using the trans-PCR method described previously (Neve et al., 1991), except with Pfu thermostable DNA polymerase instead of Taq. The T268A mutant was cloned into pcDNA1, whereas the S380A mutant was cloned into pcDNA3. The PCR-amplified portion of each mutant was sequenced to confirm the presence of the desired mutation and to identify random errors introduced by PCR. The untagged T268A had two additional mutations that changed Ile34→Val in the first transmembrane domain and Gln436→His close to the C terminus.
Creation and Maintenance of Cell Lines.
Both tagged and untagged wild-type and mutant receptors were stably expressed in NS20Y neuroblastoma cells by calcium phosphate coprecipitation, and after selection for G418-resistance, pooled populations of D1-EGFP-, T268A-, and S380A-expressing cells were isolated using a flow cytometry system (BD FACSVantage SE; BD Biosciences, San Jose, CA) with excitation at 488 nm. Cells were maintained at 37°C in a humidified atmosphere with 10% CO2 in Dulbecco's modified Eagle's medium (DMEM; Sigma Chemical, St. Louis, MO) supplemented with 5% fetal bovine serum and 5% calf bovine serum, with 600 μg/ml G418 sulfate (Calbiochem, San Diego, CA). Clonal C6 cell lines expressing untagged wild-type and mutant D1 receptors were created by cotransfection with pBabe-Puro (Morgenstern and Land, 1990) and selection with puromycin (wild-type D1 and T268A, 2 μg/ml) or by transfection (in pcDNA3) and selection with G418 (S380A, 600 μg/ml), exactly as described previously (Cox et al., 1995). C6 cells expressing wild-type and mutant D1 receptors were maintained in DMEM supplemented with 2% fetal bovine serum, 3% iron-supplemented calf bovine serum, 50 μg/ml penicillin, 50 μg/ml streptomycin, and either puromycin or G418.
[3H]SCH 23390 Saturation Binding.
Confluent cells in 10 cm-diameter tissue culture plates were lysed by replacing the medium with ice-cold hypotonic buffer (1 mM Na+-HEPES, pH 7.4, 2 mM EDTA). After swelling for 10 to 15 min, the cells were scraped off the plate and centrifuged at 24,000g for 20 min. The crude membrane fraction was resuspended in Tris-buffered saline (50 mM Tris-HCl, pH 7.4, 155 mM NaCl) with a Polytron homogenizer (Brinkmann Instruments, Westbury, NY) at setting 6 for 10 s. Aliquots of the membrane preparation (∼30 μg of protein) were added to duplicate assay tubes containing 50 mM Tris-HCl, pH 7.4, 155 mM NaCl, 0.001% bovine serum albumin, and six concentrations of [3H]SCH 23390 (75.5 Ci/mmol; PerkinElmer Life Sciences, Boston, MA) ranging from 0.05 to 2.0 nM in a final volume of 0.5 ml. Nonspecific binding was determined in the presence of 2 μM (+)-butaclamol. Incubations were carried out at 30°C for 60 min and terminated by filtration with a 96-well Tomtec Mach II cell harvester (Tomtec, Orange, CT). Filters (BetaPlate filtermat A) were dried and 50 μl of BetaPlate scintillation cocktail was added to each sample. Radioactivity on the filters was determined using an LKB/PerkinElmer Wallac BetaPlate 1205 scintillation counter (Gaithersburg, MD).
Agonist Binding.
The crude membrane fraction was resuspended in competition binding assay buffer (20 mM HEPES, pH 7.5, 6 mM MgCl2, 1 mM EDTA, 1 mM EGTA, and 1 mM dithiothreitol), incubated at 37°C for 15 to 20 min, recentrifuged, and resuspended a second time in assay buffer. An aliquot of the membrane preparation was added to duplicate assay tubes containing assay buffer with 0.025% ascorbic acid, 0.001% bovine serum albumin, ∼0.8 nM [3H]SCH 23390, and 24 concentrations of dopamine ranging from 1 × 10−12 to 1 × 10−3 M. Incubations were carried out and filtered as detailed above.
Cyclic AMP Accumulation Assay.
NS20Y cells stably expressing wild-type or mutant D1 receptors were plated out in 12-well tissue culture clusters and used to assess desensitization 36 to 48 h later, when they were at a density of ∼150,000 cells/well, whereas C6 cells stably expressing untagged wild-type and mutant D1 receptors were plated in 48-well clusters at a density of 100,000 to 150,000 cells/well and used 3 to 4 days later. Cells were preincubated for 10 min with 200 μl of assay buffer (Earle's balanced salt solution containing 0.02% ascorbic acid, 2% bovine calf serum, and 500 μM 3-isobutyl-1-methylxanthine). The cells were then placed on ice and drugs added to triplicate wells as indicated. For measurement of dopamine-stimulated cyclic AMP accumulation in C6cells, 1 μM propranolol was added to prevent potential stimulation of endogenous β-adrenergic receptors by dopamine. After incubation for 15 min at 37°C, the assay buffer was decanted and cells were lysed with 3% trichloroacetic acid (100 μl/well). The cyclic AMP in each well was quantified using a competitive binding assay modified as described previously (Watts and Neve, 1996). Ten microliters of cell lysate was incubated in duplicate tubes containing 1.5 nM [3H]cyclic AMP and crude adrenal extract (∼100 μg of protein) in 500 μl of cyclic AMP binding buffer (100 mM Tris-HCl, pH 7.4, 100 mM NaCl, 5 mM EDTA) for 2 h on ice, and then harvested by filtration as described for radioligand binding assays, except using filtermat B. Unknown samples were plotted against a standard curve (0.1–100 pmol) to calculate the amount of cyclic AMP.
Metabolic Labeling.
NS20Y neuroblastoma cells stably expressing the wild-type or mutant D1-EGFP receptors were plated in six-well tissue culture clusters and used 36 to 48 h later, when they were at a density of ∼320,000 cells/well. On the following day, plates were rinsed twice with 1 ml of phosphate-free DMEM, and then incubated for 1 h at 37°C with 200 μCi/ml of [32P]PO4 in 1 ml of phosphate-free DMEM containing 25 mM Na-HEPES, pH 7.4, and 0.02% ascorbic acid. Cells were then treated with 25 μM dopamine or vehicle for 5 or 20 min, followed by cell lysis and affinity purification of D1-EGFP.
Affinity Purification.
The medium was aspirated off and replaced with ice-cold lysis buffer [10 mM Tris-HCl, pH 7.4, 10 mM NaF, 10 mM disodium pyrophosphate, and protease inhibitor cocktail (Roche Applied Science, Mannheim, Germany)]. Cells were then loaded into 50-ml centrifuge tubes and centrifuged at 500 rpm for 5 min at 4°C. The supernatant was decanted, and cells were resuspended in lysis buffer before centrifugation at 40,000g at 4°C for 20 min. The supernatant was removed and the pellet resuspended in 3 ml of ice-cold solubilization buffer (500 mM NaCl, 20 mM Tris-HCl, pH 7.5, 1% n-dodecylmaltose, 10 mM NaF, 10 mM disodium pyrophosphate, and 1 tablet of protease inhibitor cocktail/50 ml). The resuspended pellet was sonicated on ice for 20 s, and the contents were transferred to a 15-ml conical centrifuge tube and placed on a tilt shaker at 4°C for 1 h. After solubilization, the solution was centrifuged at 40,000g at 4°C to remove nonsolubilized material. The His-tagged receptor was then purified using nickel-charged Chelating Sepharose Fast Flow beads (Amersham Biosciences AB, Uppsala, Sweden) as described previously (Kobilka, 1995). The eluate (1 ml) from this procedure was collected in a 1.7-ml microfuge tube and 10 μl of 1 M phosphate buffer, pH 4.5, was added for a final concentration of 10 mM. Next, 20 μl of Protein G Sepharose 4 Fast Flow (Amersham Biosciences AB) and 1 μg of monoclonal anti-autofluorescent protein (Quantum Biotechnologies Inc., Montreal, QC, Canada) were added. After incubating for 2 h at 4°C on a tilt rocker, the tube was centrifuged at 2000gand the supernatant removed by aspiration. The pellet was resuspended in 1 ml of radioimmunoprecipitation assay buffer (10 mM Tris, 150 mM NaCl, 1 mM EDTA, 0.1% SDS, 1% Triton X-100, and 1% sodium deoxycholate) and centrifuged at 2000g; this step was repeated three times. The washed pellet was resuspended in 30 μl of Laemmli sample buffer and incubated at 65°C for 20 min to dissociate the D1-EGFP receptor from the protein G Sepharose. After centrifugation at 2000g for 5 min, the supernatant protein was separated by SDS-PAGE and transferred to a PVDF membrane. The D1-EGFP receptor was detected by immunoblotting as described below using monoclonal anti-GFP (CLONTECH). Membranes were scanned using a Storm 840 imaging system (Molecular Dynamics, Sunnyvale, CA), and then juxtaposed to Kodak X-OMAT film for 24 to 48 h. The film was developed, and the resulting autoradiograph was digitally captured with a Hewlett Packard Scan Jet LP. The bands showing D1-EGFP immunoreactivity were quantified by densitometry with ImageQuant software (Molecular Dynamics).
Immunoblotting.
Proteins were separated by SDS-PAGE through a 10% polyacrylamide gel and transferred to PVDF membranes (Costar, Cambridge, MA). The membranes were blocked overnight with 5% nonfat dry milk, washed with Tris-buffered saline, and incubated with the indicated primary antibody for 2 h. The PVDF membranes were washed, incubated with secondary antibody (fluorescein-linked anti-mouse IgG; Amersham plc, Little Chalfont, Buckinghamshire, UK), and immunodetection was accomplished using an ECF Western blotting kit (Amersham plc). In some cases membranes were stripped and reblotted by using a Western blot recycling kit (Chemicon International, Temecula, CA) and a monoclonal antibody recognizing phosphothreonine (Santa Cruz Biotechnology, Santa Cruz, CA).
Fluorescence Microscopy.
Sterile coverslips (Fisherbrand #1; Fisher Scientific, Fair Lawn, NJ) were placed in 12-well tissue culture dishes and seeded with cells stably expressing D1-EGFP. Imaging of live cells was performed the next day by using either a Leica TCS SP scanning confocal microscope or a Nikon TE200 inverted fluorescent microscope with a CH350L camera. Images captured by the Nikon microscope were deconvolved using software by API Delta Vision (Applied Precision Co., Issaquah, WA). Cells were maintained at 37°C using a Delta T stage adapter (Bioptechs Co., Butler, PA) for the Nikon TE200 microscope and an RC 26 open bath imaging chamber on a PH-1 heater platform (Warner Instruments, Hamden, CT) for the Leica TCS SP. Image quantification was done using NIH Image (version 1.62B; http://rsb.info.nih.gov/nih-image/).
Biotinylation of Membrane Proteins.
Cells grown to 80% confluence on 10-cm tissue culture plates were treated with 25 μM dopamine or vehicle (0.02% ascorbic acid) in DMEM for 20 min, after which the medium was decanted and the plates were placed on ice. The remaining cell surface proteins were then biotinylated and separated from nonbiotinylated proteins by ImmunoPure immobilized streptavidin (Pierce Chemical, Rockford, IL) as described previously (Melikian and Buckley, 1999). Protein was eluted from the streptavidin beads with 30 μl of Laemmli sample buffer and constant mixing for 20 min before separation by SDS-PAGE and transfer to PVDF membranes for Western blotting with monoclonal anti-GFP (CLONTECH). D1-EGFP immunoreactivity was quantified by densitometry using ImageQuant software (Molecular Dynamics).
Data Analysis.
Saturation isotherms, radioligand binding inhibition curves, and dose-response curves for cyclic AMP accumulation were analyzed by nonlinear regression with Prism 3.0 (GraphPad Software, San Diego, CA). K I,K D, and EC50values are geometric means from three or more experiments followed by the limits defined by the asymmetric standard error of the mean. For displacement of radioligand binding by dopamine, curves were analyzed assuming the presence of one and two classes of binding sites. The assumption of two classes of binding sites was accepted whenp < 0.05 for the improvement of the fit, determined using an F test. Statistical comparisons between cells expressing mutant and wild-type receptors or between treated and untreated cells were made using the Student's t test.
Results
Expression of D1-EGFP Receptors in NS20Y Cells.
We constructed a tagged dopamine D1 receptor (D1-EGFP) with a polyhistidine tag attached to the amino terminus and enhanced green fluorescent protein fused to the carboxyl terminus of the receptor (Fig. 1). To eliminate two potential sites of phosphorylation by PKA, alanine substitution mutants of D1-EGFP were made for Thr268 (T268A) and Ser380 (S380A). The untagged D1 receptor as well as tagged wild-type and mutant receptors were stably expressed in NS20Y mouse neuroblastoma cells. For wild-type or mutant D1-EGFP receptors, fluorescence-activated cell sorting-enriched cell populations were selected that expressed each receptor at approximately 2 pmol/mg of membrane protein (Table1). Because the cell line selected for untagged D1 receptor expressed that receptor at a lower density (0.75 pmol/mg of protein) a second population of cells expressing D1-EGFP at a lower density was also used for comparison. Mutant and wild-type receptors had similar affinities for the antagonist radioligand [3H]SCH 23390 (K D = 0.3–0.6 nM; Table 1). Similarly, curves for inhibition of the binding of [3H]SCH 23390 by the agonist dopamine were best fit by assuming the presence of two classes of binding sites for each wild-type and mutant receptor, and the proportions of sites with high affinity for dopamine (24–37%) were similar among the receptors, as were the K H andK L values for dopamine. Thus, the antagonist and agonist binding properties of the D1 receptor were not altered by tagging the receptor or by mutation of Thr268 or Ser380.
To investigate further the functional properties of the tagged and mutant receptors, stimulation of cyclic AMP accumulation by dopamine and the agonist 6-chloro-7,8-dihydroxy-1-phenyl-2,3,4,5-tetrahydro-1H-3-benzazepine (SKF 81297) was assessed. Maximal stimulation of cyclic AMP accumulation was similar in cells expressing wild-type D1, D1-EGFP, T268A, or S380A receptors (Table 2). In contrast, the mean EC50 values for the tagged wild-type and mutant receptors were lower than the value for the untagged wild-type receptor. This was particularly evident for dopamine, which was 3.5- to 7-fold more potent at the EGFP-tagged receptors than at the untagged D1 receptor. The difference was not due to the higher level of expression of most of the EGFP-tagged receptors compared with the untagged receptor, because the difference was also observed for a cell line expressing D1-EGFP at a lower density (Table 2). Thus, neither His- and EGFP-tagging nor mutation of Thr268 or Ser380 greatly altered ability of the D1 receptor to maximally stimulate adenylate cyclase. Tagging the receptor caused a modest leftward shift in the potency of agonists that was similar for wild-type and mutant receptors.
Agonist-Induced Phosphorylation.
To quantify agonist-induced incorporation of phosphate into the D1-EGFP receptor, the receptor was purified by sequential nickel affinity chromatography and immunoprecipitation from NS20Y cells metabolically labeled with 32P. As shown in the Western blot in Fig. 2A, the sequential purification procedure produced a cleaner receptor preparation than did nickel affinity chromatography alone. The incorporation of32P into the D1-EGFP receptor was increased after treatment with 25 μM dopamine for 5 min (76 ± 30%) or 20 min (137 ± 46%) compared with untreated cells (Fig. 2, B and C). Dopamine treatment also increased the incorporation of 32P into S380A, although neither the effect at 5 min (p = 0.2) nor the effect at 20 min (p = 0.08) was statistically significant. Collapsing the data for S380A across time resulted in a 73 ± 31% increase in dopamine-stimulated phosphorylation of the mutant (p< 0.05, n = 24). In contrast, dopamine treatment caused a slight reduction in the phosphorylation of T268A after 5 min, with no difference from untreated cells after 20 min of dopamine treatment (Fig. 2, B and C). Reprobing a subset of the filters with anti-phosphothreonine demonstrated that 20-min treatment with dopamine increased phosphothreonine immunoreactivity in D1-EGFP by 41 ± 15%, whereas the same treatment caused a nonsignificant decrease in phosphothreonine immunoreactivity in T268A (Fig. 3).
To address whether dephosphorylation might be occurring at the plasma membrane, we pretreated cells with 0.45 M sucrose to block receptor internalization. Sucrose pretreatment reduced basal receptor phosphorylation and also prevented dopamine-stimulated incorporation of phosphate (Fig. 4).
To determine whether the agonist-induced phosphorylation of the receptor was mediated by PKA, metabolically32P-labeled cells were pretreated with the PKA inhibitor H-89 (5 nM) for 30 min, followed by 5- or 20-min treatment with 25 μM dopamine. Pretreatment with H-89 prevented agonist-induced phosphorylation of D1-EGFP (Fig. 2, B and C).
Characterization of Agonist-Induced Desensitization.
Desensitization of the D1 receptor was induced by pretreatment of NS20Y cells with 25 μM dopamine before measuring cyclic AMP accumulation. The responsiveness of either the untagged (D1) or tagged (D1-EGFP) receptor decreased rapidly, so that maximal cyclic AMP accumulation was decreased by 40 to 50% within 5 min, with no further decrease observed after pretreatment for up to 60 min (Fig. 5A; Table3). Thus, tagging the D1 receptor with His6 and EGFP did not alter acute desensitization of the receptor.
Because the T268A mutant was resistant to dopamine-induced phosphorylation, we determined whether acute desensitization of the mutant was altered. The magnitude of desensitization of T268A was similar to that of D1-EGFP (Table 3; Fig. 5B), suggesting that PKA-catalyzed phosphorylation of the D1 receptor on Thr268 is not required for maximal acute desensitization. The responsiveness of S380A was also decreased by ∼50% within 5 min of dopamine pretreatment (Table 3; Fig. 5B).
Mutation of Thr268 has been reported to decrease the rate of desensitization of the D1 receptor expressed in C6 glioma cells (Jiang and Sibley, 1999), so it is possible that phosphorylation on this residue regulates desensitization in a cell type-dependent manner. In the present studies, C6 cells expressing untagged D1, T268A, or S380A receptors were treated for 1, 2, or 4 h with 10 nM, 100 nM, or 1 μM dopamine before measuring dopamine-stimulated cyclic AMP accumulation. Pretreatment of C6-D1 cells with dopamine caused a time- and concentration-dependent decrease in subsequent maximal stimulation of cyclic AMP accumulation by dopamine, with the greatest decrease of ∼60% being evident after treatment with 1 μM dopamine for 2 to 4 h (Fig. 6A). As observed for the tagged mutant receptors expressed in NS20Y cells, the time and dose dependence of desensitization of the untagged T268A and S380A receptors stably expressed in C6 cells was similar to that of the D1 receptor (Fig. 6, B and C).
Intracellular Trafficking.
NS20Y cells expressing either D1-EGFP or T268A were used to evaluate agonist-induced trafficking of the D1 receptor. Cells were grown on coverslips that were placed in a heated (37°C) chamber on the microscope stage. The addition of 25 μM dopamine to the medium markedly increased the intensity of fluorescence in the perinuclear region of cells expressing D1-EGFP (Fig.7A). The perinuclear and plasma membrane fluorescence was measured at the same coordinates of images captured every 2 min for 20 min, and the ratio of the two values was calculated for each time point. The ratio of perinuclear/membrane fluorescence increased linearly for up to 20 min of dopamine treatment of cells expressing D1-EGFP, at which time the ratio had more than doubled (Fig. 7C). In contrast, although intracellular fluorescence increased in dopamine-treated cells expressing T268A (Fig.7B), the accumulation of fluorescence in the perinuclear region was not observed, and the ratio of perinuclear fluorescence to fluorescence in the plasma membrane was only slightly enhanced compared with untreated cells (Fig. 7C).
If the altered trafficking of T268A reflects a lack of phosphorylation by PKA, and then inhibition of PKA should similarly alter the trafficking of wild-type D1-EGFP. Cells expressing wild-type D1-EGFP were pretreated with the PKA inhibitor H-89 (5 nM) for 30 min, followed by the addition of 25 μM dopamine. Inhibition of PKA prevented the dopamine-induced increase in perinuclear fluorescence (Fig. 7C).
Agonist-Induced Receptor Internalization.
Biotinylation of cell surface proteins was used to quantify agonist-induced internalization of D1 receptors. NS20Y neuroblastoma cells expressing either wild-type D1-EGFP or T268A receptors were incubated with or without 25 μM dopamine, and then placed on ice for biotinylation of the remaining surface receptors. Membranes were solubilized, and biotinylated proteins isolated by avidin gel matrix purification. The abundance of surface receptors before and after dopamine treatment was quantified by Western blot with an antibody directed against green fluorescent protein (Fig. 8A). Treatment with dopamine for 20 min reduced the abundance of D1-EGFP receptors on the surface of the membrane by 29 ± 11% compared with cells not treated with dopamine (Fig. 8B). Similarly, dopamine decreased the abundance of T268A on the surface of the cell membrane by 32 ± 12%, indicating that phosphorylation on Thr268 is not necessary for rapid receptor internalization.
To determine whether PKA regulates rapid internalization of the D1 receptor, dopamine-induced internalization was assessed after 30-min pretreatment with 5 nM H-89. Under these conditions, dopamine treatment reduced the abundance of D1-EGFP surface receptors by 27 ± 9%, indicating that phosphorylation by PKA is not a requirement for internalization.
The lack of effect of the T268A mutation on desensitization of cyclic AMP accumulation and on receptor internalization suggested that phosphorylation by PKA regulates a late step in the endocytotic trafficking of the D1 receptor to the perinuclear region, but that agonist-induced trafficking close to the cell membrane is not regulated by PKA-dependent phosphorylation of Thr268. The agonist-induced increase in T268A-containing endocytic vesicles shown in Fig. 7B suggested a similar conclusion. To evaluate this more closely, we tracked the appearance and disappearance of EGFP-containing vesicles in NS20Y cells expressing D1-EGFP or T268A. Dopamine treatment of either wild-type (Fig. 9A) or mutant (Fig. 9B) receptors caused the formation of vesicles that remained close to the plasma membrane and seemed to re-merge with the membrane within several minutes of vesicle formation.
Discussion
The mechanisms of desensitization and resensitization of the dopamine D1 receptor are not well understood. Although the high homology of the D1 receptor with the β2-adrenergic receptor might indicate that similar mechanisms regulate the responsiveness of the receptors, even subtypes of β-adrenergic receptors differ substantially in this respect (Shiina et al., 2000). For β-adrenergic receptors, a distinction has been made between desensitization that is homologous (resulting from stimulation of the same receptor) or heterologous (resulting from stimulation of a different receptor subtype, or nonreceptor-mediated activation of a second messenger) (Harden, 1983). A mechanism for heterologous desensitization is phosphorylation of the receptor by PKA (Clark et al., 1989; Hausdorff et al., 1989), whereas homologous desensitization of the β2-adrenergic receptor involves phosphorylation of the agonist-occupied receptor by both GRK (Benovic et al., 1986) and PKA (Post et al., 1996; Moffett et al., 2001). Heterologous desensitization of the β2-adrenergic receptor is mediated by PKA-dependent phosphorylation of Ser262, located in the carboxyl terminal portion of the third cytoplasmic loop, a position similar to that of Thr268 in the D1 receptor (Clark et al., 1989; Hausdorff et al., 1989; Yuan et al., 1994). In contrast, the PKA-dependent component of homologous (agonist-dependent) desensitization of the β2-adrenergic receptor involves phosphorylation of the PKA site Ser345,346 in the cytoplasmic tail of the receptor (Moffett et al., 2001). For the D1 receptor, too, PKA and GRK may both contribute to homologous desensitization (Zhou et al., 1991). The aim of this study was to determine the role of PKA-catalyzed phosphorylation of the D1 receptor in agonist-stimulated receptor desensitization and internalization.
To facilitate purification and imaging of the receptor, we attached a polyhistidine tag to the amino terminus and EGFP to the carboxyl terminus of the D1 receptor. We chose NS20Y neuroblastoma cells for stable expression of the receptors because the line has some characteristics of striatal neurons (Amano et al., 1972), suggesting that it may be more physiologically relevant than non-neuronal cells for the characterization of a neuronal receptor such as the D1 receptor. Although NS20Y cells have been reported to express endogenous dopamine D1receptors at low levels (Barton and Sibley, 1990), in our hands both D1 receptors and dopamine-stimulated cyclic AMP accumulation (data not shown) were barely detectable in untransfected cells.
The utility of any tagged molecule depends on the degree to which the tagged protein retains the functional characteristics of the wild-type protein. The polyhistidine- and EGFP-tagged D1receptor was similar to the untagged D1 receptor in several respects. First, the receptors had similar affinities for the antagonist [3H]SCH 23390 and for the agonist dopamine. Second, the receptors had similar abilities to couple to G proteins, as indicated by the proportion of receptors with high affinity for dopamine in competition binding assays. Third, the subcellular localization of D1-EGFP in NS20Y cells was comparable with that determined in other cell types by using immunocytochemistry (Ng et al., 1994, 1995; Ariano et al., 1997;Vickery and von Zastrow, 1999). Finally, both receptors mediated agonist-stimulated cyclic AMP accumulation to a similar extent, although dopamine, in particular, was more potent at D1-EGFP than at the untagged D1 receptor.
Agonist-induced phosphorylation of the dopamine D1 receptor has been demonstrated in Sf9 (Ng et al., 1994), human embryonic kidney 293 (Tiberi et al., 1996), and C6 (Gardner et al., 2001) cells. We report herein that dopamine-induced phosphorylation of D1-EGFP expressed in NS20Y neuroblastoma cells occurred within 5 min, and that dopamine treatment increased the phosphorylation of the receptor approximately 2-fold. The increased incorporation of phosphate was accompanied by increased phosphothreonine immunoreactivity. The primate D1 receptor has three potential sites of PKA-catalyzed phosphorylation in cytoplasmic domains. Although the predicted rank order of preference of the sites for phosphorylation by PKA is Thr136 > Thr268 = Ser380 (Kennelly and Krebs, 1991), we mutated the two residues suggested by previous work to be involved in desensitization of the D1 receptor (Thr268;Jiang and Sibley, 1999), or to be a site of PKA-catalyzed phosphorylation (Ser380; Zamanillo et al., 1995). Whereas mutation of Ser380 had no effect on dopamine-stimulated phosphorylation of the receptor, mutation of Thr268 prevented the dopamine-induced incorporation of phosphate. These data suggest strongly that occupancy of the D1 receptor by dopamine stimulates phosphorylation of the receptor at Thr268. Pretreatment with the PKA inhibitor H-89 abolished dopamine-induced receptor phosphorylation, supporting the hypothesis that the phosphorylation was due to activation of PKA. Dopamine treatment tended to decrease the phosphorylation of T268A and of the D1 receptor in the presence of H-89, particularly after short (5 min) dopamine treatments. It is possible that dopamine treatment enhances phosphatase activity, an effect that would be detectable only when dopamine-induced phosphorylation is prevented.
Although mutation of Thr268 prevented PKA-dependent phosphorylation of the D1 receptor, the same mutation had no detectable effect on desensitization of dopamine-stimulated cyclic AMP accumulation in either NS20Y or C6 cells, suggesting that desensitization in either cell line does not depend on PKA-dependent phosphorylation of this residue. The lack of effect of the PKA inhibitor H-89 on desensitization might suggest that PKA does not play any role in acute desensitization, but this result should be interpreted with caution because of the considerable evidence implicating PKA in homologous desensitization of the D1 receptor (Zhou et al., 1991; Jiang and Sibley, 1999; Ventura and Sibley, 2000; Gardner et al., 2001). In particular, previous work has demonstrated that mutation of Thr268 reduces the rate more than the extent of desensitization of the D1receptor (Jiang and Sibley, 1999). It may be significant that D1 receptor desensitization in C6 cells occurred more slowly and to a lesser maximal extent in the present work than in previous work with this cell line (Lewis et al., 1998; Jiang and Sibley, 1999), which could explain the discrepancy in the observed role of Thr268 in receptor desensitization. Furthermore, because desensitization in NS20Y cells was maximal at the first time point studied, it is possible that prevention of PKA-dependent phosphorylation caused a reduction in the rate of desensitization that could not be detected by this experimental design. Species differences may also be important, because the rat D1 receptor has several Ser/Thr residues in the C terminus and a potential PKA phosphorylation site at Ser229 that are not shared by the primate D1 receptor.
It is surprising that inhibition of receptor phosphorylation by PKA, either by treatment with H-89 or by mutation of Thr268, prevented detectable agonist-induced phosphorylation of the D1 receptor without preventing receptor desensitization. Work with the D1 receptor (Tiberi et al., 1996; Gardner et al., 2001) and other closely related receptors suggests, on the one hand, that GRK-mediated phosphorylation would be detectable even when PKA is inhibited and, on the other hand, that the rate and extent of homologous desensitization would be greatly reduced in the complete absence of agonist-induced phosphorylation. One possibility is that the D1 receptor is not phosphorylated by GRKs in our line of NS20Y cells. An alternative explanation for these results is that the GRK-phosphorylated D1 receptor is rapidly dephosphorylated in NS20Y cells, so that our inability to detect agonist-induced phosphorylation when PKA was inhibited was a false negative result. This hypothesis is consistent with the observation that dephosphorylation of the D1 receptor can occur in the plasma membrane, in the absence of receptor internalization (Gardner et al., 2001), and also with preliminary data indicating that sucrose-induced inhibition of receptor internalization greatly decreases basal and dopamine-induced phosphorylation of the D1receptor, and that phosphorylation of T268A or of the D1 receptor in the presence of H-89 is enhanced by dopamine if cells are pretreated with the phosphatase inhibitor calyculin (J. N. Mason and K. A. Neve, unpublished observations). An alternative explanation is that phosphorylation by PKA is a prerequisite for phosphorylation by GRK, so that preventing the former also prevents the latter. This could be similar to the PKA-induced enhancement of β2-adrenergic receptor phosphorylation by GRK (Moffett et al., 2001), but would also suggest that agonist-induced phosphorylation is not necessary for D1 receptor desensitization, as has been suggested for several other GPCR subtypes (Malecz et al., 1998; Sadeghi et al., 1998; Olivares-Reyes et al., 2001). Finally, as noted above, the particular characteristics of desensitization observed in the present studies might have precluded detection of an effect of PKA inhibition on D1receptor desensitization.
In agreement with previous studies of the D1receptor (Ng et al., 1995; Ariano et al., 1997; Vickery and von Zastrow, 1999), agonist treatment caused a rapid accumulation of D1-EGFP in the cytoplasm of NS20Y cells. In cells expressing the wild-type receptor, the accumulation was particularly evident in a perinuclear region that may represent an accumulation of recycling vesicles. Comparing agonist-induced cellular redistribution of the wild-type D1-EGFP with that of the mutant T268A receptor, we found there to be significantly less accumulation of T268A in the perinuclear region. Consistent with the hypothesis that the effect of mutation of Thr268 was due to the loss of a PKA phosphorylation site, treatment of cells expressing wild-type D1-EGFP with H-89 also prevented the accumulation of fluorescence in the perinuclear region.
Vesicles that form at the plasma membrane rapidly (<1 min) fuse to early endosomes, which are comprised of sorting endosomes and the recycling compartment (for review, see Mukherjee et al., 1997). After initial fusion with sorting endosomes, internalized receptors are thought to be sorted to the recycling compartment or to late endosomes, from which they are recycled to the membrane or shuttled to lysosomes, respectively. Depending on the cell type, the recycling compartment may be concentrated in the juxtanuclear region, or dispersed throughout the cell, whereas late endosomes are mainly perinuclear and close to thetrans-Golgi network. Agonist-induced trafficking of D1-EGFP to the perinuclear region could represent accumulation in the recycling compartment or sorting to late endosomes, suggesting that PKA-dependent phosphorylation of the D1 receptor regulates trafficking into one of these compartments. The perinuclear accumulation of fluorescence showed little colocalization with fluorescent markers for lysozomes or the Golgi apparatus (data not shown), indicating that D1-EGFP was not accumulating in either of those compartments.
Although trafficking of the D1 receptor to the juxtanuclear region was inhibited by mutation of Thr268, the T268A mutant desensitized normally and seemed to accumulate intracellularly during treatment with dopamine. We used two approaches to evaluate internalization and recycling of the D1-EGFP receptor. To quantify receptor internalization, cell surface receptors were biotinylated after dopamine treatment. Under these conditions, a decreased amount of biotinylated receptor is presumed to represent receptor that has been internalized by dopamine treatment. We found that dopamine treatment reduced the surface density of receptors by a similar amount in cells expressing D1-EGFP or T268A. In the second approach, we used high-resolution fluorescence microscopy to confirm that, in cells expressing wild-type or T268A mutant receptors, EGFP-containing vesicles form and recycle to the membrane during treatment with dopamine. Thus, as suggested for the cholecystokinin receptor (Roettger et al., 1995), the D1-EGFP receptor may be resensitized through rapid dephosphorylation in a vesicular compartment adjacent to the plasma membrane. On the other hand, because inhibition of endocytosis of the D1 receptor in C6glioma cells does not prevent receptor dephosphorylation (Gardner et al., 2001), endocytosis may not be required for resensitization.
The signals that regulate the passage of internalized receptors to different endosomal compartments are not known, although receptor phosphorylation sites may contribute to this sorting. Phosphorylation of the epidermal growth factor receptor by protein kinase C shunts the receptor from the late endosome/lysosome pathway into a recycling pathway (Bao et al., 2000).N-Methyl-d-aspartate-induced internalization of the α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor is associated with dephosphorylation of a PKA site on the α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor GluR1 subunit, and inhibition of PKA reduces the rate of reinsertion of the receptor into the cell membrane (Ehlers, 2000). Our data demonstrate that inhibiting PKA-catalyzed phosphorylation of the D1-EGFP receptor pharmacologically or by mutation of Thr268 prevented agonist-induced trafficking of the receptor to a perinuclear region, without altering receptor internalization or desensitization. These results are consistent with a model in which phosphorylation of Thr268 of the D1 receptor by PKA regulates a late step in the sorting of the receptor to the recycling compartment or to late endosomes.
Footnotes
- Received May 9, 2001.
- Accepted December 14, 2001.
-
This work was supported by the Department of Veterans Affairs Career Scientist and Merit Review Programs and by grants R01-MH45372, T32-DA07262, and F31-MH12435 from the National Institutes of Health. Portions of this work were presented previously in abstract form (Soc Neurosci Abstr 26:1422).
Abbreviations
- GPCR
- G protein-coupled receptor
- PKA
- cyclic AMP-dependent protein kinase
- GRK
- G protein-coupled receptor kinase
- D1-EGFP
- D1 receptor with enhanced green fluorescent protein attached to the C terminus and a polyhistidine tag at the amino terminus
- H-89
- N-[2-(p-bromocinnamylamino)ethyl]-5-isoquinolinesulfonamide
- PCR
- polymerase chain reaction
- DMEM
- Dulbecco's modified Eagle's medium
- PAGE
- polyacrylamide gel electrophoresis
- PVDF
- polyvinylidene difluoride
- GFP
- green fluorescent protein
- SCH 23390
- R(+)-7-chloro-8-hydroxy-3-methyl-1-phenyl-2,3,4,5-tetrahydro-1H-3-benzazepine
- U.S. Government