JOURNAL OF BONE AND MINERAL RESEARCH Volume 6, Number 3, 1991 Mary Ann Liebert, Inc., Publishers

Interspecies Comparison of Renal Cortical Receptors for Parathyroid Hormone and Parathyroid Hormone-Related Protein JOHN J. ORLOFF,',' DOUGLAS GOUMAS,' TERENCE L. WU,' and ANDREW F. STEWART','

ABSTRACT Parathyroid hormone (PTH) and PTH-related proteins (PTHrP) interact with a common receptor in rat bone cells and in canine renal membranes with similar affinity, but PTHrP are substantially less potent than PTH in stimulating adenylate cyclase in canine renal membranes; in contrast, PTH and PTHrP are equipotent in stimulating adenylate cyclase in rat bone cells. This discrepancy has been largely viewed as reflecting differences in the relative efficiency of signal transduction for PTHrP between bone and kidney assay systems. To test the alternative (but not mutually exclusive) hypothesis that these differences could reflect interspecies differences in PTH receptors, we have characterized the bioactivity of amino-terminal PTHrP and PTH in rat and human renal cortical membranes (RCM) and compared them to results we previously reported in canine RCM. The stability of PTH and PTHrP peptides under binding and adenylate cyclase assay conditions was greater than 80% for each species. Competitive inhibition of [1251](Tyr36)hPTHrP-(1-36)NH2 binding to rat RCM by bPTH-( 1-34) and (TyP)hPTHrP-( 1-36)NH2 yielded nearly identical binding dissociation constants (3.7 and 3.6 nM, respectively), and binding to human RCM demonstrated slightly greater potency for PTHrP (0.5 nM) than for PTH (0.9 nM). Similarly, adenylate cyclase stimulating activity was equivalent for the two peptides in rat RCM, but PTHrP was twofold more potent than PTH in human RCM. Covalent photoaffinity labeling of protease-protected rat RCM yielded an apparent 80 kD receptor protein, and crosslinking of human RCM labeled an 85 kD receptor, indistinguishable in size from the canine renal PTH receptor. We conclude that rat, canine, and human renal cortical PTH receptors exhibit species specificity. The previously observed differences between rat bone cells and canine renal membranes in the efficiency of signal transduction by PTHrP may be explained, at least in part, by these species differences.

INTRODUCTION amino acid sequencing, and molecular cloning of parathyroid hormone-related proteins (PTHrP)I1-" has allowed the characterization of receptor interactions in kidney and bone using synthetic amino-terminal analogs. Several groups have demonstrated that amino-terminal PTHrP interact with the PTH receptor in

T

HE PumrcATIoN,

these t i ~ s u e s . ( Although ~ ~ - ~ ~ ~the affinity of PTHrP for receptors in rat bone cells and canine kidney membranes is similar to that of PTH, PTHrP have been reported to be less potent thant PTH in stimulating adenylate cyclase activity in canine renal membranes but equipotent or more potent than PTH in rat bone cells.'n) This discrepancy in receptor-effector coupling efficiency for PTHrP in canine renal membranes compared to rat bone cells has been gen-

'Divisions of Endocrinology and Metabolism, West Haven Veterans Administration Medical Center, West Haven, C T 0 6 5 16. 'Yale University School of Medicine, New Haven, CT 06510.

279

280

ORLOFF ET AL.

erally viewed as reflecting tissue-specific differences between bone and kidney assay systems. However, it could also reflect differences in intact compared to broken cell preparations, which may affect coupling to second messenger systems, and/or species-specific differences in the P T H receptor (rat versus canine). To begin to distinguish among these possibilities and to test the hypothesis that species-specific differences among renal PTH/PTHrP receptors exist, we have investigated the receptor binding, adenylate cyclase stimulating, and photoaffinity labeling properties of PTHrP and PTH in rat and human renal cortical membranes (RCM) and compared them to findings observed previously in canine RCM. Although pharmacologic characterization of the P T H receptor has been described in detail for bovine, canine, rabbit, and chicken kidney membranes,f'6-'9) opossum and monkey kidney cells,(20.21) and rat and avian bone cells, ( 2 2 - 2 4 ) little information is available describing PTH/ P T H r P receptor binding in rat and human kidney. Rat kidney was selected for examination because of the available information on PTH/PTHrP binding and adenylate cyclase stimulation in rat bone cells (UMR and ROS). Human kidney was selected since events observed in this tissue would most closely mirror events in humans in vivo. Our results indicate that signal transduction by PTHrP is impaired in canine renal membranes but not in membranes from rat or human kidney and suggest that the previously reported differences between rat bone and canine kidney assay systems, at least in part, may be accounted for by species-specific differences in the efficiency of signal transduction.

MATERIALS AND METHODS

Pept ides (TyP)human PTHrP-( 1-36)amide [PTHrP-( 1-36)] was prepared by solid-phase synthesis as previously described. ( I 5 ) Synthetic (Nlee~1sTyr34)humanPTH-( 1-34) [NNT-hPTH-(1-34)] and bovine PTH-(bPTH) (1-34) were purchased from Bachem Inc. (Torrance, CA). The peptide concentration for all peptides used is given as the value determined by amino acid analysis, not as the dry weight of the peptide. The same batches of peptides were used in these studies as in the previously reported studies using canine RCM.('2)

Radioiodination Radioiodination of PTHrP-( 1-36) and NNT-hPTH-( 134) was performed using a modification of the lactoperoxidase method as previously described.(") Purification of radioligands was accomplished by reverse-phase high-performance liquid chromatography (HPLC) using a 30 cm pBondapak C18 column (Waters Associates, Milford, MA).(12' We have shown previously that both radioligands prepared and purified in this manner are composed almost exclusively of the monoiodinated form.' 1 2 ) The specific ac-

tivity, calculated as described previously,c1z1was 379 pCi/pg for PTHrP and 397 pCi/pg for PTH at the time of iodination. Both radioligands displayed full biologic activity in the canine renal adenylate cyclase assay"') when compared to the corresponding unlabeled peptide.

Preparation of renal membranes Highly purified rat renal cortical membranes (rat RCM) were prepared using discontinuous sucrose gradient ultracentrifugation as described previously for canine renal membranes.'lz1 All steps were performed in the presence of the following protease inhibitors: aprotinin (10 kIU/ml), pepstatin ( 5 gg/ml), leupeptin (25 or 45 pg/ml), and phenylmethanesulfanylfluoride (PMSF, 10 pg/ml), with or without N-ethylmaleimide (NEM, 1 .O mM). Membranes prepared with NEM, which has been previously shown to prevent degradation of the canine renal PTH re~ e p t o r , 'were ~ ~ ) used only for cross-linking studies to assess the size of the receptor in its undegraded form. Since NEM abolishes adenylate cyclase activity,(26'it could not be used in the membranes prepared for binding and adenylate cyclase studies. Highly purified human renal cortical membranes (hRCM) were prepared as for rat kidney, employing the following protease inhibitor combination in all steps: leupeptin (45 pg/ml), pepstatin ( 5 pg/ml), aprotinin (10 klU/ml), and PMSF (10 pg/ml). Normal human kidney cortex was obtained with Yale Human Investigation Committee approval from a nephrectomy specimen removed for localized transitional cell carcinoma. The specimen was promptly placed on ice at the time of removal and transported to the laboratory. The kidney was obtained from an otherwise healthy 28-year-old female. Renal function as assessed by serum creatinine and renal computed tomographic (CT) scan with pyelography was normal. Canine renal membranes for use in cross-linking studies were prepared in the same manner utilizing the protease inhibitors leupeptin ( 5 pg/ml), aprotinin (10 klU/ml), pepstatin ( 5 pg/ml), N-ethylmaleimide (1.0 mM), and PMSF (10 pg/ml) in all steps as previously described.'") Protein was measured by the method of Lowry using bovine serum albumin (BSA) as standard.(27)

Receptor binding studies Binding assays"') for rat and human RCM were conducted in siliconized 12 x 75 mm borosilicate glass test tubes at 30°C in a final volume of 0.2 ml. The binding buffer consisted of 50 mM Tris-HCI (pH 7 . 9 , 4.2 mM MgCI,, 0.3% BSA, 26 mM KCI, bacitracin (200 pg/ml), chymostatin (50 pg/m1),f28)approximately 60-80 x lo3 cpm per tube of radioligand, and, when appropriate, unlabeled peptides. Binding was initiated by adding 20 jig rat membranes or 18 pg human membranes to a final concentration of 100 or 90 pg/ml, respectively. At the end of the incubation period 50 81 triplicate aliquots were layered onto 300 pl of iced binding buffer containing 0.3% BSA in 500 pl

RENAL PT'H/PTHrP RECEPTORS

28 1

polypropylene tubes. The tubes were centrifuged at ap- (SDS), and 20% glycerol and analyzed by fluorography proximately 16,000 x g for 3 minutes at 4°C in a micro- using an 8-l8% SDS-polyacrylamide gel as previously decentrifuge. The supernatant was aspirated, and the tip of scribed. ( ' I ) the tube containing the membrane-associated radioligand was cut off. Radioactivity in both the pellet and superna- Adenylate cyclase assay tant was measured. Recovery of both radioligands from inAdenylate cyclase stimulating activity was examined cubation and wash tubes was routinely in excess of 95%. Total binding (TB) of [ L'"I](Tyr36)hPTHrP-(1-36)amide using a guanyl nucleotide-amplified PTH-sensitive ade(['251]PTHrP-(1-36)) to rat RCM varied between 18 and nylate cyclase assay, performed as previously described in with the following 22% of total counts added and nonspecific binding (NSB) detail for canine renal ranged from 3.5 to 5 % . Total binding in human RCM modifications. Briefly, synthetic PTHrP-( 1-36) or bPTHranged from 11 to 20% and NSB ranged from 2.4 to 4%. (1-34) was added in duplicate to assay tubes containing rat Specific binding of [12SI]PTHrP-( 1-36) reached equi- or human RCM, and the conversion of [cY-~'P]ATPto librium by 60 minutes in rat RCM and 30 minutes in hu- [32PP]cAMPwas examined. The incubations were carried man RCM at 30°C. These incubation times were used for out under conditions identical to those employed in the binding assays for each species (including the presence of subsequent equilibrium binding competition studies. bacitracin and chymostatin). Results are expressed as the percentage increment in adenylate cyclase activity in tubes Stability of radioligand containing the peptides compared with tubes containing Degradation or loss of radioligand in the presence of rat vehicle only. or human RCM was examined in two ways. After incubating [12s1]NNT-hPTH-(1-34)or [1zsI]PTHrP-(1-36)(4 x Data analysis los cpm/ml) with membranes under equilibrium binding Dissociation constants (I(&) were determined by Scatassay conditions, the suspension was centrifuged and the supernatant was collected for rebinding studies and HPLC chard analysis'3o)of the data obtained from competitive analysis. Rebinding of radioligand in the supernatant was binding experiments using radioligand and increasing concompared to binding of identical concentrations of "fresh" centrations of unlabeled ligand. In competition studies radioligand in another aliquot of the same membrane using an unlabeled competitor different from the radiopreparation. RP-HPLC analysis of supernatant radioactiv- ligand, binding affinities ( K , ) were derived from the ICso ity from binding experiments was compared to that of un- (concentration of unlabeled ligand displacing 50% of speStaused radioligand, and the percentage of radioligand coelut- cific radioligand binding), as described ing with the authentic, fresh radioligand was calculated. tistical differences were assessed by a paired two-tailed Student's t-test. Further analysis of competition curves was Values for specific rebinding for [1z51](Nles.'8Tyr34)hPTHcarried out with the LIGAND computerized least-squares (1-34) ([""I]PTH) and for ['zSI](Tyr36)hPTHrP-(1-36)NH, (['2sI]PTHrP) were 89.5 =t 8.7 and 80.4 f 1.9% for rat nonlinear curve-fitting program of Munson and Rodrenal membranes, respectively, and 84.0 + 7.0 and 95.3 + bard. ( 3 2 ) 6.8% for human membrnaes, respectively. HPLC coelution of supernatant radioactivity with fresh radioligand was 81.9% for [""I]PTH and 80.2% for ['z51]PTHrPin rat renal membranes and 99.2 and 86.8%, respectively, for RESULTS human membranes. These observations indicate that Binding studies neither amino-terminal PTHrP nor PTH analog was substantially degraded under the assay conditions. Binding of [12s1]PTHrP-(1-36)to canine RCM was competitively inhibited by increasing concentrations of unlabeled PTHrP-( 1-36) or bPTH-( 1-34) under equilibrium Photoaffinity labeling conditions, as previously reported'") (Fig. 1). Both pepAliquots of the membrane preparation (75 pg) were mi- tides displayed nearly equivalent affinity (Table 1). For crocentrifuged at 4°C for 4 minutes at 16,000 x g and re- purposes of comparison to published results in canine suspended in photoaffinity buffer (125 mM NaCI, 1.2 mM renal membranes, these peptides were also chosen for comCaCI,, 1.25 mM MgCI, 0.1 Vo BSA, and 50 mM HEPES at petition studies in rat and human RCM. PTHrP-(1-36) a final pH of 7.5) to a final volume of 300 pl. Membranes and bPTH-(1-34) were equipotent in inhibiting binding of were incubated with 500 x 103 cpm of either [1251]NNT- [1251]PTHrP-( 1-36) to rat RCM, with mean equilibrium hPTH-(1-34) or ['251]PTHrP-(I-36) for 60 minutes at binding dissociation constants of 3.6 f 0.8 and 3.7 f 0.7 25"C, with or without unlabeled hormone. The photoreac- nM, respectively (Fig. 1 and Table 1). Competitive inhibitive cross-linking reagent hydroxysuccinimidyl-p-azidoben- tion of ['z51]PTHrP-(l-36) binding to human RCM (Fig. 1) zoate (HSAB, Pierce Chemical Co., Rockford, IL), freshly yielded binding constants of 0.5 f 0.1 nM for PTHrPdissolved in Me2S0, was added 150 (vol/vol) to a final (1-36) and 0.9 f 0.1 nM for bPTH-(1-34) (Table I). This concentration of 500 pM.(25)After photolysis and centrifu- difference (1 .&fold), although small, was statistically siggation the pellets were resuspended in gel buffer containing nificant (p < 0.01) and was observed in three separate ex125 mM Tris-HC1 (pH 6.8), 4% sodium dodecyl sulfate periments.

282

ORLOFF ET AL.

Human

Rat

Canine

_.

1201

1201

0

Log[Cornpetitor] M

8

,

-10

-9

-8

-7

-6

Log(Cornpetitor1M

Log[Cornpetitor] M

0201L

0 301

0301

I

I

0 1 2 h

l

h

000

004

OW 0 10

000'

00

08

04

12

16

20

008

012

Bound inMl

Bound (nM)

FIG. 1. (Top) Competitive inhibition of ['zSI](Tyr36)hPTHrP-( 1-36)NH2 binding under equilibrium conditions to canine, rat, and human renal membranes by unlabeled (TyP)hPTHrP-(1-36)NH2 (0)or bPTH-(1-34) ( 0 ) .Values are the mean f SEM of triplicate determinations for a single experiment. Results are representative of at least three separate experiments for each species. (Bottom) Corresponding Scatchard plots of representative binding experiments are shown for each species. The results for canine renal membranes have been reported previously'") and are shown for comparison to results in rat and human renal membranes.

TABLE 1. INTERSPECIES COMPARISON OF RENAL PTH/PTHrP RECEPTORS~

Assay /species Binding ( K d / K i ) ,nM Canine Rat Human Adenylate cyclase (K,,,), nM Canine Rat Human Photolabeling (Mr), kD Canine Rat Human

bPTH-(I-34) 10.5 3.7 0.9b 0.13c 0.09 0.02d 85 80 85

Tyr36hPTHrP-(I-36)NH2 14.0 3.6 0.9

2.wc 0.10 0.01d

85 80 85

aIn vitro activity of (TyP)hPTHrP-(l-36)NHZcompared to bPTH-(1-34) in renal cortical membranes from three species using ['251)(Tyr'6)hPTHrP-(1-36)NHZ as radioligand. The Kd values were determined by Scatchard analysis, and the K; values

were derived from the ICsovalues as described previously. l3O) Peptide stability was documented under assay conditions for each species. Values represent the mean + SEM for two or more separate experiments. The results for canine renal membranes have been previously reported'") and are shown for comparison to rat and human data. bP < 0.01 (PTH versus PTHrP). CP < 0.05 (PTH versus PTHrP). dP < 0.01 (PTH versus PTHrP).

283

RENAL PTH/PTHrP RECEPTORS Corresponding Scatchard plots of representative binding experiments for each species are shown in the lower panel of Fig. 1. Analysis of the data with the LIGAND computer program was compatible with a single class of high-affinity receptor sites for each species. The receptor numbers calculated from B,,, values were 5.08 + 0.56, 7.10 1.90, and 0.84 f 0.02 pmol/mg membrane protein for canine, rat, and human RCM, respectively.

*

Adenylatt? cyclase stimulating activity Stimulation of adenylate cyclase by bPTH-( 1-34) and PTHrP-(1-36) was assessed for each species in the same membrane preparations and incubation conditions used in the binding assay and under conditions in which both peptides were shown to be stable (see Materials and Methods). In the canine assay, as previously reported,(") PTHrP-(1--36) ( K , = 2.00 + 0.17 nM) was significantly less potent (15-fold) than bPTH-(1-34) ( K , = 0.13 f 0.01 nM; P < 0.05; Fig. 2 and Table 1). In contrast, PTHrP-( 1-36) and bPTH-(1-34) were equipotent in the rat renal adenylate cyclase assay, with K,, values of 0.10 f 0.03 and 0.09 + 0.03 nM, respectively (Fig. 2 and Table 1). These results are analogous to their equivalent potency in the binding assay. In the human renal adenylate cyclase 0.002 nM for PTHrP-(1-36) and assay the K , were 0.01 0.02 f 0.001 nM for bPTH-(l-34). PTHrP-(1-36) was 2-fold more potent than bPTH-(1-34) (p < 0.01; Fig. 2 and Table l), a relationship nearly identical to the relative potency of the two peptides observed in the binding assay (1.8-fold).

Photoaffinity labeling Photoaffinity cross-linking of [IzsI]PTHrP-(1-36) to receptors in RCM from each species isolated under full protease protection was visualized by autoradiography after

SDS-polyacrylamide gel electrophoresis. A slight difference in electrophoretic mobility was observed between the canine P T H receptor (M,. G 85 kD) and the rat PTH receptor ( M , 1 80 kD) isolated in the presence of multiple protease inhibitors (Fig. 3). In the absence of NEM but in the presence of leupeptin in high concentrations (25 or 45 pg/ml), the major band in rat membranes remains at 80 kD, with modest degradation to a 65-70 kD form (not shown). When leupeptin is present in the concentrations described, the 85 kD receptor in canine renal membranes is also preserved in the absence of NEM (not shown)."" The size of the human renal PTH receptor is indistinguishable from that of the canine renal PTH receptor and is also slightly larger than the rat renal PTH receptor (Fig. 3). Photoaffinity labeling of RCM by ["'I]NNT-hPTH-( 1-34) (canine) or ["SI]PThrP-(l-36) (rat and human) in the presence of increasing concentrations of unlabeled PTHrP(1-36) or bPTH-( 1-34) demonstrated progressive inhibition of binding to the 80-85 kD receptor for each species (Fig. 4), corroborating the binding affinities for these peptides in the conventional binding analysis and indicating that the labeled protein represents the PTH receptor.

DISCUSSION Substantial evidence exists to support a pathophysiologic role for PTHrP in most instances of humoral hypercalcemia of malignancy. A variety of bioassay systems have shown PTHrP to mimic the effects of PTH, including activation of kidney and bone adenylate cyclase, stimulation of fetal or neonatal bone resorption, promotion of renal phosphate excretion, stimulation of renal Icu-hydroxylase, and induction of hypercalcemia in vivo.'sl Further, PTHrP bind to receptors in bone(I314' and kidney'" I I ) that are indistinguishable from PTH receptors. There is no firm evidence at the present time to support the existence

m

Canine

Human

400 1

700 -

400 300 200 100 -

600-

500

500

400. 300

~

200 -

100-

0 12

~

-11

-10

-9

-8

Log[Compelitor] M

-7

-12

1

-7

-11 -10 -9 -8 Log[CompetitorJM

0

1 -12

-11 -10 -9 Log[Competitor] M

-8

FIG. 2. Stimulation of adenylate cyclase activity by ( T Y ~ - ' ~ ) ~ P T H ~ P - ( I - ~(U) ~)N orHby, bPTH-(1-34) ( 0 )in canine, rat, and human renal membranes. Assays were performed in the same membrane preparations and under the same conditions employed in the binding assays for each species. Values are the mean SEM of duplicate determinations for a single experiment, and results are representative of those obtained in at least two separate experiments for each species. The results in canine renal membranes have been reported previously'"' and are shown for comparison purposes.

*

ORLOFF ET AL.

284

Canine

Rat

CANINE

Human

946a-

31 41

- - - - PTHRP

- - - - - -

0.1 I

10 lo2 lo3

21 -

I4 -

PTH

- - - -

I

10 102

PTHRP FIG. 3. Photoaffinity cross-linking of [1zsI](Tyr'6)hPTHrP-(1-36)NHz to receptors in canine, rat, and huHUMAN man renal membranes isolated in the presence of multiple protease inhibitors was visualized by autoradiography after SDS-polyacrylamide gel electrophoresis. Photolabeling was performed in the presence (+) or absence (-) of excess unlabeled (Tyr3')hPTHrP-( 1-36)NHL M). The M,. standards ( x 1ooO) are indicated. The arrows point out the slight difference in electrophoretic mobility between the PTH 0.1 I 10 102 canine and rat renal PTH/PTHrP receptors. The human PTHRP 0.1 I lo2 renal PTH/PTHrP receptor is indistinguishable in size from the canine receptor. Identical results were obtained with [1zsl](Nle818Tyr34)hPTH-( 1-34) as radioligand (not FIG. 4. Photoaffinity labeling of canine, rat, and human shown). 1-34) (carenal membranes with [Izsl](Nle8~'8Tyr"')hPTH-( 1-36)NHz (rat and human) nine) or [1zsI](Tyr3b)hPTHrP-( in the presence of increasing concentrations (nM) of bPTH-(1-34) ( P T H ) or (Tyr'")hPTHrP-(l-36)NH2 of unique receptors for PTHrP in these classic PTH target (PTHrP). The M , standards ( x 1ooO) are indicated, and organs. However, striking relative differences in efficiency the arrows indicate the bands of interest. Similar results in of signal transduction between PTH and PTHrP have each species were obtained when using either [IzSI]PTHor been demonstrated by several groups: whereas synthetic [iz51]PTHrPas radioligand (not shown). amino-terminal PTHrP are equipotent to PTH in receptor binding affinity in canine renal membranes and are equipotent to or more potent than PTH both in binding and activation of adenylate cyclase in rat bone cell protease activity in the rat renal cortex.(1i.3J-"8) Similarly, a assays, 19-1 1 , 13. IS) they are 5- to 15-fold less potent in stimu- report describing stimulation of CAMP production by lating adenylate cyclase activity in canine renal mem- P T H in primary cultures of rat proximal convoluted tubule branes. < 1 1 . 1 3 , l S ) We have been interested in the possibility cells demonstrated a high K,,, value (100 nM).[.l9) Two that species-specific differences in P T H receptors account studies have reported high-affinity P T H binding (2-4 nM) for these relative differences in signal transduction. The in rat renal membranes,[".'01 although in these no attempt current study was undertaken to address these possible in- was made to control for proteolytic degradation of terspecies differences and because no information on the ligands. The ability of PTHrP-(1-34) to stimulate adeinteractions of P T H r P with human renal receptors exists. nylate cyclase activity in rat renal membranes was equivaWe sought to characterize PTH/PTHrP receptor proper- lent to that of hPTH-(1-34) in one report'"' ( K , = 100 ties in rat and human RCM and compare them using the nM) but less potent in another("8L( K , = 30 versus 5 nM). same methods to results we previously published employ- In these studies, again, precautionary measures to avoid ing canine RCM.'Iz1 the proteolytic metabolism of peptides were not taken and Previous attempts at characterizing the PTH-sensitive may have led to the non-physiologically high K , observed. adenylate cyclase assay system in rat renal cortical mem- Characterization of PTH/PTHrP receptor binding and branes demonstrated adenylate cyclase systems with little stimulation of adenylate cyclase in human kidney memsensitivity ( K , = 10-lo00 nM), reflecting the abundant branes has not been reported.

- - - lo

- - - -

RENAL PTH/PTHrP RECEPTORS In this report we describe the receptor binding, adenylate cyclase activation, and photoaffinity labeling characteristics of amino-terminal PTH and PTHrP analogs in highly purified rat and human RCM and compare them to results observed previously using canine RCM. The use of the protease inhibitors bacitracin and chymostatin prevented significant proteolytic degradation of PTH and PTHrP by canine, rat, and human membranes under the conditions employed for binding studies, as well as for the adenylate cyclase assays. This allowed direct comparison of potency in these assays without the potential artifact introduced by degradation of the peptides. The receptor binding properties of bPTH-( 1-34) and PTHrP-( 1--36) in the rat renal cortical membrane system using ['251]PTHrP-(1-36) as radioligand demonstrates superimposable competition curves and nearly identical binding affinity constants for the two peptides (3.7 and 3.6 nM, respectively). These values are nearly identical to the values for P T H binding reported in two recent studies emand are similar to the ploying rat renal relative affinity of these peptides in the canine RCM binding assay.f81z 1 3 ) Each unlabeled peptide reduced the binding of radioligand to the same degree, providing no evidence for the existence of receptor subtypes. Scatchard analysis was consistent with a single class of high-affinity receptor sites. Competitive inhibition of [1z51]PTHrP-(l-36)binding to human renal membranes by bPTH-(1-34) and PTHrP(1-36) yielded small but statistically significant differences in binding affinity. PTHrP-(1-36) (Kd = 0.5 nM) was 1.8-fold more potent than bPTH-(1-34) (K, = 0.9 nM) in inhibiting radiolabeled PTHrP binding. Scatchard analysis again indicated a single class of high-affinity receptors. The receptor number calculated from the B,,, value in human RCM was lower than that calculated for canine and rat renal membranes. This difference may relate to a higher affinity receptor in human membranes, to varying degrees of membrane purification, and/or to inherent differences in the number of receptor sites among species. A direct comparison of specific binding is misleading, since receptor number as calculated from Scatchard analysis is also dependent on receptor affinity and membrane protein concentration, which differed among the three species. Differences in binding affinities among species were apparent: bPTH-( 1-34) and PTHrP-( 1-36) displayed higher binding potency in human compared to canine or rat renal membranes. This difference is not due to greater proteolytic degradative activity in canine or rat membranes, since stability of these peptides under binding assay conditions was demonstrated. A preliminary report from another laboratory suggests that P T H analogs are also more potent in human than in rat bone cells.'41) The relative potency of bPTH-(1-34) and PTHrP-(1-36) in the rat and human binding assays was also reflected in their coupling to postreceptor events and contrasted with results observed in canine RCM. Both peptides were equipotent in stimulating adenylate cyclase in rat renal membranes, but PTHrP-(1-36) was slightly more potent (2fold) than bPTH-( 1-34) in human renal membranes. The greater potency for agonist stimulation of adenylate cy-

285 clase than for receptor binding has been previously observed for P T H receptors and other receptor-cyclase coupling systems''2 4 1 ) and may reflect lowering of the K , by guanyl nucleotide amplification or a "spare receptor" The results in rat and human membranes, demonstrating nearly equivalent receptor-effector coupling between PTH and PTHrP, contrast with those observed in canine renal membranes. In canine renal membranes less efficient signal transduction is apparent for PTHrP, since P T H r P and PTH bind equivalently, yet PTHrP are 5- to 15-fold less potent in stimulating adenylate cyclase.'8 I ' PTH/PTHrP receptor interactions previously described in bovine renal membranes differ from those in canine, rat, and human membranes in that PTHrP-( 1-34) was substantially (approximately 10-fold) less potent than bPTH-(1-34) in both the binding and adenylate cyclase assays. l o ) However, the efficiency of signal transduction appears to be roughly equivalent for both peptides in bovine renal membranes, an observation similar to those in rat and human but different from those in canine RCM. Thus receptors from several different species d o not bind and transduce human and bovine peptides with the same relative affinity and potency. The use of other P T H and PTHrP analogs in these assays might reveal more pronounced interspecies differences among renal PTH/PTHrP receptors. Nonetheless, these studies indicate that differences are apparent when using two standard bPTH-( 1-34) and amino-terminal hPTHrP analogs. Interestingly, the results observed in rat renal membranes demonstrating equivalent relative potency for the two peptides in the binding and adenylate cyclase assays are similar to those described in rat osteosarcoma cells. '-I5) These findings suggest that bone and renal PTH receptor heterogeneity may not exist in the rat. Further, these observations strongly suggest that species-specific differences in PTH receptor signaling mechanisms account for the previously described differences in signal transduction between rat bone cells and canine kidney assay systems. Photoaffinity covalent labeling of rat and human PTH/ PTHrP receptors was performed using the heterobifunctional cross-linking reagent, HSAB, as previously reported in canine RCM. In the presence of a combination of protease inhibitors including NEM, which along with leupeptin appears to be important in protecting the intact canine renal PTH receptor from degradation,'") the rat receptor appears as an 80 kD protein. This is slightly smaller in size than the 85 kD canine receptor but analogous to the 80 kD receptor protein reported in rat osteosarcoma 411 The human renal PTH receptor protein labeled in the same manner appears as an 85 kD band, indistinguishable in size from the canine renal PTH receptor. Competitive inhibition by unlabeled PTHrP or P T H confirms that the labeled protein in each species demonstrates the high affinity expected of a physiologically relevant receptor. In summary, species-specific differences in the relative affinity and potency of amino-terminal bovine PTH and human PTHrP peptides are observed in canine, rat, and human renal cortex. The impairment of signal transduction by PTHrP in canine renal membranes but not in

286

ORLOFF ET AL.

membranes from other species may account, at least in part, for the previously reported differences between rat bone cells and canine renal membranes. These studies document the existence of species specificity among PTH receptor and signal transduction systems, but they do not exclude the possibility of tissue-specific differences in P T H receptors. These findings suggest that an entirely homologous assay system employing human renal and skeletal tissues and human peptides would most faithfully reflect events related to the physiology and pathophysiology of PTH and P T H r P and to the humoral hypercalcemia of malignancy.

1 1 . Yates AJP, Guttierez GE, Smolens P , et al. 1988 Effects of a

synthetic peptide of PTH-related protein on calcium homeostasis, renal tubular calcium reabsorption, and bone metabolism in vivo and in vitro in rodents. J Clin Invest 81:932-938. 12. Orloff J J , Wu TL, Heath HW, Brady TG, Brines ML, Stewart A F 1989 Characterization of canine renal receptors for the parathyroid hormone-like protein associated with humoral hypercalcemia of malignancy. J Biol Chem 264: 6097-6103. 13. Nissenson RA, Diep D, Strewler G J 1988 Synthetic peptides

comprising the amino-terminal sequence of a PTH-like protein from human malignancies. J Biol Chem 263:1286612871. 14. Jueppner H , Abou-Samra AB, Uneno S, Gu WX, Potts JT, Segre GV 1988 The PTH-like peptide associated with hu-

REFERENCES We thank Ms. Kristie Vollono and Ms. Charleen Stewart for expert secretarial assistance. Supported by the Veterans Administration, West Haven, CT, and NIH Grant 1 KO8 DK 01936-01. Presented in part at the Annual Meeting of the American Federation for Clinical Research, May 4-7, 1990, Washington, DC.

5.

6.

17.

REFERENCES 1. Burtis WJ, Wu T, Bunch C , et al. 1987 Identification of a

2.

3.

4.

5.

6.

7.

8.

novel 17,000-dalton PTH-like adenylate cyclase stimulating protein from a tumor associated with humoral hypercalcemia of malignancy. J Biol Chem 262:7151-7156. Stewart AF, Wu T , Goumas D, Burtis WJ, Broadus AE 1987 N-terminal amino acid sequence of two novel tumor-derived adenylate cyclase-stimulating proteins: Identification of parathyroid hormone-like and parathyroid hormone-unlike domains. Biochem Biophys Res Commun 146:672-678. Moseley JM, Kubota M, Diefenbach-Jagger H , et al. 1987 Parathyroid hormone-related protein purified from a human lung cancer cell line. Proc Natl Acad Sci USA 845048-5052, Strewler G J , Stern P H , Jacobs JW, et al. 1987 Parathyroid hormone-like protein from human renal carcinoma cells: Structural and functional homology with parathyroid hormone. J Clin Invest 80:1803-1807. Suva LJ, Winslow GA, Wettenhall REH, et al. 1987 A parathyroid hormone-related protein implicated in malignant hypercalcemia: Cloning and expression. Science 2372393-896. Mangin M, Webb AC, Dreyer B, et al. 1988 Identification of a cDNA encoding a parathyroid hormone-like peptide from a human tumor associated with humoral hypercalcemia of malignancy. Proc Natl Acad Sci USA 85597-601. Theide MA, Strewler G J , Nissenson RA, Rosenblatt M, Rodan G A 1988 Human renal carcinoma expresses two messages encoding a PTH-like peptide: Evidence for the alternative splicing of a single copy gene. Proc Natl Acad Sci USA 85 :4605 -4609. Orloff J J , Wu TL, Stewart A F 1989 PTH-like proteins: Biochemical responses and receptor interactions. Endocr Rev 10: 476-495.

9. Kemp BE, Moseley JM, Roda C P , et al. 1987 PTH-related

protein of malignancy: Active synthetic fragments. Science 238: 1568-1570. 10. Horiuchi N, Caulfield MD, Fisher JE, et al. 1987 Similarity of synthetic peptide from human tumor to P T H in vivo and in vitro. Science 138:1566-1568.

18.

19.

moral hypercalcemia of malignancy and parathyroid hormone bind to the same receptor o n the plasma membranes of ROS 17/2.8 cells. J Biol Chem 26323557-8560. Stewart AF, Mangin M, Wu T , Goumas D, lnsogna KL, Burtis W J , Broadus AE 1988 A synthetic human parathyroid hormone-like protein stimulates bone resorption and causes hypercalcemia in rats. J Clin Invest 81596-600. Zull JE, Malbon CC, Chuang J 1977 Binding of tritiated bovine parathyroid hormone to plasma membranes from bovine kidney cortex. J Biol Chem 252:1071-1078. Segre GV, Rosenblatt M, Reiner BL, Mahaffey JE, Potts JT J r 1979 Characterization of parathyroid hormone receptors in canine renal cortical plasma membranes using a radioiodinated sulfur-free hormone analogue. Correlation of binding with adenylate cyclase activity. J Biol Chem 254:6980-6698. Nissenson RA, Arnaud C D 1979 Properties of the parathyroid hormone receptor-adenylate cyclase system in chicken renal plasma membranes. J Biol Chem 254:1469-1475. Kremer R, Bennett H P J , Mitchell J , Goltzman D 1982 Characterization of the rabbit renal receptor for native parathyroid hormone employing a radioligand purified by reversephase liquid chromatography. J Biol Chem 257:14048-

14054. 20. Teitelbaum AP, Strewler GJ 1984 Parathyroid hormone re-

ceptors coupled to cyclic adenosine monophosphate formation in an established renal cell line. Endocrinology 114980985. 21. Wright BS, Tyler GA, O'Brien R, Caporale LH, Rosenblatt M 1987 Immunoprecipitation of the parathyroid hormone receptor. Proc Natl Acad Sci USA 84:26-30. 22. Pliam NB, Nyiredy KO, Arnaud C D 1982 Parathyroid hor-

mone receptors in avian bone cells. Proc Natl Acad Sci USA 79:206-2063. 23. Rizzoli RE, Somerman M, Murray T M , Aurbach G D 1983

Binding of radioiodinated parathyroid hormone to cloned bone cells. Endocrinology 113:1832-1883. 24. Yamamoto I, Shigeno C, Potts J T Jr, Segre GV 1988 Characterization and agonist-induced down-regulation of parathyroid hormone receptors in clonal rat osteosarcoma cells. Endocrinology 122: 1208-1217. 25. Nissenson RA, Karpf D, Bambino T , et al. 1987 Covalent labeling of a high-affinity, guanyl nucleotide sensitive parathyroid hormone receptor in canine renal cortex. Biochemistry 26: 1874- 1878. 26. Teitelbaum AP, Nissenson RA, Arnaud C D 1982 Coupling

of the canine renal parathyroid hormone receptor to adenylate cyclase: Modulation by guanyl nucleotides and Nethylmaleimide. Endocrinology 111: 1524- 1533. 27. Lowry OH, Rosebrough NJ, Farr AL, Randall RS 1951 Protein measurement with the folin phenol reagent. J Biol Chem 193:265-275.

RENAL PTH/PTHrP RECEPTORS 28. Yamaguchi T, Fukase M, Nishikawa M, Fujimi T, Fujita T 1988 Parathyroid hormone degradation by chymotrypsin-like endopeptidase in the opossum kidney cell. Endocrinology 123~2812-2817. 29. Stewart AF, lnsogna KL, Goltzman D, Broadus AE 1983 Identification of adenylate cyclase-stimulating activity and cytochemical bioactivity in extracts of tumors from patients with humoral hypercalcemia of malignancy. Proc Natl Acad Sci USA 80:1454-1458. 30. Scatchard G 1949 The attraction of proteins for small molecules and ions. Ann NY Acad Sci 51:660-672. 31. Cheng Y , Prusoff W H 1973 Relationship between the inhibition constant ( K , ) and the concentration of inhibitor which causes SO per cent inhibition ( I r oof) an enzymatic reaction. Biochem Pharmacol 22:3099-3 108. 32. Munson P J , Rodbard D 1980 LIGAND: A versatile computerized approach for chracterization of ligand-binding systems. Anal Biochem 107:220-239. 33. Chu LLH, Forte LR, Anast CS, Cohn DV 1975 Interaction of parathyroid hormone with membranes of kidney cortex: Degradation of the hormone and activation of adenylate cyclase. Endocrinology 97:1014-1023. 34. Goltzman D, Peytremann A, Callahan EN, Segre GV, Potts J T J r 1976 Metabolism and biological activity of parathyroid hormone in renal cortical membranes. J Clin Invest 57:8-19. 35. DiBella FP, Arnaud CD, Brewer HB Jr 1976 Relative biologic activities of human and bovine parathyroid hormones and their synthetic, NH,-terminal (1-34) peptides, as evaluated in vitro with renal cortical adenylate cyclase obtained from three different species. Endocrinology 99:429-

._-.

Alh

36. Cloix J-F, d’Herbigny E, Ulmann A 1980 Renal parathyroid hormone-dependent adenylate cyclase in vitamin D-deficient rats. J Biol Chem 255:11280-11283. 37. Mitchell J , Tenenhouse A, Warner M, Goltzman D 1988 parathyroid hormone desensitization in renal membranes of

287 vitamin D-deficient rats is associated with a postreceptor defect. Endocrinology 122: 1834- 1841. 38. Rabbani SA, Mitchell J , Roy DR, Hendy GN, Goltzman D 1988 Influence of the amino-terminus on in vitro and in vivo biological activity of synthetic parathyroid hormone-like peptides of malignancy. Endocrinology 123:2709-27 16. 39. Rao LG, Yau CH, Kovacs K 1989 Dog and rat kidney proximal tubule cells in culture: Responses to parathyroid hormone. J Bone Miner Res 4:293-303. 40. Nickols GA, Metz-Nickols MA, Pang PKT, Roberts MS, Cooper CW 1989 Identification and characterization of parathyroid hormone receptors in rat renal cortical plasma membranes using radioligand binding. J Bone Miner Res 4:615623. 41. McKee RL, Caulfield MP, Fisher JE, Duong LT, Rosenblatt M 1989 Differences in human and rat parathyroid hormone (PTH) receptors: Agonists and antagonists are 3- to 130-fold more potent in human derived systems (abstract). J Bone Miner Res 4:S341. 42. Levitzki A 1986 Beta-adrenergic receptors and their mode of coupling to adenylate cyclase. Physiol Rev 662319-854. 43. Shigeno C, Hiraki Y, Westerberg DP, Potts J T Jr, Segre GV 1988 Photoaffinity labeling of parathyroid hormone receptors in clonal rat osteosarcoma cells. J Biol Chem 263:38643871.

Address reprint requests to: John J . Orloff, M . D . Research 151 West Haven VA Medical Center 950 Campbell Ave. Wesl Haven, CT 06516 Received for publication May 15, 1990; in revised form September 14, 1990; accepted October 29, 1990.