Expression cloning and pharmacological characterization of a human hippocampal neuropeptide Y/peptide YY Y2 receptor subtype.

The pancreatic polypeptide family includes neuropeptide Y (NPY), one of the most abundant neuropeptides in the mammalian nervous system, as well as peptide YY (PYY) and pancreatic polypeptide (PP). This peptide family is involved in numerous physiological processes such as memory, pain, blood pressure, appetite, anxiety, and circadian rhythm. Of the multiple Y-type receptors proposed for PP family members, only the Y1 subtype was cloned previously. We now report the isolation of a human Y2 (hhY2) receptor cDNA by expression cloning from a human hippocampal cDNA library, using a 125I-PYY binding assay. hhY2 cDNA encodes a predicted protein of 381 amino acids with low amino acid identity to the human Y1 receptor (31% overall; 41% transmembrane). 125I-PYY binding to transiently expressed hY2 receptors was saturable (pKd = 10.17) and displaceable by human PP family members in rank order: PYY (pKi = 9.47) ∼ NPY (pKi = 9.27) PP (pKi < 6) and by peptide analogs: NPY2-36 (pKi = 8.80) ∼ NPY13-36 (pKi = 8.55) ∼ C2-NPY (pKi = 8.54) > NPY26-36 (pKi = 6.51) ∼ [Leu31,Pro34]NPY (pKi = 6.23). Human PYY decreased [cAMP] and increased intracellular [Ca2+] in hY2-transfected 293 cells.

The pancreatic polypeptide family includes neuropeptide Y (NPY), one of the most abundant neuropeptides in the mammalian nervous system, as well as peptide YY (PYY) and pancreatic polypeptide (PP). This peptide family is involved in numerous physiological processes such as memory, pain, blood pressure, appetite, anxiety, and circadian rhythm. Of the multiple Y-type receptors proposed for PP family members, only the Y1 subtype was cloned previously. We now report the isolation of a human Y2 (hhY2) receptor cDNA by expression cloning from a human hippocampal cDNA library, using a 125 I-PYY binding assay. hhY2 cDNA encodes a predicted protein of 381 amino acids with low amino acid identity to the human Y1 receptor (31% overall; 41% transmembrane). 125 I-PYY binding to transiently expressed hY2 receptors was saturable (pK d ‫؍‬ 10.17) and displaceable by human PP family members in rank order: PYY (pK i ‫؍‬ 9.47) ϳ NPY (pK i ‫؍‬ 9.27) > > PP (pK i < 6) and by peptide analogs: NPY 2-36 (pK i ‫؍‬ 8.80) ϳ NPY 13-36 (pK i ‫؍‬ 8.55) ϳ C2-NPY (pK i ‫؍‬ 8.54) > NPY 26 -36 (pK i ‫؍‬ 6.51) ϳ [Leu 31 ,Pro 34 ]NPY (pK i ‫؍‬ 6.23). Human PYY decreased [

cAMP] and increased intracellular [Ca 2؉ ] in hY2-transfected 293 cells.
The pancreatic polypeptide family includes neuropeptide Y (NPY), 1 peptide YY (PYY), and pancreatic polypeptide (PP), all of which are 36 amino acid peptides characterized by an NH 2terminal polyproline helix and a COOH-terminal ␣-helix brought together by a hairpin loop (1). NPY functions primarily as a neurotransmitter and is widely distributed throughout the central and peripheral nervous system, with additional local-ization in adrenal gland and certain non-neuronal cells (1,3). NPY modulates numerous physiological processes, including appetite, anxiety, blood pressure, and circadian rhythm (1,2). PYY is localized primarily in intestinal endocrine cells and functions as a circulating hormone with increasing levels postprandially; small amounts are also found in central and peripheral neurons (1,2,4). PYY is known to regulate intestinal secretion and motility, as well as emesis (5,6). PYY and NPY act similarly in a majority of physiological models (e.g. to stimulate feeding and increase blood pressure), but exceptions have been noted (1). PP is localized primarily in endocrine cells of pancreatic islets and exerts regulatory effects on gastrointestinal processes such as pancreatic exocrine secretion, gall bladder contraction and gastric emptying (7,8).
NPY and related family members are proposed to activate at least five receptor subtypes (1,2) (9 -13). We describe here the expression cloning and pharmacological characterization of a human hippocampal Y2 receptor.

MATERIALS AND METHODS
Cloning and Sequencing-Total RNA was prepared by a modification of the guanidine thiocyanate method (14) from 6 g of human hippocampus. Poly(A) ϩ RNA was purified with a FastTrack kit (Invitrogen Corp., San Diego, CA). Double-stranded (ds) cDNA was synthesized from 4 g of poly(A) ϩ RNA according to Gubler and Hoffman (15), except that ligase was omitted in the second strand cDNA synthesis. After size selection, high molecular weight fractions were ligated in pEXJ.BS (an Okayama and Berg expression vector) cut by BstXI, as described by Aruffo and Seed (16). The ligated DNA was electroporated in Escherichia coli MC 1061 (Gene Pulser, Bio-Rad). The library (2.2 ϫ 10 6 cfu; 3-kb average insert size) was plated on Petri dishes (ampicillin selection) in pools of 0.4 -1.2 ϫ 10 4 independent clones. After 18-h amplification, the bacteria from each pool were scraped, resuspended in 4 ml of LB medium, and 1.5 ml processed for plasmid purification by the alkali method (17). 1-ml aliquots of each bacterial pool were stored at Ϫ85°C in 20% glycerol. DNA from pools of Ϸ5000 independent clones was transfected into COS-7 cells by a modification of the DEAE-dextran procedure (18). 48 h after transfection, the binding assay was performed directly on the cell culture slides (1 chamber, Permanox slide from Nunc Inc., Naperville, IL). After two washes with PBS, positive pools were identified by incubating the cells with 1 nM (3 ϫ 10 6 cpm/slide) of porcine 125 I-PYY (DuPont NEN; specific activity ϭ 2200 Ci/mmol) in binding buffer (20 mM Hepes-NaOH, pH 7.4, 1.26 mM CaCl 2 , 0.81 mM MgSO 4 , 0.44 mM KH 2 PO 4 , 5.4 mM KCl, 10 mM NaCl, 0.1% BSA, and 0.1% bacitracin) for 1 h at room temperature. After washing in binding buffer without ligand, the monolayers were fixed, dipped in liquid photoemulsion and exposed in the dark as described except that gelatin dipping was omitted (23). Slides were developed, fixed, and mounted with Aqua-Mount (Lerner Laboratories, Pittsburgh, PA). Slides were screened at ϫ 25 total magnification. From the first 200 pools tested, three gave rise to positive cells in the screening assay (pools 145, 158 and 189). The last 220 pools tested were all negative. DNA from pools 145, 158, and 189 was analyzed by PCR using human Y1 receptor subtype-specific primers. Pools 145 and 158 turned out to contain cDNAs coding for the Y1 receptor subtype, but DNA from pool 189 was negative in the PCR assay. A single clone, hhY2 (human hippocampal Y2), was isolated from pool 189, as described (19). ds DNA was sequenced with a Sequenase kit (U. S. Biochemical Corp.) according to the * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EMBL Data Bank with accession number(s) U36269.
manufacturer. Nucleotide and peptide sequences analysis were performed with GCG programs (Genetics Computer Group, Madison, WI).
Northern and Southern Blots-Multiple tissue Northern blots (brain MTN, MTN, and MTN II, Clontech, Palo Alto, CA) carrying mRNA purified from various human brain areas or from various peripheral tissues were hybridized at high stringency according to the manufacturer's specifications. The probe was a 1.15-kb DNA fragment generated by PCR, corresponding to the entire coding region of the human Y2 receptor. The probe was labeled with [ 32 P]dATP and a random-primed DNA labeling kit (specific activity: 2 ϫ 10 9 cpm/g; Pharmacia). A Southern blot (Geno-blot, Clontech) containing human genomic DNA cut with five different restriction enzymes was hybridized at high stringency according to the manufacturer's specification. The probe, corresponding to the TM1-TM5 coding region, was labeled with [ 32 P]dATP and a random-primed DNA labeling kit (specific activity: 1 ϫ 10 9 cpm/g; Pharmacia).
Receptor Expression-COS-7 cells were transiently transfected with the unmodified hhY2 construct (expression vector containing the hhY2 gene) or with human Y1 receptor (hY1) DNA by a DEAE-dextran method (18) or without plasmid for mock-transfection. For stable expression, plasmid hhY2 was co-transfected with a G418-resistant plasmid into 293 cells by the calcium phosphate method (20). Stably transfected cells were selected with G418 and screened as membrane preparations for specific 125 I-PYY binding.
Membrane Preparation-Membranes from COS-7 cells were prepared as described (22). Protein was measured by the Bradford method (21) using Bio-Rad reagent.
125 I-PYY Binding-Membrane suspensions, porcine 125 I-[Tyr 36 ]PYY and peptides were diluted in binding buffer supplemented with 0.1% BSA (Sigma catalog number A-7511) and distributed into 96-well polypropylene microtiter plates. Samples were incubated at 30°C with shaking for 120 min. Membranes were collected by filtration over Whatman GF/C filters (precoated with 0.5% polyethyleneimine and air-dried before use) and counted for 125 I. Nonspecific binding was defined by 100 nM human NPY. Data were analyzed by nonlinear regression.
cAMP Measurements-96-well microtiter plates were precoated with poly-D-lysine (0.05 mg/well), dried, and seeded with stably transfected 293 cells. Serum, potentially desensitizing, was reduced to 1.5% 4 -16 h before the assay. Cells were pre-equilibrated in Hanks' buffered saline plus supplements (HBSP: 150 mM NaCl, 20 mM Hepes, 1 mM CaCl 2 , 5 mM KCl, 1 mM MgCl 2 , and 10 mM glucose plus 0.1% BSA and 5 mM theophylline) for 20 min at 37°C in 5% CO 2 , then stimulated 5 min with 10 M forskolin and peptides prepared in HBSP. Intracellular cAMP was extracted with 100 mM HCl at 4°C for 30 min and quantified by radioimmunoassay. The final antigen-antibody complex was collected by filtration through a 96-well Durapore filter plate (Millipore) and counted for 125 I. Data were analyzed by nonlinear regression.
[Ca 2ϩ ] Measurements-Intracellular free [Ca 2ϩ ] was measured by microspectrofluorometry (22). Stably transfected 293 cells were seeded into 35-mm culture dishes with glass coverslip inserts. Cells were loaded with 10 M fura-2/AM in HBS for 20 -40 min, then equilibrated with HBS alone for an additional 10 -20 min. Cells were visualized under a ϫ 40 microscope objective, and fluorescence emission was determined at 510 nM with excitation alternating between 340 and 380 nM. Raw fluorescence data were converted to [Ca 2ϩ ] using standard [Ca 2ϩ ] curves and reported as mean Ϯ S.E.

RESULTS AND DISCUSSION
In order to clone a human NPY/PYY Y2 receptor subtype, we used an expression cloning strategy in COS-7 cells (23)(24)(25). Because the Y2 receptor is described as a presynaptic receptor, it is difficult to locate cell bodies that actually contain this specific mRNA in restricted brain areas. We reasoned that human hippocampus was a good source of mRNA, because it contains both a large number of interneurons and has been shown to carry a particularly dense population of Y2 receptors (26 -29). A single clone encoding a Y2 receptor (designated hhY2) was isolated from a human hippocampal cDNA library (see "Materials and Methods"). The isolated clone carries a 4.2-kb cDNA. This cDNA contains an open reading frame between nucleotides 1003 and 2145 that encodes a 381 amino acid protein. The flanking sequence around the putative initiation codon conforms to the Kozak consensus sequence for optimal translation initiation (30,31). The hydrophobicity plot displayed seven hydrophobic, putative membrane-spanning regions which make the human hippocampal Y2 receptor a member of the G protein-coupled superfamily. The deduced hhY2 amino acid sequence and an alignment with the human Y1 and Y4 receptors (see accompanying paper by Bard et al. (46)) are shown in Fig. 1. The hhY2 receptor presents features common to most members of the GPCR family (32, 33), including a potential N-linked glycosylation site and multiple potential phosphorylation sites in the putative intracellular loop regions. In addition, it carries transmembrane amino acid residues highly conserved within the neuropeptide receptor family (34). Interestingly, the two amino acid residues "RY" downstream of transmembrane domain 3, found in almost all known GPCR sequences (32), are replaced by "RH" in all three Y1, Y2, and Y4 receptor amino acid sequences (Fig. 1). As seen in Fig. 1, the hhY2 amino acid sequence shows a surprisingly low overall identity of 31% with the Y1 receptor. The alignment scores for the TM domains are 41 and 43% identity with the Y1 and the Y4 receptors, respectively. When compared with other neuropeptide GPCR sequences, the amino acid transmembrane domains identities are very low, ranging from 26% (human bradykinin (2,35)) to 33% (human neurokinin 1 and PR4, a Y2-like Drosophila receptor (36, 37)) with the notable exception of the orphan sequence MUSGIR score at 43% (38). Using the human Y2 probe, Northern hybridizations reveal a unique band at 4.3 kb in human brain after a 3-day exposure ( Fig. 2A). This is in good agreement with the 4.2-kb cDNA that we isolated by expression cloning and indicates that our cDNA clone is full length or nearly full length. The mRNA encoding the human Y2 receptor subtype is present in significant amounts in amygdala, corpus callosum, hippocampus, and subthalamic nucleus. A faint band is detectable in caudate nucleus, hypothalamus, and substantia nigra. No signal could be detected in thalamus. It should be noted that the Clontech brain MTN blot does not contain any mRNA from cortex or brain stem. No detectable signal was observed on Northern blots containing mRNA extracted from human peripheral tissues after an 8-day exposure (data not shown). Southern hybridizations to human genomic DNA followed by high stringency washes (Fig. 2B) suggest that the human genome contains a single Y2 receptor gene (single band with EcoRI, HindIII, BamHI, and PstI). The faint bands at 9 and 12 kb observed with BglII can be explained by the presence of two BglII restriction sites in the coding region of the Y2 sequence and are also consistent with a single Y2 receptor gene.
Characterization of the novel receptor cloned from pool 189 was accomplished using radioligand binding and functional assays. 125 I-PYY (0.06 nM) bound specifically to membranes from hhY2-transfected COS-7 cells (but not from mock-transfected cells) at 30°C. The association curve was biphasic, with approximately 55% of the specific binding following a rapid time course (k obs ϭ 1.28 Ϯ 0.02 min Ϫ , t1 ⁄2 ϭ 0.5 min) and 45% following a slower time course (k obs ϭ 0.02 Ϯ 0.00 min Ϫ1 , t1 ⁄2 ϭ 37 min). Equilibrium binding composed of both phases was 95% complete within 120 min and 100% complete within 240 min. The biphasic time course suggests the possibility of multiple conformations for the receptor/ligand complex. hY1-transfected COS-7 cell membranes, when studied under the same conditions, yielded a monophasic association curve with k obs ϭ 0.06 Ϯ 0.02 min Ϫ1 , t1 ⁄2 ϭ 12 min, and 100% complete equilibrium binding within 90 min (n ϭ 3). Subsequent 125 I-PYY binding assays involving both hY1 and hY2 receptors were conducted for 120 min. 125 I-PYY binding to the transiently expressed hhY2 receptor was specific and saturable at 125 I-PYY concentrations ranging from 0.5 pM to 3.0 nM. Binding data were fit to a one-site model with an apparent pK d ϭ 10.17 Ϯ 0.05 (0.067 nM) and B max ϭ 7.7 Ϯ 0.7 pmol/mg membrane protein (n ϭ 5). The transiently expressed hY1 receptor bound 125 I-PYY with an apparent pK d ϭ 10.19 Ϯ 0.04 (0.065 nM) and B max ϭ 4.0 Ϯ 0.7 pmol/mg membrane protein (n ϭ 9).
Native Y2 receptors are coupled to three second messenger pathways in the human neuroblastoma cell line SMS-KAN: cAMP, intracellular Ca 2ϩ mobilization, and K ϩ -induced Ca 2ϩ influx through -conotoxin-sensitive channels (41). Two of these pathways, cAMP and Ca 2ϩ mobilization, were evaluated in hhY2 stably transfected 293 cells. Incubation of intact 293 cells with 10 M forskolin generated an average 12-fold increase in [cAMP] (n ϭ 10). Simultaneous incubation with human PYY decreased the forskolin-stimulated [cAMP] with an E max of 52 Ϯ 5% and with high potency (pEC 50 ϭ 9.47 Ϯ 0.15, n ϭ 10) in 293 cells stably transfected with hY2 (Fig. 3A), but not in untransfected cells. The inhibitory response was mimicked by NPY and COOH-terminal fragments, whereas Pro 34 analogs were relatively inactive (Table I) (Fig. 3B). In comparison, Y2 receptors inhibited bradykinin and angiotensin II-induced Ca 2ϩ flux in SMS-KAN cells (39). These data reveal a complexity in Ca 2ϩ signaling that can be further evaluated by studying hhY2 and species homologs in a variety of transfected cells. This is the first report of a Y2 receptor clone whose distribution and pharmacology support the potential for in vivo activation by NPY and PYY or fragments. A previous report described the cloning of a Y2-like receptor (PR4) from Drosophila and functional characterization in oocytes (37). PR4 was activated by mammalian peptides (PYY, NPY, C2-NPY, [Pro 34 ]NPY, and PP) at concentrations between 0.03 and 3 M with a Y2-like rank order, but functional invertebrate ligands were not identified. As there have been no published reports of an NPY analog in Drosophila, classification of PR4 as Y2-like could reasonably be viewed as tentative.
In summary, we have cloned the gene for a novel human hippocampal Y-type receptor and generated a pharmacological profile using NPY and related peptide family members. hhY2 closely resembles pharmacologically defined Y2 receptors in a variety of cell and tissue models (1). Y2 receptors were first proposed to exist in rat vas deferens, for example, based on the ability of both NPY and NPY  to block field-stimulated contraction through presynaptic inhibition of norepinephrine release (42). Y2 receptors in hippocampus exhibit higher binding affinity for [Ile 31 ,Gln 34 ]PP than for PP and [Leu 31 ,Pro 34 ]NPY (43). Hippocampal Y2 receptors are proposed to play a role in memory based on behavioral effects induced by NPY and fragments (44). Other functions (e.g. analgesia and antisecretory effects) are indicated by widespread distribution of Y2 receptors in dorsal root ganglia, intestinal enterocytes, and elsewhere (1,2,45). As such, the hhY2 receptor is likely to be involved in a number of physiological processes that can be further studied using selective agonists and antagonists.