Periplakin interferes with G protein activation by the melanin-concentrating hormone receptor-1 by binding to the proximal segment of the receptor C-terminal tail.

In mice genetic ablation of expression of either melanin-concentrating hormone or the melanin-concentrating hormone-1 receptor results in alterations in energy metabolism and a lean phenotype. There is thus great interest in the function and regulation of this receptor. Using the yeast two-hybrid system we identified an interaction of the actin- and intermediate filament-binding protein periplakin with the intracellular C-terminal tail of the melanin-concentrating hormone-1 receptor. Direct association of these proteins was verified in pull-down and coimmunoprecipitation experiments. Truncations and internal deletions delineated the site of interaction to a group of 11 amino acids proximal to transmembrane helix VII, which was distinct from the binding site for the melanin-concentrating hormone-1 receptor-interacting zinc finger protein. Immunohistochemistry demonstrated coexpression of periplakin with melanin-concentrating hormone-1 receptor in specific cells of the piriform cortex, amygdala, and other structures of the adult mouse brain. Coexpression of the melanin-concentrating hormone-1 receptor with periplakin in human embryonic kidney 293 cells did not prevent agonist-mediated internalization of the receptor but did interfere with binding of (35)S-labeled guanosine 5'-3-O-(thio)triphosphate ([(35)S]GTPgammaS) to the G protein Galpha(o1) and the elevation of [Ca(2+)](i). Coexpression of the receptor with the interacting zinc finger protein did not modulate receptor internalization or G protein activation. The interaction of periplakin with receptors was selective. Coexpression of periplakin with the IP prostanoid receptor did not result in coimmunoprecipitation nor interfere with agonist-mediated binding of [(35)S]GTPgammaS to the G protein Galpha(s). Periplakin is the first protein described to modify the capacity of the melanin-concentrating hormone-1 receptor to initiate signal transduction.

Even before identification of the MCH receptors a clear role for MCH in feeding and energy balance and homeostasis had been established (1). A solid body of genetic and pharmacological evidence now supports a role for MCH and the MCH-1 receptor in the modulation of food intake and energy expenditure. Key data derive from the genetic ablation of both MCH (13) and the MCH-1 receptor (14,15) in mice. These alterations produced animals that were both lean and resistant to dietinduced obesity. Mice lacking the MCH-1 receptor have been reported to be hyperphagic (14,15) but to be both hyperactive and to have altered metabolism, and this may explain the apparent paradox of being resistant to obesity but eating more. Studies also indicate that acute and chronic administration of MCH enhances food intake and body weight (16) and that overexpression of MCH in transgenic mice leads to obesity and insulin resistance (17). A role for the MCH-2 receptor in these effects is unlikely because genetic evidence indicates that mice do not express a functional form of this receptor (18). Furthermore, the distribution of the MCH-1 receptor in the central nervous system (19) also suggests that antagonists at the MCH-1 receptor might provide a treatment for obesity (for review, see Refs. 20 and 21). High throughput screening efforts have led to the identification of small molecule MCH-1 receptor antagonists with diverse structural features and drug-like properties. In vivo results with two of these antagonists indicate efficacy in several animal models of body weight regulation and feeding behavior (21). At least when expressed in heterologous cell lines, the MCH-1 receptor is able to regulate signal transduction through both G i -and G q -coupled pathways (2)(3)(4)(5)(6). Despite this, little is known about the regulation of this receptor.
GPCRs generally do not exist in isolation, and there is increasing information on proteins that interact with GPCRs and by so doing alter their cellular distribution or function (22)(23)(24)(25). The intracellular C-terminal tail of GPCRs has been particularly well studied in this regard and interacting proteins identified by the application of both proteomic (26,27) and yeast two-hybrid (28,29) techniques. We have previously used the MCH-1 receptor C-terminal tail as the bait in yeast two-hybrid screens to identify interactions with a zinc finger protein named MCH-1 receptor-interacting zinc finger protein (MIZIP) (30). Using the same approach we now report the interactions of the MCH-1 receptor with the actin and intermediate filament-binding protein, periplakin (PPL) (31). PPL is a 1,756amino acid polypeptide that is expressed widely and at considerable levels in both rodent and human brain (32,33). Although most actively studied for its role in production of the cornified epithelium in keratinocytes (34 -37), we have previously detailed its capacity to interact with splice variants of the human MOP opioid receptor (32). We now map the site of interaction of PPL with the MCH-1 receptor and show that this is distinct from the interactions of the MCH-1 receptor with MIZIP. Interactions between PPL and the MCH-1 receptor are selective but do not modulate MCH-1-mediated internalization of the receptor. This interaction does, however, reduce the capacity of the receptor to active G proteins and hence initiate signal transduction. This is the first protein-protein interaction demonstrated to alter agonist-mediated function of the MCH-1 receptor and, as we also demonstrate coexpression of these proteins in specific neurones, is likely to have significance for MCH-mediated functions in vivo.

EXPERIMENTAL PROCEDURES
Materials-All materials for tissue culture were supplied by Sigma. Oligonucleotides were purchased from ThermoHybaid (Ulm, Germany). Mouse monoclonal anti-vesicular stomatitis virus glycoprotein (VSV-G) antibody was from Roche Applied Science. Rabbit polyclonal anti-His antiserum was from Santa Cruz Biotechnology. The sheep anti-GFP antiserum was raised in-house against recombinantly expressed eGFP. The polyclonal sheep anti-PPL antiserum was generated against the C-terminal 208 amino acids of PPL. The mouse monoclonal anti-Myc tag antibody was from Sigma. Both a rabbit polyclonal anti-MCH-1 receptor antiserum (Alpha-diagnostic) and an affinity-purified chicken anti-MCH-1 receptor antibody (Chemikon) were used in various studies.
Production of VSV-G-tagged forms of the hMCH-1 receptor constructs was performed by PCR with a 5Ј-oligonucleotide primer introducing the VSV-G epitope (YTDIEMNRLGK) adjacent to the sequence of codon 2. For the C-tail truncated mutants, 3Ј-oligonucleotide primers introducing stop codons after residues 314, 327, and 340 were used in the reactions. PCR fragments were subcloned into PcDNA3.1(ϩ). Equivalent eYFP-tagged versions of the hMCH-1 receptor constructs were made by insertion of the PCR-amplified cDNA fragments (minus stop codons) into plasmid eYFP-Ni (Clontech).
Yeast Two-hybrid Screen-Transformation of the yeast reporter strain CG1945, screening of a human brain cDNA library in the yeast vector pACTII, and the ␤-galactosidase filter lift assay were performed as described previously (30).
Cell Culture and Transient Transfection-HEK293 cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% (v/v) newborn calf serum and 2 mM glutamine. Prior to transfection with plasmid DNA, cells were seeded onto 10-cm 2 or 6-cm 2 dishes at 50 -70% confluence. Cells were transfected using Lipofectamine reagent (Invitrogen). 48 h later cells were harvested by removal of the growth medium, rinsing twice with ice-cold PBS and scraping in 5 ml of PBS. Cells were pelleted by centrifugation at 2,000 rpm at 4°C and then stored at Ϫ80°C until membrane preparation.
Preparation of Membranes-Cell pellets were resuspended in TE buffer (1 mM Tris-HCl, 0.1 mM EDTA, pH 7.5) and ruptured with 50 strokes of a Teflon-on-glass homogenizer. Unbroken cells and nuclei were removed by centrifugation at 1,500 rpm for 5 min. The supernatant fraction was passed through a 25-gauge syringe needle 20 times and then centrifuged at 50,000 rpm for 30 min. The pellets were resuspended in TE buffer and stored at Ϫ80°C until use. [Ca 2ϩ ] i and ERK MAP Kinase Assays-After expression in HEK293 cells of the hMCH-1 receptor in the absence or presence of PPL, the capacity of MCH to elevate [Ca 2ϩ ] i was monitored in single cells as described by Liu et al. (40). Total and phosphorylated/activated forms of the ERK1 and ERK2 MAP kinases in such transfected HEK293 cells were detected by immunoblotting lysates of cells treated for 15 minutes with or without 1 M MCH with nonselective and phosphospecific antibodies. Cells were washed in situ twice with ice-cold PBS and lysed in radioimmune precipitation assay buffer for 1 h at 4°C. After centrifugation (15 min, 14,000 ϫ g at 4°C) to remove cell debris, 10 g of protein/sample was boiled in 2ϫ Laemmli buffer and resolved by 4 -12% BisTris NuPAGE gels and transferred onto nitrocellulose membranes. Total and phosphorylated/activated forms of the ERK1 and ERK2 MAP kinases in these samples were detected by immunoblotting using antiphospho-ERK and anti-ERK antisera (Cell Signaling) followed by detection with horseradish peroxidase-conjugated secondary antibodies and visualized by ECL (Pierce).
Coimmunoprecipitation of His-MIZIP and PPLC-GFP with VSV-G-hMCH-1 Receptor-24 h post-transfection cells from 6-cm 2 dishes were split into 6-well plates. The following day the cells were incubated in the presence or absence of agonist (1 M MCH for VSV-G-hMCH-1 receptortransfected cells; 1 M iloprost for equivalent experiments with the VSV-G-IP prostanoid receptor) for 30 min. Cells were then washed three times with ice-cold PBS and lysed in radioimmune precipitation assay buffer (50 mM HEPES, pH 7.5, 150 mM NaCl, 1% Triton X-100, 0.5% deoxycholic acid, 0.1% SDS, 5 mM EDTA, 10 mM NaF, 5% ethylene glycol, and a mixture of protease inhibitors). Insoluble material was removed by a 15-min centrifugation at 14,000 rpm at 4°C. Equalized amounts of protein were incubated with protein G-Sepharose beads conjugated to mouse monoclonal anti-VSV-G antibody (Roche Applied Science) or sheep polyclonal anti-GFP antiserum for 2 h at 4°C. The immune complexes were eluted from the beads by the addition of 2ϫ Laemmli buffer. Proteins were resolved by SDS-PAGE using 4 -12% BisTris NuPAGE gels (Invitrogen) and transferred to nitrocellulose membranes. A rabbit polyclonal anti-His antiserum (Santa Cruz Biotechnology) was used to detect His-MIZIP. VSV-G-hMCH-1 receptors were detected using the anti-VSV-G antibody.
Receptor Internalization Assay-Biotin Labeling-HEK293 cells transfected with the eYFP-tagged hMCH-1 receptors were split into 6-well plates and cultured for a further 24 h. Cells were incubated with or without 1 M MCH for 60 min and then immediately washed twice with ice-cold PBS. The alcohol groups on the cell surface glycoproteins were oxidized to aldehydes by a 30-min incubation at 4°C with 10 mM sodium m-periodate in PBS containing 1 mM MgCl 2 , 0.1 mM CaCl 2 . After further washing with PBS, the cells were incubated for 30 min with acetate buffer (0.1 M sodium acetate, pH 5.5, MgCl 2 , 0.1 mM CaCl 2 ) containing 1 mM Biotin-LC-Hydrazide (Pierce) that reacts with the newly formed aldehyde groups thereby labeling cell surface glycoproteins with biotin. Labeling was terminated by the removal of the biotin solution and washing with PBS. Cells were then solubilized for receptor immunoprecipitation with the sheep polyclonal anti-GFP antiserum as described earlier. Biotin-labeled receptors were detected using horseradish peroxide-conjugated streptavidin (Pierce) and visualized by ECL.
Purification of His-tagged Proteins-PQE30 plasmids containing the His fusion inserts of PPLC or MIZIP were transformed into Escherichia coli BL21 cells. A 10-ml starter culture was used to inoculate 400 ml of LB medium containing 100 g/ml ampicillin and grown at 37°C until the culture reached an A 600 of 0.5. Expression of the His-tagged proteins was induced by the addition of 1 mM isopropyl-␤,D-thiogalactopyranoside for 4 h before cells were harvested by centrifugation at 10,000 rpm for 15 min at 4°C. Pellets were resuspended in denaturing lysis buffer containing 6 M urea, and the His-tagged proteins were purified following the manufacturer's instructions (Qiagen). Eluted proteins were dialyzed against three changes of PBS containing 5% glycerol at 4°C over 2 days before storage at Ϫ80°C.
Purification of GST Fusion Proteins and GST Pull-down Assays-BL21 cells transformed with the GST fusion constructs were cultured and harvested as described earlier. Pellets were lysed in 20 ml of STE buffer (10 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1 mM EDTA) containing 0.1 mg/ml lysozyme and incubated with rotation at 4°C for 1 h. Dithiothreitol to a final concentration of 5 mM and 10% sarkosyl in STE buffer to a final concentration of 1.5% were added to lysates before sonication for 1 min. The removal of insoluble material was performed by a 15-min centrifugation at 10,000 rpm at 4°C. 10% Triton X-100 in STE buffer to a final concentration of 2% and a mixture of protease inhibitors were added to the cleared supernatants before use in GST pull-down assays.
For pull-down assays 0.5-5 ml of lysates was incubated with a 100-l slurry of 50% (v/v) PBS-washed glutathione-Sepharose beads (Amersham Biosciences) for 2 h at 4°C. The beads were pelleted by centrifugation (14,000 rpm, 1 min) and washed twice with PBS containing 1% Triton X-100 before the addition of 50 g of the required His-tagged protein in a 1-ml solution of PBS and Triton X-100 containing 5 mM dithiothreitol. After a 2-h incubation at 4°C, the beads were collected and washed three times with PBS and Triton X-100 solution and eluted in 100 l of 2ϫ Laemmli buffer. Eluates were resolved by SDS-PAGE, and the His fusion proteins were detected by immunoblotting.
Confocal Immunofluorescence Staining-Cells were transfected on coverslips in 6-well plates and the following day were fixed in 4% (v/v) paraformaldehyde in PBS containing 5% sucrose for 10 min at room temperature. Cells were then permeabilized for 10 min in TM buffer (0.15% Triton X-100 and 3% nonfat milk in PBS). Coverslips were incubated for 1 h at room temperature with either 2.5 g/ml anti-HA antibody (Roche Applied Science) to detect HA-PPL or 2 g/ml anti-His antibody for His-MIZIP staining. After washing with TM buffer and PBS, coverslips were incubated for a further 1 h with an Alexa 594conjugated complementary secondary antibody. After further washing with PBS, coverslips were mounted onto glass slides and observed using a laser-scanning Zeiss LSM510 confocal microscope.
Fluorescence Immunohistochemistry-All incubations were performed at room temperature or as indicated. Free floating 50-m cryostat sections from 4% paraformaldehyde-fixed adult C57BL6 mouse brains were permeabilized for 10 min with 0.3% Triton X-100 in PBS, washed with PBS-T (0,05% Triton X-100), postfixed for 30 min with methanol-acetone (1:1) at 4°C, and incubated in blocking solution (5% horse serum in PBS-T) for 1 h. Sections were incubated for 24 h at 4°C with blocking solution complemented with an affinity-purified chicken anti-MCH-1 receptor antibody (Chemikon) and the polyclonal sheep anti-PPL antiserum (both diluted 1:1000). After washing with PBS-T, sections were incubated for 2 h with Cy2-conjugated donkey antichicken (1:800, Dianova) and Cy3-conjugated donkey anti-sheep (1:400, Dianova) antisera, washed with PBS-T, and mounted in glycerol gelatin (Sigma). Images were captured with a CCD camera (Hamamatsu) mounted on a Leica Aristoplan fluorescence microscope and analyzed with the Openlab software (Improvision). Brain regions were identified using the mouse brain atlas of Paxinos and Franklin (41) and the rat brain atlas of Paxinos and Watson (42).
RNA Expression Analysis-RNA in situ hybridization was performed as described previously (43). In brief, rat postnatal and adult brains were fixed overnight with 4% paraformaldehyde in PBS and sectioned using a cryomicrotome. Radiolabeled antisense and sense RNA probes were generated by in vitro transcription using [␣-35 S]UTP and linearized cDNAs with nucleotides 1114 -1705 of human PPL (cloned from yeast clone L16) and with the full-length open reading frame of rMCH-1 receptor, respectively. Sections were hybridized overnight at 54°C in a humid chamber, washed with decreasing salt concentrations, dehydrated using increasing ethanol concentrations, and exposed to Beta-Max x-ray film (Amersham Biosciences) for 3-7 days.

RESULTS
A human brain cDNA library was employed in yeast twohybrid screens using the sequence corresponding to amino acids 299 -353 of the rat MCH-1 receptor as bait. This consists of the 5 amino acids upstream of the highly conserved NPXXY motif of transmembrane region VII and the entire C-terminal tail. From a total of 2.7 ϫ 10 6 transformed cDNAs, 20 were positive after selection, two of which encoded C-terminal fragments of human PPL (L16, amino acids 1114 -1705; L25Ј, amino acids 1415-1705) (Fig. 1a). The interaction of the C terminus of the rMCH-1 receptor in pAS2 with the yeast clones L16 and L25Ј was specific because no growth of yeast was detected using either pAS2 control or the C-terminal regions of hMCH-2 receptor or the rat somatostatin receptor 5 cloned into pAS2 (Fig. 1b). Using deletion mutants of the C terminus of rMCH-1 receptor, we confined the binding of PPL to the prox-FIG. 1. Identification of the interaction of the C terminus of human PPL with the proximal C terminus of the MCH-1 receptor. Using yeast two-hybrid analyses, two clones from a human brain cDNA library, L16 and L25Ј, were identified to interact with the C terminus of the rMCH-1 receptor. Both clones encoded the C-terminal region of PPL, the domain structure of which is shown on top (a). In b, the growth of yeast on selection medium after transfection with different cDNAs, cloned in the yeast bait vector pAS2, and the yeast fish vector pACTII, is shown.
imal C-terminal region of the MCH-1 receptor: growth of yeast on selection medium was not affected by the deletion of the last 28 amino acids of rMCH-1 receptor (MCH-1 (299 -325) ). However, the deletion of the proximal C terminus (MCH-1 (319 -353) ) abolished binding because no growth of yeast on selection medium was detected (Fig. 1b).
To extend these studies an N-terminally VSV-G epitope-tagged form of the full-length hMCH-1 receptor was expressed in HEK293 cells along with PPLC tagged at the C terminus with GFP. Immunoprecipitation of PPLC-GFP with an anti-GFP antiserum resulted in coimmunoprecipitation of the 32-kDa VSV-G-hMCH-1 receptor as monitored by immunoblotting such samples with the anti-VSV-G antibody following SDS-PAGE (Fig. 5a). Expression of PPLC-GFP without VSV-Gtagged hMCH-1 receptor did not result in coimmunoprecipitation of the 32-kDa polypeptide (Fig. 5a). This interaction was selective. Coexpression of PPLC-GFP with an N-terminally VSV-G epitope-tagged form of the full-length IP prostanoid receptor followed by immunoprecipitation of PPLC-GFP did not result in coimmunoprecipitation of this VSV-G-tagged receptor (Fig. 5a). Interaction of PPLC-GFP with the VSV-G-tagged IP prostanoid receptor was not promoted by treatment of the cells with the selective IP prostanoid receptor agonist iloprost (1 M, 30 min) (Fig. 5a). Similar results were obtained with coexpression of His-MIZIP and the VSV-G-tagged forms of the hMCH-1 and IP prostanoid receptors. Immunoprecipitation of VSV-G-hMCH-1 with the anti-VSV-G antiserum resulted in coimmunoprecipitation of His-MIZIP (Fig. 5b), but coexpression of His-MIZIP with the VSV-G-IP prostanoid receptor followed by immunoprecipitation of the receptor did not result in coimmunoprecipitation of His-MIZIP (Fig. 5b). Again, addition of the IP prostanoid receptor agonist iloprost did not promote interactions between this receptor and His-MIZIP.
To explore the functional consequence of interactions of the hMCH-1 receptor with either PPL or MIZIP, the hMCH-1 receptor C-terminally tagged with eYFP was expressed with or without either N-terminally HA-tagged full-length PPL or His-MIZIP in HEK293 cells. Confirmation of successful expression of PPL and MIZIP was obtained by immunoblotting cell lysates to detect these polypeptides (Fig. 7a). Cell surface expression of the hMCH-1 receptor-eYFP was detected in cell surface biotin- ylation experiments (Fig. 7a). Treatment of the cells with 1 M MCH for 60 min resulted in internalization of a proportion of the receptor as monitored by the reduction of cell surface receptors available to be biotinylated (Fig. 7a). Coexpression of neither PPL nor MIZIP prevented agonist-mediated receptor internalization (Fig. 7a). By contrast, although an eYFP-tagged form of hMCH-1 receptor 340Stop was internalized in response to the addition of MCH, further deletion to produce eYFPtagged forms of hMCH-1 327Stop and hMCH-1 314Stop eliminated agonist-mediated internalization (Fig. 7b). Cell surface localization of the eYFP-tagged full-length hMCH-1 receptor and its internalization in response to treatment of the cells with MCH was confirmed further in confocal imaging studies (Fig. 7c).
Key signals from the MCH-1 receptor are mediated via members of the G i family of G proteins. When coexpressed with a form of the ␣ subunit of G o1 , which is the most highly expressed member of this group of G proteins in the mammalian central nervous system, MCH caused a large, concentration-dependent, enhancement of binding of [ 35 S]GTP␥S in G␣ o1 immunoprecipitates which was achieved with EC 50 ϭ 350 nM (Fig. 8a). The extent of MCH-mediated activation of G␣ o1 was unaffected by expression of MIZIP (Fig. 8b), but this was reduced substantially in membranes of cells coexpressing PPL (Fig. 8b). The IP prostanoid receptor functions predominantly via activation of G␣ s and elevation of cAMP levels. Coexpression of the IP prostanoid receptor with G␣ s resulted in a substantial increase in binding of [ 35 S]GTP␥S in G␣ s immunoprecipitates upon addition of 1 M iloprost (Fig. 9). As anticipated from the lack of coimmunoprecipitation of this receptor with PPL, this was unaffected by the coexpression of HA-PPL (Fig. 9). The MCH-1 receptor is also able to elevate intracellular [Ca 2ϩ ], at least partially via pertussis toxin-insensitive G proteins. Although MCH had no effect in nontransfected, control HEK293 cells (not shown), when the MCH-1 receptor was expressed in these cells, the addition of 0.1 M MCH resulted in a substantial, transient elevation of intracellular [Ca 2ϩ ] (Fig. 10). This effect of MCH was reduced substantially in cells coexpressing the MCH-1 receptor and HA-PPL (Fig. 10), indicating that PPL was also able to compete with members of the G q /G 11 G protein family to bind to the MCH-1 receptor. By contrast, although MCH produced phosphorylation of the MAP kinases ERK1 and ERK2 in HEK293 cells expressing the MCH-1 receptor, the ability of MCH to cause phosphorylation of these kinases was not compromised by coexpression of either PPL or MIZIP (Fig. 11).
For these experiments to have potential physiological relevance requires the MCH-1 receptor and PPL to be coexpressed in native tissues. Northern blot and reverse transcription-PCR techniques have shown previously that PPL and the MCH-1 receptor are highly expressed within the central nervous system of rodent species (19,33). To look for overlap of the expression patterns of mRNAs encoding PPL and the MCH-1 receptor in the brain we applied RNA in situ hybridization techniques to analyze the expression of PPL and MCH-1 in adjacent sections of the postnatal rat brain (Fig. 12). At p14, PPL mRNA expression was detected at high levels throughout the brain, especially in the cortex and hippocampus (Fig. 12). In contrast, only low levels of mRNA encoding the MCH-1 receptor were detectable at p14 (Fig. 12). However, in the adult brain, overlap of distribution was detectable in all regions of the brain, especially in the hippocampus and cerebellum (Fig. 12). No signals were detected when using sense control probes for hybridization (not shown). Such studies, although informative, do not demonstrate coexpression of the MCH-1 receptor and PPL polypeptides in specific cells. We therefore applied fluorescence immunohistochemical techniques to look for colocalization of the MCH-1 receptor and PPL in the adult mouse brain using a sheep polyclonal antiserum raised against the C-terminal 208 amino acids of PPL and an affinity-purified antiserum against the MCH-1 receptor raised in chicken. PPL immunoreactivity was largely restricted to the cytoplasm and plasma membrane of labeled cells (Fig. 13); however, labeling was in addition observed in dendrites, for example, of neurones within the amygdala (arrowheads in Fig. 13d). Colocalization of PPL with the MCH-1 receptor was detected in several regions of the brain, for example, in the cortex, especially at the piriform cortex (Fig. 13, a-c), within the amygdala (Fig. 13, d-f), and the stratum pyramidale of the hippocampus formation (Fig. 13,  g-i). In the cerebellum, colocalization of PPL and the MCH-1 receptor was detected in the Purkinje cells, but not in granular cells that express high levels of the MCH-1 receptor but showed only low levels of PPL immunoreactivity (Fig. 13, k-m). DISCUSSION It is becoming increasingly clear that GPCRs do not exist in isolation but rather within protein complexes (22)(23)(24)(25). Interactions within such complexes can maintain the receptor at a specific location, alter its trafficking properties, or modulate function. The C-terminal tail of many GPCRs encompasses binding sites for interacting proteins, and this has resulted in this region being described as the "magic tail" of GPCRs (44) and "as an anchorage for functional protein networks" (44). A significant number of identified GPCR-interacting proteins also link to the actin cytoskeleton and thus may provide frameworks to define subcellular localization.
PPL is a member of the plakin family of cytolinker proteins (31). Although best studied in keratinoctyes, where it plays a key role in providing the basal layer for the construction of the cornified epithelium (36 -37), it is both highly expressed and widely distributed in the central nervous system of both human and rodents (32). We recently described the interactions of PPL with the MOP-1 and MOP-1A splice variants of the human MOP opioid receptor (32) and demonstrated this to be a selective interaction because it did not bind to either the ␤ 2 -adrenoreceptor or the ␣ 2A -adrenoreceptor (32). In the current study we demonstrate selective and high affinity interactions between PPL and the C-terminal tail of a second GPCR, the MCH-1 receptor. The MCH-1 receptor is a well validated target for potential therapeutic intervention in obesity because knock- out studies in mice have resulted in animals with altered metabolism and a lean phenotype after elimination of expression of either MCH or the MCH-1 receptor (13)(14)(15). Treatment of animal models of obesity and type 2 diabetes with either MCH or small molecule antagonists of the MCH-1 receptor also modulate the expression and levels of other key regulators of appetite and energy balance (45,46). Furthermore, because early antagonists at the MCH-1 receptor have been found to have antidepressant and anxiolytic properties (45, 47) they may find further clinical uses. MCH-1 receptor mRNA and protein are expressed in the ventromedial and dorsomedial nuclei of the hypothalamus, consistent with a role for this GPCR in mediating the effects of MCH on feeding. Equally, because the MCH-1 receptor is expressed in several brain regions (19), in particular those involved in olfactory learning and reinforcement mechanisms, therapies targeting the MCH-1 receptor should act on the neuronal regulation of food consumption.
Identification of PPL as a partner protein for the MCH-1 receptor C-terminal tail in yeast two-hybrid screens was confirmed in both pull-down and coimmunoprecipitation assays. We then defined the region of the receptor responsible as a short section predicted to be proximal to the end of transmembrane helix VII. This region is similar to the section of the MOP opioid receptor identified as a PPL interacting domain (32). There are, however, a number of differences in this area between the MCH-1 receptor and the MOP opioid receptor. There is a cysteine residue located in the MOP opioid receptor Cterminal tail in a location that in other rhodopsin-like GPCRs generally results in thioacylation (48). At least based on the crystal structure of bovine rhodopsin (49), acylation at this location (50) produces the "fourth intracellular loop" or "helix VIII" of receptors. Data on post-translational thioacylation of the MOP opioid receptor are unclear as to the location of the modification (51), and it should be noted that this cysteine is FIG. 11. Lack of effect of PPL on MCH-1 receptor-mediated phosphorylation of the ERK1 and ERK2 MAP kinases. HEK293 cells were transfected to express the VSV-G-tagged hMCH-1 receptor alone or in combination with either His-MIZIP or HA-PPL. Serumstarved cells were exposed to 1 M MCH for 15 min) (ϩ) or vehicle (Ϫ), and after SDS-PAGE lysates were immunoblotted to detect phospho-ERK1 and phospho-ERK2 (P-ERK; a), total levels of ERK1 and ERK2 (b), and expression of the VSV-G-tagged hMCH-1 receptor (MCH-1-R; c).

FIG. 12.
Overlapping distributions of mRNA encoding the MCH-1 receptor and PPL in rat brain. RNA expression of MCH-1 receptor and PPL was analyzed by RNA in situ hybridization of radiolabeled MCH-1 receptor (a and b) and PPL (c and d) riboprobes to adjacent horizontal sections of a p14 rat brain (a and c) and sagittal sections of an adult rat brain (b and d). The autoradiographic signals obtained with Beta-Max x-ray film are shown. At p14 of postnatal brain development MCH-1 receptor expression is low and restricted to the hippocampus formation and cerebellum (a). In contrast, high levels of PPL expression are visible in the cortex and hippocampus and a lower level of expression in the cerebellum (c). In the adult rat brain, coexpression of MCH-1 receptor (b) and PPL (d) is visible within the whole brain, especially the hippocampus and cerebellum. ce, cerebellum; co, cortex; hi, hippocampus. Bar, 1 mm. reported not to be the target site (51). However, in studies of the interactions between PPL and the MOP opioid receptor we described the interaction as occurring at helix VIII (32). By contrast, the MCH-1 receptor does not have any cysteine residues in the C-terminal tail which are likely sites for thioacylation. Thus, although this region may adopt a helical structure, such a helix will not be terminated by an acyl chain burying into the membrane bilayer. However, it is noteworthy that both the MCH-1 receptor and the MOP opioid receptor have a proline residue (328 in the hMCH-1 receptor) located immediately at the end of the region we map as the PPL binding site. This helix breaker may define the end of helix VIII in rhodopsin-like GPCRs that are not thioacylated. Deletion (⌬316 -320) of the five-amino acid stretch ETFRK within this region eliminated binding of PPLC to the MCH-1 receptor C-terminal tail. The previously characterized region of interaction of the MOP opioid receptor with PPL contains a sequence, ENFKR, that is highly homologous to ETFRK, and it is interesting to note that the completely unrelated Fc␥R1 (also called CD64) that has also recently been reported to bind PPL (52) has the sequence ELRRK within the binding region. It is probably too simplistic, however, to define the PPL binding motif as a triplet of hydrophobic-basic-basic amino acids because such a sequence is present in helix VIII of many rhodopsin-like GPCRs, including the IP prostanoid receptor, which we show not to bind PPL. This section alone therefore cannot define interaction selectivity and may be more directly involved in the contribution of this region to G protein interaction and activation (53,54).
The stepwise truncation of 13-amino acid blocks from the C terminus of the MCH-1 receptor also defined interaction between the C-terminal tail of this GPCR and the zinc finger protein MIZIP to require sequences proximal to transmembrane helix VII. However, because both the ⌬316 -320 and ⌬322-326 segments bound MIZIP, it is clear that MIZIP and PPL do not share a common binding site on the MCH-1 receptor, although they may overlap to some degree. MIZIP was unable to interfere with either agonist-mediated internalization of the full-length MCH-1 receptor or G protein activation. By contrast, coexpression of PPL with the MCH-1 receptor resulted in a substantial loss of G protein activation. The MCH-1 receptor is known to interact with and activate members of the G i and G q G protein subfamilies and by monitoring both agonist-mediated loading of [ 35 S]GTP␥S onto the ␣ subunit of G o1 and the elevation of intracellular [Ca 2ϩ ] we demonstrate that the presence of PPL is able to limit activation of both G protein classes. As noted earlier, a rational explanation for this can be suggested based on the known contribution of the proximal section of C-terminal tail, i.e. helix VIII, in G protein interaction and activation (53). Mutation of the two basic residues within the ETFRK sequence located between 316 and 320 of the MCH-1 receptor has recently been reported to interfere with agonist-mediated calcium elevation (54), and we now show both that elimination of the ETFRK sequence prevents interaction with PPL and that coexpression of PPL with the hMCH-1 receptor interferes with MCH-mediated elevation of intracellular [Ca 2ϩ ]. It is thus likely that PPL competes with both families of G proteins for a common binding site on this receptor. In contrast to these results, after expression of the hMCH-1 in HEK293 cells, MCH-mediated phosphorylation of the MAP kinases ERK1 and ERK2 was not abrogated by coexpression of PPL. Virtually all GPCRs are able to promote phosphorylation and activation of these kinases, and this can occur via a wide range of mechanisms, some of which appear not to require G protein activation (55,56). Moreover, ERK1 and ERK2 activation in HEK293 cells is a highly sensitive and amplified response in which only a very small fraction of the total G i -G protein pool requires activation to generate a maximal signal (for example, see Ref. 57). Thus, if hMCH-1 receptor stimulation of ERK phosphorylation is a G protein-mediated effect in these cells, expression of PPL would have to be extremely high to compete sufficiently to modulate or block this response.
Overlapping distributions of mRNAs encoding the MCH-1 receptor and PPL were observed in rat brain, and, more directly, coexpression of the two proteins was also demonstrated in neurones in a range of brain regions and structures, suggesting likely interactions of these proteins in vivo. These interactions may therefore modify the physiological actions of MCH and the MCH-1 receptor.