(Received for publication, November 11, 1996, and in revised form, January 2, 1997)
From the Endocrine Unit, Massachusetts General Hospital and Department of Medicine, Harvard Medical School, Boston, Massachusetts 02114
To define the structural requirements of the
parathyroid hormone (PTH)/PTH-related protein (PTHrP) receptor
necessary for activation of phospholipase C (PLC), receptors with
random mutations in their second cytoplasmic loop were synthesized, and
their properties were assessed. A mutant in which the wild type (WT)
rat PTH/PTHrP receptor sequence EKKY (amino acids 317-320) was
replaced with DSEL had little or no PTH-stimulated PLC activity when
expressed transiently in COS-7 cells, but it retained full capacity to
bind ligand and to generate cAMP. This phenotype was confirmed in
LLC-PK1 cells stably expressing the DSEL mutant receptor, where both
PTH-stimulated PLC activity and sodium-dependent phosphate
co-transport were essentially abolished. Individual mutations of these
four residues point to a critical role for Lys-319 in receptor-G
protein coupling. PTH-generated IPs were reduced to 27 ± 13%
when K319E, compared with the WT receptor, and PLC activation was fully
recovered in a receptor revertant in which Glu-319 in the DSEL mutant
cassette was restored to the WT residue, Lys. Moreover, the WT receptor and a mutant receptor in which K319R had indistinguishable properties, thus suggesting that a basic amino acid at this position may be important for PLC activation. All of these receptors had unimpaired capacity to bind ligand and to generate cAMP. To ensure adequacy of
Gq-subunits for transducing the receptor signal,
G
q was expressed in HEK293 and in LLC-PK1 cells together
with either WT receptors or receptors with the DSEL mutant cassette.
PTH generated no inositol phosphates (IPs) in either HEK293 or LLC-PK1
cells, when they expressed DSEL mutant receptors together with
G
q. In contrast, PTH generated 2- and 2.5-fold increases
in IPs, respectively, when these cells co-expressed both the WT
receptor and G
q. Thus, generation of IPs by the
activated PTH/PTHrP receptor can be selectively abolished without
affecting its capacity to generate cAMP, and Lys-319 in the second
intracellular loop is critical for activating the PLC pathway.
Moreover,
-subunits of the Gq family, rather than
-subunits, transduce the signal from the activated receptor to
PLC, and the PLC, rather than the adenylyl cyclase, pathway mediates
sodium-dependent phosphate co-transport in LLC-PK1
cells.
Signals from the parathyroid hormone (PTH)1/PTH-related peptide (PTHrP) receptor are transduced by G proteins. The lack of sequence homology between PTH/PTHrP receptors and all but a few of the other G protein-coupled receptors, however, justifies classifying them as members of a distinct family (1), which includes two mammalian receptors each for PTH or PTHrP (2, 3), vasoactive intestinal peptides (4, 5), and corticotrophin-releasing hormone (6), and receptors for secretin (7), calcitonin (8), glucagon-like peptide 1 (9), growth hormone-releasing hormone (10), glucagon (11), pituitary adenylyl cyclase-activating peptide (12), gastric inhibitory peptide (13), and an orphan receptor in brain similar to the calcitonin receptor (14). Receptors from this family are not limited to vertebrates, as two insect diuretic hormone receptors (15) and a partial genomic clone from Caenorhabditis elegans (16) also are homologous. Rat, opossum, and human PTH/PTHrP receptors activate multiple signal transduction pathways that include both adenylyl cyclase and phospholipase C (PLC) (1, 17); they bind two ligands, PTH and PTHrP, with nearly equal efficacy (1, 2), and they are widely distributed in various tissues including the major PTH targets, kidney and bone (18, 19). Little, however, is known concerning the structural features of PTH/PTHrP receptors that enable them to activate these multiple effectors.
We previously showed that the rat PTH/PTHrP receptor with truncations of its C-terminal intracellular tail had markedly enhanced capacity to increase intracellular cAMP, but unaltered capacity to generate inositol phosphates (IPs), when expressed at levels comparable with the WT receptor (20). These observations extended those of Nissenson et al. (21), who showed that the PTH/PTHrP receptor's C-terminal tail could be deleted or mutated without major loss in its capacity to increase intracellular cAMP. PTH-stimulated increases in cAMP and IPs are both dependent on the level of receptors expressed, but we showed that adenylyl cyclase responsiveness is elicited at lower PTH concentrations and when fewer receptors are expressed, compared with the PLC response (20, 22).
To determine portions of the PTH/PTHrP receptor important for activation of PLC, we first constructed receptors with random clustered mutations in the putative second intracellular (2i) loop of the rat PTH/PTHrP receptor and screened them for ligand binding and adenylyl cyclase stimulating properties. Then, mutant receptors whose maximal radioligand binding and maximal PTH-stimulated cAMP response were at least 50% as great as those of the WT receptor were further characterized for their capacity to stimulate PLC, as assessed by measuring PTH-stimulated accumulation of IPs. This initial screening identified a receptor containing a random cassette mutation, DSEL (amino acids 317-320) in the 2i loop rather than the WT sequence EKKY, which had little or no capacity to stimulate PLC, but its capacity to bind the ligand and to stimulate adenylyl cyclase was unimpaired.
We then systematically examined which amino acid in this 2i loop
sequence accounted for PLC activation by testing the functional properties of PTH/PTHrP receptors with single- and double-point mutations of the WT sequence, EKKY, and single-point "revertants" of receptors with DSEL mutant cassette. The studies reported here highlight the critical role of a portion of the receptor's 2i loop,
especially of lysine at position 319, for selective coupling of the
PTH/PTHrP receptor to member(s) of the Gq family of
-subunits that results in PLC activation. Furthermore, we extend our
previous observations to show that sodium-dependent
phosphate co-transport in LLC-PK1 cells that stably express PTH/PTHrP
receptors is independent of cAMP and that it is mediated through a PLC
pathway.
Rat PTH-(1-34)-NH2 (rPTH), human
[Tyr36]PTHrP-(1-36)-NH2 (PTHrP), and
oligonucleotides were synthesized by Dr. H. T. Keutmann, Endocrine
Unit, Massachusetts General Hospital (Boston). Reagents of the highest
purity available were obtained from Sigma and Fisher. Na125I (2,125 Ci/mmol), goat anti-rabbit
125I-IgG, 35S-dATP (1000-1500 Ci/mmol), and
[32P]orthophosphate (3,000 Ci/mmol) were purchased from
DuPont NEN. [3H]Myoinositol (17.7 Ci/mmol) was obtained
from Amersham Corp. T4 ligase, T4 polynucleotide kinase, T7 polymerase
(unmodified), and single-stranded binding protein for mutagenesis were
purchased from United States Biochemical Corp. Restriction enzymes were purchased from New England BioLabs (Beverly, MA) and Promega (Madison, WI). DEAE/dextran was obtained from Pharmacia Biotech Inc. COS-7 cells
and porcine kidney LLC-PK1 cells were generous gifts from Drs. B. Seed
and S. R. Goldring of the Laboratory of Molecular Biology and Arthritis
Unit, respectively, of the Massachusetts General Hospital (Boston).
Human embryonic kidney HEK293 cells were purchased from American Tissue
Culture Collection, and cDNA encoding Gq was
provided by Dr. M. I. Simon (California Institute of Technology,
Pasadena, CA).
COS-7 and LLC-PK1 cells were cultured in Dulbecco's modified Eagle's medium (DMEM, Mediatech, Washington, D.C.) supplemented with 7% fetal bovine serum (Sigma), 1.2 mM glutamine, and 4.5 g/liter glucose at 37 °C, in a humidified atmosphere containing 95% air and 5% CO2. HEK293 cells were grown in minimum essential Eagle's medium supplemented with 10% horse serum (Life Technologies, Inc.). Medium was replenished every 2-3 days, and cells were passaged when nearly confluent.
Construction of Mutant Rat PTH/PTHrP ReceptorsMutant rat PTH/PTHrP receptors were constructed by site-directed mutagenesis, using a minor modification of the method of Kunkel et al. (23). Oligonucleotides were designed to either randomly mutate amino acids in the WT receptor's 2i loop (amino acids 316-320, SEKKY) or to modify individual amino acids between 317 and 320. The codons targeted for mutagenesis were replaced by NNG/C, where N is an equal mixture of all four nucleosides and G/C is an equal mixture of G and C. Briefly, the mutant oligonucleotides in a sense orientation were phosphorylated by T4 polynucleotide kinase at 37 °C for 45 min in a buffer (pH 7.8), containing Tris (100 mM), dithiothreitol (10 mM), ATP (0.4 mM), and MgCl2 (10 mM), and the mixture then was heated at 65 °C for 10 min. The phosphorylated oligonucleotides and the single-stranded antisense template of the rat PTH/PTHrP receptor cDNA, carried on the pcDNA1 plasmid vector (Invitrogen, San Diego, CA), were heated to 70 °C for 2 min and then allowed to anneal by slowly cooling at room temperature. Complementary DNA was synthesized by sequentially incubating the mixture on ice for 5 min, at room temperature for 5 min, and then at 37 °C for 90 min in the presence of single-stranded binding protein, T4 ligase (2.5 units), and unmodified T7 DNA polymerase (1 unit), in a buffer (pH 7.4) containing Tris (17.4 mM), ATP (0.75 mM), dNTP (0.4 mM, each), MgCl2 (0.4 mM), and dithiothreitol (21.5 mM). The reaction was terminated by adding 10 µl of "stop" buffer (pH 8.0), containing Tris (10 mM) and EDTA (10 µM). The synthesized cDNA was then introduced into Escherichia coli strain MC1061/p3 by electroporation (25 microfarads/2.5 kV/400 ohm). The sequence of the each mutant cDNA was verified by sequence analysis.
Transient Transfection of COS-7 and HEK293 CellsAll
plasmid DNAs were prepared by the CsCl method (24) and stored at
20 °C in sterile buffer (pH 7.5) containing Tris (10 mM), EDTA (1 mM) at a concentration of 1 µg/µl, as established by absorbance at 260 nm. Their quality and
quantity also were assessed by agarose gel electrophoresis. Before
transfection, a given amount of the plasmid DNA was diluted in 190 µl
of phosphate-buffered saline (PBS, pH 7.42), and this mixture was
further diluted in 760 µl of a DEAE/dextran (10 mg/ml) solution. The
empty pcDNA1 plasmid or pcDNA1 containing the WT or mutant DNAs
was introduced into COS-7 or HEK293 cells by a slight modification of
the DEAE/dextran transfection method (20). In brief, cells in 150-mm
dishes were gently washed (2 ×) with pre-warmed PBS and exposed to
DMEM medium (19 ml) containing 10% Nuserum (Life Technologies, Inc.)
and chloroquine (100 µM). The plasmids in the
DEAE-dextran solution were added to the cells for 3 h at 37 °C;
the transfection medium was aspirated, and the cells then were shocked
by exposure to PBS containing 10% dimethyl sulfoxide at room
temperature for 2 min. The cells were cultured in fresh DMEM medium
supplemented with 7% FBS. Twenty-four h later, they were replated into
24- or 6-multi-well plates at a density of 50-100,000
cells/cm2. In some experiments, cells in multi-well plates
were directly transfected under the same condition with comparable
concentrations of cDNA and transfection medium. Results from the
two methods were indistinguishable. Cells were studied 60-72 h after
transfection.
WT and mutant PTH/PTHrP receptor constructs were introduced into LLC-PK1 cells, together with a neomycin gene, by the calcium phosphate method (22, 25), and clones expressing these receptors were selected by growth in G418 (Life Technologies, Inc.). In brief, cells in 100-mm dish were rinsed (2 ×) with warm PBS and preincubated at 37 °C for 30-60 min in DMEM (5 ml), containing Hepes (pH 7.12, 20 mM) and 7% FBS. Calcium phosphate precipitates (500 µl) of the mixture containing receptor cDNAs and pSV2neo plasmid DNA, in Hepes buffered salt solution, were added to the dish and further incubated at 37 °C for an additional 6 h. Cells were shocked at 22 °C for 3 min with 15% glycerol/Hepes buffered salt solution and washed gently (2 ×) with warm PBS, and the medium then was renewed. Eighteen to 20 h later, cells were replated into two 150-mm dishes in DMEM containing 7% FBS and 1 mg/ml G418. Selected colonies were propagated, and cells were further characterized for their capacity to bind 125I-PTHrP and a receptor-specific polyclonal antibody, G48 (see below), and for their signal transduction properties. Cells were routinely stored in liquid nitrogen.
Radioreceptor Assay and Scatchard AnalysisTransfected
cells, grown in 24-well plates, were washed (× 2) with binding buffer
(pH 7.7), which contained Tris (50 mM), NaCl (100 mM), KCl (5 mM), CaCl2 (2 mM), 5% heat-inactivated horse serum, and 0.5% FBS. For
radioreceptor assays, cells were incubated in binding buffer
with
125I-[Tyr36]PTHrP-(1-36)-NH2
(PTHrP, 100,000 cpm/500 µl/well), in the absence or presence of
unlabeled PTHrP (1011 to 10
6 M)
at 15 °C for 4 h. These incubation conditions and conditions for preparing the radioiodinated ligand have been described (20). Cells
then were washed (3 ×) with ice-cold binding buffer and solubilized in
0.5 ml of 0.1% SDS solution. Radioactivity in the cell lysate was
measured by
-counting (Micromedic System Inc., Model 6/400 plus).
Scatchard analysis was performed as described previously (20). Protein
concentration was determined (Bio-Rad) in an aliquot of the lysate
using bovine serum albumin as the standard. Cells per well were
quantified using a hemocytometer, after cells in two wells from each
plate had been suspended by trypsinization.
Cells grown in 24-well plates were washed (3 ×) with PBS (pH 7.4), containing 5% FBS (FBS/PBS), and then incubated at room temperature for 2 h in FBS/PBS in the presence of sheep receptor-specific antibody G48 (1:500), or preimmune sheep serum (1:500) (20). G48 was developed against a synthetic 18-amino acid peptide (DKGWTPASTSGKPRKEKA) of the rat PTH/PTHrP receptor's extracellular N-terminal region, and it was purified by affinity chromatography with antigen immobilized to Sepharose 4B by the CNBr method (26). Cells were then washed thoroughly (4 ×) with 5% FBS/PBS, incubated first for 1 h at room temperature with affinity-purified rabbit anti-sheep IgG (H + L) (Kirkland & Perry Laboratories, Inc., Gaithersburg, MD), and then for an additional hour with 100,000 cpm/250 µl/well of goat anti-rabbit 125I-IgG (DuPont NEN). The incubation was terminated by washing the cells with PBS (3 ×); the cells were solubilized in 1 N NaOH, and the radioactivity was counted.
Measurement of Intracellular cAMP AccumulationCells grown
in 24-well plates were incubated in serum-free DMEM medium, containing
Hepes (pH 7.42, 10 mM) with 3-isobutyl-1-methylxanthine (1 mM) and 0.1% bovine serum albumin at room temperature for
10 min. Then, COS-7 and LLC-PK1 cells were incubated at 37 °C in the
absence or presence of rPTH (1012 to 10
6
M) for an additional 15 and 5 min, respectively. The
reaction was terminated by aspirating the medium, washing the cells
with ice-cold PBS (2 ×), and freezing the cells in 500 µl of HCl (50 mM). The HCl extracts were stored at
80 °C until
intracellular cAMP was measured by radioimmunoassay, as described
previously (20). Cyclic AMP accumulation (pmol/well) is expressed as
stimulated minus basal values.
Accumulated
IPs were measured in cells grown in 6-well plates by methods described
previously (20, 22). Briefly, cells were prelabeled at 37 °C for
8-12 h with 3 µCi/ml [3H]myoinositol in inositol- and
serum-free DMEM (Life Technologies, Inc.), supplemented with 0.1%
bovine serum albumin. Cells then were washed (2 ×) with prewarmed
inositol- and serum-free DMEM containing LiCl (30 mM) and
treated with rPTH (1011 to 10
6
M) or with vehicle alone (10 mM acetic acid) at
37 °C for an additional 30 min. The reactions were terminated by
aspirating the incubation medium and adding 1 ml of ice cold 5%
trichloroacetic acid. Trichloroacetic acid extractions and separation
of the IPs by AG 1X8 anion exchange column chromatography (formate
form, 100-200 mesh, Bio-Rad) were performed as described previously in
detail (20). Specific PTH-stimulated phosphoinositide hydrolysis was
determined by subtracting basal values from the total values obtained
by PTH stimulation. Radioactivity in each fraction was counted by
liquid scintillation counting (Beckman, Model LS 6000IC).
LLC-PK1 cells stably expressing WT
or mutant PTH/PTHrP receptors were grown in 24 multi-well plates, and
subsequently they were incubated with rPTH-(1-34) (1010
to 10
6 M) at 37 °C for 6 h, prior to
measurements of sodium-dependent phosphate co-transport, as
described previously in detail (22, 25). Briefly, cells were washed (2 ×) with warm phosphate-free Hepes buffer (pH 7.42, 10 mM)
that contained NaCl (150 mM), KCl (5 mM),
MgCl2 (1.8 mM), CaCl2 (1.0 mM), and glucose (5 mM) and then further
incubated in the same buffer which now contained 0.2 mM
phosphate (pH 7.42) and 0.5 µCi/250 µl/well
[32P]orthophosphate at 37 °C for 5 min. The reaction
was terminated by placing the plate on ice and washing the cells (3 ×)
with ice-cold Hepes stop buffer (pH 7.42, 10 mM), which
contained choline chloride (150 mM) and sodium arsenate (5 mM). Cells then were solubilized in NaOH (1 N),
and the radioactivity in the lysates was measured by liquid
scintillation counting. Protein concentration was measured in an
aliquot of each lysate. The initial velocity of phosphate uptake was
expressed as nmol/5 min/mg protein.
Means ± S.D. of the three to six replicates were calculated from each of two or more independent experiments. One-way analysis of variance followed by Student's t test were used to determine significance (p < 0.05).
The
capacity of the rat WT PTH/PTHrP receptor, with SEKKY at positions
316-320 in the 2i loop, to increase intracellular accumulation of cAMP
and IPs was compared with those of four receptors with random clustered
mutations at the same site (SDSEL, GRELG, RTKKS, and SKEKY) in COS-7
cells (Fig. 1). Expression of all receptors also was
measured by ligand binding and by binding of the receptor-specific antiserum, G48. PTH (100 nM)-stimulated accumulation of IPs
was severely impaired in cells expressing receptors with the DSEL clustered mutation, whereas cAMP accumulation was indistinguishable from that of cells expressing WT receptors (Fig. 1). Cell-surface expression of this mutant was only slightly impaired, if at all, and
Scatchard analysis showed that the binding affinities of WT and DSEL
mutants were indistinguishable, with apparent Kd values of 2.3 ± 0.3 and 2.5 ± 0.4 nM,
respectively (data not shown). Results from cells expressing receptors
with the three other clustered mutations were less informative.
Receptors with either RTKKS or SKEKY clustered mutations showed no
significant impairment in their signal transduction properties. PTH
(100 nM)-generated IPs in cells expressing receptors with
the GRELG mutation was not totally abolished, although it was markedly
impaired, compared with the WT receptor (20 ± 8%,
n = 4). We thus focused attention on receptors with the
DSEL mutant cassette.
Since we had previously shown that PTH-stimulated accumulation of IPs
was strikingly dependent on the density of receptors expressed on the
cell surface (20, 22), we next examined the relationship between
PTH-stimulated accumulation of IPs and cAMP, with expression of WT and
DSEL mutant receptors. Receptor expression was proportional to the
amount of cDNA used for transfection, up to about 10 µg/10-cm
dish, and expression of PTH/PTHrP receptors with the DSEL mutant
cassette was only slightly less efficient than that of WT receptors
(Fig. 2A). PTH (100 nM)-generated
IPs increased with increased expression of WT PTH/PTHrP receptors, whereas no detectable accumulation of IPs was observed in cells transfected with the DSEL mutant receptor at any expression level (Fig.
2B). In contrast, PTH (100 nM)-generated cAMP
increased with increased expression of either WT or mutant
receptor (Fig. 2C).
Point Mutations in the Rat PTH/PTHrP Receptor 2i Loop Diminish, but Do Not Abolish PTH-stimulated PLC Activity
To define more precisely the amino acids essential for PTH-generated IPs, we constructed mutant receptors in which each amino acid of the WT sequence, EKKY, from residues 317 to 320 in the 2i loop was replaced by alanine or the corresponding amino acid in the DSEL sequence. Glu-317, Lys-318, and Lys-319 were also replaced by other charged amino acids, and Tyr-320 was replaced with Phe. All these mutant receptors were relatively well expressed; G48 binding was 79-107% and maximal radioligand binding (B0) was 92-112%, respectively, compared with the WT receptor (Table I). Although maximal PTH-generated cAMP was minimally altered by mutations in these four amino acids, accumulation of IPs differed strikingly (Table I and Fig. 3). The signal transduction properties of the PTH/PTHrP receptor were most influenced by modifications at Lys-319. K319E selectively decreased the receptor's capacity to generate IPs to 27 ± 13%, compared with the WT receptor, but it did not eliminate it. In contrast, K319A modestly reduced the receptor's capacity to signal through both signal transduction pathways. Interestingly, the signaling properties of a mutant receptor in which K319R were indistinguishable from those of the WT receptor. Modifications of Lys-318 modestly changed the receptor's capacity to transduce signals, and these changes were nonselective: K318S reduced signaling through both adenylyl cyclase and PLC pathways; K318E increased signaling through both pathways; and K318A did not affect receptor signaling. Furthermore, the phenotype of a double mutant receptor in which Lys-318 and Lys-319 both were Glu closely resembled that of K319E, rather than K318E, thus reflecting the greater importance of the amino acid at position 319 (Table I, Fig. 3).
|
Mutating E317A did not affect PTH-stimulated accumulation of IPs, but PTH-generated IPs increased slightly in cells expressing receptors with E317D or E317K. None of these mutations at position 317 affected PTH-generated cAMP accumulation. Changing Tyr-320 to Ala, Phe, or Leu did not affect receptor signaling (Table I, Fig. 3).
PTH-stimulated PLC Activity in Rat PTH/PTHrP Receptors with Mutations in the 2i Loop Is Restored by a Single-point Revertant in Which the Glutamic Acid at Position 319 Is Replaced with Lysine, the WT Amino AcidWe then modified the rat PTH/PTHrP receptor with the
DSEL mutant cassette by individually introducing the WT amino acids
back at their original positions in the sequence and assessing their functional properties. Replacing the mutant E319K (DSKL) fully restored
its capacity to increase accumulation of IPs (Fig. 4, Table I). Furthermore, PTH-generated cAMP and IPs in cells expressing this revertant had dose-response characteristics that were
indistinguishable from the WT receptor: the EC50 values of
the WT and this mutant receptor, respectively, for accumulation of IPs
were 25 ± 3 and 23 ± 4 nM (Fig.
5A) and for cAMP accumulation were 0.32 ± 0.04 and 0.37 ± 0.05 nM (Fig. 5B). PTH
treatment of cells expressing the receptor with the DSEL mutant
cassette failed to generate IPs even at doses as high as 1 µM (Fig. 5A), whereas cAMP responsiveness to
incremental doses of PTH was indistinguishable from that of the WT
receptor.
On the other hand, replacing the other amino acids in the DSEL mutant cassette with the corresponding WT residue partially restored the receptor's capacity to generate IPs; PTH-generated IPs in cells expressing receptors with D317E (ESEL), S318K (DKEL), and L320Y (DSEY) were 26 ± 15, 22 ± 9, and 12 ± 4%, respectively, compared with the WT receptor (Fig. 4, Table I). These data thus confirm the importance of Lys-319 for signaling PLC, but they also suggest a minor role for neighboring amino acids in the 317-320 sequence.
Co-expression of GTo
establish whether the mutant PTH/PTHrP receptor with the DSEL mutant
cassette was incapable of generating IPs because the levels of
Gq were inadequate, we compared the properties of the WT
with this mutant receptor in cells co-transfected with
G
q. PTH treatment of cells that co-expressed the DSEL
mutant receptor together with G
q did not significantly
increase IPs in either HEK293 cells or LLC-PK1 cells. In contrast,
PTH-stimulated accumulation of IPs in both HEK293 and LLC-PK1 cells was
enhanced by 2- and 2.5-fold, respectively, in the cells expressing WT
PTH/PTHrP receptor and G
q, compared with PTH-treated
cells expressing only the WT rat PTH/PTHrP receptor (Fig.
6). Overexpression of G
q did not increase
basal PLC activity in either cell. Thus, G
q
overexpression did not overcome the signaling defect of PTH/PTHrP
receptors with the DSEL mutant cassette.
A PLC-dependent Biological Function in LLC-PK1 Cells Is Lost in Cells Stably Expressing PTH/PTHrP Receptors with the DSEL Mutant Cassette
We previously reported that PTH augments
sodium-dependent phosphate co-transport in LLC-PK1 cells
that stably express the rat WT PTH/PTHrP receptor and further that this
response appeared to depend on activation of PLC and protein kinase C
(22, 25). To determine whether mutant PTH/PTHrP receptors with the DSEL mutant cassette would exhibit the same properties in LLC-PK1 cells as
they did in COS-7 cells and whether sodium-dependent
phosphate co-transport was linked to the PLC response, we isolated
LLC-PK1 subclones following stable transfection with the WT receptors or receptors with the DSEL mutant cassette. As we previously had shown
that PLC stimulation and phosphate transport were strongly dependent on
the level of surface receptors expressed (20, 22), we compared the
function of the WT and the mutant receptor in subclones that expressed
comparable numbers of receptors (approximately 200,000 sites per cell).
PTH increased both IP3 accumulation and phosphate
uptake, dose-dependently, over a similar range of PTH concentrations (0.1 to between 100 and 1000 nM) only in
cells expressing the WT receptor. Even at PTH doses as high as 1 µM, however, neither intracellular IP3
accumulation nor phosphate uptake was affected in cells expressing
receptors with the DSEL mutant cassette (Fig. 7).
PTH-stimulated cAMP accumulation was indistinguishable, however, in
cells expressing either WT receptors or receptors with the DSEL mutant
cassette (data not shown).
These studies provide a clear demonstration that the capacity of the PTH/PTHrP receptor to signal through the PLC pathway can be selectively abolished, without affecting its capacity to activate adenylyl cyclase. Furthermore, they define some of the structural requirements necessary for signaling by this receptor through members of the Gq family. Preliminary analysis revealed that cells expressing the rat PTH/PTHrP receptor, in which four sequential amino acids, EKKY (amino acids 317-320) in its 2i loop were mutated to DSEL, totally lacked the capacity to generate IPs when treated with PTH, although the mutant receptor was well expressed, and its PTH-stimulated cAMP response was indistinguishable from the WT receptor.
Extensive investigation of the role of each amino acid in this region pointed to residue 319 as the site whose sole modification caused the greatest reduction in PLC signaling. Replacing Lys-319 with Glu maximally decreased PTH-generated IPs to 27 ± 13%, compared with the response of the WT receptor, but, importantly, no single substitution completely abolished signaling through the PLC pathway. It seems likely, in retrospect, that PTH-generated IPs was markedly reduced in receptors with the clustered mutation, GRELG at position 316-320, because of K319L mutation. The importance of Lys-319 for generating IPs was further substantiated by showing that a revertant in which Glu-319 in the DSEL mutant cassette was replaced with WT amino acid, Lys, restored full activation of PLC by PTH. However, PTH/PTHrP receptor-Gq interactions are more complex, as evidenced by our finding that revertants in which individual WT amino acids replaced the mutant residues at positions 317, 318, and 320 in the DSEL mutant cassette (D317E, S318K, L320Y) also partially restored PTH-generated IPs. The contributions of any individual amino acid for activating the PLC pathway, therefore, must be considered within the context of elaborate interactions between PTH/PTHrP receptors and Gq heterotrimers.
The rat WT PTH/PTHrP receptor's PLC-stimulating capacity was indistinguishable from that of receptors with K319R. It was also modestly reduced (64 ± 11% of WT), however, when Lys-319 was replaced with the neutral amino acid, Ala. These data, together with experiments in which full PLC activation was restored by changing Glu-319 in the DSEL sequence back to Lys, the WT amino acid, suggest that a positively charged amino acid at position 319 may be important, although not absolutely necessary, for activating the relevant member(s) of the Gq family.
The lysine at position 318 in the WT rat PTH/PTHrP receptor appeared to be of only minor importance for coupling to Gq proteins. Replacing Lys-318 with Ala, Ser, or Glu affected the receptor's phenotype only modestly, and these changes were nonselective, that is they simultaneously decreased or increased accumulation of both IPs and cAMP. Support for the notion that Lys-319 is more important than Lys-318 for signaling through the PLC pathway also comes from examining the properties of the PTH/PTHrP receptor with both K318E/K319E mutations. The capacity of mutant receptor to generate second messengers closely resembled that of receptors with K319E, rather than K318E. Furthermore, these data demonstrate that in contrast to position 319, a positively charged amino acid at position 318 is not critical for receptor function.
The properties of receptors with single-point mutations, in which Glu-317 was replaced with Ala, Asp, or Lys, or Tyr-320 was replaced with Phe, Ala, or Leu, were closely similar to or indistinguishable from those of the WT receptor. These data indicate that the amino acids at these positions probably are not integrally involved in coupling to Gq.
Our data are consistent with the notion that receptors containing the
DSEL mutant cassette might couple to Gq proteins but that
this interaction is only less efficient. If this were the case,
overexpression of Gq, together with this mutant
receptor, might overcome the signaling defect in this mutant receptor.
However, PTH failed to generate IPs when receptors with the DSEL mutant cassette were expressed together with G
q in either
HEK293 or LLC-PK1 cells. This sharply contrasted with the 2- and
2.5-fold increase in the PLC response in HEK293 and LLC-PK1 cells that co-expressed both the rat WT PTH/PTHrP receptor and G
q.
Since a PTH/PTHrP receptor with the DSEL mutant cassette had unimpaired cAMP responsiveness and thus must release
-subunits from the activated heterotrimeric Gs, these data strengthen the argument that
stimulation of PLC by PTH in cells expressing the WT receptor is
mediated by G
subunit(s), rather than via activation of
, as
has been reported with some G protein-coupled receptors (27-30). Presumably, the DSEL mutant cassette in the PTH/PTHrP receptors' 2i
loop strongly impairs interaction with G
q. Furthermore,
our findings in cells co-expressing the WT PTH/PTHrP receptor and G
q extend our earlier observations that co-expression of
any one of four G
q proteins, G
q,
G
11, G
14, and G
16,
together with the WT rat PTH/PTHrP receptor, markedly increased
PTH-stimulated accumulation of IPs (31) and that the PTH-stimulated
sodium-dependent phosphate co-transport in LLC-PK1 cells
stably expressing the rat PTH/PTHrP receptor is cAMP-independent and
apparently mediated through PLC (25).
Treating all members of the PTH/PTHrP/secretin/calcitonin receptor
family with their cognate agonists generates cAMP. In many, and perhaps
all, of these receptors agonist occupancy also activates the PLC
pathway. For example, activated receptors for pituitary adenylyl
cyclase-activity peptide (PACAP) (32), glucagon (33), and calcitonin
(34, 35) increased both intracellular IPs and free calcium. Also,
activated corticotropin-releasing hormone (CRF) receptors generated IPs
(36), and activated receptors for both gastric inihibitory peptide
(GIP) and glucagon-like peptide 1 (GLP-1) increased intracellular free
[Ca2+] (33, 37, 38). Aligning the amino acids of members
of this receptor family reveals that the EKKY sequence is identical in all five mammalian PTH/PTHrP receptors, and it is highly homologous with the sixth from Xenopus laevis, where the only change is
Asp, rather than Glu, at the corresponding position (39). Three of these amino acids, Asp, Lys, and Tyr, also are conserved in the PTH2
receptor sequence, including the Lys corresponding to Lys-319 in the
rat PTH/PTHrP receptor (3). Among the other members of this receptor
family, EKKY in the PTH/PTHrP receptor aligns with ERKY in the
vasoactive intestinal peptide (VIP) and secretin receptors and with
ERRY in the PACAP receptor; Lys-318 and Lys-319 align with positively
charged amino acids, RR, in the growth hormone-releasing peptide (GHRH)
receptor. Homology is less evident, however, with other members; EQR in
the glucagon receptor correspond with EKK in the PTH/PTHrP receptor,
and ER in the GIP receptor are positioned identically with Glu-317 and
Lys-318 in the PTH/PTHrP receptor, although Ser aligns with Lys-319 in
the PTH/PTHrP receptor. There is no apparent sequence homology among
calcitonin, GLP-1, and PTH/PTHrP receptors in this portion of the 2i
loop, except for Glu, which corresponds to Glu-317 (Fig.
8). The residues important for PLC signaling by these
other receptors have not yet been defined.
All G protein-coupled receptors are thought to have seven
membrane-spanning helices. Many of them couple to Gq(s),
including adrenergic, muscarinic, and glycoprotein hormone receptors.
Activation of glycoprotein hormone receptors, like activation of
PTH/PTHrP receptors, also stimulates both adenylyl cyclase and PLC
(40-42). Therefore, relevant features that enable them to couple to
Gq possibly might be gleaned from examining their primary
structure, despite their lack of amino acid homology with PTH/PTHrP
receptors. Few paradigms have emerged, however, concerning determinants
important for coupling these receptors to Gq proteins. All
three intracellular loops, especially 2i and 3i, as well as the
N-terminal portion of the cytoplasmic tail have been implicated in PLC
activation by adrenergic, muscarinic, and glycoprotein hormone
receptors, although the putative interactive sites vary among the
individual receptors (43-47). There are several examples where
agonist-stimulated PLC activity of a G protein-coupled receptor is
severely affected, when a single amino acid is mutated. The TSH
receptor's capacity to generate IPs, for example, was markedly
impaired when alanine at position 623 in the 3i loop was mutated to
lysine or glutamic acid, but its capacity to activate adenylyl cyclase
was unimpaired (48). It is puzzling, however, that the 1B-adrenergic
receptor became constitutively activated with respect to stimulating
PLC, when alanine at position 293, which is located at the same
relative position as alanine 623 in the TSH receptor, was replaced with any amino acid (49). Mutating a highly conserved leucine, putatively located in the middle of the 2i loop of the GnRH, m1, and m3 muscarinic receptors to alanine or to a polar amino acid also markedly reduced agonist-generated IPs, although like mutations of Lys-319 in the PTH/PTHrP receptor, PLC activity was not totally eliminated (50-52). Moreover, based on evidence that a Phe to Ala mutation at the same
relative position in the
2-adrenergic receptor markedly reduced isoproterenol-generated cAMP, these authors concluded that the
a bulky hydrophobic residue at this site played a nonspecific role in
coupling G protein-coupled receptors to more than one G protein (50).
In recent studies of mutant
1B-adrenergic receptors, in which
charged residues in the highly conserved 2i loop sequence, DRY, were
substituted, Scheer et al. (53) showed that D142A conferred
high constitutive activity, whereas R143A or Asn markedly reduced
agonist-stimulated accumulation of inositol phosphate. These findings
were consistent with predictions from molecular dynamic simulations and
led the authors to suggest that the main role of Arg-143 is to mediate
receptor activation, by allowing several amino acids in the 2i and 3i
loops to attain a configuration that facilitates docking with the G
protein.
Our results establish that polyphosphoinositol production by the
activated PTH/PTHrP receptor can be selectively abolished by mutations
in the 2i loop, without affecting its capacity to generate cAMP, and
that Lys-319, or perhaps any basically charged amino acid at this
position, is important for coupling the PTH/PTHrP receptor to the
relevant Gq(s). Moreover, activation of the PLC pathway by
the PTH/PTHrP receptor is mediated by -subunits of the
Gq family, rather than by
-subunits, and
sodium-dependent phosphate co-transport in LLC-PK1 cells is
mediated through the PLC, rather than the adenylyl cyclase, pathway.
Future characterizations of receptor mutants together with structural
analyses will ultimately be necessary to define these complex
receptor-G protein interactions.
We thank Dr. T. Usa and T. Iida for their technical help, Drs. E. Schipani, H. M. Kronenberg, and T. Gardella for their valuable suggestions.