A Prostaglandin E Receptor Coupled to a Pertussis Toxin-sensitive Guanine Nucleotide Regulatory Protein in Rabbit Cortical Collecting Tubule Cells*

of binding to receptor preparations cells.

to inhibit arginine vasopressin (AVP)-induced adenylate cyclase activity in freshly isolated rabbit cortical collecting tubule (RCCT) cells. This inhibitory effect, which was blocked by pertussis toxin, was not detectable in RCCT cells which had been grown in primary culture, although an an-adrenergic-, G,-linked response was detectable in the cultured cells. We also found that PGE derivatives, with the notable exception of sulprostone, act at somewhat higher concentrations (~10~~ M) to stimulate CAMP synthesis and that a stimulatory response to PGE, was observed not only with freshly isolated RCCT cells, but also with cultured cells.
These results suggested that PGE, exerts its effects on CAMP metabolism in the rabbit cortical collecting tubule by interacting with two distinct PGE receptors: (a) a high affinity, pertussis toxin-sensitive inhibitory PGE receptor which binds both PGE, and sulprostone and is present only in freshly isolated RCCT cells, and (b)  Binding assays were performed in duplicate at 30 "C for 2 h as described in the text in the presence of varying concentrations of PGE,. The solid line is a computer-generated regression (6) line for the binding site. Binding assays were performed in duplicate at 30 'C for 2 h as described in the text in the presence of varying concentrations of PGEp. The solid line is a computer-generated regression line (6)  Freshly isolated RCCT cells were pretreated for 4 h at 37 "C with vehicle or pertussis toxin as described in the text. Membranes were prepared and binding assays were performed in the presence (hatched bars) or absence (open bars) of 100 pM GTP-yS for 2 h at 30 "C as described in the text. Values represent the mean of triplicate determinations + SE. Asterisk, significantly different than control values (p < 0.05).
the binding of 2 nM [3H]PGE2 to membranes from freshly isolated RCCT cells (Fig. 4). Importantly, this stimulatory effect of GTPrS on [3H]PGE2 binding was eliminated and a small inhibitory effect was observed when, prior to preparation of membranes, freshly isolated RCCT cells were pretreated with pertussis toxin; the toxin treatment conditions (4 h at 37 "C with 1 pg of toxin/ml) were previously shown to block the inhibitory effect of both sulprostone and PGE2 on AVP-induced CAMP accumulation (1); moreover, following this pretreatment, membranes isolated from the treated, but not the control cells, were refractory to radiolabeling of a 41-kDa species when incubated with pertussis toxin and ["'PI 9"LPROSTOWE 0 10-9 10-7 10-g 0 10-s D-7 10-a 0 10-g 10-7 IUS Binding assays were performed in duplicate at 30 "C for 2 h as described in the text in the presence of the indicated concentrations of PGE, (---), sulprostone, PGE,, or 16,16-dimethyl-PGE2 (DM-PG&).  ICk values were determined as described in the text as the concentration of prostanoid necessary to cause 50% inhibition of binding of 2 nM [3H]PGE, to membranes prepared from the indicated sources. Data for canine outer medulla, except sulprostone (Fig. S) PGEz and sulprostone both cause inhibition of hormone-induced CAMP formation by freshly isolated RCCT cells, neither PGEz nor sulprostone produce an inhibitory response in RCCT cells which have been grown in primary culture; however, PGEz (but not sulprostone) stimulates CAMP formation by both freshly isolated and cultured RCCT cells (1). Some properties of [3H]PGEz binding to membranes from freshly isolated and cultured RCCT cells are compared in Fig. 7. Specific binding of 2 nM [3H]PGE2 to membranes prepared from RCCT cells which had been grown in culture for 5-7 days was only lo-20% of that seen with membranes from freshly isolated cells and was near the limit of our detection; in fact, nonspecific binding of [3H]PGE2 to membranes from cultured RCCT cells averaged about 50% of total binding (uersus 15% with membranes from freshly isolated RCCT cells). Moreover, although GTPrS stimulated [3H]PGE2 binding to membranes from freshly isolated cells, GTPyS actually inhibited binding of [3H]PGEz to membranes from cultured cells by about 65% (Fig. 7). Pertussis toxin pretreatment of cultured RCCT cells did not affect the ability of GTP$S to inhibit [3H]PGEz binding to membranes prepared from these cells (data not shown). Although sulprostone (10m6 M) completely inhibited [3H]PGE2 binding to membranes from freshly isolated RCCT cells, sulprostone had no significant effect on [3H]PGE2 binding to membranes from cultured cells (Fig. 7).
Binding to Membranes from Rabbit and Canine Outer Medulla-We previously described the characteristics of a PGE receptor from canine renal outer medulla (2). This latter receptor could be solubilized in a complex with a Gi, and GTP derivatives were found to stimulate [3H]PGEZ binding by decreasing the apparent & (2). As shown in Fig. 8A, sulprostone was as effective as PGEz in inhibiting [3H]PGE2 binding to membranes from canine renal outer medulla. A PGE receptor with properties very similar to the receptor present in membranes from canine renal outer medulla was also found to be present in membranes prepared from rabbit outer medulla. The KD value for [3H]PGEz binding to the rabbit outer medullary receptor was determined by Scatchard Membranes were prepared and ["H]PGEz binding assays performed in duplicate at 30 "C as described in the text in the presence of the indicated concentrations of sulprostone or PGE,.
A. FRESHLY ISOLATED RCCT CELLS analysis to be 2 nM; moreover, GTPyS routinely stimulated binding 1.5-to 3.0-fold (data not shown). Sulprostone and PGEz were comparably effective in inhibiting the binding of [3H]PGE2 to preparations of rabbit outer medullary membranes (Fig. 8B). Table I compares the abilities of a series of prostanoids to compete with [3H]PGE2 for binding to membranes from freshly isolated RCCT cells and the GTP-stimulatable PGE receptors present in canine and rabbit outer medulla membranes. In all cases, PGEi, PGEz and sulprostone were comparably effective, while PGF2, was about lo-fold less effective, and PGD2, 30-to loo-fold less effective. Thus, the [3H]PGEz binding activities of membranes from freshly isolated RCCT cells and rabbit and canine renal outer medullae all exhibit similar behavior with respect both to their binding specificities and to their unusual behavior upon addition of GTP+.  (A) or cultured (B) RCCT cells were prepared and assayed for specific 13HlPGE, binding in the presence or absence of either 100 ;M GTprS-or 1 pM sulpr&tone at 30 "C for 2 h as described in the text. Values represent the mean of triplicate determinations + S.E. Asterisk, significantly different than control values @ < 0.05).

Prostanoids
of the E-series acting at concentrations of about lo-' M inhibit AVP-induced water reabsorption in the rabbit cortical collecting tubule while at higher concentrations (>lOO nM), PGEl and PGE, by themselves can stimulate water reabsorption (9, 10). As described in an earlier report (l), these responses can be rationalized on the basis of changes in cellular CAMP levels. E-series prostaglandins acting at low concentrations (0.1-10 nM) function via a pertussis toxinsensitive mechanism to inhibit AVP-induced adenylate cyclase in freshly isolated RCCT cells, and at higher concentrations PGEi and PGEz acting alone can stimulate CAMP synthesis (1). In work described in this article, we have correlated prostanoid binding to RCCT cell membranes with the effects of prostanoids on CAMP metabolism in freshly isolated and cultured RCCT cells.
Our results indicate that there is a single, detectable [3H] PGE, binding activity associated with membranes from freshly isolated RCCT cells. Kn values determined for PGE2 binding to this site in the presence of GTPrS were in the range of l-5 nM, consistent with a concentration of 3 nM PGE, being able to cause half-maximal inhibition of AVPinduced CAMP formation in freshly isolated RCCT cells (1). The [3H]PGE2 binding activity of membranes from freshly isolated RCCT cells has a collage of properties expected for a Gi-linked PGE receptor which is coupled to inhibition of AVPinduced CAMP formation in the rabbit renal cortical collecting tubule. First, the order of potency for inhibition of AVPinduced CAMP formation by various prostanoids (1) is the same as the order of potency with which these same prostanoids inhibit the binding of [3H]PGE2 (i.e. sulprostone = PGE, = PGEl = 16,16-dimethyl-PGE2 > carbacyclin = PGF2, > PGD,); moreover, this rank order of potency differs from that seen for prostanoid-induced CAMP synthesis by RCCT cells (i.e. PGE, = PGEl > 16,16-dimethyl-PGE2, with sulprostone, PGFZa, carbacyclin, and PGDz being inactive (1)). Second, treatment of freshly isolated RCCT cells with pertussis toxin both blocks the ability of sulprostone and PGE to inhibit AVP-induced CAMP formation (1) and eliminates the GTP-yS effect on [3H]PGE2 binding to RCCT cell membranes. Third, the [3H]PGE2 binding activity of membranes from freshly isolated RCCT cells and the PGE receptor from canine renal medullary receptor, which has previously been solubilized in a complex with a Gi, have very similar properties: (a) GTPyS stimulates [3H]PGE2 binding, (b) the stimulatory effect of GTP-&S is blocked by pertussis toxin, and (c) the order of affinities for prostanoids are the same. And fourth, cultured RCCT cells, unlike fresh cells, lack an inhibitory Gilinked response to PGE2 and sulprostone (l), and cultured RCCT cells, unlike fresh cells, lack a pertussis toxin-sensitive sulprostone-inhibitable [3H]PGE2 binding activity. Based on these observations, we conclude that the [3H]PGE2 binding activity of freshly isolated RCCT cells represents the Gilinked PGE receptor which mediates inhibition of AVP-induced CAMP synthesis in the renal cortical collecting tubule.
The effect of GTPyS on [3H]PGEz binding to membranes from freshly isolated RCCT cells is unusual and warrants a comment. Although guanine di-and trinucleotides typically decrease the affinity of G protein-linked receptors for their corresponding agonists in vitro (11, 1'2), GTPrS actually stimulates binding of [3H]PGE2 to membranes from freshly isolated RCCT cells by decreasing the KD value. However, it is unlikely that this effect of GTP derivatives has much significance in terms of regulating receptor responses. It is sometimes assumed for intact cells (a) that receptors normally cycle between different affinity states depending on whether or not the G protein is occupied by a guanine nucleotide and (b) that this cycling has some kind of regulatory significance in duo. However, binding of muscarinic agonists to intact cells appears to occur via a single affinity site with a KD corresponding to that expected for a receptor interacting with a G protein occupied with a guanine nucleotide (13). Thus, KD values measured in the presence of guanine nucleotides probably best reflect the KD values for receptors of intact cells.
Our model for the actions of PGE, in the collecting tubule indicates that there are both high and low affinity PGE receptors coupled to inhibition and activation of adenylate cyclase, respectively (1). We have found no compelling evidence for the presence of a [3H]PGE2 binding activity other than the Gi-linked PGE receptor of freshly isolated RCCT cells. There is, however, a GTPyS-inhibitable pertussis toxininsensitive [3H]PGE2 binding activity associated with membranes from cultured RCCT cells. This binding activity could be distinguished from nonspecific binding, but the level of binding was too low to permit us to determine binding constants accurately. The fact that sulprostone fails to inhibit [3H]PGE2 to membranes from cultured RCCT cells raises the possibility that this binding activity corresponds to a PGE receptor coupled via G. to the activation of adenylate cyclase (14).