Apical targeting of the P2Y₄ receptor is directed by hydrophobic and basic residues in the cytoplasmic tail

Abstract

The P2Y₄ receptor is selectively targeted to the apical membrane in polarized epithelial cell lines and has been shown to play a key role in intestinal chloride secretion. In this study, we delimit a 23 amino acid sequence within the P2Y₄ receptor C-tail that directs its apical targeting. Using a mutagenesis approach, we found that four hydrophobic residues near the COOH-terminal end of the signal are necessary for apical sorting, whereas two basic residues near the NH₂-terminal end of the signal are involved to a lesser extent. Interestingly, mutation of the key hydrophobic residues results in a basolateral enrichment of the receptor construct, suggesting that the apical targeting sequence may prevent insertion or disrupt stability of the receptor at the basolateral membrane. The signal is not sequence specific, as an inversion of the 23 amino acid sequence does not disrupt apical targeting. We also show that the apical targeting sequence is an autonomous signal and is capable of redistributing the normally basolateral P2Y₁₂ receptor, suggesting that the apical signal is dominant over the basolateral signal in the main body of the P2Y₁₂ receptor. The targeting sequence is unique to the P2Y₄ receptor, and sequence alignments of the COOH-terminal tail of mammalian orthologs reveal that the hydrophobic residues in the targeting signal are highly conserved. These data define the novel apical sorting signal of the P2Y₄ receptor, which may represent a common mechanism for trafficking of epithelial transmembrane proteins.

nucleotides are ubiquitous small molecules involved in a wide variety of biological processes. In addition to playing essential roles in phosphorylation, energy utilization and metabolism, and synthesis of nucleic acids and enzymatic cofactors, nucleotides are also released from cells where they serve as extracellular ligands for transmembrane receptors involved in signal transduction (33, 34). Nucleotide receptors have been found in all cell types, where they mediate a broad range of cell activities. There are two distinct families of nucleotide receptors, P2X and P2Y. P2X receptors are ion channels that open in response to extracellular ATP, while the P2Y receptors are a family of G protein-coupled receptors (GPCRs) that respond to extracellular nucleotides (8, 66). Eight P2Y receptors have been identified to date: P2Y₁, P2Y₂, P2Y₄, P2Y₆, P2Y₁₁, P2Y₁₂, P2Y₁₃, and P2Y₁₄. The numbering is not sequential because several receptors were reported to be P2Y receptors but later determined to either not respond to nucleotides or to be orthologs of existing mammalian P2Y receptors (21, 27, 35, 50).

The P2Y receptors can be divided into two subclasses based on their downstream signaling properties. The P2Y₁-like receptors (P2Y₁, P2Y₂, P2Y₄, P2Y₆, and P2Y₁₁) signal primarily through G_αq, while the P2Y₁₂-like receptors (P2Y₁₂, P2Y₁₃, and P2Y₁₄) signal primarily through G_αi. The P2Y₁₁ receptor also signals through G_αs to stimulate adenylyl cyclase (9, 51). The receptors can also be characterized by their activating ligands: P2Y₁, P2Y₁₁, P2Y₁₂, and P2Y₁₃ receptors are activated by adenine nucleotides (either ADP or ATP); P2Y₄, P2Y₆, and P2Y₁₄ receptors are activated by uracil nucleotides (UDP or UTP); and the P2Y₂ receptor is activated equally by ATP and UTP. The P2Y₁₄ receptor is unusual as it is activated by both UDP and UDP-sugars (4).

P2Y receptors are expressed in a variety of tissue types, and the subtype and density of the receptors vary significantly. P2Y₄ receptor mRNA is widely distributed and most abundant in the intestine (46), where it has been shown to play a role in luminal Cl⁻ secretion (53). Since bacterial invasion can induce nucleotide release (11, 45, 64) and Cl⁻ secretion is the known mediator of diarrhea symptoms (13, 14, 32), antagonists of the P2Y₄ receptor may have therapeutic value for the treatment of infectious diarrhea. Agonists of the P2Y₄ receptor may also be useful as it has been hypothesized that stimulation of the Cl⁻ secretory pathway may alleviate intestinal abnormalities associated with cystic fibrosis (53). Unfortunately, there are currently no selective ligands for the P2Y₄ receptor available. The P2Y₄ receptor has also been implicated in the control of K⁺ secretion in vestibular dark cell epithelium and in mouse colon (40, 43).

Epithelial cells are specialized cells that form water- and ion-impermeable barriers between organ lumens and underlying cells and tissues (3). This impermeable barrier is created by tight junctions, which are multiprotein complexes that surround and connect the cells that form the epithelial monolayer, as well as separate the epithelial cell plasma membrane into two distinct compartments (19). The apical membrane faces the lumen and contacts the external milieu, while the basolateral membrane faces the underlying cells and is in contact with the interstitial fluid. Formation of this barrier allows epithelial cell monolayers to regulate absorption and excretion of water, salts, nutrients, and toxins by selective expression and distribution of transmembrane channels, pumps, and receptors. The proper targeting of these proteins to the correct membrane surface is critical for the proper function of the monolayer. Our lab has investigated the targeting of the entire family of P2Y receptors and shown that seven of the eight P2Y receptors are strongly polarized when expressed in epithelial cell lines (71).

Despite the fact that aberrant protein sorting is often associated with disease states, the mechanisms by which epithelial cells establish and maintain these polarized distributions are poorly understood (30, 31, 41, 42, 55). It is well established that many proteins contain amino acid sequences that act as trafficking signals to direct the protein to the appropriate membrane (15, 54, 67). Tyrosine- and di-hydrophobic-based signals have been shown to interact with clathrin adaptor complex proteins and to direct basolateral targeting (16, 25, 26, 44, 47). In contrast, posttranslational modification, such as glycosylation or glypiation, is often sufficient to confer apical targeting (38, 56, 65, 69, 72). Oligomerization and lipid raft association have also been suggested to direct apical trafficking (37, 48, 49, 58).

Beyond these conserved signals, a wide array of seemingly unrelated sorting signals have been described (15, 67). Our lab has described such signals for the P2Y₁ and P2Y₂ receptor. The P2Y₁ receptor is directed to the basolateral membrane by a 25 amino acid signal located within its COOH-terminal tail, whose function depends on nine charged residues. The signal functions even after inversion or scrambling of the sequence in the context of the full receptor, indicating that it is sequence independent (70). The P2Y₂ receptor is sorted to the apical membrane of type II Madin-Darby canine kidney cells [MDCK(II)] cells and contains an apical targeting signal in its first extracellular loop that is dependent on four amino acids: Arg95, Gly96, Asp97, and Leu108 (52). This signal is highly unusual because, following receptor synthesis, it is located on the inside of vesicles and therefore inaccessible to intracellular sorting machinery. This is the only protein-based extracellular apical targeting signal identified in any protein to date.

The P2Y₄ receptor is expressed almost exclusively at the apical surface at steady state in MDCK(II), Caco-2, and 16HBE14o− cells (71). Even though P2Y₂ and P2Y₄ receptors are 52% identical and both are targeted to the apical membrane in MDCK(II) cells, they do not share a common sorting signal. We have previously shown that the apical sorting signal of the P2Y₄ receptor, in contrast to the P2Y₂ receptor, is located within its COOH-terminal tail (52). When the COOH-terminal tail of the P2Y₄ receptor is deleted, the truncated receptor is unsorted, whereas fusion of the P2Y₄ C-tail to the unsorted BK2 receptor just after transmembrane segment 7 (TM7) results in its expression at the apical membrane. In this study, we delimit the P2Y₄ targeting signal, identify key amino acids within this signal that are responsible for apical targeting, and demonstrate its dominance over a basolateral signal in the main body of the P2Y₁₂ receptor. These data add to the known targeting signals by which P2Y receptors are sorted to distinct membrane surfaces and may represent a novel mechanism for polarized trafficking.¹

METHODS

Construction of mutant and chimeric receptors.

Construction and cloning of HA-tagged P2Y₄ and BK2/Y₄ receptors into the retroviral vector pLXSN was accomplished as described previously (52, 71). Targeted mutations were introduced into the P2Y₄ receptor C-tail of these constructs by overlap-extension PCR (22), whereas P2Y₄ receptor truncation constructs were made using PCR with a 5′-vector primer and 3′-primers containing a stop codon at the appropriate location followed by a XhoI site to facilitate cloning.

The P2Y₁₂/P2Y₄ receptor chimera was constructed using overlap extension PCR. One set of primers amplified the P2Y₁₂ receptor coding sequence from the second codon (with an MluI site to facilitate cloning) through the codon for Ser304, the start of the C-tail, and also included the first seven codons of the P2Y₄ receptor C-tail starting at Asp311. The second set of primers amplified the C-tail of the P2Y₄ receptor starting at Asp311 through the end of the gene and contained a XhoI site at the end of the downstream primer. After the initial amplification, the two PCR products were isolated, then combined and amplified with only the outside primers. The resulting product was digested with MluI and XhoI and ligated into a similarly digested modified pLXSN vector that added an HA-tag to the NH₂ terminus of the chimera.

The P2Y₄ C-tail invert construct was created using two long primers (68 and 69 nucleotides) whose respective 3′-ends overlapped by 30 nucleotides. These primers encoded the P2Y₄ receptor C-tail in which the codons for the targeting signal were inverted N→C and included SalI and SbfI restriction sites at the ends of the primers. The primers were annealed, filled in with the Klenow fragment of DNA polymerase I (New England Biolabs, Beverly, MA), and digested with SalI and SbfI. The digested fragment was cloned into a modified pLXSN-HA-P2Y₄ plasmid in which SalI and SbfI sites were introduced by incorporating silent mutations in the codons for Arg314/Arg315 and Ser345, respectively.

We initially created receptor mutants in the context of either the BK2-P2Y₄ C-tail chimeric receptor or the P2Y₄ receptor. While confocal microscopy revealed essentially identical targeting as observed with modified P2Y₄ receptor constructs described below, cell surface expression was often too low for accurate quantification of apical versus basolateral receptor distribution (data not shown). Therefore, we introduced a cleavable signal sequence (MKTIIALSYIFCLVPA) and FLAG epitope tag (DYKDDDDA) immediately upstream of the HA-tag to increase receptor expression (20). All of the constructs shown in Fig. 3A, with the exception of the BK2/P2Y₄-C321S receptor, contained this addition. Inclusion of the signal sequence increased steady-state receptor levels, which facilitated imaging and quantification, without having an appreciable effect on localization in polarized monolayers.

Cell culture.

All cells were grown in a humidified incubator at 37°C in a 5% CO₂-95% air atmosphere. MDCK(II) cells (ATCC, Rockville, MD) were maintained in 1:1 DMEM/F-12 medium containing 5% fetal bovine serum and 1× penicillin and streptomycin (pen/strep). PA317 cells were maintained in DMEM containing 10% fetal bovine serum and 1× pen/strep.

Retroviral infection.

Recombinant retroviruses were produced by calcium phosphate-mediated transfection of PA317 cells with pLXSN plasmids as previously described (10). Retroviral particles in the culture supernatant were harvested 3 days after transfection and used to infect MDCK(II) cells. Infected cells were selected for 7–10 days in medium containing 1 mg/ml G418. After selection, cells were maintained in medium containing 0.4 mg/ml G418. Surface expression of receptors was confirmed by radioimmunoassay with an antibody directed against the HA-epitope tag as described previously (2).

Confocal fluorescence microscopy.

MDCK(II) cells were seeded at 6 × 10⁵ cells/well in 12-mm Transwell inserts (Corning Life Sciences, Acton, MA) and grown for 5–7 days with daily medium changes to allow the cells to form polarized monolayers. Cells were prepared for confocal microscopy as described previously (71). Briefly, cells were washed and fixed in 2% PFA in PBS with 2 mM CaCl₂ and 2 mM MgCl₂ for 30 min at 4°C. After fixation, cells were permeabilized with cold (−20°C) methanol for 30 s. Cells were then quenched by three washes of 150 mM sodium acetate in 1% nonfat dry milk (NFDM) and blocked by three more washes in 1% NFDM. Cells were incubated with a 1:1,000 dilution of mouse monoclonal anti-HA antibody (HA.11; Covance, Berkeley, CA) and a 1:500 dilution of rabbit anti-ZO-1 antibody (Zymed, South San Francisco, CA) in 1% NFDM overnight at 4°C. Cells were then washed three times and incubated with both goat anti-mouse Alexa-488 and goat anti-rabbit Alexa-594 (Molecular Probes, Eugene, OR), each diluted 1:500 in 1% NFDM, for 1 h at room temperature. After being washed five times in PBS and once in Molecular Probes Equilibration Buffer, the polyester membranes were removed from the Transwell inserts with a scalpel and mounted under coverslips in Slow Fade A mounting media (Invitrogen, Carlsbad, CA).

Confocal images were acquired using an Olympus Fluoview 300 laser scanning microscope system equipped with a PlanApo ×60 oil-immersion objective. Multiple representative XY (parallel to the apical cell membrane) and XZ (vertical cross section) images were acquired from each monolayer. For apically sorted or unsorted receptors an XY slice through the apical membrane is shown, whereas for basolaterally sorted receptors an XY slice below the level of the apical membrane is shown. In both instances, the XZ image shows the relative receptor expression at the two membrane regions. The brightness and contrast of the resulting images were adjusted in Adobe Photoshop with the goal of highlighting membrane expression while minimizing background fluorescence. In most cases, autofluorescence of the polyester membrane has been removed for clarity. Representative images are shown for each receptor construct.

Polarized cell surface biotinylation.

MDCK(II) cells were seeded on 12-mm (6 × 10⁵ cells/well) or 24-mm (1.2 × 10⁶ cells/well) Transwell inserts and grown for 5–7 days with daily medium changes to allow the cells to form a polarized monolayer. A polarized biotinylation assay was used to quantify cell surface expression of HA-tagged receptors essentially as described previously (71). Briefly, cells were carefully cooled to 4°C and kept cold for the entire assay to avoid potential nucleotide release and redistribution due to receptor activation. Cells were washed twice in PBS++ (phosphate-buffered saline, pH 8.0, plus 2 mM CaCl₂ and 2 mM MgCl₂), then PBS++ containing 2 mg/ml Sulfo-NHS-SS-Biotin (Pierce, Rockford, IL) was applied to either the apical or the basolateral surface, and the reaction was allowed to proceed for 20 min. Following aspiration of the biotinylation solution, the cells were incubated with PBS++ containing 100 mM glycine (pH 8.0) for 10 min to quench the reaction and were then washed three times with PBS++.

Protein extracts were prepared by adding RIPA lysis buffer (50 mM Tris·HCl pH 8.0, 100 mM NaCl, 5 mM EDTA, 1% Triton X-100, 0.5% deoxycholate, and 0.1% SDS) to washed cells and passing the lysate through a 25-gauge needle 7–10 times to ensure complete disaggregation. Insoluble materials were removed by centrifugation at 13,000 g for 30 min, and the resulting supernatant was incubated for 90 min with 35 μl (for 12-mm inserts) or 50 μl (for 24-mm inserts) of immobilized NeutrAvidin resin (Pierce). The resin was washed twice with Tris-Triton buffer (50 mM Tris·HCl, pH 7.4, 100 mM NaCl, 5 mM EDTA, and 1% Triton X-100), and biotinylated proteins were released from the resin by incubation with 30 μl of 2× SDS-PAGE sample buffer containing 100 mM dithiothreitol for 10 min at 37°C. Eluted proteins were separated on a 10% SDS-PAGE gel and transferred to a PVDF membrane overnight at 4°C. HA-tagged proteins were detected with a peroxidase-conjugated monoclonal antibody (3F10, Roche Biochemicals, Indianapolis, IN) at 1:500 dilution and visualized with SuperSignal West Pico chemiluminescent substrate (Pierce). Bands were quantified from scanned images with ImageJ software (57).

RESULTS

Delimitation of the apical targeting signal of the P2Y₄ receptor.

We showed previously that the human P2Y₄ receptor is sorted to the apical membrane of three different epithelial cell types [MDCK(II), 16HBE14o⁻, and Caco-2 cells], and that the apical targeting sequence is located within the C-tail of the receptor (52, 71). To delimit the COOH-terminal end of the apical targeting signal, we constructed a series of HA-tagged human P2Y₄ receptors in which increasing amounts of the COOH terminus were truncated. The truncated receptors were expressed in MDCK(II) cells, and their steady-state localization was determined by confocal microscopy (Fig. 1). XZ scans of the full-length human P2Y₄ receptor and P2Y₄ receptors truncated at amino acids 355 or 343 revealed that these receptors were localized to the apical membrane at steady state, indicating that the signal remained intact (Fig. 1). In contrast, the P2Y₄ receptor truncated at amino acid 332 resulted in an unsorted phenotype. These data indicate that the COOH-terminal end of the apical targeting sequence is no further downstream than amino acid 343.

Fig. 1.

The P2Y₄ apical targeting signal ends before Asp343. A: HA-tagged wild-type and truncated P2Y₄ receptors with the C-tail sequences shown were stably expressed in MDCK(II) cells. B: subcellular localization of the receptor constructs in polarized monolayers was analyzed by confocal microscopy. The receptor tags are marked by green fluorescence while red fluorescence marks ZO-1, a tight junction protein. For all fluorescence micrographs in this study, the larger images are scans of the XY plane (parallel to the apical membrane) and the smaller images are XZ scans along the white line shown in the XY scan.

The C-tail of the P2Y₄ receptor confers apical targeting to the normally unsorted BK2 receptor when fused to the receptor just past TM7 (52). To delimit the NH₂-terminal end of the apical targeting sequence, we created chimeric receptors composed of the body of the BK2 receptor (NH₂ terminus through TM7) fused with progressively smaller portions of the P2Y₄ C-tail (Fig. 2). The longest of these constructs has a C-tail that starts at the putative beginning of the P2Y₄ C-tail (Asp311) and ends at the end of the apical targeting signal (Asp343) as determined in Fig. 1. The other two chimeric receptors have shorter lengths of the P2Y₄ C-tail fused to the body of the BK2 receptor, beginning at Cys321 and Lys325, respectively. Fusion of amino acids 311–343 or 321–343 of the P2Y₄ C-tail was sufficient to confer polarized sorting of the BK2 receptor to the apical surface of MDCK(II) cells (Fig. 2B), and the apical localization of these chimeric receptors revealed by confocal microscopy was confirmed by an established polarized cell surface biotinylation assay [Fig. 2C; (29, 52, 71)]. In contrast to the other chimeras, the BK2 receptor with the fewest number of amino acids (325–343) of the P2Y₄ C-tail was not sorted to the apical membrane. This receptor construct was presumably unstable and failed to reach the cell surface, as we were unable to pull down sufficient quantities of receptor to produce a visible band in our biotinylation assay. These data indicate that the NH₂-terminal end of the apical targeting signal of the P2Y₄ receptor is no farther upstream than amino acid 321.

Fig. 2.

The P2Y₄ apical targeting signal begins after C-321. A: HA-tagged BK2-P2Y₄ chimeric receptors with the indicated C-tail sequences were stably expressed in MDCK(II) cells. B: subcellular localization of the receptor constructs in polarized monolayers was analyzed by confocal microscopy. C: surface expression of the receptor constructs was determined by a polarized biotinylation assay (see methods). Quantification is based on densitometric analysis of apical (AP) and basolateral (BL) bands (n ≥ 3). Error bars represent the standard deviation of each data set. The BK2-Y₄CT 325–343 construct was not detected in our biotinylation assay, presumably because it does not reach the cell surface in sufficient quantities.

Taken together, these experiments defined a 23 amino acid sequence (Cys321 to Asp343) that is both necessary and sufficient to target the P2Y₄ or the BK2 receptor to the apical surface. This sequence is unique to the P2Y₄ receptor, bears no similarity to any known sorting signal, and does not contain any known binding motifs or conserved domains.

Identification of key amino acids in the P2Y₄ apical targeting sequence.

Because its length made single residue mutational analysis unlikely to yield useful information, we mutagenized amino acid groups (e.g., charged, polar, nonpolar, etc.) to discern their role in apical targeting (Fig. 3A) as we have done previously for the basolateral targeting sequence of the P2Y₁ receptor (70). We began by analyzing a P2Y₄ receptor construct in which all of the serine and threonine residues in the C-tail were mutated to alanine. Previous work had established that agonist-dependent phosphorylation of Ser333 and Ser334 within the apical targeting sequence is required for P2Y₄ receptor desensitization and internalization (2), which potentially could affect steady-state receptor distribution. However, these mutations had no effect on the polarized expression of the P2Y₄ receptor (Figs. 3 and 4).

Fig. 3.

Subcellular localization of receptor constructs with mutant P2Y₄ C-tails. A: mutant receptors with the C-tail sequences shown were stably expressed in MDCK(II) cells. The delimited apical targeting sequence is marked in green and the amino acid mutations of each construct are marked in red. B: subcellular localization of the receptor constructs in polarized monolayers was analyzed by confocal microscopy.

Fig. 4.

Surface expression of receptor constructs with mutant P2Y₄ C-tails. Mutant receptors (see Fig. 3A for specific sequences) were stably expressed in MDCK(II) cells. A: surface expression of the mutant receptor constructs was determined by polarized biotinylation. B: quantification is based on densitometric analysis of apical and basolateral bands (n ≥ 3). Error bars represent the standard deviation of each data set.

We next targeted the single cysteine residue (Cys321) in the apical targeting sequence, as palmitoylation of cysteines has been implicated in protein stability as well as trafficking (24, 36). We found that a C321S mutation had no effect on receptor localization of the BK2/P2Y₄ C-tail chimeric receptor (Figs. 3 and 4). Likewise, mutation of three consecutive glycines or the PQP triad near the beginning of the sequence to alanines in the context of the P2Y₄ receptor had no effect on apical targeting (Figs. 3 and 4).

Charged residues within the basolateral targeting sequence of the P2Y₁ receptor are critical to its function (70). To address the role of charged residues in the P2Y₄ receptor apical targeting signal, we mutated the acidic and/or basic residues of the apical targeting sequence to alanine. Mutation of the glutamate and aspartate residues to alanine had no effect on targeting, whereas mutation of lysine and arginine residues to alanine reduced apical polarization by 24% (61% apical vs. 85% apical for wild-type P2Y₄). Mutation of both the basic and acidic amino acids to alanine closely matched the results of mutating the basic residues alone.

We also made mutations to a small hydrophobic region in the latter half of the apical targeting sequence (one valine and three leucine residues within six residues). Mutation of these four amino acids to alanine markedly disrupted the apical targeting of the P2Y₄ receptor and in fact promoted pronounced (but not complete) basolateral targeting (63% basolateral). These data highlight the importance of these four hydrophobic residues for proper apical targeting of the P2Y₄ receptor.

The apical targeting signal of the P2Y₄ receptor is sequence independent.

We previously showed that the 25 amino acid basolateral targeting signal of the P2Y₁ receptor was functional either in its normal N→C direction or inverted in a C→N direction (i.e., where the last amino acid of the signal becomes the first amino acid in the inverted construct's signal sequence), suggesting that the basolateral targeting signal was sequence independent (70). To determine whether there is any sequence or directional specificity to the apical targeting sequence of the P2Y₄ receptor, the signal was inverted in the context of the full-length receptor and its localization was determined by confocal microscopy (Figs. 3 and 4). Surprisingly, the inverted signal was still able to direct complete apical targeting to the P2Y₄ receptor. This lack of sequence- and direction-specificity may indicate that only a few key amino acids within the sequence and/or a particular structure of the P2Y₄ receptor C-tail (which is preserved upon inversion) are necessary for proper polarization.

The P2Y₄ receptor apical targeting sequence redistributes a basolaterally targeted receptor.

There are numerous examples of membrane proteins that have more than one targeting signal, and these signals can be either redundant or opposing. For opposing signals, it has been generally assumed that when both apical and basolateral signals reside within the same protein, the basolateral signal is dominant (44). For example, we have shown that the P2Y₁ receptor contains both an apical signal in the main body of the receptor and a basolateral targeting signal in the C-tail, with the basolateral targeting signal completely dominant over the apical signal (70). However, dominance of a basolateral signal over an apical signal is not always observed (7).

To address whether the apical targeting signal of the P2Y₄ receptor can be dominant over a basolateral targeting signal, we fused the C-tail of the P2Y₄ receptor to the main body of the P2Y₁₂ receptor. Unpublished work in our lab has shown that the P2Y₁₂ receptor contains two redundant basolateral sorting signals; one in its C-tail, and one in the main body (NH₂ terminus to TM7) of the receptor. The HA-tagged P2Y₁₂ receptor truncated at the start of its C-tail is targeted to the basolateral membrane (P2Y₁₂ΔCT; 95% basolateral; Fig. 5). Fusion of the C-tail of the P2Y₄ receptor to the P2Y₁₂ receptor results in a protein that is mostly, but not completely, redirected to the apical surface of MDCK(II) cells (76% apical; Fig. 5).

Fig. 5.

The P2Y₄ apical targeting signal is dominant over a basolateral signal. The HA-tagged P2Y₁₂ receptor lacking its C-tail and P2Y₁₂-P2Y₄ C-tail chimera were stably expressed in MDCK(II) cells. A: subcellular localization of the receptor constructs in polarized monolayers was analyzed by confocal microscopy. B: surface expression of the receptor constructs was determined by polarized biotinylation. Quantification is based on densitometric analysis of apical and basolateral bands (n ≥ 3). Error bars represent the standard deviation of each data set.

DISCUSSION

Seven of the eight known P2Y receptors are sorted to distinct membrane surfaces when expressed in polarized epithelial cells, including the P2Y₄ receptor. We show here that the P2Y₄ receptor contains a 23 amino acid sequence in its C-tail that is both necessary and sufficient to direct apical targeting in MDCK(II) cells. The signal is autonomous, as it is functional even when fused to other GPCRs. The signal depends primarily on the presence of 4 hydrophobic residues near the COOH-terminal end of the sequence, and less so on basic amino acids. Other amino acids in this sequence appear to have little to no effect on apical targeting. Surprisingly, the apical targeting signal functions normally even when inverted and is capable of redirecting a basolaterally targeted GPCR to the apical membrane. The properties of this novel apical sorting signal have not been described for any apical signal identified to date.

Alignment of the COOH-terminal tail of the P2Y₄ receptor from a broad range of mammalian orthologs indicates that the most important residues of the signal (i.e., basic and hydrophobic residues) are conserved. For example, the four key hydrophobic residues are highly conserved, with only slight variations [L→R in Monodelphis domestica (opossum) and Sarcophilus harrisii (Tasmanian devil), and V→G in Ailuropoda melanoleuca (panda)] (Fig. 6). Conservation of the basic amino acids is not as strong as with the hydrophobic residues, but when they are not conserved, other nearby residues are changed to basic amino acids such that there is at least one basic amino acid within the NH₂-terminal region of the signal [the exceptions, Tupaia belangeri (tree shrew) and Oryctolagus cuniculus (European rabbit), have no basic residues within the same region]. This may indicate that the specific position of these amino acids is not critical, as long as positive charges are located nearby. Consistent with this idea, inversion of the targeting sequences in the P2Y₄ receptor (Fig. 3) and the P2Y₁ receptor (70), or scrambling the basolateral targeting sequence in the P2Y₁ receptor, had little to no effect on steady-state localization, indicating that the absolute positions of important residues are not critical and that these sorting signals function in a sequence-independent manner. Taken together, these data suggest that other orthologs of the P2Y₄ receptor are also trafficked to the apical membrane of epithelial cells in their respective hosts by a similar mechanism involving the C-tail of the receptor.

Fig. 6.

Conservation of the apical targeting sequence of the P2Y₄ receptor. Known mammalian P2Y₄ C-tail amino acid sequences are shown. Sequences were aligned using ClustalX 2.0.10 and shaded using Boxshade 3.21, both available on the World Wide Web (http://www.ch.embnet.org/). Residues that are conserved in >50% of the sequences are shaded black. Positions with similar residues in >50% of the sequences are shaded gray. The key hydrophobic amino acids for apical targeting are marked with arrows. Protein sequence accession numbers or ensembl.org identification numbers are as follows: Homo sapiens (human), P51582.1; Pan troglodytes (chimpanzee), XP_003317560.1; Pongo abelii (orangutan), XP_002834732.1; Gorilla gorilla (gorilla), ENSGGOP00000000615; Macaca mulatta (macaque), NP_001185632.1; Otolemur garnettii (bush baby), ENSOGAP00000021789; Callithrix jacchus (marmoset), XP_002807938.1; Sus scofra (pig), NP_001231097.1; Bos taurus (cow), NP_001243486.1; Canis lupus familiaris (dog), XP_003640305.1; Felis catus (cat), ENSFCAP00000007557; Sarcophilus harrisii (Tasmanian devil), ENSSHAP00000006770; Ailuropoda melanoleuca (panda), ENSAMEP00000021321; Rattus norvegicus (rat), NP_113868.1; Mus musculus (mouse), NP_065646.1; Oryctolagus cuniculus (rabbit), XP_002720124.1; Tupaia belangeri (tree shrew), ENSTBEP00000001285; Spermophilus tridecemlineatus (squirrel), ENSSTOP00000015075; Cricetulus griseus (Chinese hamster), XP_003509162.1; Pteropus vampyrus (megabat), ENSPVAP00000013535; Myotis lucifugus (microbat), ENSMLUP00000018666; and Monodelphis domestica (opossum), XP_001362915.1.

The apical sorting signal of the P2Y₄ receptor is a member of a large group of heterogeneous sequences of various lengths with little sequence homology, characteristic motifs, or known common-binding partners (17, 63). Several studies with three apically expressed sodium-dependent acid transporters suggest that a β-turn secondary structure motif may be important in apical sorting. A 14 amino acid sequence in the C-tail of the rat ileal sodium-dependent bile acid transporter (Abst) was shown to direct apical targeting, and NMR studies indicated that a 4 amino acid tetramer (NKGF) within this sequence formed a β-turn that was suggested to be important in apical targeting (61). Apical targeting of two other members of this family, the human sodium-dependent vitamin C transporter and the excitatory amino acid transporter-3, was shown to depend on similar sequences for apical sorting (5), and computer modeling suggested that these sequences also form β-turns (59). However, both potential phosphorylation sites within the apical signal of Abst and its interaction with the 16-kDa subunit c of the vacuole proton pump were also shown to be important in apical targeting, suggesting a more complex sorting mechanism than simple recognition of a secondary structure element (60, 61).

The mechanism of apical sorting directed by the apical signal in the P2Y₄ receptor appears to be distinct from that of the acid transporters. Hydrophobic and basic amino acids are the most important for proper apical targeting of the P2Y₄ receptor, and our experiments rule out a role of phosphorylation in this process. We have not ruled out the possibility that the P2Y₄ apical targeting sequence forms a β-turn (albeit with a different sequence of amino acids than the acid transporters) or other secondary structure that is selectively stabilized by lipids or proteins at the apical membrane, although modeling programs do not predict a β-turn in the region comprising the hydrophobic cluster (data not shown).

Four members of the G protein-coupled receptor family (including the P2Y₄ receptor) have been shown to contain apical sorting sequences within their C-tails. Even among receptors from the same superfamily, there is no apparent common mechanism for apical targeting. The final 32 amino acids of rhodopsin, the best characterized of the four apical signals, interact with the Tctex-1 dynein light chain, a minus-end microtubule motor, presumably directing rhodopsin to the apical membrane along the cell's cytoskeleton (62). The exact amino acids responsible for targeting rhodopsin are unknown, but there are no hydrophobic clusters or sequence similarity to P2Y₄. While we have shown that apical targeting of the P2Y₄ receptor is most dependent on hydrophobic residues, apical targeting of the metabotropic glutamate receptor 1b isoform is dependent on three basic residues (17). The C-tail of the serotonin 5-HT_1B receptor operates as an apical targeting signal through an as yet unknown mechanism (28). This sequence also appears to be distinct from the P2Y₄ receptor, but there is a group of two hydrophobic residues flanked by two charged residues.

Taken together, the diversity of cytoplasmic apical targeting signals may simply reflect a diverse set of adaptor proteins that may be resolved by common sorting mechanisms, as others have speculated (23). However, relatively few sorting signals and even fewer binding partners and mechanisms are currently known. It is possible that as more signals are discovered and characterized, more commonalities will begin to appear. What is emerging from our studies of the sorting signals in the P2Y₁ and P2Y₄ receptors is that there is a set of sorted proteins that do not interact with proteins in the sorting machinery in a manner that depends on specific protein-protein interactions; instead, these sorting proteins likely recognize a secondary structure or physiochemical state, but how this is achieved is not clear. If the apical targeting sequence does interact directly with another protein, it seems more likely that it would resemble the interaction of a GPCR with arrestin proteins, which are involved in receptor internalization (1, 39). Arrestin proteins bind to a broad range of different receptors in a manner that is dependent on charged residues (usually through phosphorylation) but is sequence independent (12).

The two P2Y receptors with the highest identity (52%), P2Y₂ and P2Y₄, are both located at the apical membrane, but their targeting sequences are markedly different (52). The P2Y₂ receptor has an apical targeting signal located in the first extracellular loop that is absent in the P2Y₄ receptor. In contrast, the COOH terminus of the P2Y₄ receptor directs apical targeting, whereas the COOH terminus of the P2Y₂ receptor does not contain any targeting information (52). This difference has important mechanistic implications. For example, once the P2Y₂ receptor is synthesized and inserted into the endoplasmic reticulum, the apical signal in the extracellular loop becomes inaccessible to cytosolic sorting proteins, whereas the apical signal in the COOH terminus of the P2Y₄ receptor is available. Given the distinct locations of the apical signals, it will be informative to examine the mechanistic similarities (and differences) by which these receptors reach steady state.

Curiously, many membrane proteins contain more than one targeting sequence, and these signals can be either redundant or opposing. In at least two proteins, mGluR1 and PMCA2, alternative splicing determines their distribution, indicating a functional and physiological reason for multiple targeting signals (6, 17). Within the P2Y family, we have shown that the P2Y₁ receptor contains a cryptic apical targeting sequence in addition to its dominant basolateral targeting sequence (70), and the P2Y₁₂ receptor contains redundant basolateral sequences (unpublished observations). It was thought that when opposing targeting sequences are present in the same protein, the basolateral signal is always dominant, as is the case for the P2Y₁ receptor. However, we show here that the P2Y₄ receptor apical targeting signal is able to override most (but not all) of the basolateral targeting encoded in the P2Y₁₂ receptor lacking its C-tail. These data are consistent with those from Chuang et al. (7), who showed that the apical targeting sequence in the C-tail of rhodopsin can overcome the basolateral signal in the single pass protein, CD7. Thus, both apical and basolateral targeting signals can overcome an opposing signal elsewhere in the protein, and likely depend on context and location.

Our studies of the P2Y₄ receptor have shown strong apical polarization in cell lines derived from kidney, lung, and colon epithelia, suggesting that the sorting mechanism for this receptor is common to all epithelial cells (71). Even so, it is possible that epithelial cells exist that do not sort the P2Y₄ receptor in the same manner or to the same extent as in MDCK(II) cells. Indeed, one group has shown P2Y₄ receptor-mediated functional responses at both the apical and basolateral membranes in a study of nucleotide-stimulated Cl⁻ secretion in mouse jejunum and colon (18). There are caveats to these experiments—the measurements were taken using only a single, saturating dose of UTP in a “leaky” membrane preparation—but they may demonstrate a variation in the location of the P2Y₄ receptor in different epithelial cells. In contrast, using P2Y₂ and P2Y₄ receptor knockout mice, members of this same group found no evidence of basolateral P2Y₄ receptor function in mouse colon when studying nucleotide-induced K⁺ secretion (43). Unfortunately, the lack of selective, high-affinity antibodies makes it impossible to detect the native P2Y₄ receptor directly.

Our studies have provided important information about the steady-state localization of P2Y receptors, but how they achieve steady state is not clear. Previous work showed that the P2Y₄ receptor undergoes internalization and recycling much faster than other P2Y receptors (2), but the lack of an effect on apical targeting from mutating the two Ser residues (Ser333 and Ser334) responsible for desensitization and internalization strongly suggests that internalization and recycling are not important in the establishment of the steady-state apical localization. It is further unclear when and where the sorting of P2Y receptors and most other polarized transmembrane proteins occurs. Recent improvements in fluorescence microscopy technology should make it possible to answer not only these questions of spatial regulation, but also the regulation of these proteins over time, and further experiments are warranted.

GRANTS

This work was supported by National Institutes of Health Grant R01 HL-71131 (to R. A. Nicholas).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

D.R.D., S.C.W., A.-D.Q., and I.N. performed the experiments; D.R.D., S.C.W., A.-D.Q., and I.N. analyzed the data; D.R.D., S.C.W., A.-D.Q., I.N., and R.A.N. interpreted the results of the experiments; D.R.D. and S.C.W. prepared the figures; D.R.D. drafted the manuscript; D.R.D. and R.A.N. edited and revised the manuscript; R.A.N. conception and design of the research; R.A.N. approved the final version of the manuscript.