The splice variant D3nf reduces ligand binding to the D3 dopamine receptor: evidence for heterooligomerization

Molecular Brain Research 80 (2000) 63–74 www.elsevier.com / locate / bres

Research report

The splice variant D3nf reduces ligand binding to the D3 dopamine receptor: evidence for heterooligomerization a

a

a,c

Jennifer L. Elmhurst , Zhidong Xie , Brian F. O’Dowd , Susan R. George a

a,b,c ,

*

Department of Pharmacology, Room 4358, Medical Sciences Building, University of Toronto, 1 King’ s College Circle, Toronto, Ontario, Canada M5 S 1 A8 b Department of Medicine, University of Toronto, 1 King’ s College Circle, Toronto, Ontario, Canada M5 S 1 A8 c Centre for Addiction and Mental Health, Toronto, Canada Accepted 23 May 2000

Abstract The D3 dopamine receptor belongs to the D2-like family of dopamine receptors. As with other members of this group, the D3 dopamine receptor gene contains introns which allow for alternative splicing of gene products. The best characterized of the human D3 dopamine receptor mRNA splice variants encodes a truncated protein called D3nf. The D3 dopamine receptor and D3nf were epitope-tagged and expressed in Sf9 insect cells by recombinant baculovirus infection. The D3 dopamine receptor showed saturable, high affinity binding of agonists and antagonists, consistent with reported D3 dopamine receptor pharmacology. When the D3 dopamine receptor and D3nf were co-expressed, the apparent density of D3 dopamine receptor expression, as determined by radioligand binding, was significantly lowered compared to D3 dopamine receptor expressed alone. This effect of D3nf was specific for the D3 dopamine receptor, since co-expression with the D2 dopamine receptor or b2-adrenoceptor had no effect on binding. Confocal immunofluorescence studies were used to confirm that both D3 dopamine receptor and D3nf were well expressed on the cell surface and densitometric analysis of cell surface membrane protein confirmed that D3nf did not significantly alter the amount of D3 dopamine receptor expressed. Photoaffinity labelling with [ 125 I]azidonemonapride showed that the amount of ligand bound by membranes co-expressing D3 dopamine receptor and D3nf was significantly less than that bound by membranes expressing D3 dopamine receptor alone. The greatest decrease in binding was observed in the D3 dopamine receptor oligomeric forms. Ligand binding to dimers and tetramers was reduced by 69 and 46%, respectively, indicating effects of a protein–protein interaction. Co-immunoprecipitation confirmed that the D3DR and D3nf interact with each other. These data indicate that D3nf heterodimerizes with the D3 dopamine receptor and decreases the capacity of D3 dopamine receptor to bind ligand.  2000 Elsevier Science B.V. All rights reserved. Theme: Neurotransmitters, modulators, transporters and receptors Topic: Catecholamine receptors Keywords: Dopamine D3 receptor; D3nf; Splice variant; Dimers; Oligomers

1. Introduction Dopamine receptors are of interest clinically because they are targets in the treatment of certain neurological disorders including schizophrenia and Parkinson’s disease. Dopamine receptors are classified into two main groups, D1-like and D2-like, based on their pharmacology, cou*Corresponding author. Tel.: 11-416-978-3367; fax: 11-416-9712868. E-mail address: [email protected] (S.R. George).

pling to adenylyl cyclase (AC) and gene structure. The D3 dopamine receptor belongs to the D2-like subfamily and its limbic distribution in the brain implies that drugs selective for this receptor may provide clinical efficacy in the treatment of the above disorders without extrapyramidal side effects [24]. As with other D2-like receptors, the D3 dopamine receptor gene contains introns which allow for the possibility of alternative splicing of gene products. Indeed, several splice variants of the D3 dopamine receptor have been reported in the rat, mouse and human

0169-328X / 00 / $ – see front matter  2000 Elsevier Science B.V. All rights reserved. PII: S0169-328X( 00 )00120-0

64

J.L. Elmhurst et al. / Molecular Brain Research 80 (2000) 63 – 74

[3,7,8,13,17,20,22], in contrast to the D2 dopamine receptor which is known to exist in only two alternatively spliced forms. Many of the D3 dopamine receptor splice variants are predicted to encode truncated receptors, the most extensively characterized of which is D3nf [20]. The D3nf cDNA differs from the cDNA for the D3 dopamine receptor by the deletion of 98 bp in the portion of the gene which encodes the putative third intracellular loop. The deletion results in a nucleotide shift in the open reading frame, leading to novel sequence and a premature stop codon. Hence, D3nf differs from the D3 dopamine receptor in its predicted carboxyl-terminal amino acid sequence and does not contain the D3 dopamine receptor transmembrane domains 6 or 7. D3nf mRNA has been shown to be as abundant as D3 dopamine receptor message and is expressed in the same regions of human brain [11]. Further, antibodies raised against the unique carboxyl region of D3nf specifically immunoreact in human brain, demonstrating that D3nf mRNA is translated into protein [11]. The physiological function of D3nf is not yet known; however, enhanced production of D3nf mRNA has been reported in brains of schizophrenics [19]. One possible role for truncated receptors, such as those arising through mRNA splicing or mutation, may be to modulate the function of the full-length version of the receptor. Some truncated receptors have been shown to interact with their full-length counterparts and attenuate the function of the full-length receptor in a dominant negative fashion through inhibition of cell surface expression [1,2,9,18] or coupling [18]. The implications of this dominant negative effect on cellular function can be profound. For example, a mutant truncated form of the CCR5 receptor prevents the CCR5 receptor, a co-receptor for HIV cell entry, from being expressed at the surface of cells thereby reducing the cells’ susceptibility to HIV-1 infection [1]. However, the D3nf is not a mutant form of the D3 dopamine receptor but rather a physiologically present truncated splice variant which co-localizes with the D3 dopamine receptor in human brain [20]. The physiological presence of D3nf in brain and its co-localization with the D3 dopamine receptor may indicate that it has an important functional role yet to be elucidated. In this study, we investigated the possibility that D3nf modulates the pharmacology of the full-length D3 dopamine receptor through the formation of heterodimers.

2. Materials and methods

2.1. Construction of D3 dopamine receptor and D3 nf expression vectors The cDNA encoding the human D3 dopamine receptor in the expression vector pRC / CMV was obtained as a gift from Dr. C. Schmauss (Mt. Sinai School of Medicine, NY). The D3 dopamine receptor was tagged with amino

acids 411–421 of the human c-myc protein (EQKLISEEDL) and D3nf was tagged with FLAG, an eightamino acid synthetic peptide (DYKDDDDK), to facilitate immunodetection. The TransformerE site-directed mutagenesis method (Clontech, Palo Alto, CA) was used to insert the DNA code for the c-myc epitope tag after the ATG start codon of the cDNA encoding the D3 dopamine receptor. All enzymes for DNA manipulation were purchased from Pharmacia Biotech (Baie d’Urfe, Quebec). The primer sequences are as follows, with mutated sequences indicated in bold and the sequence encoding the c-myc epitope is underlined: mutagenic primer; CGCGGCCGCGGTGAG AA CC TG TTT ATG GAACAAAAGCTTATTTCTGAA ]]]]]]]]]] GAAGACTTGGCATCTCTGAGTCAGCTGAGTAGCC ]]]]] and selection primer; ACGGATCGGGAAATATCCCGATCCCCTATG. The desired mutations were introduced via a mutagenic primer and mutated plasmids were selected for via a selection primer which eliminated a unique restriction site (for BglII) from the vector. The altered sequence was confirmed by sequencing in both directions. The polymerase chain reaction (PCR) was used to create the cDNA encoding D3nf, based on the published sequence [20]. The strategy exploited a unique EcoRI site in the D3 dopamine receptor cDNA located 59 of the point at which 98 pairs are deleted from D3nf cDNA. Two PCR fragments were generated. Fragment 1 consisted of the amino portion of the receptor and was created using primer A (CGCGGCCGCGGTGAGAACCTGTTTAAGCTTATG GACTACAAGGACGACGACGACAAG ]]]]]]]]]]]] GCATCTCTGAGTCAGCTGAGTAGCCACCTG), which introduced a HindIII site (bolded) for subcloning and the FLAG code (underlined), and primer B (CTCAGGGAATTCCGAGTC TTCTCCTCTC), which terminated just before the deletion point in D3nf and contained the EcoRI site (bold). Fragment 2 consisted of the carboxyl portion of the receptor and was created using primer C (CTCGGAATTCCCTGAGT ↑ GCCACTTCGGGAGAAGAAGGCAACCCAAATGGTGGCCATTGTGCTTGGG), which annealed completely to the portion of the gene 39 of the deletion point (↑) and contained sequence 59 of the deletion until the EcoRI site (bold), and primer D (GACCCTCGAGTCAGCAAGACAGGATCTTGAGGAAG), which introduced a XhoI site (bold) for subcloning. Fifty ml PCR reactions consisting of 23 Buffer A (with 1 mM MgCl 2 ), dNTPs (250 mM each), eLONGase (GibcoBRL, Burlington, ON, 1 ml), primers (A and B or C and D, 170 ng each) and 0.5 mg pRC / CMV/ D3 DNA were pre-denatured at 948C for 5 min and subjected to 35 cycles of denaturing at 948C for 30 s, annealing at 688C for 30 s and extension at 688C for 1 min. Reactions were soaked at 688C for 7 min for final product elongation and stored at 2208C until use. A control reaction substituted water for the template DNA. PCR products (4 ml) were run on a 1% agarose / ethidium bromide (0.5 mg / ml) gel for confirmation of product size.

J.L. Elmhurst et al. / Molecular Brain Research 80 (2000) 63 – 74

The PCR products were extracted from the reaction mix using a QIAquick PCR purification kit (Qiagen, Chatsworth, CA). The PCR products were digested by EcoRI and subcloned into the HindIII and XhoI sites of the vector pcDNA3 using standard techniques. The entire cDNA sequence was confirmed by sequencing in both directions.

2.2. Generation of D3 dopamine receptor and D3 nf recombinant baculovirus The cDNAs for the D3 dopamine receptor and D3nf were subcloned into the HindIII and KpnI site of pFastBac1. The generation of recombinant baculovirus was obtained following the Bac-to-BacE baculovirus expression system protocol (Gibco-BRL, Burlington, ON). Briefly, competent DH10Bac cells were transformed with 0.5 mg of vector DNA and plated onto LB selection plates. A helper plasmid in DH10Bac cells facilitated the transposition of a mini-Tn7 region of pFastBac1 to the miniattTn7 attachment site on the bacmid genome. Insertion disrupted the lacZa gene thereby allowing recombinant bacmid identification on selection plates. Candidate white colonies were grown overnight and the recombinant bacmid DNA was isolated by mini-prep for high-molecular weight DNA (Gibco-BRL, Burlington, ON). PCR was used to verify the insertion of cDNA into the bacmid using specific primers (M13 / pUC reverse and forward, GibcoBRL). Recombinant bacmid DNA was transfected into Spodoptera frugiperda (Sf9) cells using cellFECTIN, according to Gibco-BRL’s BEVS protocol. The virus containing supernatant was collected after 72 h and stored at 48C until use.

2.3. Sf9 cell culture and viral amplification Sf9 cells were from American Type Culture Collection (Rockville, MD). All materials for Sf9 cell culture and recombinant baculovirus production including media, antibiotics and supplements were purchased from Gibco-BRL (Burlington, ON). Sf9 cells were grown in a monolayer or suspension culture in Grace’s Insect Medium supplemented with penicillin (50 units / ml), streptomycin (50 mg / ml), qualified heat-inactivated fetal bovine serum (10%), and Pluronic F-68 (Sigma, St. Louis, MO, 1%). Cells were maintained in a monolayer culture in T-125 flasks at 278C for no more than 30 generations. For infection, approximately 2310 5 viable cells / ml were seeded in a 50–100-ml culture. Flasks were kept at 278C on an orbital shaker set at 135 rpm. After 2–3 days, once cells reached mid-log phase growth (at a density of approximately 1.5–2310 6 ml), they were infected with the appropriate recombinant baculovirus at a multiplicity of infection of approximately 5. The number and viability of the cells was assessed using a hemocytometer and trypan blue dye exclusion. Cell membranes were harvested after 48 h as described below. Cell viability when infecting was

65

.98% and after 48 h of viral infection was between 70 and 100%.

2.4. Sf9 membrane preparation Membranes were prepared at 48C. Infected cells were pelleted by centrifugation and washed with phosphatebuffered saline. Cells were resuspended in hypotonic lysis buffer (5 mM Tris–HCl (pH 7.4), 2 mM EDTA and the protease inhibitors benzamidine (10 mg / ml), leupeptin (5 mg / ml) and soybean trypsin (10 mg / ml)). The cells were homogenized with a Beckman polytron for 60 s. The homogenates were centrifuged at 1003g to pellet unbroken cells and nuclei. The supernatant was centrifuged at 27 0003g for 20 min to obtain a P2 pellet which was resuspended in lysis buffer and stored at 2708C until use or used immediately for radioligand binding assays. Protein concentration was determined using a Bio-Rad Assay Kit (Toronto, ON) with bovine serum albumin as the standard.

2.5. SDS–PAGE and immunoblot analysis Fifty mg of membrane protein were solublized in SDS sample buffer (125 mM Tris–HCl, pH 6.8, 4% SDS, 20% glycerol, 0.005% bromophenol blue and 5% b-mercaptoethanol) and electrophoresed on 12% pre-cast acrylamide gels (Novex, Encinitas, CA). The proteins were transferred onto nitrocellulose using a wet-dry apparatus. The nitrocellulose was blocked in a solution of 5% milk protein in TTBS for 1 h. The blot was then rinsed with TTBS for 10 min and incubated overnight at 48C in a 1:1000 dilution of primary antibody. The 9E10 antibody (Santa Cruz Biotechnology, Santa Cruz, CA) was used to detect D3 dopamine receptor and the M5 antibody (Sigma) was used for D3nf. The blot was washed 3 times over the course of an hour and incubated for 2–3 h with a goat anti-mouse IgG coupled to alkaline phosphatase (Bio-Rad, Toronto, ON), diluted 1:1000 in TTBS. The blot was washed and developed in 0.1 M Tris–HCl, pH 9.5, containing BCIP/ NBT (5-bromo-4-chloro-3-indolyl phosphate / nitroblue tetrazolium) substrates (Bio-Rad). The relative intensities of bands detected by immunoblot were determined using densitometry. Gels were digitized and analyzed using the computer program MCID-MR (Imaging Research, St. Catherines, Ontario).

2.6. Immunoprecipitation P2 membranes from infected Sf9 cells were prepared as above and solubilized overnight with gentle agitation in freshly prepared solubilization buffer containing 2% digitonin, 100 mM NaCl, 10 mM Tris–HCl, pH 7.4, 5 mM EDTA and protease inhibitors. Non-solubilized membranes were pelleted at 27 0003g for 20 min. Digitonin was removed from the supernatant by dialysis with Centriprep-

66

J.L. Elmhurst et al. / Molecular Brain Research 80 (2000) 63 – 74

30 concentrators (Amicon) against ice-cold buffer S consisting of 100 mM NaCl, 10 mM Tris–HCl, pH 7.4, with protease inhibitors. The washed and concentrated fractions were pre-cleared at 48C with a 1 / 40 dilution of agarose fixed goat anti-mouse IgG overnight. The solubilized receptors (c-myc D3DR or Flag-D3nf) were immunoprecipitated with the mouse monoclonal 9E10 or Flag antibody at a 1 / 50 dilution in buffer S for 6 h and agitated gently overnight with a 1 / 40 dilution of agarose-fixed goat anti-mouse IgG. The immunoprecipitates were washed with cold-buffer S, solubilized in SDS sample buffer, and electrophoresed on SDS–PAGE as described.

2.7. Photoaffinity labelling Twenty-five mg of membrane protein was resuspended in antagonist binding buffer (see below) with protease inhibitors and incubated in the dark with saturating concentrations (1–5 nM) of [ 125 I]azidonemonapride (NIMH, custom synthesis program) for 1.5 h at 228C. The membranes were placed in an ice water slurry and exposed to ultraviolet light (370 nm) for 2 min. Membranes were pelleted by centrifugation and subjected to SDS–PAGE. The gel was fixed and dehydrated in a 40% methanol / 10% acetic acid solution for 30 min, dried on a gel dryer and exposed to Kodak XAR film with an intensifying screen at 2708C for 2–7 days. Non-specific binding was defined in the presence of 1 mM (1)-butaclamol.

2.8. Receptor binding assays Saturation or competition binding experiments were done using P2 membrane proteins prepared fresh from infected Sf9 cells or from membranes stored at 2708C as described. Binding of [ 3 H]spiperone or [ 3 H]nemonapride (NEN Life Science Products, Boston, MA, 32–6000 pM final concentration) was assayed in duplicate, after 2 h incubation at 228C in antagonist binding buffer (50 mM Tris–HCl, 5 mM EDTA, 1.5 mM CaCl 2 , 5 mM MgCl 2 , 5 mM KCl, 120 mM NaCl, pH 7.4) with protease inhibitors. Nonspecific binding was defined as binding in the presence of 1 mM (1)-butaclamol. Competition experiments were done in triplicate with increasing concentrations (10 212 to 10 23 M) of ligand. The concentration of radioligand used in the assays was approximately equivalent to its KD . Tubes were incubated for 2 h at 228C in a final volume of 1 ml with antagonist binding buffer for antagonists or buffer without NaCl for agonists. When dopamine was used, incubation was performed in the dark and the binding buffer contained 0.1% ascorbic acid to prevent oxidation. Each reaction tube contained approximately 5 mg protein which was determined to not contain enough receptor protein to significantly deplete the ligand. Bound ligand was separated from free by rapid filtration through a Brandel 48-well cell harvester onto Whatman GF / C filters and washed with 10 ml of ice-cold 50 mM Tris–HCl, pH

7.4. Filters were placed into glass vials and incubated overnight with scintillation fluid. Tritium was counted using a Beckman LS 6500 scintillation counter.

2.9. Adenylyl cyclase assay The assay was performed as described previously [14]. The assay mix contained 0.012 mM ATP, 0.1 mM cAMP, 0.053 mM GTP, 2.7 mM phosphoenolpyruvate, 0.2 units of pyruvate kinase, 1 unit of myokinase, 0.13 mCi of [ 33 P]ATP (NEN Life Science Products, Boston, MA) in a final volume of 50 ml. Cyclase activity in the presence of dopamine (10 212 to 10 23 M) was determined in triplicate. The reaction was initiated by the addition of 25 mg of freshly prepared membrane protein, allowed to progress for 20 min at 278C and stopped by the addition of 1 ml of an ice-cold solution containing 0.4 mM ATP, 0.3 mM cAMP and [ 3 H]cAMP (25 000 cpm). cAMP was isolated by sequential column chromatography using Dowex cation exchange resin and aluminum oxide.

2.10. Immunocytochemistry Immunofluorescent staining of whole or permeabilized Sf9 cells infected with D3 dopamine receptor, D3nf or both recombinant baculoviruses was performed in parallel. Control experiments utilized cells infected with wild-type baculovirus. Cells were obtained 48 h post-infection with .85% viability as assessed by trypan blue staining. The cells were pelleted and fixed in freshly prepared 4% paraformaldehyde in phosphate-buffered saline, pH 7.3, for 15 min in Eppendorf tubes, followed by two washes in phosphate-buffered saline. Cells were either resuspended in phosphate-buffered saline, for whole cell labelling, or permeabilized with methanol at 2208C for 3 min. To reduce non-specific staining, the cells were incubated for 30 min in a blocking solution of 1% bovine serum albumin and 5% heat-inactivated goat serum. Cells were incubated for 60 min with 9E10 (mouse) and / or M2 (rabbit, Sigma) primary antibodies (diluted 1:50 in phosphate-buffered saline), followed by three 5-min washes. The samples were covered and further incubated for 60 min with goat antimouse fluorescein isothiocyanate-conjugated secondary antibody and / or goat anti-rabbit Texas Red-conjugated secondary antibody (diluted 1:500 in phosphate-buffered saline, Calbiochem, San Diego, CA), followed by three 5-min washes. The cells were resuspended in the mounting media Mowiol-88 (Hoechst, Montreal, Quebec) with 2.5% 1,4-diazobicyclo-octane (Sigma) to reduce photobleaching. Several drops of the media were placed on a slide, covered by a coverslip and allowed to polymerize overnight. Slides were wrapped in foil and stored at 48C until use. Visualization of the cells was achieved using a Zeiss LSM516 laser scanning microscope (Carl Zeiss Jena) equipped with a krypton–argon laser.

J.L. Elmhurst et al. / Molecular Brain Research 80 (2000) 63 – 74

2.11. Data analysis Radioligand binding data were analyzed by nonlinear least-squares regression using the computer program Prism (GraphPad Software, San Diego, CA). Data were fitted to the better of one or two binding sites. The significance of the resolution in the two components of a competition curve was performed using a F-test, where F is the ratio of the deviation of mean squares obtained in the one-site model and that in the two-site model. Statistical comparisons between the binding parameters determined for D3 dopamine receptor expressed alone or in combination with D3nf were performed using an unpaired Student t-test with a level of significance of P,0.05. pKi and pKD were used in statistical comparisons as they are normally distributed.

3. Results

3.1. Binding characteristics of the D3 dopamine receptor and D3 nf expressed alone in Sf9 cells To facilitate immunological analysis of the D3 dopamine receptor and D3nf, the proteins were epitope-tagged with the c-myc or FLAG epitope, respectively. The radioactive antagonists [ 3 H]nemonapride and 3 [ H]spiperone were used to assess the saturation binding characteristics of the D3 dopamine receptor expressed in Sf9 cells. As shown in Fig. 1, the binding of [ 3 H]nemonapride to membranes expressing D3 dopamine receptor was saturable and specific. No specific binding of the antagonists was observed in uninfected Sf9 cells or in cells infected with wild-type baculovirus (data not shown).

67

The binding curves for [ 3 H]spiperone and 3 [ H]nemonapride were best fitted to a single state of the receptor with KD values of 4866126 and 145628 pM, respectively (Table 1). To define the pharmacology of the D3 dopamine receptor expressed in Sf9 cells, various dopaminergic antagonists and agonists were used to compete for [ 3 H]spiperone or [ 3 H]nemonapride binding. The inhibition of [ 3 H]spiperone or [ 3 H]nemonapride binding by agonists and antagonists was monophasic and best fitted to a singleclass of binding sites. The Ki values are shown in Table 2 and are consistent with results from other systems [4,12,23]. The primary amino acid sequence of the D3nf diverges from that of the D3 dopamine receptor after transmembrane domain 5 in the third intracellular loop for 55 amino acids, after which it is prematurely truncated. Hydrophobicity analysis suggests that this divergent region contains a span of hydrophobic amino acids which is predicted to cross the membrane forming a non-homologous transmembrane domain 6 (Fig. 2). No specific binding of [ 3 H]nemonapride, [ 3 H]spiperone or 3 [ H]dopamine was observed in Sf9 cells infected with the D3nf (Fig. 1 and results not shown).

3.2. Binding characteristics of D3 dopamine receptor co-expressed with D3 nf in Sf9 cells Competition and saturation binding experiments were used to compare the pharmacology of the D3 dopamine receptor expressed alone or together with the D3nf in a 1:1 ratio. These experiments were performed in parallel using the same drug dilutions to minimize variation. The results summarized in Table 2 show that the relative affinities (Ki ) of dopamine, 7-OH-DPAT, sulpiride and haloperidol for D3 dopamine receptor were not significantly altered in the presence of D3nf. The agonist competition curves were best fitted to a single affinity state of the receptor. Table 1 Characterization of [ 3 H]nemonapride and [ 3 H]spiperone binding to Sf9 cell membranes expressing D3 dopamine receptor or co-expressing D3 dopamine receptor and D3nf a Antagonist

3

Fig. 1. Saturation isotherm of [ H]nemonapride binding to the Sf9 cell membranes expressing D3 dopamine receptor (m), D3 dopamine receptor co-expressed with the D3nf (j) or D3nf alone (d). Specific binding was defined in the presence of 1 mM (1)-butaclamol. Each point is the mean of duplicate determinations in a single representative experiment which was repeated at least 3 times. The data were analyzed as described in Section 2.

[ 3 H]Spiperone [ 3 H]Nemonapride a

D3 dopamine receptor

D3 dopamine receptor1D3nf

KD (pM)

Bmax (pmol / mg protein)

KD (pM)

Bmax (pmol / mg protein)

4866126 145628

1061* 1262**

5456168 104615

561* 561**

Results are expressed as the arithmetic mean of the dissociation constant (KD ) or maximal binding capacity (Bmax )6standard error of the mean from four independent trials. Statistical significance was determined by student t-tests. The Bmax was significantly lower when the D3 dopamine receptor was co-expressed with the D3nf than when expressed alone (*,**P,0.05).

68

J.L. Elmhurst et al. / Molecular Brain Research 80 (2000) 63 – 74

Table 2 Characterization of the inhibition constant (Ki ) for competition of various ligands for [ 3 H]spiperone and [ 3 H]nemonapride binding to the D3 dopamine receptor expressed alone and co-expressed with the D3nf in Sf9 cells a Radioligand

Ligands

D3 dopamine receptor Ki (nM)

D3 dopamine receptor 1D3nf Ki (nM)

[ 3 H]Spiperone

Dopamine (1)7-OH-DPAT Sulpiride Haloperidol

27.06614.41 1.7260.64 59.61611.37 14.5061.79

20.8569.48 2.0760.57 49.63614.84 14.3962.07

[ 3 H]Nemonapride

Dopamine (1)7-OH-DPAT Sulpiride Haloperidol

58.73614.57 6.9961.12 82.9766.67 54.2269.27

57.60611.13 4.2660.73 78.11614.23 54.96610.01

a

Results are expressed as the arithmetic mean of the inhibition constant (Ki )6standard error of the mean. IC 50 values were determined from three to four independent experiments. Calculation of the Ki is as described in Section 2. No significant differences for the affinity of the different ligands was found between D3 dopamine receptor expressed alone or with the D3nf (P.0.05).

The KD values of [ 3 H]spiperone or [ 3 H]nemonapride binding were not significantly altered when D3 dopamine receptor was co-expressed with D3nf (Table 1). There was, however, a 38% decrease in the apparent receptor density (Bmax ) recognized by both radioligands when the D3nf and D3 dopamine receptor were co-expressed. The increased viral load during co-expression was controlled for by co-infecting cells with D3 dopamine receptor and wildtype baculovirus at equal MOIs. The alteration in Bmax was

only observed when D3 dopamine receptor and D3nf were expressed in the same cell and not when membranes separately expressing either construct were mixed before binding analysis (data not shown). To confirm that the amount of D3 dopamine receptor immunoreactive protein expressed at the cell surface was not altered in the presence of D3nf, the membranes of cells expressing D3 dopamine receptor alone or with the D3nf were analysed by immunoblot and the relative amounts

Fig. 2. Hydropathy plots of the predicted amino acid sequences of the D3 dopamine receptor (A) and D3nf (B). Shown below are cartoons, representing the transmembrane spanning domains as predicted by hydropathy. The D3 dopamine receptor sequence contains seven hydrophobic regions, consistent with the topology of a G protein-coupled receptor. The D3nf differs in sequence from the D3 dopamine receptor after intracellular loop 3 and the divergent sequence is predicted to contain a non-homologous transmembrane domain 6.

J.L. Elmhurst et al. / Molecular Brain Research 80 (2000) 63 – 74

compared by densitometry. The amount of D3 dopamine receptor did not vary significantly between membranes expressing D3 dopamine receptor alone or co-expressing D3 dopamine receptor and D3nf (see below and Table 3). D3nf was co-expressed with another catecholamine receptor, c-myc-b2-adrenoceptor, to ensure that the effect of co-expression of D3nf on radioligand binding was specific for the D3 dopamine receptor. The apparent receptor density recognized by [ 3 H]dihydroalprenolol for membranes co-expressing D3nf and c-myc-b2-adrenoceptor was not significantly different from the receptor density detected when the c-myc-b2-adrenoceptor was expressed alone (2164 versus 1864 pmol / mg protein, respectively). Similarly, co-expression of D3nf with the homologous D2 dopamine receptor, had no effect on

69

apparent D2 dopamine receptor density (361 pmol / mg protein alone, versus 361 pmol / mg protein in the presence of D3nf).

3.3. Coupling of the D3 dopamine receptor and D3 nf in Sf9 cells Treatment of Sf9 cells with forskolin, an activator of AC, caused in increase in the levels of cAMP. We have previously shown that the D2 dopamine receptor couples to inhibition of AC in Sf9 cells [15]. To determine if the D3 dopamine receptor also couples to AC in this system, we tested the effect of dopamine on forskolin-stimulated cAMP production in cells infected with the D3 dopamine receptor. Dopamine had no significant effect on the levels of cAMP in cells infected with the D3 dopamine receptor alone or co-infected with D3nf (data not shown).

3.4. Immunoblot and photolabelling analyses of the D3 dopamine receptor and D3 nf expressed alone and together in Sf9 cells Western blots of membranes from Sf9 cells expressing the D3 dopamine receptor revealed predominant species at approximately 40, 80 and 160 kDa (Fig. 3A, lane 2). These species are consistent in size with D3 dopamine receptor monomers, dimers and tetramers since the predicted molecular mass of the D3 dopamine receptor with the c-myc epitope is |45 kDa. Similarly, membranes from cells infected with the D3nf showed major species at approxi-

Fig. 3. Detection of D3 dopamine receptor and D3nf oligomerization by immunoblotting. (A) Immunoblot probed with the anti-c-myc (9E10) antibody. Fifty mg of total cellular protein from P2 membrane fractions of Sf9 cells harvested 48 h post-infection were run per lane. Lane 1, membranes from cells infected with wild-type baculovirus. Lane 2, membranes from cells expressing D3 dopamine receptor. Lane 3, membranes from cells expressing D3 dopamine receptor and D3nf. Lane 4, membranes from cells expressing D3nf. Immunoreactive bands in lanes 2 and 3 represent D3 dopamine receptor species and are observed at |40, |80 and |160 kDa, consistent in size with receptor monomers, dimers and tetramers. (B) Immunoblot probed with the anti-FLAG (M5) antibody. Fifty mg of protein from P2 membrane fractions of Sf9 cells harvested 48 h post-infection were run per lane. Lane 1, membranes from cells infected with wild-type baculovirus. Lane 2, membranes from cells expressing D3nf. Lane 3, membranes from cells co-expressing D3 dopamine receptor and D3nf. Lane 4, membranes from cells expressing D3 dopamine receptor. Immunoreactive bands in lanes 2 and 3 represent D3nf species and are observed at |36, |72 and |144 kDa, consistent in size with protein monomers, dimers and tetramers. The results shown are representative of at least three independent experiments.

Fig. 4. Photoaffinity labelling of Sf9 cell membranes expressing c-myc D3 dopamine receptor. Autoradiogram showing photolabelling by [ 125 I]azidonemonapride of Sf9 cell P2 membranes after 4 days exposure. Lanes 1 and 2, membranes from cells expressing D3 dopamine receptor. Lanes 3 and 4, membranes from cells co-expressing D3 dopamine receptor and D3nf. The results shown are representative of at least three independent experiments. Twenty-five mg membrane protein were run per lane. Species of the same molecular weights are revealed as observed on the immunoblot (A). Lanes 1 and 3 represent total binding in the presence of vehicle and lanes 2 and 4 represent non-specific binding in the presence of 1 mM (1)-butaclamol.

J.L. Elmhurst et al. / Molecular Brain Research 80 (2000) 63 – 74

70

Table 3 Densitometric comparison of immunoreactive and photolabelled D3 dopamine receptor species in the presence and absence of D3nf a Analysis

Species

Optical density units

Change

D3 dopamine receptor

D3 dopamine receptor1D3nf

(%)

Immunoblot

Monomer Dimer Tetramer

411.46176.2 183.1640.9 23.367.3

352.96118.1 192.7681.3 25.4624.8

214 5 9

Photoaffinity label

Monomer Dimer Tetramer

549.1626.3 53.1611.1 225.2639.1

405.7661.8* 16.665.3** 121.1626.3***

226 269 246

a

Results are expressed as the arithmetic mean6standard error of the mean. Optical density units are arbitrary. Values were determined from three to four independent experiments. The density of the species observed in the presence of D3nf was significantly lowered as determined by photoaffinity labelling (*,**,***P,0.05), but not by immunoblot.

mately 36, 72 and 144 kDa, consistent with monomers, dimers and tetramers of the D3nf based on its predicted size with the FLAG epitope of 38 kDa (Fig. 3B, lane 2). Membranes co-expressing D3 dopamine receptor and D3nf showed the same protein species when probed with the c-myc or FLAG antibody as those observed in membranes expressing either construct alone (Fig. 3A, lane 3 and Fig. 3B, lane 3). Densitometry confirmed that the relative amount of D3 dopamine receptor protein was not significantly altered in the presence of D3nf (Table 3). These results suggest that the reduction of Bmax observed in the presence of D3nf is not due to a decreased amount of D3 dopamine receptor on the cell surface. Using the photoreactive ligand [ 125 I]azidonemonapride with membranes infected with D3 dopamine receptor and / or D3nf, three major species were resolved on SDS–PAGE at approximately 40, 80 and 160 kDa (Fig. 4). That binding to all three species was displaced in the presence of excess (1)-butaclamol confirms that they represented D3 dopamine receptor proteins. Furthermore, the protein species revealed via photoaffinity labelling were identical in size to those observed on the immunoblot and implies that these species represent functional D3 dopamine receptor monomers, dimers and tetramers. Notable is the lesser intensity of photolabelling in membranes co-expressing D3 dopamine receptor and D3nf (Fig. 4, lane 2) which corresponds to the decrease in binding capacity documented by the saturation binding studies (Fig. 1). In all cases, immunoblotting was performed in parallel to ensure that the decreased labelling of D3 dopamine receptor species did not correspond to a decrease in immunoreactive D3 dopamine receptor protein. Densitometry showed that the binding capacity of the oligomeric species of D3 dopamine receptor were most affected by co-expression of D3nf (Table 3). Particularly sensitive was the dimeric species, in which binding was reduced by 69%, and the tetramer, in which binding was reduced by 46%. The monomer species lost approximately 26% of binding compared to when the D3 dopamine receptor was ex-

pressed alone. Membranes from Sf9 cells expressing D3nf alone showed no specific binding of [ 125 I]azidonemonapride, consistent with the D3nf’s inability to bind dopaminergic ligands.

3.5. Co-immunoprecipitation of D3 dopamine receptor and D3 nf A co-immunoprecipitation strategy was used to determine if a physical interaction between the D3 dopamine receptor and D3nf resulted in heterodimer formation. Immunoprecipitating D3 dopamine receptor using c-myc antibody and probing with c-myc antibody on Western blot, led to the detection of protein which corresponded to receptor monomers and oligomers (Fig. 5, lane 2). C-myc antibody was also able to detect D3 dopamine receptor momomers and oligomers in membranes containing protein immunoprecipitated with the FLAG antibody (Fig. 5, lane 3), suggesting that D3 dopamine receptor co-immunoprecipitated with the D3nf. Similarly, D3nf protein could also immunodetected in protein which had been immunoprecipitated by c-myc antibodies.

3.6. Confocal microscopy Immunocytochemical studies were performed to compare the expression and subcellular distribution of the D3 dopamine receptor and D3nf in Sf9 cells. When expressed individually, both the D3 dopamine receptor and D3nf were found abundantly on the surface of Sf9 cells, as evidenced by the ring of fluorescent staining seen in whole cell preparations (Fig. 6A,C,E). When expressed together in the same cell, there was no alteration in the overall pattern of surface staining of the cells. In permeabilized cells, both proteins could also be seen in the intracellular regions (Fig. 6B,D,F). No fluorescent labelling of cells was observed in Sf9 cells which were infected with wild-type baculovirus.

J.L. Elmhurst et al. / Molecular Brain Research 80 (2000) 63 – 74

Fig. 5. Co-immunoprecipitation of D3 dopamine receptor and D3nf. Cells co-expressing D3 dopamine receptor and D3nf were immunoprecipitated (IP) with the c-myc 9E10 or FLAG antibody and run on SDS–PAGE and described in Section 2. The blot was probed with the c-myc 9E10 antibody (IB). Lane 1, control membranes which were not immunoprecipitated. Lane 2, membranes immunoprecipitated with the c-myc antibody. Lane 3, membranes immunoprecipitated with the FLAG antibody. In all lanes as shown, dominant species are observed at |80 kDa, which corresponds in size to a D3 dopamine receptor dimer. The fact that D3 dopamine receptor protein, as detected by the c-myc antibody, is observed in membranes immunoprecipitated with either the c-myc or FLAG antibody suggests that the D3 dopamine receptor and D3nf physically interact.

4. Discussion Our results show that co-expression of D3 dopamine receptor and D3nf led to a significant reduction in the number of D3 dopamine receptor binding sites compared to when the D3 dopamine receptor was expressed alone, as assessed by conventional saturation binding assays and by photoaffinity labelling. The decrease in D3 dopamine receptor binding in the presence of D3nf did not result from a decreased amount of D3 dopamine receptor immunoreactive protein at the cell surface, as confirmed by immunoblotting and densitometry. By photoaffinity labelling of the membranes and resolving the D3 dopamine receptor species on SDS–PAGE, we were able to determine that ligand binding to D3 dopamine receptor dimers and tetramers was most inhibited by co-expression of D3nf. The co-immunoprecipitation of D3 dopamine receptor with D3nf shows that they physically interact to form heterodimers or heterooligomers. The effect of D3nf was specific to the D3 dopamine receptor as co-expression of D3nf had no effect on D2 dopamine receptor or b 2 adrenoceptor binding densities. The D3 receptor expressed

71

in Sf9 cells was detected by dopaminergic agonists as a single affinity site (Table 2), and agonists had no effect on forskolin-stimulated adenylyl cyclase activity, indicating the likely absence of the appropriate G protein(s) necessary for D3 receptor coupling, in contrast to the robust responses with dopamine D1 [14] or D2 [15] receptors in this cell line. The molecular sizes of the protein species detected by immunoblot analysis are consistent in size with D3 dopamine receptor monomers, dimers and oligomers (Fig. 3). Photoaffinity labelling was able to confirm, for the first time, that these species consisted of functional D3 dopamine receptor protein capable of binding ligand (Fig. 4). The relative proportion of monomer:dimer:tetramer detected by the photolabel was not the same as observed on the immunoblot, suggesting that the ligand binding capability of each species was not equivalent (Table 3). The tetramer species bound more ligand than the dimer and the dimer bound less ligand than monomer. Even though only a small amount of D3 dopamine receptor protein exists as a tetramer on SDS–PAGE, it bound a relatively large amount of ligand. The tetrameric form of the D3 dopamine receptor is reported to be the predominant form in human brain [16] so these observations may suggest that oligomers play an important role in D3 dopamine receptor pharmacology and function. The decreased photolabelling of D3 dopamine receptor species in the presence of D3nf is consistent with the saturation binding data and showed that the greatest reduction in binding was to the oligomeric dimers and tetramers. Co-immunoprecipitation of D3 dopamine receptor protein with D3nf confirmed that the D3 dopamine receptor and D3nf physically interact to form heterodimers (Fig. 5). The decreased ligand binding to dimers and tetramers observed by photolabelling suggests that heterooligomers of the D3 dopamine receptor and D3nf are unable to bind ligand and, therefore, non-functional. Interestingly, binding of photoligand to the monomeric species was also decreased in the presence of D3nf (Fig. 4). This observation supports the idea that D3 dopamine receptor species at the cell surface may exist in an oligomeric array which is disrupted into monomeric, dimeric and tetrameric species when separated on SDS– PAGE. The resistance of dimers and tetramers to SDS has been noted before with many other G protein-coupled receptors [5,6,10,16]. It may be speculated that D3nf, through its homology to the D3 dopamine receptor, integrates into the D3 dopamine receptor array and interferes with the binding of ligand to the D3 dopamine receptor. The decreased binding to the D3 dopamine receptor would then be represented in all former members of the D3 dopamine receptor array: monomers, dimers and tetramers. The importance of a biological process regulating the interaction between D3 dopamine receptor and D3nf is evidenced by the fact that mixing of cell lysates

72

J.L. Elmhurst et al. / Molecular Brain Research 80 (2000) 63 – 74

J.L. Elmhurst et al. / Molecular Brain Research 80 (2000) 63 – 74

73

Fig. 6. Immunofluorescence staining of Sf9 cells. Cells expressing D3 dopamine receptor alone (A,B) were probed with the monoclonal c-myc 9E10 antibody and those expressing D3nf alone (C,D) were detected with the polyclonal FLAG M2 antibody. Fluorescence images were obtained with a confocal microscope, using a fluorescein isothiocyanate-linked secondary antibody to detect the 9E10 antibody (green) or a Texas Red-linked secondary antibody to detect the M2 antibody (red). Overlapping signals from cells coexpressing D3 and D3nf (E,F) appear yellow. Images were obtained from non-permeablized (A,C,E) and permeabilized (B,D,F) cells. The horizontal bar represents 10 mm. Each picture is representative of three independent experiments and were obtained under identical illuminating conditions.

containing D3 dopamine receptor and D3nf separately expressed prior to binding analysis does not have any affect on D3 dopamine receptor binding. Confocal microscopy showed that both the D3 dopamine receptor and D3nf were well expressed at the cell surface, implying that each protein was properly inserted into the membrane and recognized for trafficking from the endoplasmic reticulum (Fig. 6). As well, immunoblot analysis of cell surface membranes showed that the amount of D3 dopamine receptor and D3nf at the surface was not significantly altered in the presence of the other (Fig. 3 and Table 3). It has been noted in the past that truncated forms of receptors cause retention of their full-length counterparts in the endoplasmic reticulum, resulting in an attenuation of the function of the full-length, wild-type receptor [1,9,25]. However, other truncated forms of G proteincoupled receptors have been shown to be trafficked to the cell surface including artificially truncated forms of the m3 muscarinic receptor [21] and, in this study, the D3nf. It has been proposed that D3nf functions to promote D3 dopamine receptor oligomer formation in GH3 cells since D3 dopamine receptor oligomers were only observed in the presence of D3nf [16]; however, we consistently observed D3 dopamine receptor and D3nf homodimers and oligomers when the proteins were expressed alone. No heterodimers of D3 and D3nf were ever visualized (Fig. 3), but hetero-oligomerization occurred, as indicated by the coimmunoprecipitation studies (Fig. 5). In this report, we have shown that the D3nf, a physiologically expressed receptor splice variant which does not have any D3 dopamine receptor-like pharmacology on its own, may function to decrease the capacity of D3 dopamine receptor on the cell surface to bind ligand. D3nf co-immunoprecipitated with D3 dopamine receptor and was most effective at inhibiting ligand binding to oligomeric forms of the D3 dopamine receptor, suggesting that direct protein–protein interactions were responsible for D3nf’s modulatory effect. The effect of D3nf on D3 dopamine receptor was specific as it did not affect ligand binding to the b 2 -adrenergic receptor or the homologous D2 dopamine receptor. D3nf did not affect expression of the D3 dopamine receptor and was itself well expressed at the cell surface. It may be speculated that differences in the cellular ratio of D3 dopamine receptor:D3nf protein in vivo may modulate the activity of D3 dopamine receptor expressing neurons. This idea is supported by the fact that there are differences in the distribution of D3 dopamine receptor and D3nf along a single neuron [16]. Hence, a variable ratio of D3 dopamine receptor:D3nf along a

neuron may modulate the capacity of the D3 dopamine receptor to bind dopamine in vivo and may impart differing sensitivity of the neuron to D3 dopamine receptor activation.

Acknowledgements We are grateful to Dr. Claudia Schmauss for her kind gift of the cDNA encoding the D3 dopamine receptor. This research was supported by grants from the Medical Research Council of Canada, NIH-National Institute on Drug Abuse, the Smokeless Tobacco Research Council Inc., and a Natural Sciences and Engineering Research Council of Canada Scholarship awarded to J.L.E.

References [1] M. Benkirane, D.-Y. Jin, R.F. Chun, R.A. Koup, K.-T. Jeang, Mechanism of transdominant inhibition of CCR5-mediated HIV-1 infection by ccr5D32, J. Biol. Chem. 272 (1997) 30603–30606. [2] N.J. Colley, J.A. Cassill, E.K. Baker, C.S. Zuker, Defective intracellular transport is the molecular basis of rhodopsin-dependent dominant retinal degeneration, Proc. Natl. Acad. Sci. USA 92 (1995) 3070–3074. [3] C.S. Fishburn, D. Belleli, C. David, S. Carmon, S. Fuchs, A novel short isoform of the D3 dopamine receptor generated by alternative splicing in the third cytoplasmic loop, J. Biol. Chem. 268 (1993) 5872–5878. [4] S.B. Freedman, S. Patel, R. Marwood, F. Emms, G.R. Seabrook, M.R. Knowles, G. McAllister, Expression and pharmacological characterization of the human D3 dopamine receptor, J. Pharmacol. Exp. Ther. 268 (1994) 417–426. [5] Y. Fukushima, T. Asano, T. Saitoh, M. Anai, M. Funaki, T. Ogihara, H. Katagiri, N. Matsuhashi, Y. Yazaki, K. Sugano, Oligomer formation of histamine H2 receptors expressed in Sf9 and COS7 cells, FEBS Lett. 409 (1997) 283–286. [6] S.R. George, S.P. Lee, G. Varghese, P.R. Zeman, P. Seeman, G.Y.K. Ng, B.F. O’Dowd, A transmembrane domain-derived peptide inhibits D1 dopamine receptor function without affecting receptor oligomerization, J. Biol. Chem. 273 (1998) 30244–30248. [7] B. Giros, M.-P. Martres, C. Pilon, P. Sokoloff, J.-C. Schwartz, Shorter variants of the D3 dopamine receptor produced through various patterns of alternative splicing, Biochem. Biophys. Res. Commun. 176 (1991) 1584–1592. [8] N. Griffon, M.-A. Crocq, C. Pilon, M.P. Martres, A. Mayerova, G. Uyanik, E. Burgert, F. Duval, J.P. Macher, F. Javoy-Agid, C.A. Tamminga, J.-C. Schwartz, P. Sokoloff, Dopamine D3 receptor gene: organization, transcript variants and polymorphism associated with schizophrenia, Am. J. Med. Gen. 67 (1996) 63–70. [9] R. Grosse, T. Schoneberg, G. Schultz, T. Gudermann, Inhibition of gonadotropin-releasing hormone receptor signaling by expression of

74

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

J.L. Elmhurst et al. / Molecular Brain Research 80 (2000) 63 – 74 a splice variant of the human receptor, Mol. Endocrinol. 11 (1997) 1305–1318. T.E. Hebert, S. Moffett, J.-P. Morello, T.P. Loisel, D.G. Bichet, C. Barret, M. Bouvier, A peptide derived from a b2-adrenergic receptor transmembrane domain inhibits both receptor dimerization and activation, J. Biol. Chem. 271 (1996) 16384–16392. K. Liu, C. Bergson, R. Levenson, C. Schmauss, On the origin of mRNA encoding the truncated dopamine D3-type receptor D3nf and detection of D3nf-like immunoreactivity in human brain, J. Biol. Chem. 269 (1994) 29220–29226. R.G. MacKenzie, D. VanLeeuwen, T.A. Pugsley, Y.H. Shih, S. Demattos, L. Tang, R.D. Todd, K.L. O’Malley, Characterization of the human dopamine D3 receptor expressed in transfected cell lines, Eur. J. Pharmacol. 266 (1994) 79–85. Y. Nagai, S. Ueno, Y. Saeki, F. Soga, T. Yanagihara, Expression of the D3 dopamine receptor gene and a novel variant transcript generated by alternative splicing in human peripheral blood lymphocytes, Biochem. Biophys. Res. Commun. 194 (1993) 368–374. G.Y.K. Ng, B. Mouillac, S.R. George, M. Caron, M. Dennis, M. Bouvier, B.F. O’Dowd, Desensitization, phosphorylation and palmitoylation of the human dopamine D1 receptor, Eur. J. Pharmacol. 267 (1994) 7–19. G.Y.K. Ng, G. Varghese, H.T. Chung, J. Trogadis, P. Seeman, B.F. O’Dowd, S.R. George, Resistance of the dopamine D2L receptor to desensitization accompanies the up-regulation of receptors on to the surface of Sf9 cells, Endocrinology 138 (1997) 4199–4206. E.A. Nimchinsky, P.R. Hof, W.G.M. Janssen, J.H. Morrison, C. Schmauss, Expression of dopamine D3 receptor dimers and tetramers in brain and in transfected cells, J. Biol. Chem. 272 (1997) 29229–29237. S. Pagliusi, A. Chollet-Daemerius, C. Losberger, A. Mills, E.

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

Kawashima, Characterization of a novel exon within the D3 receptor gene giving rise to an mRNA isoform expressed in rat brain, Biochem. Biophys. Res. Commun. 194 (1993) 465–471. R.J.M. Ross, N. Esposito, X.Y. Shen, S. Von Laue, S.L. Chew, P.R.M. Dopson, M.-C. Postel-Vinay, J. Finidori, A short isoform of the human growth hormone receptor functions as a dominant negative inhibitor of the full-length receptor and generates large amounts of binding protein, Mol. Endocrinol. 11 (1997) 265–273. C. Schmauss, Enhanced cleavage of an atypical intron of dopamine D3-receptor pre-mRNA in chronic schizophrenia, J. Neurosci. 16 (1996) 7902–7909. C. Schmauss, V. Haroutunian, K.L. Davis, M. Davidson, Selective loss of dopamine D3-type receptor mRNA expression in parietal and motor cortices of patients with chronic schizophrenia, Proc. Natl. Acad. Sci. USA 90 (1993) 8942–8946. T. Schoneberg, J. Liu, J. Wess, Plasma membrane localization and functional rescue of truncated forms of a G protein-coupled receptor, J. Biol. Chem. 270 (1995) 18000–18006. L.A. Snyder, J.L. Roberts, S.C. Sealfon, Alternative transcripts of the rat and human dopamine D3 receptor, Biochem. Biophys. Res. Commun. 180 (1991) 1031–1035. P. Sokoloff, M. Andrieux, R. Besancon, C. Pilon, M.-P. Martres, B. Giros, J.-C. Schwartz, Pharmacology of human dopamine D3 receptor expressed in a mammalian cell line: comparison with D2 receptor, Eur. J. Pharmacol. 225 (1992) 331–337. P. Sokoloff, M.-P. Martres, B. Giros, M.-L. Bouthenet, J.-C. Schwartz, The third dopamine receptor (D3) as a novel target for antipsychotics, Biochem. Pharmacol. 43 (1992) 659–666. X. Zhu, J. Wess, Truncated V2 vasopressin receptors as negative regulators of wild-type V2 receptor function, Biochemistry 37 (1998) 15773–15784.