INTRODUCTION

Neuropeptide Y (NPY),¹ peptide YY (PYY), and pancreatic polypeptide (PP) belong to a family of structurally related 36-amino acid peptides which have functions in both neural and endocrine signaling. These peptides exert their actions via receptors on the targeted neurons and peripheral cells. Several receptor subtypes have been defined by their ability to bind NPY, PYY, PP, and derivatives of these peptides. Earlier studies classified this receptor family into at least three receptor subtypes, Y1, Y2, and Y3 (1). The existence of additional receptors were also proposed, ``atypical'' Y1 receptor (mediating the feeding response of NPY in hypothalamus), PP-preferring receptor (exerting PP activity), and PYY-preferring receptor (2, 3). Molecular cloning studies have revealed the structure of Y1, Y2, and Y4/PP1 receptors (4, 5, 6, 7, 8, 9). These are all heptahelix (seven-transmembrane regions) receptors which couple to G-proteins. The Y4/PP1 receptor has higher affinity for PP than PYY and NPY, suggesting this receptor to be a member of PP receptors.

As yet unknown members of this NPY receptor family are expected to be identified by molecular cloning. Very recently, during the course of the present work, the cloning and characterization of mouse Y5 (10) and rat Y5 (11) receptors was reported. The mouse Y5 receptor has significant homology with the Y1 receptor, but the rat Y5 receptor has little identity with the previously cloned NPY receptors, showing that these receptors are not species orthologues, despite having the same name, Y5. In the present report, we describe the cloning of a novel gene sharing the highest homology with the Y1 receptor from humans, rabbits, and several other species. Our study revealed that this novel gene is the orthologue of the mouse Y5 receptor gene (10) and encodes functional NPY/PYY receptor in rabbits, which we have named Y2b, but has been inactivated for receptor function in primates by a frameshift mutation occurring early in primate evolution.

EXPERIMENTAL PROCEDURES

Human caudate nucleus and hypothalamus cDNA libraries were purchased from Clontech. Approximately 700,000 plaques for each library were lifted with nylon membranes, and hybridization was done with human Y1 cDNA fragments (position 42-1412 refers to HUMNEYPEPY in the GenBank^TM data base) as a probe. The membranes were hybridized with ³²P-labeled probe (1 × 10⁶ cpm/ml) for 20 h at 34 °C in buffer containing 5 × SSPE, 5 × Denhardt's, 0.1% SDS, 100 µg/ml denatured salmon sperm DNA, and 30% formamide. The membranes were washed twice in 1 × SSC, 0.1% SDS at room temperature, and twice for 30 min in 1 × SSC, 0.1% SDS at 34 °C. Twelve and eight individual clones were selected from the caudate nucleus and hypothalamus cDNA library, respectively. Sequences were determined with an automated fluorescent dye DNA Sequencer (ABI). The existence of a frameshift in the human Y2b gene and in the transcripts was confirmed by PCR using independent genomic DNAs or by RT-PCR using total brain, hypothalamus, heart, and skeletal muscle poly(A)⁺ RNA. Human Y2b reverted mutant was produced using PCR, and the reading frame was confirmed by sequencing.

Using primers 5-TCTGTGTGCATAGTGGAGAT-3 and 5-CTGCTGTCGAAAGATATCAG-3 produced from human Y2b cDNA, the orthologous gene fragments of the primate and rabbit were amplified by PCR at 94 °C for 1 min, 50 °C for 1 min, and 72 °C for 2 min for 34 cycles (12). Primers 5-TCCCTTCTGTTGTCCATTCC-3 and 5-AACCATAGCAACCAAGTGGC-3 produced from the rabbit sequence also amplified the mouse and hamster orthologue as well. The PCR products were sequenced directly to avoid the influence of PCR error. To obtain ORF of the rabbit cDNA, the 5- and 3-RACE method was employed using a Marathon cDNA Amplification Kit (Clontech) according to the manufacturer's protocol. Briefly, the cDNA of rabbit brain and skeletal muscle was ligated to the cDNA adaptor. For RACE, the adaptor-ligated cDNA was amplified with nested primers made from the partial rabbit DNA in combination with Marathon adaptor primers. Using primers 5-GAGAGTTTAGGAAATTTCTCAG-3 and 5-TAATAACTCTCATGGATAGTCC-3 (underlined sequences represent artificial XbaI site) made from the sequence of the RACE products, the ORF was amplified from rabbit brain and skeletal muscle cDNA and genomic DNA as described previously (13).

Rabbit Y2b, human Y2b, and the reverted-mutant cDNA were ligated into mammalian expression vector pEF-BOS(dhfr) containing dhfr gene as a selective marker (14, 15). For transient expression study, the plasmid constructs and no inserted vector were separately transfected into COS-1 cells as described previously (13). To establish a rabbit Y2b expressed cell line, the plasmid construct was transfected into CHO(dhfr) cells using LipofectAMINE (Life Technologies, Inc.) as described previously (15). The transfected cells were divided into single cells, and a monoclonal cell line showing specific ¹²⁵I-porcine PYY binding was used for pharmacological characterization. For binding study, cells were washed in phosphate-buffered saline and homogenized in 10 volumes of ice-cold hypotonic buffer (5 mM Tris-HCl, pH 7.4) using a motor-driven Teflon homogenizer, then centrifuged at 40,000 × g at 4 °C for 30 min. The pellet was resuspended into ice-cold binding buffer (20 mM HEPES, 10 mM NaCl, 0.22 mM KH₂PO₄, 1.26 mM CaCl₂, 0.81 mM MgSO₄, pH 7.4). Cell membranes (170 µg/ml) were incubated with the radiolabeled peptides and competitor peptides in a total volume of 0.5 ml of binding buffer supplemented with 0.1% bovine serum albumin and 0.1% bacitracin for 120 min at 25 °C. Nonspecific binding was defined as binding in the presence of 1 µM unlabeled peptide. cAMP formation was assayed in intact rabbit Y2b-transfected CHO cells as described previously (15).

Total RNA from various rabbit tissues and brain regions was treated with DNase (RNase-free, Nippongene) to eliminate the contamination of genomic DNA. Then the RNA was reverse-transcribed as described previously (13). Using primers 5-TTCCCTTGTTGCTTTCCTACC-3 and 5-GCACCACCAGATCTTTCTGG-3, Y2b cDNA was amplified from the cDNA pool or the corresponding quantity of RNA by PCR at 94 °C for 30 s, 60 °C for 30 s, and 72 °C for 1 min for 25 cycles. The amplified products were transferred to nylon membrane and hybridized with ³²P-labeled probe made from the full-length Y2b cDNA. The G3PDH cDNA was amplified using human G3PDH primers (Clontech) by PCR at 94 °C for 30 s, 50 °C for 30 s, and 72 °C for 1.5 min for 30 cycles. Northern membranes containing 2 µg of poly(A)⁺ RNA from different human tissues and brain areas were purchased from Clontech. Northern analysis of human Y2b transcripts was performed twice under high stringency conditions as described previously (16) using distinct probes, the N terminus fragment (position 1-628 in Fig. 1) and C terminus fragment (position 629-1066 in Fig. 1). The figure shows results using the N terminus probe. The membrane was exposed to x-ray film for 7 days at 80 °C with an intensifying screen. The second analysis using the C terminus probe detected the same signals (data not shown), confirming the signals to be Y2b transcripts. Additional 3-RACE experiments using human skeletal muscle poly(A)⁺ RNA indicated the main size of human Y2b transcripts to be 3 kilobases, the same size as shown in the figure (data not shown).

RESULTS AND DISCUSSION

To identify novel members of the NPY/PYY/PP receptor family, human caudate nucleus and hypothalamus cDNA libraries were screened using human Y1 cDNA (4, 5) as a probe. Two overlapping clones from the caudate nucleus library were found to have the highest nucleotide sequence identity to the Y1 receptor (54%), indicating that this cDNA encoded a novel heptahelix receptor for the NPY/PYY/PP family. Although these clones covered the entire ORF, alignment with human Y1 indicated the existence of a frameshift between the sixth and seventh putative transmembrane domains (TMD). Subsequent RT-PCR and 3-RACE experiments revealed no intron in the ORF and no RNA editing or alternative splicing in the tissues determined, confirming that this gene has a premature stop codon downstream from the sixth TMD, giving a 290-amino acid protein (Fig. 1). PCR experiments of human/rodent somatic cell hybrids and Southern blot analysis indicated that the human genome contains a single copy of this gene on chromosome 5 (data not shown).

The presence of only one putative frameshift and the conserved homology with the Y1 receptor prompted us to investigate the orthologous gene of other species. One primer pair made from the human cDNA sequence amplified the orthologue of the primate species chimpanzee, gorilla, and tamarin (a New World monkey), and of a nonprimate species, the rabbit. Using the rabbit sequence, the mouse and hamster orthologues were amplified. Southern blot analysis supported an orthologous relationship among these genes and indicated that they are single-copy genes (data not shown). Comparison of the sequence of primates with that of nonprimates revealed a common single-base deletion in all primate sequences generating the supposed frameshift in the latter half of the sixth TMD (Fig. 2A). This finding indicates that the frameshift occurred in primate ancestors before the separation of the New World monkey lineage from the Old World monkey and ape-human lineage. Another frameshift mutation was found in the chimpanzee and tamarin gene between the forth and fifth TMD (Fig. 1), suggesting that this gene was universally inactivated in primate species, probably due to the first common frameshift mutation, and that the additional mutations have been accumulated in the inactivated gene. In fact, in transiently transfected COS-1 cells, neither the human cDNA nor the reverted-mutant (encoding a heptahelix protein of 370 amino acids restored by a single-base insertion, Figs. 1 and 2B) produced any specific binding sites for ¹²⁵I-porcine PYY, ¹²⁵I-human NPY, PYY, or PP at concentrations up to 2 nM.

Using the 5- and 3-RACE methods, the rabbit ORF was amplified from the brain and skeletal muscle poly(A)⁺ RNA (Fig. 2B). The gene was revealed to be intronless and to encode a novel heptahelix receptor of 371 amino acids, sharing highest identity with human Y1 (49%) followed by Y4/PP1 (41%) and Y2 (24%) receptors. ¹²⁵I-Porcine PYY specifically bound with high affinity (K_d = 50 pM) to membranes from CHO cells stably transfected with the rabbit cDNA. This receptor bound NPY, PYY, PP, and their derivatives with the distinctive rank order of PYY  NPY_2-36  NPY_13-36  NPY > [Leu³¹,Pro³⁴]NPY  PP (Fig. 3A). Simultaneous incubation with human PYY decreased the forskolin-stimulated cAMP formation in the transfected CHO cells with an E_max of 94% (EC₅₀ = 0.94 nM) (Fig. 3B). Although this novel receptor shares higher homology with the Y1 than the Y2 receptor, the rank order of affinity shows its preference for NPY_13-36 rather than for [Leu³¹,Pro³⁴]NPY, namely the Y2-like profile (1, 2, 3, 6, 7). As this pharmacological profile was not consistent with any NPY/PYY/PP receptors proposed to date (2, 3), we termed this receptor Y2b. During the course of this work, Weinberg et al. (10) reported a novel mouse NPY receptor termed Y5 which showed preference for [Leu³¹,Pro³⁴]NPY, namely the Y1-like profile (1, 2, 3, 4, 5). Interestingly, the sequence of mouse Y5 was identical to that of the amplified mouse Y2b orthologue, revealing that the rabbit Y2b gene (82% amino acid identity) and the human Y2b gene (82% nucleotide identity) are the orthologous gene of the reported mouse Y5 (Fig. 2B). The mouse and rabbit orthologues exhibit profound pharmacological differences (Y1-like versus Y2-like), suggesting either that the altered amino acids which possibly determine the difference in binding specificity between Y1 and Y2 (1, 2, 3) have no functional significance in these species or that directional selection has occurred.

In rabbit, Y2b mRNA was detected in several tissues and brain regions, including hypothalamus, by RT-PCR (Fig. 4A), but was not detectable by Northern blot analysis (data not shown). Surprisingly, despite the inactivation for receptor function in humans, Northern blot analysis revealed abundant transcripts (3 kilobases) in the heart and skeletal muscle (Fig. 4B). It is presently unknown whether the frameshift mutation of the Y2b gene in the primate ancestor was neutral, i.e. Y2b was dispensable in the ancestor, or had evolutionary advantages. However, considering the high levels of Y2b transcripts in human tissues, unlike the absent or low levels of transcription of inactivated genes observed to date (17, 18, 19, 20), it is interesting to speculate that the mutation was the first step toward not only the inactivation for receptor function but also the creation of another function in the Y2b gene, and that the human Y2b gene possesses this novel function in place of a receptor for NPY/PYY.

RT