The Carboxyl Terminus of the Human Calcium Receptor

The G-protein-coupled calcium receptor plays a key role in extracellular calcium homeostasis. To examine the role of the membrane-spanning domains and the ∼200-residue cytoplasmic carboxyl terminus of the calcium receptor in cell-surface expression and signal transduction, we transfected HEK-293 cells with a series of truncation and carboxyl-terminal missense mutants and analyzed expression by immunoblotting, glycosidase digestion, intact cell immunoassay, and extracellular calcium-stimulated phosphoinositide hydrolysis assay. Two truncation mutants terminating at residues 706 and 802 within the second and third intracellular loops, respectively, were not properly glycosylated, failed to reach the cell-surface, and showed no calcium response, indicating that mutant receptors with the full extracellular domain but only three or five transmembrane domains are improperly folded and/or processed. Truncation mutants terminating at residues 888 and 903 within the carboxyl terminus were equivalent to the wild type in all assays, whereas mutants truncated at residues 865 and 874 showed no response to calcium, despite only ∼25% reduction in cell-surface expression. Mutants with a full-length carboxyl terminus but with residues between positions 874 and 888 replaced with alanines showed either no (Ala875, Ala876, and Ala879) or significantly reduced (Ala881–Ala883) calcium response at levels of cell-surface expression equivalent to those of the wild-type receptor. These results indicate that deletion of the majority of the carboxyl terminus is compatible with normal processing, cell-surface expression, and signal transduction of the receptor. The truncation and alanine substitution mutants identify a small region between residues 874 and 888 critical for normal signal transduction by the receptor.

of G-protein-coupled receptors (GPCRs) (2). Extracellular calcium ion (Ca 2ϩ o ) activates the CaR, leading to activation of phospholipase C-␤ via the G q subfamily of G-proteins; this increases phosphoinositide (PI) hydrolysis, which in turn causes release of calcium from intracellular stores (1). CaR activation by Ca 2ϩ o is thought to control the rate of parathyroid hormone secretion from parathyroid cells and the rate of calcium reabsorption by the kidney (2). Thus, the CaR plays a critical role in the regulation of extracellular calcium homeostasis.
Although the CaR shares the GPCR-defining seven-transmembrane domain motif, it shows no sequence homology to GPCRs other than a unique subfamily composed of the metabotropic glutamate and ␥-aminobutyric acid type B receptors (1,3). The cloned bovine (1), human (4), rat (5,6), and rabbit (7) CaR cDNAs predict, in addition to the seven membrane-spanning domains, a very large (Ͼ600 residue) amino-terminal extracellular domain (ECD) and a large ϳ200-residue carboxylterminal tail. Unlike other members of the GPCR superfamily for which extensive mutagenesis studies have helped define regions and residues important for receptor synthesis and cellsurface expression, for ligand binding, and for G-protein coupling, relatively little is known about the structure-function relationships of the CaR. Several naturally occurring point mutations have been identified in the human CaR (hCaR) in subjects with autosomal dominant hypocalcemia and familial benign hypocalciuric hypercalcemia (FBHH) (2). A limited number of these putatively activating (autosomal dominant hypocalcemia) and inactivating (FBHH) mutations that occur primarily in the ECD have been expressed and tested for function in cell transfection studies (8,9). With the exception of a frameshift mutation (P747fs) involving the second extracellular loop (9) and an Alu insertion mutant at threonine 876 in the carboxyl terminus (10), presumptively inactivating truncation mutations involving the seven transmembrane domains or the carboxyl terminus of the hCaR have not been identified in FBHH.
Given the unique sequence and predicted topology of the hCaR and the dearth of naturally occurring mutations involving the transmembrane domains and the carboxyl terminus of the receptor, we sought to define the importance of these regions in receptor biosynthesis, cell-surface expression, and signal transduction. For this purpose, we created a series of mutants that truncate or alter the receptor sequence within the second and third intracellular loops and within the carboxylterminal tail. Our results indicate that all seven transmembrane helices may be necessary for proper folding, processing, and cell-surface expression of the hCaR and define a membrane-proximal portion of the carboxyl-terminal tail that contains residues critical for cell-surface expression and signal transduction.

EXPERIMENTAL PROCEDURES
Cloning hCaR cDNA from Human Pituitary-The entire coding region of the hCaR was amplified by polymerase chain reaction (PCR) * 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  from a pituitary Marathon cDNA library (Promega, Madison, WI) using specific primers (nucleotides Ϫ14 to 6 (5Ј-sense primer) and nucleotides 3256 -3277 (3Ј-antisense primer) of the hCaR cDNA) (4) and the Advantage KlenTaq polymerase mixture (CLONTECH). The PCR-amplified product was cloned into the pCR3.1 mammalian expression vector (Invitrogen, San Diego, CA) as a HindIII-XhoI fragment and sequenced completely by automated DNA sequencing using a Taq DyeDeoxy Terminator Cycle sequencing kit (Applied Biosystems, Foster City, CA) and an ABI Prizm-377 DNA sequencer, and some regions were also confirmed by the dideoxy chain termination method using a Sequenase 2.1 kit (U. S. Biochemical Corp.). The DNA sequence showed no differences from the published sequence of the hCaR cloned from a parathyroid cDNA library (4) (GenBank™/EMBL Data Bank accession number U20759).
Site-directed Mutagenesis of hCaR-Site-directed mutagenesis of the hCaR cDNA was performed using restriction fragment replacement "cassette mutagenesis" (11). For cassette mutagenesis, a 3.3-kilobase hCaR coding fragment was cut out from the pCR3.1 expression vector and recloned as a HindIII-XbaI fragment in the pBluescript KS II ϩ vector (Stratagene, La Jolla, CA) lacking NsiI, SphI, and XbaI restriction enzyme sites. Truncation mutant hCaRs are shown schematically in Fig. 1. Truncation mutants are designated by a "T" followed by the codon number corresponding to the first mutant stop codon. T706 was prepared by replacing a 356-base pair NsiI-SphI restriction fragment of the hCaR with a synthetic duplex constructed by PCR mutagenesis that substituted sequences coding for 128 amino acid residues with two stop codons at amino acids 706 and 707. T802 was generated by replacing the same NsiI-SphI restriction fragment with a synthetic duplex containing stop codons at amino acids 802 and 803. T888 and T903 were prepared by replacing a 605-base pair SmaI-XhoI restriction fragment encoding carboxyl-terminal amino acid residues 886 -1078 with two separate PCR-generated duplexes with stop codons introduced at amino acids 888 and 889 for T888 or at amino acids 903 and 904 for T903. Site-specific mutants were prepared by replacing a 224-base pair SphI-XbaI restriction fragment encoding amino acids 770 -886. For the Ala 875 -Ala 879 mutant, codons for Ser 875 , Thr 876 , and His 879 were replaced with codons for alanine. All the changes were included in the 3Ј-primer with the XbaI site at the end. For the Ala 881 -Ala 883 mutant, codons for Phe 881 , Lys 882 , and Val 883 were changed to codons for alanine. The changes were included in the 3Ј-primer with XbaI site at the end. All these 3Ј-mutant primers were used separately with the 5Јprimer containing wild-type receptor nucleotide sequence with the SphI site to PCR-amplify the 224-base pair SphI-XbaI fragment of the hCaR. The PCR fragments were gel-purified, digested with SphI and XbaI, and ligated with the SphI-XbaI-digested hCaR/pBluescript KS II ϩ DNA. All cloned synthetic sequences were confirmed by automated DNA sequencing using a Taq DyeDeoxy Terminator Cycle sequencing kit and an ABI Prizm-377 DNA sequencer. A clone for each mutant with correct DNA sequence was transferred into the eukaryotic expression vector pCR3.1 as a HindIII-XhoI restriction fragment.
Transient Expression of Wild-type and Mutant hCaRs in HEK-293 Cells-Receptor cDNAs in pCR3.1 were prepared with a QIAGEN maxiplasmid DNA preparation kit (QIAGEN Inc., Chatsworth, CA) and were introduced into HEK-293 cells by the LipofectAMINE transfection method (Life Technologies Inc.). The quality and quantity of plasmid DNA were electrophoretically assessed on 0.8% agarose gel after measuring absorbance at 260 nm. For transfection, a given amount of the plasmid DNA was diluted in Dulbecco's modified Eagle's medium (DMEM) (Biofluids Inc., Rockville, MD) and mixed with diluted Lipo-fectAMINE, and the mixture was incubated at room temperature for 30 min. The DNA-LipofectAMINE complex was further diluted in serumfree DMEM and added to 80 -90% confluent HEK-293 cells plated on 24-well plates using 0.5 or 1 g of DNA/well (for phosphoinositide hydrolysis assay), on 24-well plates using 0.5 or 2 g of DNA (for intact cell enzyme-linked immunoassay), and in 75-mm flasks using 5 or 10 g of DNA (for immunoblot analysis and intact cell enzyme-linked immunoassay). After 4 -5 h of incubation, an equal volume of DMEM ϩ 20% fetal bovine serum (Biofluids Inc.) was added, and the media were replaced 24 h after transfection with fresh DMEM ϩ 10% fetal bovine serum. Membrane protein extraction for immunoblotting, intact cell enzyme-linked immunoassay, or PI hydrolysis assay was performed 48 h after transfection.
Preparation of Membranes from Transfected Cells-Confluent cells in 75-mm flasks were rinsed with phosphate-buffered saline (PBS) and detached with 0.5 mM EDTA in PBS at 37°C. The detached cells were pelleted and suspended in buffer A containing 5 mM Tris (pH 6.8), 2 mM EDTA, and protease inhibitors (5 g/ml leupeptin, 0.7 g/ml pepstatin, 5 g/ml benzamidine, and 1 mM phenylmethysulfonyl fluoride) at 4°C. The cells were forced through a 22-gauge needle five to eight times, and the lysate was spun in a TLA-100 centrifuge at 55,000 rpm for 20 min at 4°C to collect the crude membrane pellet. The pellet was resuspended in buffer B containing 20 mM Tris-HCl (pH 6.8), 150 mM NaCl, 10 mM EDTA, 1 mM EGTA, and 1% Triton X-100 with freshly added protease inhibitors (5 g/ml leupeptin, 0.7 g/ml pepstatin, 5 g/ml benzamidine, and 1 mM phenylmethysulfonyl fluoride). Immunoblot Analysis-The protein content of each sample was determined by the modified Bradford method (Bio-Rad), and 5-20 g of protein/lane was separated on 5-15% gradient SDS-polyacrylamide gel (12). The proteins on the gel were electrotransferred to nitrocellulose membrane and incubated with 0.1 g/ml protein A-purified monoclonal anti-hCaR antibody ADD (raised against a synthetic peptide corresponding to residues 214 -235 of the hCaR protein). Subsequently, the membrane was incubated with a secondary goat anti-mouse antibody conjugated to horseradish peroxidase (Kirkegaard & Perry Laboratories, Inc., Gaithersburg, MD) at a dilution of 1:5000. The hCaR protein was detected with an enhanced chemiluminescence system (ECL, Amersham Corp.).

Treatment of Membranes from Transfected Cells with Glycosidases-
Intact Cell Enzyme-linked Immunoassay to Determine Cell-surface Expression-To compare cell-surface expression of receptors in transiently transfected HEK-293 cells, 48 h after transfection, confluent transfected HEK-293 cells on 24-well plates were detached with 1 mM EDTA in PBS containing 0.5% bovine serum albumin. Cells were incubated with 0.5 ml of DMEM containing 10% fetal bovine serum and 1 g/ml monoclonal anti-hCaR antibody 7F8 (raised against the purified extracellular domain of hCaR) 2 at 4°C for 1 h. Following incubation, cells were pelleted by low-speed centrifugation (1000 rpm, 4°C) and washed three times with PBS. Cells were then incubated with 5 g/ml peroxidase-conjugated goat anti-mouse IgG␥ (Kirkegaard & Perry Laboratories, Inc.) in DMEM containing 10% fetal bovine serum for 1 h at 4°C. After three washes with PBS, peroxidase substrate was added to each sample, and the color reaction was followed for 5-10 min; absorbance was measured at 405 nm using a Thermo max microtiter plate reader (Molecular Devices, Menlo Park, CA).
Phosphoinositide Hydrolysis Assay-Phosphoinositide hydrolysis was assayed as described (13,14). Briefly, cells were incubated with 3.0 Ci/ml [ 3 H]myoinositol (NEN Life Science Products) in DMEM for 16 -24 h, followed by a 30-min preincubation with PI buffer (120 mM NaCl, 0.5 mM CaCl 2 , 5 mM KCl, 5.6 mM glucose, 0.4 mM MgCl 2 , and 20 mM LiCl in 25 mM PIPES (pH 7.2)). After removal of PI buffer, cells were incubated for an additional 30 min with different concentrations of Ca 2ϩ o in PI buffer. The reactions were terminated by addition of 1 ml of acid methanol (167 l of HCl in 120 ml of methanol). Total inositol phosphates were purified by chromatography on Dowex 1-X8 columns.
Statistical Analysis-The mean EC 50 for each wild-type and mutant receptor in response to [Ca 2ϩ ] o was calculated from EC 50 values from all individual experiments and expressed as the mean Ϯ S.E. Comparison of EC 50 values was performed using analysis of variance to determine significance.

Immunoblot Analysis of Expression of Truncation Mutant
hCaRs-Six truncation mutant constructs of the hCaR (T706, T802, T865, T874, T888, and T903) were prepared as described under "Experimental Procedures." The sites of the six truncation mutants are schematically shown in Fig. 1. The wild-type hCaR cDNA and each of the mutant constructs were transiently transfected into HEK-293 cells, and expression was assessed by immunoblotting using a monoclonal antibody (ADD) raised against a peptide epitope from the extracellular domain of the receptor that was not disrupted by the truncations (Fig. 2). Membrane preparations of cells transfected with vector containing the wild-type hCaR cDNA showed two major bands of ϳ150 and 130 kDa and a higher molecular mass band (Ͼ197-kDa marker) that may represent dimeric forms and/or aggregates of the receptor as previously reported (8,13). Immunoblots done with ADD of membranes from HEK-293 cells transfected with vector lacking receptor cDNA showed no immunoreactivity (data not shown, but see Ref. 13). The relative intensity of the 150-and 130-kDa bands and the appearance of a smaller band beneath the 115-kDa marker varied between different immunoblot experiments. Unlike the wild-type receptor, the T706 and T802 mutants each showed a single major band between the 89-and 98-kDa marker bands on immunoblots, whereas the carboxyl-terminal tail truncation mutants T865, T874, T888, and T903 each showed two major bands between the 89-and 115-kDa marker bands. Higher molecular mass immunoreactivity was seen for each of the mutants as well. As expected, each truncation mutant showed a reduction in molecular size compared with the wild-type hCaR that was, in general, proportionate to the differences in length generated by truncations. The T865, T874, and T888 mutants, however, all showed comparable migration on SDS-polyacrylamide gel, whereas T903 ran at a slightly higher molecular mass. The two shortest carboxyl-terminal tail truncation mutants, T865 and T874, showed relatively fainter bands compared with the wildtype hCaR when equal quantities of membranes were loaded on each lane in three different immunoblot experiments; this may indicate a lower level of synthesis and expression of these mutant receptors.
Deglycosylation of Expressed Mutant hCaRs with PNGase F and Endo-H-The wild-type hCaR has been shown to undergo N-glycosylation (8,13). To evaluate whether the truncation mutants undergo this post-translational modification, a series of deglycosylation experiments were conducted. The enzyme PNGase F cleaves N-linked carbohydrates (both intermediate high mannose forms and fully processed forms) from glycoproteins (15). In contrast, Endo-H cleaves only high mannose forms (16). Membrane preparations of cells transfected with wild-type and truncation mutant receptors were treated with PNGase F and Endo-H, and immunoblotting was performed to visualize the products of digestion. As shown in Fig. 3, PNGase F treatment of membranes from HEK-293 cells transfected with the wild-type hCaR cDNA causes the disappearance of the 150-kDa band and a reduction in size of the 130-kDa band. In contrast, Endo-H digestion causes a minimal alteration of the 150-kDa band while, like PNGase F, reducing the size of the 130-kDa band. Unlike for the wild-type hCaR, in membranes expressing the T706 and T802 mutants, both PNGase F and Endo-H digestion reduced the size of the expressed band. The products of digestion corresponded roughly to the sizes predicted, 75 and 85 kDa, respectively, based on primary amino acid length (Fig. 3). For membranes expressing the carboxylterminal tail truncation mutants T865, T874, T888, and T903, which displayed a doublet pattern of immunoreactivity (Fig. 2), digestion with PNGase F caused a decrease in the size of both bands, but Endo-H digestion reduced only the size of the lower band (Fig. 4). A series of faster migrating bands seen after PNGase F digestion of T865, T874, and T888 may reflect incomplete digestion (Fig. 4). Intact Cell Enzyme-linked Immunoassay to Assess Cell-surface Expression Level of Mutant hCaRs-Expression of the extracellular N-terminal domain of the hCaR should not be disrupted in any of the truncation mutants we made. We therefore took advantage of a monoclonal antibody that binds to the extracellular domain of the native receptor expressed on intact cells to quantitate cell-surface expression of the mutant recep- tors (Fig. 5). The T706 and T802 mutants failed to show any significant cell-surface expression (absorbance values no different from that for the vector control). The absorbance values for the T865 and T874 mutants were lower than that for the wild-type hCaR, indicating a lower level of expression of these mutants at the cell surface. In contrast, cell-surface expression of the T888 mutant was comparable to that of the wild-type hCaR and that of the T903 mutant was consistently higher than that of the wild-type hCaR using equal amounts of DNA for transfection.

Analysis of Signal Transduction by Truncation Mutant Re-ceptors Using Phosphoinositide Hydrolysis Assay-The CaR responds to increases in [Ca 2ϩ
] o in the millimolar range by activating G q family proteins to stimulate phospholipase C-␤ activity with resultant PI hydrolysis. We (13) and others (17) have used this assay in transfected HEK-293 cells to measure signal transduction by expressed wild-type and mutant CaRs. Two intracellular loop truncation mutants (T706 and T803) showed greatly attenuated or absolutely no PI response even at 20 mM Ca 2ϩ o (data not shown). Likewise, the two shortest carboxyl-terminal tail truncation mutants (T865 and T874) showed no detectable PI hydrolysis response to [Ca 2ϩ ] o (Fig. 6). In contrast, the T888 and T903 mutants both showed PI hydrolysis responses to [Ca 2ϩ ] o similar to that of the wild-type hCaR. Both mutants showed similar dose-response curves with EC 50 values of 3.7 Ϯ 0.18 and 3.0 Ϯ 0.53, respectively (Fig. 6).

Analysis of Cell-surface Expression and Signal Transduction by Ala 875 -Ala 879 and Ala 881 -Ala 883
Mutant hCaRs-Since the T888 mutant behaved comparably to the wild-type hCaR, whereas the T874 mutant showed no PI hydrolysis response to [Ca 2ϩ ] o , we sought to evaluate the role of the 14 amino acid residues between T874 and T888 (Table I). We generated two site-directed mutants in which alanine residues were substituted for residues 875, 876, and 879 (Ala 875 -Ala 879 mutant; residues 877 and 878 are already alanines in the wild-type receptor) and for residues 881, 882, and 883 (Ala 881 -Ala 883 mutant; residues 884, 885, and 887 are alanines in the wildtype receptor) in the context of the full-length receptor. Because we sought to study the signaling properties of these receptors at similar expression levels, we first determined the levels of cell-surface expression for each mutant receptor after transfecting HEK-293 cells with varying amounts of plasmid DNA (0.5-2 g). Results of the enzyme-linked immunoassay showed that cell-surface expression differed among the mutant and wild-type receptors when the same amount of plasmid DNA was used for transfection (Fig. 7). Similar levels of expression were achieved by transfecting HEK-293 cells with half the amount of wild-type hCaR cDNA compared with mutant and with peroxidase-conjugated anti-mouse immunoglobin secondary antibody as described under "Experimental procedures." Peroxidase substrate was added, and absorbance was measured at 405 nm using a Thermo max microtiter plate reader. The absorbance (optical density (O.D.)) readings are expressed as a percent of the value for the wild-type hCaR (set at 100%). Data are from one experiment done in triplicate. Two independent experiments gave similar results. cDNA as shown in Fig. 7. Next, we compared the [Ca 2ϩ ] o response of the alanine mutant receptors under conditions in which cell-surface expression was equivalent to that of the wild-type hCaR using the PI hydrolysis assay (Fig. 8). The Ala 875 -Ala 879 mutant under transfection conditions in which cell-surface expression was equivalent to that of the wild-type hCaR showed no dose-dependent stimulation up to 20 mM Ca 2ϩ o . The Ala 881 -Ala 883 mutant showed a 60 -70% decrease in maximal response compared with the wild-type receptor and a right-shifted concentration-dependent increase in [Ca 2ϩ ] o response with an EC 50 value of 3.8 Ϯ 0.8.

DISCUSSION
This study provides new information on the role of membrane-spanning domains and the carboxyl-terminal tail in hCaR processing, cell-surface expression, and signal transduction. We created two truncation mutants, T706 (within the second intracellular loop) and T802 (within the third intracellular loop), predicted to generate receptor proteins with the full ECD and three (T706) or five (T802) transmembrane domains. Upon transfection in HEK-293 cells, immunoblotting of cell membranes showed that both T706 and T802 are expressed and glycosylated. Unlike transfection with wild-type receptor cDNA, which yields fully processed (Endo-H-resistant) and incompletely processed (Endo-H-sensitive) glycosylated proteins, T706 and T802 are expressed primarily as incompletely processed proteins. The latter form of the protein is thought to represent a high mannose form that fails to transit properly from the endoplasmic reticulum to the Golgi apparatus (8,16). Failure of the T706 and T802 mutant receptors to reach the cell surface as shown by intact cell immunoassay is consistent with defective processing and trafficking. A P747fs mutation (predicted to create a truncation at residue 775 at the start of the fifth transmembrane domain) identified in a subject with FBHH yielded a single major band on immunoblotting and showed no calcium response when expressed in HEK-293 cells (9). Although Endo-H sensitivity and cell-surface expression were not tested, the results for this mutant are consistent with those obtained even with the slightly longer T802 mutant. In contrast, all carboxyl-terminal tail truncation mutants we studied, including T865, which contains all seven membranespanning domains but only 2 additional amino acids of the carboxyl-terminal tail, are glycosylated and processed comparably to the wild-type receptor and are expressed at the cell surface, although for T865 and T874, at somewhat lower levels compared with the wild type.
Our results indicate that for the CaR, up to five membranespanning domains are insufficient, whereas seven membranespanning domains are sufficient for normal processing and cell-surface expression. It is likely that mutant hCaRs lacking a full seven membrane-spanning domains are misfolded or lack FIG. 6. Concentration dependence for calcium stimulation of phosphoinositide hydrolysis in transiently transfected HEK-293 cells expressing wild-type and carboxyl-terminal tail truncation mutant hCaRs. Transfections were performed, and calcium-stimulated phosphoinositide hydrolysis was assayed as described under "Experimental Procedures." Each data point is the mean value of three independent experiments. For each experiment, calcium concentrations were assayed in triplicate. Results (counts/min of labeled inositol phosphates generated) are expressed as a percent of the maximal response of the wild-type hCaR (WT). The mean EC 50 for each curve is presented as mean Ϯ S.E. The mean EC 50 values for the wild-type and mutant hCaRs were not significantly different (p Ն 0.05). NR, no response.  critical structural determinants for appropriate processing and trafficking. Although we did not explicitly measure stability of expressed receptors, lower levels of immunoreactive protein observed for the T706 and T802 mutants suggest that the mutant proteins may be degraded more rapidly than the wildtype receptor. Lower levels of expression of the T865 and T874 mutants and higher levels of expression of the T903 mutant compared with the wild type could also reflect differences in protein stability, but this will require verification by pulsechase and other studies of receptor protein turnover.
Although the CaR differs markedly in overall amino acid sequence and in size of the ECD and the carboxyl terminus from most other GPCRs, our results demonstrating a requirement for more than five membrane-spanning domains for normal receptor processing and cell-surface expression are similar to results obtained with many other GPCRs. For example, glucagon receptor mutants truncated in the second or third intracellular loops were glycosylated but incompletely processed as shown by Endo H sensitivity and failed to express at the cell surface; in contrast, carboxyl-terminal tail truncation mutants of the same receptor showed normal processing and cell-surface expression (16). Rhodopsin mutants truncated before the end of the second membrane-spanning domain failed to undergo glycosylation at all; truncation mutants with the first three or five membrane-spanning domains underwent glycosylation, but were incompletely processed and failed to exit the endoplasmic reticulum as determined by a variety of criteria (18).
The effects of carboxyl-terminal tail truncation on signal transduction differ in different GPCRs. For some, such as the glucagon receptor, most of the carboxyl-terminal tail can be truncated without reducing agonist-stimulated second messenger production (16). For others, e.g. the CCR5 chemokine re-ceptor, truncation of the carboxyl-terminal tail from 52 to either 8 or 20 residues did not impair cell-surface expression, but for the shorter mutant, abolished signal transduction response (19). For the metabotropic glutamate receptors, which are more closely related to the CaR than other GPCRs, studies of chimeric receptors have defined a role for the carboxyl-terminal tail in determining specificity of G-protein coupling (20). A preliminary study of carboxyl-terminal tail truncation mutant hCaRs expressed in Xenopus oocytes showed that truncation at residue 896 or 922 did not affect the response to [Ca 2ϩ ] o , whereas truncation at residue 867 abolished the response (21). An Alu insertion mutation identified in a kindred with FBHH results in a CaR with missense sequence and eventual truncation after residue 876. This mutant showed decreased cellsurface expression (measured by immunocytochemistry with an epitope tag-specific antibody) and was unresponsive to [Ca 2ϩ ] o ; an artificially created Ala 877 stop mutant also failed to respond to [Ca 2ϩ ] o , but showed increased cell-surface expression compared with the wild-type receptor (10).
Our results are consistent with these limited studies of carboxyl-terminal tail truncation mutant hCaRs and help to define more clearly a region of the carboxyl-terminal tail critical for normal signal transduction. While all of the carboxyl-terminal tail truncation mutant hCaRs we made are expressed at the cell surface, these mutants differ significantly in their ability to mediate increased PI hydrolysis in response to [Ca 2ϩ ] o . The T888 and T903 mutants, lacking 192 and 177 carboxyl-terminal residues, respectively, show responses to [Ca 2ϩ ] o equivalent to that of the wild-type receptor. This indicates that most of the hCaR carboxyl-terminal tail is dispensable for a normal signal transduction response in transiently transfected HEK-293 cells. While the amino acid sequence of the CaR is extraordinarily conserved (Ͼ92%) among human (4), bovine (1), rat (5,6), and rabbit (7), the region of greatest sequence divergence is the carboxyl-terminal tail from approximately residue 920 to the end. Lack of sequence conservation of this region is consistent with lack of a critical role in signal transduction response. Our results, however, should not be interpreted to indicate that this region of the hCaR has no biologic function. This region contains several consensus sites for phosphorylation by protein kinase C (4) and could play a role in receptor desensitization and/or internalization.
Interestingly, the T903 mutant showed higher expression than the wild-type receptor, suggesting the possibility that determinants in the truncated carboxyl-terminal tail could impede receptor expression or increase degradation. Indeed, Bai et al. (10) found that C-terminal truncation at residue 877 leads to increased receptor expression compared with the wild type. Their result contrasts with our finding that T874 mutant cellsurface expression is slightly reduced compared with and that T888 mutant cell-surface expression is comparable to that of the wild type. We cannot definitively explain this difference between our results and those of Bai et al., but they may relate to the differing methods used to assess cell-surface expression. Bai et al. employed immunocytochemistry using a monoclonal antibody against an epitope tag inserted in the receptor extracellular domain (10), whereas we used a more quantitative method, enzyme-linked immunosorbent assay, with a monoclonal antibody against an epitope within the endogenous extracellular domain.
Recently, the extreme carboxyl-terminal tail of the related metabotropic glutamate receptors (types 1 and 5) has been shown to interact specifically with a protein termed "homer" that serves to localize the receptor to discrete areas of neuronal membranes (22). It is conceivable that a similar protein interacts with the CaR carboxyl-terminal tail to localize this recep- In 24-well plates, HEK-293 cells were transfected with either 0.5 g of wild-type hCaR (WT)/well or 1 g of mutant plasmid DNA/well. As determined by intact cell immunoperoxidase assay (see Fig. 7), similar levels of receptors in transfected cells were achieved when HEK-293 cells were transfected with 0.5 g of wild-type or 1 g of each mutant hCaR plasmid DNA/well, respectively. Each data value shown here is the mean Ϯ S.E. of three different experiments each done in triplicate at each calcium concentration. All responses are normalized to the maximal response of the wild-type hCaR. The mean EC 50 values for the wild-type and mutant hCaRs were not significantly different (p Ն 0.05). NR, no response. tor within specific areas of neurons (6) or other cells in which the receptor is expressed. This and other hypotheses concerning the function of the carboxyl terminus can now be tested by appropriate studies comparing wild-type and carboxyl-terminal tail truncation mutant CaRs.
A comparison of the [Ca 2ϩ ] o response of the T874 and T888 mutants indicates that residues between positions 874 and 888 are critical for a normal signal tranduction response. At present, lack of a high-affinity ligand binding assay for the CaR prevents us from determining whether loss of signal transduction response reflects a defect in agonist binding, receptor activation, or G-protein coupling. Since the ECD, which has been suggested to be the site of agonist binding in the CaR (2), is predicted to be intact in the T874 mutant, it is tempting to speculate that failure to respond to [Ca 2ϩ ] o is due to defective G-protein coupling. That loss of responsiveness does not simply reflect a requirement for a minimum length of carboxyl-terminal tail is shown by the Ala 875 -Ala 879 mutant (substituting alanines for Ser 875 , Thr 876 , and His 879 ), which shows no response despite having a full-length carboxyl-terminal tail. The Ala 881 -Ala 883 mutant (substituting alanines for Phe 881 , Lys 882 , and Val 883 ) shows significantly reduced response compared with the wild type. Differences in signal transduction response by the Ala 875 -Ala 879 and Ala 881 -Ala 883 mutants do not reflect differences in cell-surface expression compared with the wild type since transfections were performed with amounts of plasmid DNA shown to give equivalent cell-surface expression. Since both the Ala 875 -Ala 879 and Ala 881 -Ala 883 mutants show significantly reduced signal transduction response, several residues between positions 874 and 888 are likely to contribute to normal signal transduction responsiveness. Further studies are needed to define the precise contribution of individual residues in this critical region and to determine their specific structural and functional roles.