Ectopic 6-Adrenergic Receptors Coupled to Adenylate Cyclase in Human Adrenocortical Carcinomas*
MICHAEL S. KATZ, THOMAS M. KELLY, ELIZABETH M. DAX, MARCO A. PINEYRO, JOHN S. PARTILLA, AND ROBERT I. GREGERMAN
Division of Endocrinology and Metabolism, Department of Medicine, The University of Texas Health Science Center at San Antonio, and Audie L. Murphy Memorial Veterans’ Hospital, San Antonio, Texas 78284; the Division of Endocrinology and Metabolism, Departments of Medicine, the LDS Hospital and University of Utah School of Medicine, Salt Lake City, Utah 84132; Gerontology Research Center, National Institute on Aging, National Institutes of Health at Baltimore City Hospitals, and the Departments of Medicine, Baltimore City Hospitals and Johns Hopkins University School of Medicine, Baltimore, Maryland 21224
ABSTRACT. The adenylate cyclase of an adrenocortical car- cinoma of the rat is activated not only by ACTH but also by ß- adrenergic agonists, which bind to ectopic ß-adrenergic receptors not present in normal rat adrenal cortex. Previous reports ex- amining possible ß-adrenergic control of adenylate cyclase in human adrenocortical carcinomas failed to demonstrate ß-ad- renergic receptor-linked enzyme activity. We studied six human adrenal carcinomas and normal adrenal cortex from three sub- jects for ß-adrenergic agonist-sensitive adenylate cyclase and ß- adrenergic binding sites. Three of the six carcinomas had ade- nylate cyclase responses to both ACTH and B-agonists. Two tumors were ACTH responsive but not -agonist responsive; one tumor responded to ß-agonists but not to ACTH. Adenylate cyclase activity of normal adrenal cortex from three subjects was stimulated by ACTH but not by ß-agonists. In membrane prep-
arations from three tumors with @-agonist-sensitive adenylate cyclase, the radiolabeled 8-adrenergic antagonist [125I]pindolol bound specifically and with high affinity (Ka = 38-83 pM) to a single class of binding sites which showed saturation with ligand concentration, reversibility of binding, pharmacological specific- ity, and stereospecificity. Normal cortex and one tumor without ß-adrenergic agonist-sensitive adenylate cyclase had no specific binding of [125I]pindolol. These results indicate that malignant transformation of adrenal cortex in man is frequently but not invariably associated with the appearance of ectopic ß-adrener- gic receptors functionally linked to adenylate cyclase. Loss of ACTH-responsive adenylate cyclase may also occur simultane- ously with the development of ß-adrenergic receptor-linked ad- enylate cyclase. (J Clin Endocrinol Metab 60: 900, 1985)
S TEROIDOGENESIS in the normal adrenal cortex is under the hormonal control of ACTH, which occupies a specific receptor on the plasma membrane of adrenocortical cells and activates a receptor-linked ade- nylate cyclase. The cAMP formed then initiates a cas- cade of metabolic events which ultimately result in en- hanced steroid synthesis. In contrast, the adenylate cy- clase of a well studied adrenocortical carcinoma of the rat responds not only to ACTH but also to ß-adrenergic agonists as well as glycoprotein hormones (1, 2). Specific ß-adrenergic receptor-binding sites in the membrane fraction from rat adrenocortical carcinoma but not nor-
Received October 23, 1984.
Address requests for reprints to: Dr. Michael S. Katz, Division of Endocrinology and Metabolism, Department of Medicine, The Univer- sity of Texas Health Science Center, 7703 Floyd Curl Drive, San Antonio, Texas 78284.
* A preliminary report of this work was presented at the Annual Meeting of the American Federation for Clinical Research, Washing- ton, D.C., May 1983, and was published in abstract form in 1983 [Clin Res 31:273A (Abstract)]. This work was supported in part by NIH Grant 5R01-AG-03168 and V. A. medical research funds.
mal rat adrenal cortex have also been found (3). From this work has emerged the concept of ectopic hormone receptors linked to adenylate cyclase in neoplastic adre- nal cortex. Several studies of human adrenal carcinomas failed to demonstrate ectopic ß-adrenergic control of adenylate cyclase. However, we now demonstrate in hu- man adrenocortical carcinomas the presence of both ß- adrenergic agonist-sensitive adenylate cyclase and 8- adrenergic receptor binding. Considerable heterogeneity of such tumors was apparent, since adenylate cyclase- linked ectopic ß-adrenergic receptors were frequent but not invariable and occurred with both preservation of ACTH-responsive adenylate cyclase and its complete loss.
Materials and Methods
Patients
Adrenocortical carcinomas from six patients (patients 1-6, Table 1) were studied. Subsequent metastatic lesions resected from two patients (no. 2 and 3) were also examined for adenyl-
ate cyclase activities. Normal adrenal cortex was obtained from three patients (no. 7-9) undergoing adrenalectomy for various reasons. Clinical and preoperative laboratory data on all pa- tients are shown in Table 1.
Materials
ATP (disodium salt), GTP (sodium salt), cAMP (sodium salt), creatine phosphate (disodium salt), creatine phosphoki- nase (rabbit muscle), (-)isoproterenol (+)bitartrate, (-)epi- nephrine (+)bitartrate, (-)norepinephrine bitartrate, D,L-pro- pranolol HCI, glucagon, theophylline, 3-isobutyl-1-methylxan- thine (IBMX), and BSA were purchased from Sigma Chemical Co. (St. Louis, MO). [a-32P]ATP, tetra (triethylammonium) salt (10-50 Ci/mmol), and [G-3H]cAMP, ammonium salt (20- 30 Ci/mmol), were obtained from New England Nuclear (Bos- ton, MA). [2,8-3H]cAMP, monosodium salt (15-30 Ci/mmol), and 5’-guanylyl imidodiphosphate (GMP-P(NH)P) were pur- chased from ICN (Irvine, CA). [8-14C]cAMP (35-50 Ci/mol) was obtained from Schwarz/Mann (Spring Valley, NY). (-)Pindolol was a generous gift from Sandoz, Ltd. (Basel, Switzerland); phenoxybenzamine was kindly provided by SK & F Co. (Philadelphia, PA), and (+)isoproterenol (+)bitartrate was supplied by Sterling-Winthrop Research Institute (Rens- selaer, NY). ACTH-(1-39) (porcine; 100-156 IU/mg) was pur- chased from Calbiochem (San Diego, CA) and Sigma; and ACTH (1-24) (Cortrosyn; 100 USP U ACTH/mg) was obtained from Organon (West Orange, NJ). TSH (bovine; 5-7 USP U/ mg) was obtained from Calbiochem. Ovine LH (1.03 U/mg) and hCG were obtained from the National Hormone and Pi- tuitary Program, NIADDK, NIH.
Tissue preparation
Each tissue examined was dissected from nonnecrotic areas adjacent to histologically confirmed adrenocortical carcinoma or normal adrenal cortex. Large blood vessels were excluded during preparation of samples for assay. Tissue specimens placed in ice-cold normal saline were either assayed immedi- ately for adenylate cyclase activity or frozen in liquid nitrogen or at -70 ℃ for later testing (see Results); all binding studies were conducted on tissue frozen at -70 ℃.
For adenylate cyclase assays, adrenal tissue was homogenized in 1-5 vol (wt/vol) medium containing 20 mM KCI, 20 mM NaCl, 1 mM MgSO4, and 2 mM glycylglycine, pH 7.4 (substi- tuted by 2 mM Tris-maleate, pH 7.4, for tissues from patients 5, 6, and 8). Tissue was routinely homogenized with a Dounce homogenizer (Kontes Co., Vineland, NJ), but when large vol- umes of tissue were available, a VirTis homogenizer (VirTis Co., Gardiner, NY) was used. Resulting homogenates were filtered through gauze or nylon mesh before assay. The partic- ulate fraction was prepared by centrifuging the homogenate at 10,000 x g for 5 min or 27,000 x g for 15 min and resuspending the washed pellet in the original volume of homogenizing medium.
For binding studies, frozen adrenal tissue was placed in 20 vol (wt/vol) of medium containing 0.154 M NaCl and 20 mM Na HEPES, pH 7.5, and homogenized in a Polytron (Brink- mann Instruments, Westbury, NY) using two 10-sec bursts at
setting 6. The homogenate was then filtered through 200 nylon mesh and centrifuged at 15,000 x g for 10 min. The resultant pellet was resuspended in the same medium at 1:10 to 1:100 (original wt/vol) and used in the binding assay.
Adenylate cyclase assay
Adenylate cyclase activities were measured as the conversion of [a-32P]ATP to [32P]cAMP. Twenty microliters of homoge- nate (60-358 µg protein) or the particulate fraction (17-340 µg protein) were routinely incubated in duplicate at 30 °℃ for 20 min in a total volume of 50 ul containing 1 mM ATP, 1 uCila- 32PJATP, 2 mM cAMP, 0.01 uCi [3H]CAMP, 5 mM MgCl2, 2 mM theophylline or 1 mM IBMX (see below), 20 mM Tris-HCI (pH 7.6), and an ATP-regenerating system consisting of 20 mM creatine phosphate and 100 U/ml creatine phosphokinase. In- cubation was terminated by the addition of 100 ul of a stopping solution (34 mM sodium lauryl sulfate, 40 mM ATP, and 12 mM cAMP, pH 7.6) and 50 ul containing 0.005 Ci [14C]CAMP and heating the total 200-ul volume in boiling water for 5 min. cAMP was then isolated by the two column chromatographic method of Salomon et al. (4). The routine protocol was simpli- fied in later studies (i.e. of tissues from patients 5, 6, 8, and 9). The incubation volume was increased to 100 ul, and incubation was terminated by the addition of 200 ul 10% perchloric acid. In these latter experiments, 0.04 uCi [3H]cAMP was added in 50 ul before application of the sample to the column; [14C] cAMP was omitted. Column recovery in all experiments ranged from 50-75%; the precise value in each tube was used to correct the cAMP yield to 100%.
In a number of studies (i.e. of tissues from patients 1-4 and 7), [14C]cAMP and [3H]cAMP recoveries were used as respec- tive monitors for losses during chromatography and incubation of the [32P]cAMP product, as previously described (5). Identical recoveries of [3H]CAMP and [14C]CAMP in any single tube indicated that there was no destruction of cAMP product during the incubation and that all product loss occurred during chro- matographic isolation. In several tissues, however, dispropor- tionately low [3H]cAMP recoveries indicated substantial cAMP destruction during incubation. For example, in tumor homog- nates from patient 2A (see Table 1), 3H recovery under routine assay conditions was so low (<15% of 14℃ recovery) that no [32P]cAMP production was measurable. 3H recovery was greater during assay of the particulate fraction. Recovery was increased by incubating homogenates in the presence of IBMX, a more potent phosphodiesterase inhibitor than theophylline, which was routinely used. These findings provide indirect evidence for high levels of phosphodiesterase activity in the soluble fraction. It is of interest that the metastatic tumor from the same patient (no. 2B) also had very high levels of phosphodi- esterase activity, as suggested by the dual recovery method. Similar destruction of cAMP was found during adenylate cy- clase assays of the adrenal tumor from patient 4 and of the metastatic but not primary tumor of patient 3 (cf. patient 3B vs. 3A). Adenylate cyclase activities reported in Results were obtained under conditions that minimized any apparent incu- bation losses of cAMP product. In all experiments, replicate values of [32P]cAMP varied from the mean by 5% or less.
Adenylate cyclase activities were expressed as picomoles of
| Patient no. | Age/ sex | Adrenal tissue specimen | Clinical syn- drome | 24-hr urinary 17-OHCS/17-KS (mg/24 h) | 24-h uri- nary cor- .tisol (ug/ 24 h) | Plasma cortisol (ug/dl) | Other steroid conc. | ||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Basal | After dexa- meth- asoneª | After ACTH | After metyr- aponec | AM/PM | After dex- ametha- soneª | ||||||
| 1 | 66/F | R. carcinoma | Hirsutism, hyper- tension | 52/68 | €/44 | /29 | 23 | ||||
| 2Af | 22/F | R. carcinoma | Cushing's syn- drome, hirsu- tism | 21-65/9-21 | 42/20 | 758/- | 15h | 606-2650 | 33.8-70/15.7-37.7 | ||
| 2B/ | 24/F | Hepatic metas- tasis of adre- nocortical carcinoma | Asymptomatic; hepatic masses on 6 g o,p'- DDD, 0.75 mg dexamethasone daily | 3.9-4.2/- | |||||||
| 3A | 47/M | R. carcinoma | Gynecomastia, decreased li- bido, impotence | 1.3-8/40 | 17/- | 14/- | E1 (413 pg/ml) E2 (25-109 pg/ml) DHA (461-1110 ng/dl) A4A (311-356 ng/dl) | ||||
| 3B | 50/M | Hepatic metas- tasis of adre- nocortical carcinoma | Recurrent ab- dominal mass on 2 g o,p'- DDD, 7.5 mg prednisone daily; 5-fluo- rouracil (800 mg, im, weekly) | 3.2-8.5/- | E1 (<12-17 pg/ml) E2 (<12-24 pg/ml) DHA (<25-30 ng/dl) 44A (<10 ng/dl) | ||||||
| 4 | 27/M | L. carcinoma | Cushing's syn- drome | -/146 | 1020-1320 | 19-23/20-21 | |||||
| 5 | 55/M | L. carcinoma | Asymptomatic; incidental find- ing | 5.8-6.1/9 | 10.7/1.6 | 0.7 | |||||
| 6 | 30/M | R. carcinoma | Cushing's syn- drome | 43-66/120-121 | 46/115 | 24.9-28.2/20.8-26.7 | |||||
| 7 | -/F | Normal R. cor- tex | Breast cancer | ||||||||
E1 (12-70 pg/ml) E2 (10-58 pg/ml)
DHA (160-700 ng/dl) 44A (80-210 ng/dl)
6-36/-
9-80
4.6/7
2-12/6-22
hypokalemia; L. aldosterone- producing ade- noma
43/M Normal R. cor- Asymptomatic; R. tex
suprarenal
mass (R. adre- nal myeloli- poma)
56/M Normal L. cor- Hypertension,
tex
9
Normal values
17-OHCS, 17-Hydroxycorticosteroids; 17-KS, 17-ketosteroids; E1, estrone; E2, estradiol; DHA, dehydroepiandrosterone; 44A, androstenedione. ” Values on second day of 2 mg dexamethasone, orally, every 6 h for eight doses. b Values on day of 40 U (400 µg) ACTH-(1-24) (Cortrosyn), infused iv for 8 h. ” Value on day after 750 mg metyrapone, orally, every 4 h for six doses.
d Value on the morning after 1-2 mg oral dexamethasone at about 2300 h. e Dash designates undetermined value.
” A and B designate, respectively, primary and subsequent metastatic lesions in the same patient. $ 65 mg/24 h on day before ACTH infusion. h 48 mg/24 h on day before metyrapone.
cAMP per 20 min/mg protein. Stimulation of enzyme activity was also expressed as fold stimulation, i.e. the ratio of stimu- lated to unstimulated (basal) activities. Protein was determined according to the method of Lowry et al. (6), using BSA as the standard.
Binding assay
B-Adrenergic receptor-binding sites in particulate fractions were measured in an equilibrium binding assay using the radi- oiodinated 8-adrenergic antagonist (-)[125I]iodopindolol ([125]] pindolol) as the radioligand. [125]] Pindolol was prepared from unlabeled (-)pindolol by the method of Maguire et al. for [125]] hydroxbenzylpindolol (7), as modified for pindolol by Barovsky and Brooker (8). Binding assays were carried out by the method of Weiland et al. (9). Unless otherwise stated, 100-ul aliquots of adrenal particulate (6-182 ug protein) were incubated in duplicate at 30 ℃ for 20 min in a total volume of 250 ul. Reactions were terminated by adding 4 ml wash buffer (9) at room temperature, pouring the tube contents over Whatman glass fiber filters (GF/C; Whatman, Inc., Clifton, NJ), and washing with an additional 16 ml wash buffer. The dried filters were counted at an efficiency of 60%. Nonspecific binding of [125I]pindolol was defined as the amount of radioligand bound in the presence of 10-4 M (-)isoproterenol. Specific binding was about 50-90% of the total binding at radioligand concen- trations in the range of the dissociation constant (Ka). Specific [125I]pindolol binding was also linear with respect to the quan- tity of tissue added. In the binding assays, protein was deter- mined according to the method of Bradford (10), using the reagent provided by Bio-Rad (Richmond, CA).
Results
Adenylate cyclase activities of normal human adrenal cortex
Figure 1 shows that the adenylate cyclase of homoge- nates from normal adrenal cortex (patients 7 and 9) was stimulated by ACTH in a concentration-dependent man- ner, while the ß-adrenergic agonists epinephrine and isoproterenol had no effect over a wide range of concen- trations. In these experiments and others (see below), the adenylate cyclase of normal and malignant adrenal tissues responded with greater sensitivity to ACTH-(1- 24) than to ACTH-(1-39). ACTH-(1-24) (10-5 M) stim- ulated adenylate cyclase maximally (2.3-fold the basal activity) in the normal tissue from patient 9, while acti- vation by ACTH-(1-39) in normal tissue from patient 7 was not maximal until at least a concentration of 10-4 M (see Fig. 1). In homogenized normal cortex from a third patient (no. 8), ACTH-(1-24) produced concentration- dependent stimulation, with a maximal effect (3.1-fold stimulation) at 10-5 M, but epinephrine had no stimula- tory effect (not shown). Stimulation by ACTH-(1-24) was half-maximal at 10-6 M in normal cortex from pa- tients 8 and 9. The adenylate cyclase of normal adrenal cortex (homogenate from patient 7) was unresponsive to
800
No. 7
300
Adenylate Cyclase Activity (pmol cyclic AMP/20 min /mg protein)
No. 9
ACTH1-39
600
ACTH1-24
200
400
·
0
8
O
0
a
0
0
Epinephrine
o
o
0
100
.
0
Isoproterenol
200
0
8
7
6
5
4
3
0
8
7
6
5
4
[Agent] (-log M)
a number of other hormones (TSH, LH, and glucagon) added in a range of concentrations. Sodium fluoride (F -; 5 mM) stimulated adenylate cyclase activity by 8.2- and 17.3-fold in cortex from patients 7 and 8, respectively.
Adenylate cyclase activities of human adrenocortical car- cinomas
Figure 2 shows the effects of increasing concentrations of ACTH (1-39 and/or 1-24) and 0-adrenergic agonists (epinephrine or isoproterenol) on adenylate cyclase ac- tivity of adrenocortical carcinoma preparations from six patients. In contrast to normal human adrenal cortex, adrenocortical carcinomas did exhibit ß-adrenergic stim- ulation of adenylate cyclase. Four tumors (from patients 1, 2A, 5, and 6) clearly responded to epinephrine; two tumors (from patients 3A and 4) did not respond. In the adrenergic responsive tumors, ß-agonist stimulation was concentration dependent, with maximal stimulation (1.2- to 2.5-fold) at 10-5-10-4 M epinephrine and half-maximal effect at 0.5-1.6 × 10-6 M. Stimulation of adenylate cyclase from one tumor (no. 1) by increasing concentra- tions of several adrenergic agonists revealed an order of potency of isoproterenol > epinephrine > norepineph- rine, which is characteristic of stimulation at a 8-adre- nergic receptor of the 02-subtype. In three tumors (no. 1, 5, and 6), epinephrine-stimulated adenylate cyclase ac- tivity was specifically inhibited by the ß-adrenergic an- tagonist propranolol but not by the a-adrenergic antag- onist phenoxybenzamine. Additionally, the GTP analog 5’-guanylyl imidodiphosphate [GMP-P(NH)P] in- creased 3-adrenergic stimulation of adenylate cyclase in two tumors (no. 1 and 6), but did not elicit ß-agonist
stimulation in tissue (no. 3A) showing no adrenergic response without added nucleotide.
Figure 2 also shows that ACTH activated the adenylate cyclase of five tumors (from patients 1, 2A, 3A, 4, and 6) in a manner comparable to the ACTH effect on normal cortex (cf. Fig. 1). Specifically, ACTH-(1-24) stimulated the tumor enzyme maximally (5.9- and 1.2-fold in no. 4 and 6, respectively) at 10-5-10-4 M, and half-maximal stimulation occurred at about 10-6 M (Fig. 2, no. 4 and 6). As in normal cortex, enzyme sensitivity to ACTH- (1-24) was greater than that to ACTH-(1-39) (see Fig. 2, no. 4). In one tumor (no. 1) exhibiting adenylate cyclase stimulation by both ACTH-(1-39) and epineph- rine, the effects of the two agonists were not additive (data not shown).
In general, the ACTH-sensitive adenylate cyclase ac- tivity of the six tumors was unrelated to the presence or degree of response to B-agonist. Three tumors (no. 1, 2A, and 6) of the six studied had adenylate cyclase responses to both ACTH and epinephrine. One of these (no. 6) had similar quantitative activation by the two agonists; a second (no. 1) had a greater response to epinephrine than to ACTH; in the third (no. 2A), epinephrine stimulated adenylate cyclase far less than did ACTH. Moreover, two tumors (no. 3A and 4) were ACTH responsive but not 8- agonist responsive, while one tumor (no. 5) surprisingly responded to epinephrine but not to ACTH.
Considerable variability of absolute enzyme activities and hormone responsiveness was apparent among the six adrenocortical carcinomas examined. As noted in Fig. 2, some tissues were assayed for adenylate cyclase im- mediately after tumor resection, while others were frozen before assay. Freeze-thawing of tumor tissue diminished stimulation of adenylate cyclase by both ACTH and 8- adrenergic agonists (data not shown). However, the var- iability of tumor activities shown in Fig. 2 was at least in part independent of tissue preparation. The tumor with the lowest enzyme activities (no. 1) was assayed immediately after resection, and the tumor with the highest activities (no. 5) was tested after freezing. The failure of the latter tumor to respond to ACTH may have been due to decreased ACTH-sensitive adenylate cyclase activity after freezing, but this possibility seems unlikely. Other tumors (no. 2A, 4, and 6) retained ACTH respon- siveness after freezing. Furthermore, whereas ACTH and epinephrine responses in one tumor (no. 6) were both lost after repeated freeze-thawing but were restored by GMP-P(NH)P, no ACTH stimulation could be elicited in the tumor from patient 5 even in the presence of GMP-P(NH)P. Of the two tumors (no. 3A and 4) without ß-adrenergic agonist-stimulated adenylate cyclase activ- ity, only one (no. 4) was frozen before assay. Lack of adrenergic response in this tissue was probably not due to freezing, since other tumors (no. 2A, 5, and 6) retained
No. 1
400
No. 2A
400
No. 3A
60
Epi
ACTH1-39
ACTH1-39
300
300
Adenylate Cyclase Activity (pmol cyclic AMP/20min /mg protein)
40
ACTH1-39
200
200
Iso
20
100
Epi
100
0
0
0
100
No. 4
1250
No. 5
Epi
120
No. 6
ACTHI-24
Epi
75
1000
110
8
50
750
ACTH1-24
ACTH1-24
500
100
25
ACTH1-39
0
250
Epi
90
0
8
7
6
5
4
0
8
7
6
5
4
0
8
7
6
5
4
[Agent] (-log M)
adrenergic response after freezing.
Homogenates and particulates from one adrenocortical carcinoma (no. 1) were examined for adenylate cyclase responses to a number of other hormones. Glucagon and hCG added in a wide range of concentrations had no stimulatory effects. On the other hand, TSH was stim- ulatory, producing 1.8-fold the basal activity when added at 0.2 U/ml (data not shown). Stimulation by TSH could not be accounted for by the degree of ACTH contami- nation in the impure TSH preparation, but enzyme ac- tivation by TSH was not further characterized. F- (5 mM) stimulated adenylate cyclase activity in all six pri- mary adrenocortical carcinomas, with activation ranging between 3.0-fold (no. 5) and 39.8-fold (no. 4).
Adenylate cyclase activities of metastatic adrenocortical carcinomas
Patients 2 and 3 underwent resection of metastatic adrenocortical carcinomas 2-3 yr after removal of their primary tumors (cf. patients 2A vs. 2B and 3A vs. 3B in Table 1). In these cases, adenylate cyclase activities of the metastatic lesions were compared to those of the primary tumors. Whereas the primary tumor from pa- tient 2 responded to both ACTH and epinephrine (no. 2A in Fig. 2), neither hormone stimulated adenylate cyclase activity of the metastatic tumor (not shown). The metastatic tumor from patient 3 also lost the response to ACTH shown earlier by the primary lesion (no. 3A in Fig. 2), although freezing of the metastatic, but not the
primary, tumor before adenylate cyclase assay could have contributed to the loss of ACTH response. F- (5 mM) stimulated adenylate cyclase activity of both metastatic tumors (21.0-fold in no. 2B; 9.3-fold in no. 3B). Both patients were treated with the adrenolytic agent o,p’- DDD before resection of the metastatic lesions (see Table 1), and it is possible this therapy might have influenced the in vitro enzyme activities.
Characterization of ß-adrenergic receptor-binding sites in adrenocortical carcinomas
Binding of [125I]pindolol was examined in the partic- ulate fraction of four adrenocortical carcinomas (patients 2A, 3A, 5, and 6) and two normal adrenal cortex speci- mens (patients 7 and 9). The radioligand bound specifi- cally only to particulates of the three tumors exhibiting ß-adrenergic agonist-sensitive adenylate cyclase activity (no. 2A, 5, and 6; see Fig. 2). No specific binding of [125]] pindolol was found in tissues lacking demonstrable 8- adrenergic agonist-sensitive adenylate cyclase, namely in normal adrenal cortex from patients 7 and 9 and the tumor from patient 3A (see Figs. 1 and 2 and above).
Figure 3 shows that specific [125I]pindolol binding to a tumor particulate preparation from a representative pa- tient (no. 6) was saturable with respect to the concentra- tion of added radioligand. Scatchard analysis (11), shown in the inset to Fig. 3, indicates a single class of binding sites with high affinity for [125I]pindolol. The three tu- mors (no. 2A, 5, and 6) that had specific [125I]pindolol
60
[125]] Pindolol Bound (fmol/mg protein )
5.0
4.0
0.08
Kp=56 PM
Q.06
Bmax = 6.0
3.0
Bound / Free
fmol/mg protein
0.04
2.0
0.02
1.0
0
1.0
2.0
3.0
4.0
5.0
[ 125 ]] Pindolol Bound(pM)
0
50
100
150
200
250
300
350
[125]] Pindolol (PM)
binding all had similar receptor affinity for [125I]pindolol, as indicated by similar dissociation constants (Ka) for the radioligand (see Fig. 3). The receptor density (Bmax) in the tumor from patient 5 was much greater than in the other two tumors,which had similar Bmax values (see Fig. 3). Relative receptor density of the three tumors was correlated with total epinephrine-stimulated adenylate cyclase activity, which was similar in tissues from pa- tients 2A and 6 and greatest in the tumor from patient 5. However, the receptor density was not correlated with the degree of stimulation of adenylate cyclase by epi- nephrine, which was equivalent in tissues from patients 2A and 5 and less in the tumor from patient 6 (see Fig. 2). This lack of correlation is difficult to interpret, es- pecially since the tissues were frozen before assay and may have lost some enzyme activity (see above). Ka and Bmax determinations in all three tumors were within the ranges that we found for [125I]pindolol-binding sites in particulate preparations from a variety of nonneoplastic rat tissues (Dax, E. M., et al., unpublished data).
In addition to saturability with increasing ligand con- centration, binding of [125I]pindolol to tumor particulates also satisfied other ß-adrenergic binding criteria. As shown in Fig. 4, competition by increasing concentra- tions of adrenergic agonists for [125]]pindolol binding demonstrated pharmacological specificity, with an order of potency of (-)isoproterenol > (-)epinephrine > (-)norepinephrine. This order of potency, which is iden- tical to that for adrenergic agonist-stimulated adenylate cyclase activity (see above), is characteristic of a ß2- adrenergic receptor. The Ki of epinephrine competition of [125I]pindolol binding in the tumor from patient 6 was
[125 ] ] Pindolol Bound (% Control)
100
A
(+) Iso
4
80
(-) Norepi
(-) Iso
60
(-) Epi
40
20
O
8
7
6
5
4
3
[ Adrenergic Agonist ] (- log M)
1.2-3.1 x 10-6 M, which is higher than the concentration of epinephrine (0.5 × 10-6 M) required for half-maximal stimulation of adenylate cyclase in the same tumor (Fig. 2). This apparent discrepancy may be related to the different conditions under which enzyme activity and receptor binding were measured (see Materials and Meth- ods). Figure 4 also shows stereospecificity of binding to tumor particulates, since increasing concentrations of (-)isoproterenol caused much greater displacement of [125I]pindolol than did equivalent concentrations of (+)isoproterenol. Additionally, Fig. 5 demonstrates that [125I]pindolol binding was saturable with respect to time and was reversible in the presence of excess (-)isopro- terenol. Other experiments examined the effects of GMP-P(NH)P on competition by increasing concentra- tions of the agonist (-)isoproterenol for specific [125]] pindolol binding. Addition of the nucleotide shifted the agonist competition curve to the right, i.e., decreased the affinity of the 3-receptor for agonist, and raised the Hill coefficient toward unity (data not shown).
Discussion
Our results demonstrate for the first time a 3-adrener- gic receptor-linked adenylate cyclase in adrenocortical carcinoma in man which is not present in normal human adrenal cortex. Three of six adrenal carcinomas (no. 2A, 5, and 6) had ß-adrenergic agonist-sensitive adenylate cyclase activity and high affinity binding sites for the radioiodinated Ø-adrenergic antagonist [125I]pindolol. An additional tumor (no. 1) had 8-agonist-sensitive enzyme activity, although no tissue was available for binding
2.0
(-)Iso
[125 ]] Pindolol Bound (pM)
1.5
1.0
0.5
0
10
20
30
40
50
Time (min)
studies. In contrast, one adrenal tumor (no. 3A) and normal adrenal cortex from two patients (no. 7 and 9) demonstrated neither catecholamine-stimulated adenyl- ate cyclase activity nor specific binding of [125I]pindolol. Finally, the adenylate cyclase of a sixth tumor (no. 4) and normal cortex from an additional patient (no. 8) failed to respond to ß-agonists. These findings, therefore, indicate that malignant transformation of the adrenal cortex in man is, in at least some instances, associated with the appearance of B-adrenergic receptors coupled to adenylate cyclase. The origin of these ectopic receptors is unclear, but their appearance is consistent with the idea that gene derepression during malignant transfor- mation in tissues sometimes results in the production of proteins atypical to the tissue of origin (1, 12).
The ß-adrenergic agonist-sensitive adenylate cyclase of human adrenocortical carcinomas exhibited charac- teristics common to the same enzyme complex in non- malignant tissues. In three of four adrenal carcinomas tested (see above), specific binding of [125]]pindolol sat- isfied the criteria for a ß-adrenergic receptor, including saturation with respect to time and ligand concentration, reversibility, stereospecificity, and pharmacological spec- ificity. Furthermore, the effects of GMP-P(NH)P, namely to decrease binding affinity of receptor for ago- nist and increase ß-agonist-stimulated adenylate cyclase activity, are similar to the nucleotide effects described in other ß-agonist-sensitive adenylate cyclase systems (13). Data from one tumor (no. 1) in which the stimulatory effects of ß-agonist and ACTH were not additive suggest that the two agonists bind to specific receptors which both interact with a single catalytic component. As de-
termined by addition experiments, the interaction of 8- adrenergic and other receptors with a single adenylate cyclase has also been noted in nonmalignant adenylate cyclase systems, e.g. the rat fat cell (14).
Although ectopic ß-adrenergic receptors were de- scribed previously in rat adrenal carcinoma (1-3), earlier studies of a total of seven human adrenocortical carci- nomas did not demonstrate a ß-adrenergic receptor- linked adenylate cyclase (15-17). The failure of these attempts may have been due to the small number of tumors examined and the heterogeneity of expression of catecholamine-sensitive adenylate cyclase. More likely, however, is the dependence of B-agonist responsiveness on the conditions of tissue preparation. For example, the lack of epinephrine stimulation of adenylate cyclase in preparations from four human adrenal carcinomas was associated with use of particulates (crude membranes) rather than homogenates (16). We previously reported losses of epinephrine-stimulated adenylate cyclase activ- ity during preparation of the particulate fraction from whole homogenates of normal rat liver (18). To avoid this problem, most of our studies of adrenal carcinomas were conducted with whole homogenates. With one tu- mor (no. 6), epinephrine stimulated adenylate cyclase activity in homogenates, but had no effect in the partic- ulate fraction. Particulates from two other tumors had reduced ß-agonist responsiveness relative to whole ho- mogenates (not shown). Failure to recognize high degrees of cAMP product destruction might also have contrib- uted to the failure of others to demonstrate catechola- mine-sensitive adenylate cyclase in human carcinomas.
Previous work provided evidence for ectopic ß-adre- nergic and hormone receptors linked to adenylate cyclase in some human tumors. In studies of human adrenal adenomas, catecholamines stimulated adenylate cyclase activity in three adenomas (17), and specific ß-adrenergic binding sites were present in two (12). The authors of these studies did not specify whether any one tumor demonstrated both -adrenergic binding and catechola- mine stimulation of adenylate cyclase. Adenylate cyclase responses to TSH, gonadotropins, and PRL have also been described in human adrenal carcinomas and ade- nomas (15, 17, 19). Our own studies show TSH-sensitive adenylate cyclase activity in one human adrenal carci- noma (from patient 1), but not normal adrenal cortex (patient 7). Aberrant responses of adenylate cyclase to a number of neurotransmitters and hormones have also been found in other human endocrine tumors, including pituitary adenomas (20, 21), parathyroid adenomas (20), pheochromocytoma (20), and medullary thyroid carci- noma (22). Taken together, then, our findings and those of others clearly extend to human neoplastic tissue the concept of the ectopic receptor linked to adenylate cy- clase noted previously in adrenal carcinoma of the rat
(1-3).
In contrast to human tumors with ectopic receptors linked to adenylate cyclase, other tumors apparently lose the hormone responsiveness characteristic of the non- neoplastic tissue of origin. The adenylate cyclase of one adrenocortical carcinoma (from patient 5) of the six tested did not respond to ACTH. Others also reported human adrenal carcinomas and adenomas lacking ACTH-sensitive adenylate cyclase activity (15, 16), in some cases associated with apparent alteration or loss of the functional ACTH receptor (16). Similarly, lack of TSH-sensitive adenylate cyclase activity described in undifferentiated human thyroid carcinomas may have been due to altered TSH receptors (23). Additional bio- chemical abnormalities distal to the generation of cAMP may also account for lack of appropriate cellular response to hormones in human neoplastic tissue. A defect of the cAMP-dependent protein kinase may produce deficient ACTH-induced steroidogenesis in some human adreno- cortical tumors despite the presence of ACTH-sensitive adenylate cyclase (24). Furthermore, our results suggest that substantial phosphodiesterase activity in some hu- man adrenocortical carcinomas could conceivably pre- vent the accumulation of sufficient cAMP to produce steroidogenesis. Although previous workers reported that phosphodiesterase activity in membranes of human ad- renocortical tumors was lower than that in normal ad- renals (16), our work implicates high levels of phospho- diesterase activity in the soluble fraction of tumor ho- mogenates (see Materials and Methods). It is interesting to speculate that loss of hormone-sensitive adenylate cyclase activity and acquisition of high levels of phos- phodiesterase activity may each be related to the degree of tumor dedifferentiation, since in our studies, these changes appeared to occur during metastatic spread of adrenocortical carcinomas (see Materials and Methods and Results).
Alteration of receptor-linked adenylate cyclase during malignant transformation of human adrenal cortex may be relevant to the clinical behavior of adrenocortical carcinomas. The fact that patients with adrenocortical carcinoma usually do not have a rise in urinary cortisol metabolites in response to ACTH (25) may be due to altered hormone-sensitive adenylate cyclase or a defect in steroidogenesis distal to the generation of cAMP. On the other hand, it is also possible that under conditions of maximal cAMP production by hormones or 8-adre- nergic agonists acting through ectopic receptors, infused ACTH may have no additive effects. In this regard, two observations are of interest. First, the stimulatory effects of ACTH and ß-agonist on adenylate cyclase of one adrenal carcinoma (from patient 1) were not additive. Second, of the two patients in our series who had ACTH infusions, one patient (no. 3A) responded with a doubling
of urinary 17-hydroxycorticosteroids excretion while the other (no. 2A) had no response to either ACTH or metyrapone. The tumors from both patients contained ACTH-sensitive adenylate cyclase activities, but only the tumor from the patient who had no response to ACTH had ß-adrenergic receptor-linked adenylate cy- clase. The functional implications of these observations are unknown. However, a clearer understanding of aber- rant receptor-mediated control of the function of human adrenocortical carcinoma may ultimately have practical importance in the diagnosis and treatment of this lethal disease.
Acknowledgments
The authors wish to thank the surgery and pathology staffs of the Baltimore City and Johns Hopkins Hospitals (Baltimore, MD), the Clinical Center of the NIH (Bethesda, MD), the Medical Center Hospital (San Antonio, TX), and Metropolitan General Hospital (San Antonio, TX), as well as Dr. Glenn Geelhoed (Washington, DC) for providing surgical specimens for these studies. We also thank Drs. Robert L. Ney and Michael A. Levine (Baltimore, MD) for their helpful reviews and discussion of the manuscript. The expert technical assist- ance of Mrs. Shirley Jean Schmidt and preparation of the manuscript by Mrs. Janet Gilliam are also appreciated.
References
1. Schorr I, Ney RL 1971 Abnormal hormone responses of an adre- nocortical cancer adenyl cyclase. J Clin Invest 50:1295
2. Schorr I, Rathnam P, Saxena BB, Ney RL 1971 Multiple specific hormone receptors in the adenylate cyclase of an adrenocortical carcinoma. J Biol Chem 246:5806
3. Williams LT, Gore TB, Lefkowitz RJ 1977 Ectopic 8-adrenergic receptor binding sites. Possible molecular basis of aberrant cate- cholamine responsiveness of an adrenocortical tumor adenylate cyclase. J Clin Invest 59:319
4. Salomon Y, Londos C, Rodbell M 1974 A highly sensitive adenylate cyclase assay. Anal Biochem 58:541
5. Katz MS, Kelly TM, Pineyro MA, Gregerman RI 1978 Activation of epinephrine and glucagon-sensitive adenylate cyclases of rat liver by cytosol protein factors. Role in loss of enzyme activities during preparation of particulate fractions, quantitation and par- tial characterization. J Cyclic Nucleotide Res 4:389
6. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ 1951 Protein measurement with the Folin phenol reagent. J Biol Chem 193:265
7. Maguire ME, Wiklund RA, Anderson HJ, Gilman AG 1976 Binding of [125I]iodohydroxybenzylpindolol to putative ß-adrenergic recep- tors of rat glioma cells and other cell clones. J Biol Chem 251:1221
8. Barovsky K, Brooker G 1980 (-)-[125I]Iodopindolol, a new highly selective radioiodinated Ø-adrenergic receptor antagonist: measure- ment of ß-receptors on intact rat astrocytoma cells. J Cyclic Nu- cleotide Res 6:297
9. Weiland GA, Minneman KP, Molinoff PB 1980 Thermodynamics of agonist and antagonist interactions with mammalian 8-adrener- gic receptors. Mol Pharmacol 18:341
10. Bradford MM 1976 A rapid and sensitive method for the quanti- tation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72:248
11. Scatchard G 1949 The attractions of proteins for small molecules and ions. Ann NY Acad Sci 51:660
12. Hirata Y, Uchihashi M, Sueoka S, Matsukura S, Fujita T 1981 Presence of ectopic ß-adrenergic receptors on human adrenocorti- cal cortisol-producing adenomas. J Clin Endocrinol Metab 53:953
13. Rodbell M 1980 The role of hormone receptors and GTP-regulatory proteins in membrane transduction. Nature 284:17
14. Birnbaumer L, Rodbell M 1969 Adenyl cyclase in fat cells. II. Hormone receptors. J Biol Chem 244:3477
15. Hinshaw HT, Ney RL 1974 Abnormal hormonal control in the neoplastic adrenal cortex. In: McKerns KW (ed) Hormones and Cancer. Academic Press, New York, p 309
16. Saez JM, Dazord A, Gallet D 1975 ACTH and prostaglandin receptors in human adrenocortical tumors. Apparent modification of a specific component of the ACTH-binding site. J Clin Invest 56:536
17. Matsukura S, Kakita T, Sueoka S, Yoshimi H, Hirata Y, Yokota M, Fujita T 1980 Multiple hormone receptors in the adenylate cyclase of human adrenocortical tumors. Cancer Res 40:3768
18. Katz MS, Kalish MI, Pineyro MA, Gregerman RI 1978 Quantita- tion of epinephrine- and glucagon-sensitive adenylate cyclases of rat liver. Implications of alterations of enzymatic activities during preparation of particulate fractions and membranes. Biochim Bio- phys Acta 540:205
19. Millington DS, Golder MP, Cowley T, London D, Roberts H, Butt WR, Griffiths K 1976 In vitro synthesis of steroids by a feminising adrenocortical carcinoma: effect of prolactin and other protein
hormones. Acta Endocrinol (Copenh) 82:561
20. Schorr I, Hinshaw HT, Cooper MA, Mahaffee D, Ney RL 1972 Adenyl cyclase hormone responses of certain human endocrine tumors. J Clin Endocrinol Metab 34:447
21. Matsukura S, Kakita T, Hirata Y, Yoshimi H, Fukase M, Iwasaki Y, Kato Y, Imura H 1977 Adenylate cyclase of GH and ACTH producing tumors of human: activation by non-specific hormones and other bioactive substances. J Clin Endocrinol Metab 44:392
22. Matsukura S, Kakita T, Fukase M, Fujita T 1981 Adenylate cyclase of a human medullary thyroid carcinoma. Experientia 37:523
23. Abe Y, Ichikawa Y, Muraki T, Ito K, Homma M 1981 Thyrotropin (TSH) receptor and adenylate cyclase activity in human thyroid tumors: absence of high affinity receptor and loss of TSH respon- siveness in undifferentiated thyroid carcinoma. J Clin Endocrinol Metab 52:23
24. Riou JP, Evain D, Perrin F, Saez JM 1977 Adenosine 3’5’-cyclic monophosphate dependent protein kinase in human adrenocortical tumors. J Clin Endocrinol Metab 44:413
Bertagna C, Orth DN 1981 Clinical and laboratory findings and results of therapy in 58 patients with adrenocortical tumors admitted to a single medical center (1951 to 1978). Am J Med 71:855
FASEB Summer Research Conferences
The Federation of American Societies for Experimental Biology will sponsor ten one-week Summer Research Conferences between June 9 and August 16, 1985 at the Vermont Academy, Saxtons River, VT. Attendance will be limited to 150, and will be by invitation upon application.
The Conference schedule is: June 9-14: Neurotransmitters June 16-21: Micronutrients: Trace Metals June 23-28: Smooth Muscle
June 30-July 5: Lymphocyte and Antibodies
July 7-12: Somatic Cell Genetics July 14-19: Developmental Neurobiology July 21-26: Gastrointestinal Differentiation July 28-August 2: Mechanisms of Carcinogenesis
August 4-9: Protein Kinases
August 11-16: Biology and Chemistry of Vision
For further information please contact:
Mr. Robert W. Krauss Federation of American Societies for Experimental Biology 9650 Rockville Pike Bethesda, MD 20814