Tenovus Institute for Cancer Research, Heath Park, Cardiff, CF4 4XX,
1)Queen Elizabeth Hospital, Birmingham and
2)Birmingham and Midland Hospital for Women, U. K.
IN VITRO SYNTHESIS OF STEROIDS BY A FEMINISING ADRENOCORTICAL CARCINOMA: EFFECT OF PROLACTIN AND OTHER PROTEIN HORMONES By D. S. Millington, M. P. Golder, T. Cowley, D. London1), H. Roberts1), W. R. Butt2) and K. Griffiths
ABSTRACT
The study describes the effects of ACTH, prolactin and other protein hor- mones on the synthesis and secretion of steroid hormones by tissue from a feminising adrenocortical carcinoma removed from a post-menopausal female. Steroid production by the tissue was determined by high resolution- mass fragmentography and by radioimmunoassay. Prolactin and ACTH stimulated the synthesis of oestrogens by the tissue whereas GH, LH and ACTH were more effective than prolactin in stimulating androgen syn- thesis. The effect of protein hormones, other than ACTH, on adenylate cyclase activity of this tumour tissue indicated a lack of specificity of the membrane receptor sites.
Most adrenocortical carcinomas produce endocrine symptoms. Those associated with Cushing’s syndrome and virilization are the most common (Symington & Jeffries 1962; Symington 1969) although carcinomas giving rise to primary aldo- steronism (Conn et al. 1964), hypoglycaemia (Williams 1960) and feminisation (Gabrilove et al. 1965) have also been reported. Tumours of the adrenal cortex
causing feminisation are characterised by iso-sexual precocity in the pre-puber- tal female and feminisation in the pre-pubertal (Wilkins 1948) and adult male (Gabrilove et al. 1965). Such feminising tumours are rare and Gabrilove et al. (1965) reviewed the 53 tumours, carcinomas and adenomas, which had previously been reported in the adult male to that date. Reports of feminising adrenocortical carcinomas in adult females are even more rare, presumably because of the dif- ficulties in recognising the clinical manifestations of the condition, and only Procopé (1970) and Mathur et al. (1973) have described such patients. Urinary steroid analyses (Procopé 1970) and studies in vivo and in vitro (Mathur et al. 1973) were undertaken with these cases of feminising carcinoma in post-meno- pausal women.
The present report describes some studies in vitro on tissue from a feminising adrenocortical carcinoma also removed from a post-menopausal female.
EXPERIMENTAL
Case report
The patient (66 years old in 1974) underwent an operation in 1965 for a prolapsed vaginal wall. She experienced vaginal bleeding in August, 1972 and was given a course of oestrogens for one month. In October, a D & C was performed but no abnormality was found although bleeding continued intermittently over the subsequent 5-6 weeks. Cervical cytology in January 1973 showed some evidence of early malignant change and a hysterectomy and oophorectomy was performed in February. Histological exami- nation indicated endometrial cystic hyperplasia with mitotic figures in stroma and epi- thelium, but no malignancy. On March 5, the following serum hormone values were found: progesterone, 0.7 ng/ml, oestradiol-17, 800 pg/ml and FSH < 1 mU/ml. Repeat analyses in May showed oestradiol-178, 820 pg/ml and FSH < 1 mU/ml. Plasma androstenedione concentration was approximately 30 times normal levels (140-280 nmol/l) and the testosterone concentration 11-17 nmol/l. An adrenal tumour was suspec- ted and on admission to the Queen Elizabeth Hospital, Birmingham in July, 24 h urinary steroid values were found to be as follows: 17-oxosteroids 40 mg, 17-oxogenic steroids 6 mg and total oestrogens 760 ug. A dexamethasone suppression test (0.5 mg × 4, daily for 2 days, 2.0 mg × 4 daily for 2 days) failed to influence the concentration of plasma oestradiol and the FSH level remained less than 1 mU/ml. The plasma cortisol decreased from 30 µg/100 ml to 12 ug by the end of the test. An ACTH (tetracosactin, 0.5 mg b. d. over 4 days) stimulation test resulted in an increase in the plasma cortisol con- centration from 20 to 40 ug/100 ml whereas the plasma oestradiol level on the control day was 1960 pg/ml and this decreased to 1060 pg by the fourth day of the ACTH administration. An aortogram showed the presence of a large tumour in the area of the left adrenal which was removed in August 1973. After operation, the plasma oestradiol concentration fell within 3 days to the normal post-menopausal level. Three weeks after the operation, the LH had risen to 5 mU/ml, the FSH to 31 mU/ml and the cortisol concentration in the plasma was in the normal range.
Histological examination of the tumour confirmed that it was an adrenal carcinoma which had been removed.
MATERIALS AND METHODS
The plasma and urinary steroid and protein hormone concentrations were determined at the Birmingham and Midland Hospital for Women (Dr. W. R. Butt).
Tissue culture
A portion of the adrenal tumour was transported to Cardiff in sterilised containers on ice. Tissue was cut into 1 mm3 explants under aseptic conditions for culture. Explants were cultured on lens paper rafts supported by stainless steel grids (Trowell 1959) in TC199 medium fortified with 10 % calf serum, penicillin 200 units/ml and streptomycin, 100 ug/ml, in organ culture dishes (Falcon Plastics Ltd., California, USA). Cultures were pre-incubated at 37ºC in an atmosphere of 95 % air: 5% CO2 for 24 h. After this period, medium was removed, discarded and replaced with fresh medium (control cultures) or medium in which either ACTH (Synacthen®, 81-24corticotrophin, Ciba Ltd), insulin (Weddel Pharmaceuticals Ltd), equine LH (Calbiochem Ltd), ovine prolactin (WHO, 2nd Int. Standard) or human GH (MRC 69/46) were dissolved. All cultures were prepared in triplicate and were incubated for 3 days. Medium was then removed and stored at -15℃ for steroid analysis. Tissue from each culture was carefully blotted dry with tissue paper and weighed.
Steroid extraction
Steroids were extracted from the culture medium with ethyl acetate (3 x 4 vols.), while extraction from the original non-incubated tumour tissue was effected by the procedure described previously (Millington et al. 1974).
Analysis of culture medium by radioimmunoassay
Oestradiol-17 was determined in the medium by a specific radioimmunoassay using an antibody raised against oestradiol-6-(0-carboxymethyl) oxime-BSA. The cross-reac- tion of this serum with oestrone was of the order of 1% (Exley et al. 1971). Andros- tenedione was measured using an antiserum against androstenedione-11-hemisuccinyl- BSA after alumina thin-layer chromatography (Cowley et al. 1976). Adrenocorticoste- roids were determined using the competitive protein binding assay described by Murphy (1967) and a dog plasma as the source of the binding protein.
Analysis of steroids by combined gas chromatography-high resolution mass fragmentography
Aliquots of the culture medium from selected incubations were extracted with ethyl acetate. The dried steroid extracts were dissolved in ethanol, measured volumes trans- ferred to 1 ml screw-cap vials and the ethanol removed. Bis (trimethylsilyl) acetamide (30 ul) and light petroleum (b. p. 60-80°C) (20 ul) were added and the tubes left over- night at 20℃. Combined gas liquid chromatography (GLC) high resolution-mass frag- mentography was carried out as previously described in detail (Millington 1975) using a Varian-MAT 731 mass spectrometer. The assay for each steroid derivative was then conveniently set up by monitoring the intensity of the molecular-ion peak on a poten- tiometric recorder at a resolution of 10 000 (10% valley). Under these conditions, the recorder response varied linearly with sample concentration in the range 0.1-10 ng/ul and the detection limit was approximately 3 pg/ul for the oestrogens and 25 pg/ul for the C19-steroids (Millington 1975). The two parameters of GLC retention time and
(a) Mass spectrum of oestrone trimethylsilyl ether
0
100
MI-
TMSO
80
MW - 342
Rel. int
60
40.
20
0
50
100
200
300
400
m’c
(b) Mass fragmentograms - m/e 342 . 202 (oestrone-TMS-ether)
Control incubation (1 ul = 0,90 mg tissue)
Incubation with prolactin (1 ml = 0.67 mg tissue)
780 pg/ul
300 pg//1]
Injection
Injection
1
0
5
[0
0
5
10)
Time (mins.)
Time (mins.)
0
(ii)
high resolution molecular ion detection were sufficient to detect and quantify the com- ponents regardless of the complexity of the steroid mixture and the presence of over- lapping GLC peaks.
Illustrated in Fig. 1 are the principles of high resolution-mass fragmentography (HRMF) as they apply to oestrone. Fig. 1 (a) shows the mass spectrum of oestrone-3-tri- methylsilyl ether, obtained during routine combined gas chromatography and mass spectrometry. Clearly, the molecular ion at m/e 342.202 is ideal for specific ion moni- toring and high resolution-mass fragmentograms showing the presence of oestrone in the culture medium extracts of control and prolactin containing incubations of the adrenal tumour explants are represented in Fig. 1 (b). Peak heights were recorded in arbitrary units and related to those of standard steroid solutions obtained from similar fragmentograms. Other steroids were assayed in turn by tuning the mass spectrometer to detect the appropriate molecular ions and repeating the procedure described for oestrone.
The m/e values for high resolution monitoring of the molecular ions obtained from the various steroids investigated by this technique and also the GLC retention data have been given previously (Millington 1975). Peak heights were determined for all steroids isolated from the culture medium and related to the peak height of steroid standards from similar fragmentograms.
Measurement of adenylate cyclase activity
Adrenal tumour tissue was minced finely and a 10% homogenate prepared in 0.25 M sucrose using 10 strokes of a Philpot & Stanier (1956) homogeniser. The homogenate was filtered through cheesecloth before use in the assay. All procedures were carried out at 0℃. Adenylate cyclase activity in the presence and absence of various protein hormone preparations was determined by the procedure established in these laboratories by Goldner & Boyns (1973) using [8-14C]ATP and cyclic [$H]AMP as tracer. The hor- mone preparations used were: ovine prolactin (WHO 2nd Int. Std.); §1-24corticotrophin (Synacthen®, Ciba Laboratories); HLH (IRC-2-(10.12.64); HFSH (CPDS-6-Butt); HTSH (DE-32-3-Hartree); HCS (MRC, human placental lactogen 70/194); insulin (crystalline porcine insulin 23 IU/mg, Eli Lilly).
Normal adrenal studies
Tissue from a ‘normal’ adrenal gland removed from a patient with breast cancer was cut into 1 mm3 explants and cultured as described earlier. It was not possible to attempt to separate the tissue from the fascicular and reticular zones of the cortex and the explants would therefore be composed of mixed cortical cells. With the limited tissue available, only the effect of (1-24corticotrophin and prolactin on steroid secretion was studied.
RESULTS
A relatively specific radioimmunoassay for oestradiol-178 which had been found quite adequate for the routine determination of this hormone in plasma (Came- ron & Jones 1972) was used to assess the rate of secretion of oestradiol-178 by the cultured tumour explants. Table 1 shows the values obtained. The results suggested that in vitro, neither GH nor LH at the concentrations used, effectively
| RIA determination (pg/mg wet wt. tissue) | GLC-mass fragmentography (pg/mg wet wt. tissue) | Protein binding assay (ng/mg wet wt. tissue) | ||||
|---|---|---|---|---|---|---|
| Oestradiol | Androst- enedione | Oestradiol | Oestrone | Androst- enedione | 11-Hy- droxy- cortico- steroids | |
| Control | 44 | 590 | 9 | 330 | 180 | 16.8 |
| LH (5 µg) | 54 | 31.5 | ||||
| LH (50 µg) | 33 | 2320 | 15 | 692 | 2200 | 145.3 |
| GH (2.5 µg) | 57 | 2910 | 25 | 650 | 2300 | 154.2 |
| @I-24ACTH (1.0 µg) | 204 | 1730 | 150 | 660 | 1540 | 19.2 |
| Prolactin (5.0 µg) | 134 | 1310 | 50 | 1170 | 530 | 116.7 |
| Insulin (0.08 mU) + | 152 | 3460 | 20.1 | |||
| ACTH (1.0 µg) | ||||||
| Insulin (0.08 mU) + | 303 | 260 | 1120 | 21.5 | ||
| Prolactin (1.0 µg) | ||||||
| Steroid concentration pg/mg wet wt. tissue | |
|---|---|
| Oestradiol-178 | <3 |
| Oestriol | <3 |
| Oestrone | 45 |
| DHA | 100 |
| 5a-Dihydrotestosterone | < 5 |
| Androsterone | 27 |
| Epiandrosterone | 17 |
| Testosterone | 35 |
stimulated the synthesis and secretion of oestradiol-178 by the adrenal tumour tissue. ACTH and prolactin however, with insulin appearing to influence the action of both these hormones, stimulated the output of oestradiol-178. Values obtained for the endogenous oestradiol-178 concentration of the tumour tissue determined by high resolution GLC-mass fragmentography (Table 2), indicated that steroid was being actively synthesised during the three-day culture period.
Of particular interest however, was the concentration of endogenous oestrone (45 pg/mg tissue) compared with the concentration of oestradiol-178 and oestriol (<3 pg/mg tissue of both oestrogens). It was therefore obvious that with a cross-reactivity of oestrone of 1% in the oestradiol-17฿ radioimmunoassay, further analysis was necessary to determine the oestradiol-17 and oestrone concentration present in the culture medium. Values of the oestradiol-178 and oestrone concentrations in the medium, determined by high resolution GLC- mass fragmentography are shown in Table 1. The tumour tissue actively syn- thesized and secreted oestrone as well as oestradiol-176 into the medium and the results indicated again that ACTH and prolactin stimulated this biosynthetic activity. The results again demonstrated the influence of insulin on the secretory activity of the tumour tissue.
In similar manner, Table 1 shows the results obtained after the analysis of the androstenedione content of culture medium by both radioimmunoassay and high resolution mass fragmentography. These figures suggest that LH and GH, as well as ACTH and prolactin, were capable of influencing the synthesis and secretion of androstenedione by tissue explants.
Analysis of media from selected incubations indicated that no DNA sulphate was present in the media from the tumour tissue cultures both before and after incubation with either ACTH or prolactin.
Another series of analyses to investigate the effect of various protein hormones on the secretion of 11-hydroxycorticosteroids into the culture medium are also given in Table 1. Analysis of the 11-hydroxycorticosteroid concentration of the medium by the corticosteroid binding globulin method similar to that described by Murphy (1967) tended to indicate that in vitro, LH and prolactin exerted a greater effect on the tissue than did ACTH.
The effect of various protein hormones on the adenylate cyclase activity of tumour homogenates was assessed by a double isotope procedure, using [14C]ATP as enzyme substrate (Table 3). Prolactin, LH and FSH were all more effective than ACTH in stimulating the activity of the enzyme. When insulin was added to the tumour homogenate, adenylate cyclase activity was reduced when com- pared to control incubations. Studies have previously been performed in which adenylate cyclase activity has been measured in normal guinea pig tissue in the presence of several hormone preparations. In most experiments only B1-24 corticotrophin stimulated enzyme activity (Golder & Boyns 1973) although in one case LH also resulted in stimulation (Golder, unpublished data).
| Hormone added | Concentration | Enzyme activity in tumour tissue (Base-line control activity is 100) |
|---|---|---|
| B1-24ACTH | 10.0 ng/ml | 102.4 |
| B1-24ACTH | 10.0 ug/ml | 113.4* |
| Insulin | 400.0 pg/ml | 81.8* |
| HLH | 1.5 ug/ml | 135.4 ** |
| HFSH | 1.5 µg/ml | 143.3 ** |
| HTSH | 1.5 µg/ml | 101.7 |
| HCS | 1.5 µg/ml | 145.4* |
| Prolactin | 1.5 µg/ml | 148.5* |
| Prolactin | 60.0 µg/ml | 149.1* |
* Significance of increase or decrease in enzyme activity was determined by the Wilcoxon’s Rank sum test - P < 0.05.
Results from the studies on the steroid synthesis and secretion by the ‘normal’ adrenal tissue in culture are given in Table 4. The data indicates that DNA sulphate, DHA and androstenedione were present in the culture medium after 3 days incubation. 81-24Corticotrophin stimulated the tissue to increase the output of all three steroids and also promoted oestradiol and oestrone synthesis. Prolactin also stimulated this tissue but not with the same effect as cortico- trophin.
| Oe2 | Oe1 | DHA | A | DHAS | |
|---|---|---|---|---|---|
| pg/mg wet wt. tissue | ng/mg wet weight tissue | ||||
| Control @1-24ACTH | 10 | 10 | 100 | 2.2 | 9.6 |
| (1.0 µg) | 30 | 130 | 3000 | 36.0 | 26.6 |
| Prolactin | |||||
| (1.0 µg) | 10 | 10 | 200 | 3.4 | 18.8 |
DISCUSSION
Feminising adrenocortical carcinomas are rare and there are few steroid bio- synthetic studies on such tissue. Oestrogen biosynthesis in vitro from both C21- and C19-steroid precursors has been shown (Baggett et al. 1959; Gabrilove et al. 1965; Rose et al. 1969) and recent investigations (Mathur et al. 1973) demon- strated the conversion of radioactive acetate to oestrogens, oestrone being the predominant steroid formed.
It is interesting that in this present report, the concentration of endogenous oestrone relative to oestradiol-178 and oestradiol was found to be high (45 pg/ mg wet wt. tissue). Furthermore, the tissue culture studies clearly indicate that the tumour tissue was actively synthesising oestrone and oestradiol-178 and these experiments relate well to the high urinary oestrogen excretion and plasma cestradiol-17 concentration observed in this patient. Although oestrone was the principal urinary oestrogen found in this patient, and also in a case reported by Procopé (1970), this is not a consistent pattern associated with all feminising adrenocortical carcinomas (Harrison et al. 1966; Gabrilove et al. 1970). Simi- larly, active synthesis of androstenedione by the tissue in culture was also noted and this again was consistent with the elevated androstenedione and testosterone concentrations determined in the plasma of the patient. The clinical manifesta- tions obviously were dependent on the balance of androgens and oestrogens produced, rather than on any absolute values.
Tissue culture studies also provided data on the effect of various protein hormones on steroid biosynthesis in vitro. During the clinical investigations, a dexamethasone suppression test was without effect on plasma oestradiol whereas an ACTH stimulation test caused a decrease in plasma oestradiol. The plasma cortisol was however suppressed to 50% of its control concentration after dexamethasone administration and was elevated from 20 to 40 µg/100 ml after administration of tetracosactin to the patient. In vitro, ACTH appeared to have little effect of the secretion of 11-hydroxycorticosteroids, whereas LH, GH and prolactin were found to stimulate the output of these steroids from the tissue. Androstenedione secretion was shown to be responsive in vitro to LH, GH and ACTH and, furthermore, ACTH produced the greatest effect on the synthesis of oestradiol-178. Again oestrone synthesis was stimulated more by prolactin than by ACTH. The results suggested that insulin could modify the effect of the protein hormone on this tumour tissue thereby affecting steroid synthesis. This responsiveness of the steroid biosynthetic activity of this tumour tissue to various protein hormones related well to the apparent complete lack of specificity of the adenylate cyclase enzyme system to these protein hormones. Compared to HLH, HFSH, HCS and prolactin, ACTH had a limited stimulatory effect on this enzyme system. In relation to these observations, a recent report
of Pittaway et al. (1973) described the first in vitro demonstration of gonado- trophin sensitive steroidogenesis in a human adrenal cortical adenoma.
In normal guinea pig adrenal homogenates, 81-24ACTH was shown to specifi- cally stimulate adenylate cyclase activity (Golder & Boyns 1973). Insulin alone had no significant effect on the enzyme activity although it did have a modulat- ing effect on the stimulation produced by low doses of ACTH. A similar effect has also been demonstrated by Illiano & Cuatrecasas (1972) in a study of gluca- gon and epinephrine action on liver and fat cells. Moreover, although the ade- nylate cyclase enzyme system of the normal rat adrenal is responsive only to ACTH, experiments with a transplantable rat adrenal tumour indicated that several other hormones including LH, FSH and TSH increased cyclic-AMP synthesis (Schorr & Ney 1971; Schorr et al. 1971).
₿1-24Corticotrophin stimulated the normal adrenal tissue to synthesise oestro- gens and certain C19-steroids, but DHA sulphate in particular. Addition of prolactin also resulted in an increased output of steroids although not so effect- ively as corticotrophin, an effect similar to that described previously by this group (Boyns et al. 1972).
The value of high resolution-mass fragmentography is evident from these studies. The consistently lower values for oestradiol-17 and androstenedione obtained compared with those determined by radioimmunoassay reflect the greater specificity and reliability of the former technique, especially when ap- plied to previously uninvestigated systems.
ACKNOWLEDGMENT
The authors associated with the Tenovus Institute are grateful to the Tenovus Organisa- tion for generous financial support.
REFERENCES
Baggett B., Engel L. L., Balderas L., Lanman G., Savard K. & Dorfman R. I .: Endocri- nology 64 (1959) 600.
Boyns A. R., Cole E. N., Golder M. P., Danutra V., Harper M. E., Brownsey B., Cowley T., Jones G. E. & Griffiths K. In: Boyns A. R. and Griffiths K., Eds. Proceedings of the Fourth Tenovus Workshop - Prolactin & Carcinogenesis. Alpha Omega Alpha Publishing, Cardiff (1972) p. 207.
Cameron E. H. D. & Jones D. A .: Steroids 20 (1972) 737.
Conn J. W., Knopf R. F. & Nesbitt R. M. Amer. J. Surg. 107 (1964) 159.
Cowley T., Brownsey B. G., Peeling W. B., Harper M. E. & Griffiths K .: Acta endocr. (Kbh.) 81 (1976) 310.
Exley D., Johnson M. W. & Dean P. D. G .: Steroids 18 (1971) 605.
Gabrilove J. L., Nicolis G. L., Hausknecht R. U. & Wotiz H. H .: Cancer 25 (1970) 153.
Gabrilove J. L., Sharma D. C., Wotiz H. H. & Dorfman R. I .: Medicine (Baltimore) 44 (1965) 37.
Golder M. P. & Boyns A. R .: J. Endocr. 56 (1973) 471.
Harrison M. T., Brush M. G. & Macfarlane M .: J. Endocr. 34 (1966) 61.
Illiano G. & Cuatrecasas P .: Science 175 (1972) 906.
Mathur R. S., Williamson H. O., Moody L. O. & Diczfalusy E .: Acta endocr. (Kbh.) 73 (1973) 518.
Millington D. S .: J. Steroid Biochem. 6 (1975) 239.
Millington D. S., Jenner D. A., Jones T. & Griffiths K .: Biochem. J. 139 (1974) 473.
Murphy B. E. P .: J. clin. Endocr. 27 (1967) 973.
Philpot J. St. L. & Stanier J. E .: Biochem. J. 63 (1956) 214.
Pittaway D. E., Andersen R. N. & Givens J. R .: Steroids 22 (1973) 731.
Procopé B .- J .: Acta obstet. gynec. scand. 49 (1970) 125.
Rose L. J., Williams G. H., Emerson K. & Villee D. B .: J. clin. Endocr. 29 (1969) 1526.
Schorr I. & Ney R. L .: J. clin. Invest. 50 (1971) 1295.
Schorr I., Rathnam P., Saxena B. B. & Ney R. L .: J. biol. Chem. 246 (1971) 5806.
Symington T. In: Symington T., Ed. Functional Pathology of the Human Adrenal Gland. E. & S. Livingstone Limited, Edinburgh (1969) 157.
Symington T. & Jeffries R. In: Currie A. B., Symington T. and Grant J. K., Eds. The Human Adrenal Cortex. E. & S. Livingstone Limited, Edinburgh (1962) p. 345.
Trowell O. A .: Exp. Cell Res. 16 (1959) 118.
Wilkins L .: J. clin. Endocr. 8 (1948) 111.
Williams R .: Proc. roy. Soc. Med. 53 (1960) 864.
Received on July 21st, 1975.