- Email: [email protected]
Significance of steroidogenic enzymes in the pathogenesis of hyperfunctioning and non-hyperfunctioning adrenal tumor
~UTTERWORTH IBE'~M~N~
Significance of steroidogenic enzymes in the pathogenesis of hyperfunctioning and non-hyperfunctioning adrenal tumor Hiromichi Suzuki, 1 Hirotaka Shibata,* Tatsuya Maruyama,* Yuzuru Ishimura,¢ and Takao Saruta* Departments o f *Internal Medicine and ?Biochemistry, School o f Medicine, Keio Universi~, Shinjyuku-Ku, Tokyo, Japan
To elucidate the mechanisms of abnormal steroid production in hyper#unctioning and non-hyperfunctioning adrenal tumors, we examined both the activities and amounts of steroidogenic cvtochromes P450 in the tumor and non-tumor portions of these adrenals at the posttranslational (protein) level. Adrenals from 5 patients with prima~ aldosteronism, 5 with Cushing's syndrome, 1 with deox~'corticosterone (DOC)-producing adenoma, 10 with non-hyperfunctioning adrenal adenoma, and5 subjects" with normal control adrenals (obtained from patients with renal cell carcinoma) were used in our studies. Activities of P450scc, P45011~5 and P450aldo, and P450C21 and P-45017e~ were assayed in a reconstituted enzyme system using 20a-hydroxycholesterol, DOC, and progesterone, respectively, and the substrate and the extracted products were analyzed by HPLC. Enzyme amounts were determined by immunoblot analysis with anti-bovine P450scc, P45011~5, and P450C21 lgG, and anti-porcine P45017et lgG. Human P450aldo was only detected in the tumor portion of primar3, aldosteronism adrenals, with both activities and amounts of other P-450s similar to those in the non-tumor portion of primao' aldosteronism and normal controls. In Cushing's syndrome, both activities and amounts" of P45017et and P450C21 were significantly increased in the tumor compared with those in the non-tumor portion of Cushing's syndrome and normal controls. In DOC-producing adenoma, both activities and amounts" of P45017c~ and P45011 ~ in the tumor portion of the adenoma decreased compared with normal control, while those of other P450s were similar to normal controls. On the other hand, steroidogenic enzymes including P450scc, P45011 ~5, P450aldo, P45017~, and P450C21 in the non-hyperfunctioning adenomas were not different from those of normal controls. These results demonstrate that overexpression of P450aldo in primary aldosteronism, overexpression of P45017~ and P450C21 in Cushing's syndrome, and underexpression of P45017e~ and P45011~5 in DOCproducing adenoma may play some role in the pathogenesis of these three kinds of hyperfunctioning adrenal tumors. Conversely, incidentally discovered non-hyperfunctioning adrenal adenomas possess a normal synthetic system for adrenocortical steroids, and mechanisms other than abnormal expression of steroidogenic enzymes presumably underlie tumorigenesis. (Steroids' 60:42-47, 1995) Keywords: adrenal tumor; adrenal incidentaloma; steroidogenic enzymes; cytochrome P450
Introduction The adrenal cortex produces steroid hormones and overproduction characterizes hyperfunctioning adrenal syndromes,
Address reprint requests to Takao Saruta, M.D., Department of Internal Medicine, School of Medicine, Keio University, 35 Shinanomachi, Shin-
jyuku-Ku, Tokyo, 160, Japan. Steroids 60:42-47, 1995 © Elsevier Science Inc., 1995
655 Avenue of the Americas, New York, NY 10010
such as primary aldosteronism, Cushing's syndrome and so on. Overproduction of aldosterone synthase cytochrome P450 has been found in the tumor portion of aldosteroneproducing adenoma, 1 at the m R N A level as well as at the protein level. 2 These findings suggest that tumorigenesis in the adrenal cortex may relate to the overproduction of cytochrome P450 enzymes. In patients with Cushing's syndrome, the activities and amounts of P45017c~ and P450C21 were significantly e l e v a t e d ? a finding also at the m R N A level. 4
0039-128X/95/$10.00 SSDI 0039-128X(94)00025-8
Steroidogenic enzymes in hyper- and non-hyperfunctioning adenoma: Suzuki et aL Given these results, in the present study we have compared steroidogenic cytochrome P450s and steroid contents in hyperfunctioning and non-hyperfunctioning adrenal adenoma to further elucidate the pathogenesis of adrenal adenoma, in the hope o f shedding light on our understanding of incidentally discovered adrenal adenoma.
Experimental
Patients Adrenals of five patients with primary aldosteronism, five patients with Cushing's syndrome, ten patients with non-hyperfunctioning adrenal adenomas, and normal control tissue resected from five patients with renal cell carcinoma were examined. Diagnosis of hyperfunctioning adrenal adenoma was primarily made on serum and urinary hormone levels. In patients with non-hyperfunctioning adrenal adenomas, adrenal masses were found during preoperative evaluations for gastrectomy or cholecystectomy, or at investigation for hypertension on CT scan. Before undergoing surgery, the patients were given on explanation of the study protocol and their written consent was obtained. The study protocol was approved by the Ethical Committee of Keio University Hospital. 5 The patients were chosen because 1) there was no evidence of neurogenic tumor or cyst, 2) there was suppression of the serum cortisol at 9 AM by 1 mg dexamethasone overnight, and 3) metoclopramide injection to reveal occult pheochromocytoma was negative .6
Endocrine assessment
Measurement of steroid contents in the tumors Levels of pregnenolone (PREG), l 1-deoxycorticosterone (11DOC), 18-deoxycorticosterone (18-DOC), corticosterone (B), and dehydroepiandrosterone (DHEA) were measured by RIA, with antisera against preg-3-succinate-BSA, DOC-3-(Ocarboxymethyl)-oxime-BSA, B-3-(O-carboxymethyl)-oximeBSA, 1lct-succinoyloxy-DHEA-BSA, and preceded by extraction with ether (Teikoku Hormone Mfg., Co., Ltd., Tokyo, Japan). Levels of 18-DOC were determined by RIA using an antiserum against 18-DOC-3-(O-carboxymethyl)-oxime-BSA preceded by extraction with dichloromethane (Teikoku Hormone Mfg., Co.). COR content was measured with an RIA kit (Baxter ~/coat cortisol kit, USA). The progesterone (PROG) content was estimated with another RIA kit (DPC progesterone kit, Japan DPC Co., Tokyo, Japan). ALDO content was measured with an aldosterone RIA kit II (Dainabot Co.).
Measurement of enzyme activities
The clinical and endocrine findings are summarized in Table 1. The aldosterone (ALDO) content was measured with an aldosterone RIA kit II (Dainabot Co., Tokyo, Japan). The cortisol content (COR) was estimated with a RIA kit (Baxter ~/coat cortisol kit, USA). The plasma renin activity (PRA) was measured by the RIA-coated bead method employing a kit from Dainabott Radioisotope Institute (Tokyo, Japan). ACTH was measured by RIA using kits from Mitsubishi Yuka Bio-Chemi Lab. Inc. (Tokyo, Japan). Urinary metabolites were measured by hydrolysis with 13-glucuronidase, extraction, and colorimetric quantitation employing the Porter-Silber for 17-hydroxycorticosteroids (17OHCS) and Zimmerman for 17-ketosteroids (17-KS) reactions. In patients with primary aldosteronism significantly higher lev-
Table 1
els of serum aldosterone and suppressed PRA were found. In the patients with Cushing's syndrome the levels of serum cortisol were high, and none of the five patients showed either a diurinal rhythm of serum cortisol level, or suppression of serum cortisol and urinary 17-OHCS by administration of 2 or 8 mg of dexamethasone. A 55-year-old woman with a left DOC-producing adrenal tumor had hypertension and a hypermineralocorticoid state, i.e., high levels of peripheral DOC (1.0 mg/L; normal value <0.13 mg/L) and left adrenal venous DOC (19.5 mg/L versus 1.05 mg/L on the right side), suppressed PRA (0.06 ng/L/s), normal plasma aldosterone (98 ng/L) concentration, and a low level of serum potassium (2.8 mEq/L).
Mitochondrial and microsomal fractions were obtained from the tumor and the adjacent portions of the adrenals of all patients and from the control adrenals by differential centrifugation. 7
P450scc activity assay. The mitochondrial fraction (30 p,g of protein) was disrupted by freezing and thawing. In a reconstituted enzyme system including 111 p.mol/L 20ec-hydroxycholesterol as substrate, and 14 p,mol/L adrenodoxin and 0.3 p,mol/L NADPHadrenodoxin reductase as an electron transfer system, 5 mmol/L isocitrate, 0.4 units isocitrate dehydrogenase per mL, and 5 mmol/L MgC12 as an NADPH-generating system were dissolved in 20 mmol/L potassium phosphate buffer (pH 7.4) and 0.3%
Clinical features of patients with primary aidosteronism, Cushing's syndrome, DOC-producing adenoma, and non-
hyperfunctioning tumor Blood
Number Primary aldosteronism Cushing's syndrome DOC-producing tumor Non-hyperfunctioning tumor
pressure
PRA
PAC
COR
ACTH
17-OHCS
17-OS
5 5
180/100 +- 10/10 165/95 ± 12/6
0.06 ± 0.02 0.82 ± 0.05
458 ± 50 a 72 ± 25
310 -+ 36 625 ± 52 a
2.4 ± 0.5 <5
18 ± 3 54 ± 5*
28.5 -+ 5.4 35.4 ± 6.2
1
170/100
0.06
98
180
26
12
12.4
0.17 ± 0.03
305 ± 57
290 ± 21
3.77 +- 0.92
19 ± 2
30.6 ± 3.1
10
140/80 -+ 5/3
Blood pressure (mm Hg); serum K (mEq/L; 3.8-4.5); PRA, plasma renin activity (ng/mL/s) (0.27-0.48); PAC, plasma aldosterone concentration (pmol/L) (137-685); COR, plasma cortisol concentration (nmol/L) (138-414); ACTH (pmol/L; <22.02); 17-OHCS (l~mol/ day; males:8.3-35.8, females:5.5-19.3); 17-OS (p.mol/day; males: 14-45.5, females:7-28.0); Normal values are shown in parentheses, a p < 0.05 versus control values.
Steroids, 1995, vol. 60, J a n u a r y
43
Papers Tween 20. The reaction was initiated by adding NADPH, and was terminated by adding 10% cholic acid. The product, pregnenolone, was converted to progesterone by adding cholesterol oxidase, and the final product was extracted with dichloromethane and analyzed by HPLC on a TSK gel silica 150 column (4.6 × 250 mm; Tosoh, Tokyo, Japan) with n-hexane: isopropanol (100:2) as the mobile phase. 8 The HPLC system consisted of a Trirotor 11 pump and a Uvidec-100-III detector (Jasco, Tokyo, Japan). We used 5 nmol of DOCA as the internal standard. P4501113 activity assay. The mitochondrial fraction (30 p.g) was disrupted by freezing and thawing. In a reconstituted enzyme system including t00 p~mol/L DOC as a substrate, and 14 ixmol/L adrenodoxin, 0.3 txmol/L NADPH-adrenodoxin reductase as an electron transport system, 5 mmol/L isocitrate, 0.4 units of isocitrate dehydrogenase per mL, and 5 mmol/L MgCl 2 as an NADPH-generating system were dissolved in 100 mmol/L potassium phosphate buffer (pH 7.4) and 0.1 mmol/L EDTA. The reaction was initiated by adding NADPH, and was terminated by adding dichloromethane. The reaction product was extracted with dichloromethane and analyzed by HPLC on a TSK gel silica 150 column with dichloromethane:ethanol:water (96:3.6:0.4) as the mobile phase, 9'1° P45017ol and P450C21 activity assay. The microsomal fraction (30 Ixg of protein) was reacted with 100 ixmol/L progesterone as a substrate, 5 mmol/L isocitrate, 0.4 units of isocitrate dehydrogenase per mL, and 5 mmol/L MgCI 2 as an NADPH-generating system in 1 mL of 100 mmol/L potassium phosphate buffer (pH 7.4) and 0.1 mmol/L EDTA. The reaction was initiated by adding NADPH and was terminated by adding dichloromethane. The reaction product was extracted with dichloromethane and analyzed by HPLC on a TSK gel silica 150 column with n-hexane:isopropanol:acetate (93:7:1) as the mobile phase, i~ We used 10 nmol of spironolactone as the internal standard.
Immunoblot analysis Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of mitochondrial and microsomal lysates obtained from human adrenal tumors was carried out on 7.5% gel according to the method of Laemmli. t2 lmmunoblotting onto polyvinylidene difluoride membranes was performed as described previously, t3 The membrane for each tumor was blocked with skimmed milk, and incubated with anti-bovine P450scc, anti-bovine P4501113, anti-porcine P45017a, or anti-bovine P450C21 IgG (the anti-bovine P4501113 IgG was a generous gift from Dr. F. Mitani, and the other antibodies were purchased from OXYgene Dallas, TX USA). The materials were then reacted with anti-rabbit IgG-biotin complex and visualized by reaction with diaminobenzidine. ~4 Table 2
Statistical analysis The results are expressed as the means --+ SEM. Statistical analysis was performed by Wilcoxon Signed-Rank for paired t-tests and Mann-Whitney U for unpaired t-tests. Significance was set at P < 0.05. Results
Steroid content of adrenals in primary aldosteronism, Cushing's syndrome, and the normal control adrenals The steroid contents of two kinds of hyperfunctioning adrenal tumor are shown in Table 2. In aldosterone producing adenomas the aldosterone content was much higher (14fold) than in the normal control adrenals. In cortisolproducing adenomas the cortisol content was significantly e l e v a t e d , by 2 . 3 - f o l d , w h i l e t h e a m o u n t s o f 18hydroxydeoxycorticosterone and aldosterone were significantly decreased, compared with control adrenals, to one fourth and one twenty-fifth, respectively. The steroid contents of the adrenal tumors from 5 of the l0 patients with non-hyperfunctioning adrenal adenomas were examined. No abnormal or distinctive values were found for non-hyperfunctioning adrenal adenomas, except for aldosterone which was lower than in the apparently normal adrenals.
Steroidogenic P450 activities and amounts in primao' aldosteronism, Cushing's syndrome, and normal control adrenals In primary aldosteronism, the 48.5 kDa protein was expressed only in the tumor portion of aldosterone-producing adenoma, as shown previously I (Figure 1). The expressed amounts and activities of P450scc, P45011 [3, P45017c~, and P450C21 in the tumor portion of aldosterone-producing adenoma were shown to be similar to those in the nontumor portion of the adenoma and the normal control adrenals (Figure 1 and Tables 2 and 3). In Cushing's syndrome, both P45017o~ and P450C21 activities were elevated in the tumor portion o f cortisolproducing adenoma, by approximately 3-fold compared to those in the control adrenals. In concert with these changes, amounts of these enzymes expressed were elevated in the tumor portion (Figure 1). Neither the activities nor amounts
Steroid content of adrenal from patients with primary aldosteronism, Cushing's syndrome, and non-hyperfunctioning t u m o r
Primary aldosteronism Cushing's syndrome Non-hyperfunctioning tumor
ALDO
11-DOC
18-DOC
B
COR
DHEA
PREG
PROG
810 -~ 150 a
98 ± 30
528 ± 25
25.5 -+ 12.5
50 -+ 2
660 -+ 250
7.8 ± 0.8 a
2600 -+ 300
5 + 1
60 -+ 35
145 _+ 18
18.5 ± 7.5
135 ± 25
1800 ÷ 1150
4.2 -+ 1.5
2300 ~ 300
76 +- 22
84 ± 23
350 -+ 60
29.8 ± 3.5
60 ± 22
900 ± 150
3.9 ± 0.9
5500 ± 750
ALDO, aldosterone (pmol; 153.4 -+ 10.9); 11-DOC, 11-deoxycorticosterone (pmol; 212 _+ 107); 18-DOC, 18-hydroxydeoxycorticosterone (pmol; 347 ~- 33); B, corticosterone (nmol; 18.6 ± 9.8); COR, cortisol (nmol; 49.5 ± 5.4); DHEA, dehydroxyoepiandrosterone (pmol; 1051 -+ 239); PREG, pregnenolone, (nmol; 3.85 -+ 0.56), PROG, progesterone (pmol; 13979 -- 4092). Normal values were obtained from apparently normal adrenals of patients with renal cell carcinoma (n = 5). a p < 0.05 versus control values.
44
Steroids,
1995, vol. 60, January
Steroidogenic enzymes in hyper- and non-hyperfunctioning adenoma: Suzuki et aL
P-450scc • 51K 12
T N 1
T N 2
T N 3
T N T N 1 2
i;i
T N 3
i ii
P-450~
50K 1 2
T
N 1
T N 2
T N 3
T N T N T N 1 2 3
T N
P-45017c~
55K 1 2
TN 1
T N 2
T N 3
T N T N 1 2
T N 3
T N
47K
P-450c21 1 2
NL
TN 1
T N 2
T N 3
T N T N 1 2
T N 3
T N
12
DOC
NL
12345 Non-
PA
CS
hyperfunetioning
Figure I
Immunoblot analysis of enzyme levels of P450scc, P4501113,P-45017~, and P450c21 from adrenals of patients with primary aldosteronism (PA), Cushing's syndrome (CS), DOC-producing adenoma (DOC), non-hyperfunctioning adrenal cortical tumor and normal control adrenals (NL). T and N represent the tumor and adjacent portions of the adrenal adenoma. Numbers 1 to 5 refer to individual cases.
Table 3 Enzyme activities in tumor portion of adrenals from patients with primary aldosteronism, Cushing's syndrome, DOC producing adenoma, and non-hyperfunctioning tumor
Primary aldosteronism Cushing's syndrome DOC-producing tumor Non-hyperfunctioning tumor
P450scc
P4501113
P450aldo
P45017a
P450C21
4.8 -+ 0.8 5.6 +- 1.5 1.0 6.3 +- 0.6
3.5 .+ 1.3 2.4 +- 0.3 0.9 1.8 --- 0.4
80 .+ 25a <3 <3 <3
4.2 .+ 0.2 8.6 -+ 2.0a <1 5.6 -+ 1.1
4.8 +- 0.6 9.1 .+ 1.8 3 3.4 ~ 1.0
P450scc (nmol/min/mg protein; 6.03 -+ 2.03); P4501113 (nmol/min/mg protein; 1.94 -+ 0.59); P450aldo (pmol/min/mg protein; <3); P-45017e (nmol/min/mg protein; 4.25 -+ 0.87); P450C21 (nmol/min/mg protein; 4.32 -+ 0.57). Normal values obtained from the apparently normal adrenals of patients with renal cell carcinoma (n = 5) are shown in parentheses. a p < 0.05 versus control values.
of P450scc and P4501113 in cortisol-producing adenoma were different from those in control adrenals. Both activities and amounts of P450scc, P45017ot, and P450C21 in the non-tumor portion of the adenoma were significantly lower than in the tumor portion of the adenoma. In the DOC-producing adenoma, both activities and amounts of cytochromes P45017a, P4501 ll~, and P450scc in the tumor portion of the adenoma were one third, one third, and one sixth of those in the non-tumor portion of the adenoma and the normal control adrenals, while those of other P450s in the tumor portion were not significantly different from normal control adrenals. The amounts and activities of P450scc, P45011~, P450aldo, P45017a, and P450C21 in the tumor and adja-
cent portions of the non-hyperfunctioning tumors from 5 of the 10 patients were similar to those in the normal control adrenals (Figure 1 and Tables 2 and 4). However, the amounts and activities of both P450scc and P45017a in the tumor portion in two cases (lines 1 and 5 in Figure 1) were moderately increased compared to the other 3 cases.
Histology In all 10 patients, grossly well-demarcated yellowish-brown tumors of the adrenal gland were seen. Microscopically, these tumors were surrounded by a thin, broken capsule. The surrounding cortex was not atrophic and apparently normal. The cells of both the tumor and adjacent portions
S t e r o i d s , 1995, v o l . 60, J a n u a r y
45
Papers Table 4
Enzyme activities in adjacent portion of adrenals from patients with primary aldosteronism, Cushing's syndrome, DOCproducing adenoma, and non-hyperfunctioning tumor
Primary aldosteronism Cushing's syndrome DOC-producing tumor Non-hyperfunctioning tumor
P450scc
P45011 [3
P450aldo
P45017e
P450C21
6.5 ± 1.2 1.8 -+ 0.9 4.0 4.8 -+ 0.4
2.6 ± 0.6 1.8 ± 0.3 2.0 1.6 ± 0.2
<3 <3 <3 <3
4.6 _+ 0.5 2.2 _+ 0.8 <1 2.9 ± 0.6
4.0 -+ 0.6 1.8 -+ 0.3 3 2.9 ± 1.0
P450scc (nmol/min/mg protein) (6.03 -+ 2.03); P4501113 (nmol/min/mg protein) (1.94 -+ 0.59); P450aldo (pmol/min/mg protein) (<3); P45017e (nmol/min/mg protein) (4.25 -+ 0.87); P450C21 (nmol/min/mg protein) (4.32 -+ 0.57). Units and normal values in parentheses were obtained from the apparently normal adrenals of patients with renal cell carcinoma (n = 5), respectively.
were arranged in nets and cords. No nuclear atypia or mitosis was observed,
Discussion In the tumor portion of aldosterone-producing adenoma, aldosterone synthase cytochrome P450 (P450aldo) was found to be specifically overexpressed. Our findings on the steroid content of aldosterone producing adenoma and cortisol producing adenoma are compatible with those reported by Kaplan, 15 Biglieri et al., 16 Ogo et al., 2'4 and other investigators. In Cushing's syndrome, both enzyme activities and amounts of P45017et and P450C21 were found to be significantly greater in the tumor portion of cortisol-producing adenoma than in the non-tumor portion of the adrenal and in normal control adrenals, while those of P450scc and P4501113 were unchanged. Evidence that both enzyme activities and amounts of P450scc, P45017et, and P450C21 were significantly more suppressed in the non-tumor portion than in the tumor portion may be the result of suppression of the plasma ACTH level by negative feedback inhibition of cortisol overproduction. These results suggest that overexpression of both P45017et and P450C21 may be involved in the pathogenesis of Cushing's syndrome. In DOC-producing adenoma, both activities and amounts of cytochromes P45017ct and P4501113 in the tumor portion of the adenoma were one third of those in the non-tumor portion of the adenoma and the normal control adrenals. These data suggest that underexpression as well as overexpression of steroidogenic enzymes may play a role in the pathogenesis of hyperfunctioning adrenal tumor. In the present study, apparently non-hyperfunctioning adrenal cortical adenomas produced adrenal steroids. Further, in the adrenal tissue containing the non-hyperfunctioning adrenal cortical adenoma, the adjacent portion was not atrophic and also bad steroidogenic enzymes. From these data, it is suggested that non-hyperfunctioning adrenal cortical adenomas are completely different from hyperfunctioning adenomas. There have been few reports on the steroid contents and steroidogenic enzymes in cases of non-hyperfunctioning adrenal cortical adenoma. 17'is One major reason is probably due to their heterogeneity. In the present study, we excluded metastasis, cyst, lipoma, and ganglioneuroma using CT scan and MR images, and also ruled out adrenal
46
S t e r o i d s , 1995, v o l . 60, J a n u a r y
cortical carcinoma histologically. Fukushima and Gallagher ~9 first reported that "non functional" adrenocortical adenoma was able to produce pregnenolone and progesterone. Kaplan 15 examined the steroid contents of adrenal adenomas from patients with essential hypertension and demonstrated that the contents of aldosterone and corticosterone in these patients were similar to those of normal adrenal tissue and much less than that of adenoma from patients with PA. Recently, Gr6ndel et al.~7 examined the basal and ACTH-stimulated cortisol and aldosterone release from adrenocortical adenoma of patients with primary aldosteronism, Cushing's syndrome and non-hyperfunctioning adenoma. Their study was not conclusive, given the varied histologic appearance of non-hyperfunctioning adrenal cortical adenoma. Two cases which may be similar to our data showed that ACTH increased tumor cortisol release to the same extent as in the surrounding cortex. In the present study, since all steroidogenic enzymes are found in the tumor and adjacent portion, it is possible that even nonhyperfunctioning adrenal cortical adenoma may respond to the stimuli. Recently, Ogo et al. is have documented the expression of adrenal steroidogenic P450 mRNAs in nonhyperfunctioning adrenal cortical adenoma, and found them to be not different from those in the normal adrenal gland. These data are essentially identical to the present findings on protein levels of steroidogenic P450s. In the present study, although we did not measure the steroid content of adjacent portion of non-hyperfunctioning tumor, there were some differences in enzyme activities between tumor and adjacent portion of non-hyperfunctioning tumor. In particular, both P450scc and P45017ot levels were greater in the tumor than in adjacent tissue. In our previous results of adrenal tumor from patients with Cushing's syndrome the activity of P45017a was found to be increased. 3 Moreover, the aldosterone content of the tumor portion was significantly lower. Taking these data together, we suggest that a shift of steroidogenesis from the 17-deoxypathway to 17hydroxypathway is occurring in some non-hyperfunctioning tumors. Alternatively, some of these tumors may develop pre-Cushing's syndrome. These results demonstrate that overexpression of P450aldo in primary aldosteronism, overexpression of P45017a and P450C21 in Cushing's syndrome, and underexpression of P45017o~ and P4501113 in DOC-producing adenoma may play some role in the pathogenesis of these three kinds of hyperfunctioning adrenal tumors. Con-
Steroidogenic enzymes in hyper- and non-hyperfunctioning adenoma: Suzuki et aL versely, incidentally discovered non-hyperfunctioning tumors possess a normal synthetic system for adrenocortical steroids, and mechanisms other than abnormal expression of steroidogenic enzymes may be related to tumorigenesis.
8.
9.
Acknowledgment
10.
Bovine adrenodoxin and bovine NADPH-adrenodoxin reductase were generous gifts from Dr. Fumiko Mitani, Department of Biochemistry, School of Medicine, Keio University.
11.
References
12.
1.
2.
3. 4.
5. 6. 7.
Ogishima T, Shibata H, Shimada H, Mitani H, Suzuki H, Saruta T, Ishimura Y (1991). Aldosterone synthase cytochrome P450 expressed in the adrenals of patients with primary aldosteronism. J Biol Chem 266:10731-10734. Ogo A, Haji M, Ohashi M, Nawata H (1991). Expression of cytochrome P-450 mRNAs in steroidogenesis of adrenocortical adenomas from patients with primary aldosteronism. Mol Cell Endocrinol 76:7-12. Shibata H, Suzuki H, Ogishima T, Ishimura Y, Saruta T (1993). Significance of steroidogenic enzymes in the pathogenesis of adrenal tumour. Acta Endocrinol 128:235-242. Ogo A, Haji M, Ohashi M, Nawata H (1991). Markedly increased expression of cytochrome P-45017 ct-hydroxylase (P-450c17) mRNA in adrenocortical adenomas from patients with cushing's syndrome. Mol Cell Endocrinol 80:83-89. Saruta T, Suzuki H, Handa M, Igarashi Y, Kondo K, Senba S (1986). Multiple factors contribute to the pathogenesis of hypertension in Cushing's syndrome. J Clin Endocrinol Metab 62:275-279. Kawabe H, Itaya Y, Suzuki H, Kondo K, Saruta T (1985). Metoclopramide in the diagnosis of pheochromocytoma. Jpn Heart J 26:557-566. Johnson D, Lardy H (1967). Isolation of liver or kidney mitochondria. Methods Enzymol 10:94-101.
13.
14. 15. 16. 17. 18.
19.
Sugano S, Morishima N, lkeda H, Horie S (1989). Sensitive assay of cytochrome P-450scc activity by high-performance liquid chromatography. Anal Biochera 182:327-333. Ogishima T, Mitani F, Ishimura Y (1989). Isolation of aldosterone synthase cytochrome P-450 from zona glomerulosa mitochondria of rat adrenal cortex. J Biol Chera 264:10935-10938. Momoi K, Okamoto M, Fujii S, Kim C, Miyake Y, Yamano T (1983). 19-hydroxylation of 18-hydroxy-ll-deoxycorticosterone catalyzed by cytochrome P-450 1113of bovine adrenocortex. J Biol Chem 258:8855-8860. Kominami S, Inoue S, Higuchi A, Takemori S (1989). Steroidogenesis in liposomal system containing adrenal microsomal cytochrome P-450 electron transfer components. Biochim Biophys Acta 98:293-299. Laemmli U (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage 1"4. Nature 227:680-685. Erickson P, Minter L, Lasher R (1982). Quantitative electrophoretic transfer of polypeptide from SDS-polyacrylamide gels to nitrocellulose sheets: a method for their re-use in immunoautotradiographic detection of antigens. J lmmunol Methods 51:241-249. Dellis R, Sternberger L, Mann R, Banks P (1979). Immunoperoxidase technics in diagnostic pathology. Am J Clin Pathol 71:483488. Kaplan NM (1967). The steroid content of adrenal adenomas and measurements of aldosterone production in patients with essential hypertension and primary aldosteronism. J Clin Invest 46:728-734. Biglieri EG, Hane S, Slaton PEJ, Forsham PH (1963). In vivo and in vitro studies of adrenal secretions in Cushing's syndrome and primary aldosteronism. J Clin Invest 42:516-524. Gr6ndel S, Grimelius L, Tboren M, Hamberger B (1991). Basal and ACTH-stimulated cortisol and aldosterone release from adrenocortical adenomas in vitro. Eur J Surg 157:179-183. Ogo A, Haji M, Ohashi M, Takayanagi R, Yanase T, Nawata H (1992). Disordered expression of adrenal steroidogenic P450 mRNAs in incidentally discovered nonfunctioning adrenal adenoma. Endocrinol Japon 39:73-80. Fukushima DK, Gallagher TF (1963). Steroid production in "nonfunctioning" adrenal cortical tumor. J Clin Endocrinol Metab 23: 923-927.
Steroids, 1995, vol. 60, January
47
Significance of steroidogenic enzymes in the pathogenesis of hyperfunctioning and non-hyperfunctioning adrenal tumor Hiromichi Suzuki, 1 Hirotaka Shibata,* Tatsuya Maruyama,* Yuzuru Ishimura,¢ and Takao Saruta* Departments o f *Internal Medicine and ?Biochemistry, School o f Medicine, Keio Universi~, Shinjyuku-Ku, Tokyo, Japan
To elucidate the mechanisms of abnormal steroid production in hyper#unctioning and non-hyperfunctioning adrenal tumors, we examined both the activities and amounts of steroidogenic cvtochromes P450 in the tumor and non-tumor portions of these adrenals at the posttranslational (protein) level. Adrenals from 5 patients with prima~ aldosteronism, 5 with Cushing's syndrome, 1 with deox~'corticosterone (DOC)-producing adenoma, 10 with non-hyperfunctioning adrenal adenoma, and5 subjects" with normal control adrenals (obtained from patients with renal cell carcinoma) were used in our studies. Activities of P450scc, P45011~5 and P450aldo, and P450C21 and P-45017e~ were assayed in a reconstituted enzyme system using 20a-hydroxycholesterol, DOC, and progesterone, respectively, and the substrate and the extracted products were analyzed by HPLC. Enzyme amounts were determined by immunoblot analysis with anti-bovine P450scc, P45011~5, and P450C21 lgG, and anti-porcine P45017et lgG. Human P450aldo was only detected in the tumor portion of primar3, aldosteronism adrenals, with both activities and amounts of other P-450s similar to those in the non-tumor portion of primao' aldosteronism and normal controls. In Cushing's syndrome, both activities and amounts" of P45017et and P450C21 were significantly increased in the tumor compared with those in the non-tumor portion of Cushing's syndrome and normal controls. In DOC-producing adenoma, both activities and amounts" of P45017c~ and P45011 ~ in the tumor portion of the adenoma decreased compared with normal control, while those of other P450s were similar to normal controls. On the other hand, steroidogenic enzymes including P450scc, P45011 ~5, P450aldo, P45017~, and P450C21 in the non-hyperfunctioning adenomas were not different from those of normal controls. These results demonstrate that overexpression of P450aldo in primary aldosteronism, overexpression of P45017~ and P450C21 in Cushing's syndrome, and underexpression of P45017e~ and P45011~5 in DOCproducing adenoma may play some role in the pathogenesis of these three kinds of hyperfunctioning adrenal tumors. Conversely, incidentally discovered non-hyperfunctioning adrenal adenomas possess a normal synthetic system for adrenocortical steroids, and mechanisms other than abnormal expression of steroidogenic enzymes presumably underlie tumorigenesis. (Steroids' 60:42-47, 1995) Keywords: adrenal tumor; adrenal incidentaloma; steroidogenic enzymes; cytochrome P450
Introduction The adrenal cortex produces steroid hormones and overproduction characterizes hyperfunctioning adrenal syndromes,
Address reprint requests to Takao Saruta, M.D., Department of Internal Medicine, School of Medicine, Keio University, 35 Shinanomachi, Shin-
jyuku-Ku, Tokyo, 160, Japan. Steroids 60:42-47, 1995 © Elsevier Science Inc., 1995
655 Avenue of the Americas, New York, NY 10010
such as primary aldosteronism, Cushing's syndrome and so on. Overproduction of aldosterone synthase cytochrome P450 has been found in the tumor portion of aldosteroneproducing adenoma, 1 at the m R N A level as well as at the protein level. 2 These findings suggest that tumorigenesis in the adrenal cortex may relate to the overproduction of cytochrome P450 enzymes. In patients with Cushing's syndrome, the activities and amounts of P45017c~ and P450C21 were significantly e l e v a t e d ? a finding also at the m R N A level. 4
0039-128X/95/$10.00 SSDI 0039-128X(94)00025-8
Steroidogenic enzymes in hyper- and non-hyperfunctioning adenoma: Suzuki et aL Given these results, in the present study we have compared steroidogenic cytochrome P450s and steroid contents in hyperfunctioning and non-hyperfunctioning adrenal adenoma to further elucidate the pathogenesis of adrenal adenoma, in the hope o f shedding light on our understanding of incidentally discovered adrenal adenoma.
Experimental
Patients Adrenals of five patients with primary aldosteronism, five patients with Cushing's syndrome, ten patients with non-hyperfunctioning adrenal adenomas, and normal control tissue resected from five patients with renal cell carcinoma were examined. Diagnosis of hyperfunctioning adrenal adenoma was primarily made on serum and urinary hormone levels. In patients with non-hyperfunctioning adrenal adenomas, adrenal masses were found during preoperative evaluations for gastrectomy or cholecystectomy, or at investigation for hypertension on CT scan. Before undergoing surgery, the patients were given on explanation of the study protocol and their written consent was obtained. The study protocol was approved by the Ethical Committee of Keio University Hospital. 5 The patients were chosen because 1) there was no evidence of neurogenic tumor or cyst, 2) there was suppression of the serum cortisol at 9 AM by 1 mg dexamethasone overnight, and 3) metoclopramide injection to reveal occult pheochromocytoma was negative .6
Endocrine assessment
Measurement of steroid contents in the tumors Levels of pregnenolone (PREG), l 1-deoxycorticosterone (11DOC), 18-deoxycorticosterone (18-DOC), corticosterone (B), and dehydroepiandrosterone (DHEA) were measured by RIA, with antisera against preg-3-succinate-BSA, DOC-3-(Ocarboxymethyl)-oxime-BSA, B-3-(O-carboxymethyl)-oximeBSA, 1lct-succinoyloxy-DHEA-BSA, and preceded by extraction with ether (Teikoku Hormone Mfg., Co., Ltd., Tokyo, Japan). Levels of 18-DOC were determined by RIA using an antiserum against 18-DOC-3-(O-carboxymethyl)-oxime-BSA preceded by extraction with dichloromethane (Teikoku Hormone Mfg., Co.). COR content was measured with an RIA kit (Baxter ~/coat cortisol kit, USA). The progesterone (PROG) content was estimated with another RIA kit (DPC progesterone kit, Japan DPC Co., Tokyo, Japan). ALDO content was measured with an aldosterone RIA kit II (Dainabot Co.).
Measurement of enzyme activities
The clinical and endocrine findings are summarized in Table 1. The aldosterone (ALDO) content was measured with an aldosterone RIA kit II (Dainabot Co., Tokyo, Japan). The cortisol content (COR) was estimated with a RIA kit (Baxter ~/coat cortisol kit, USA). The plasma renin activity (PRA) was measured by the RIA-coated bead method employing a kit from Dainabott Radioisotope Institute (Tokyo, Japan). ACTH was measured by RIA using kits from Mitsubishi Yuka Bio-Chemi Lab. Inc. (Tokyo, Japan). Urinary metabolites were measured by hydrolysis with 13-glucuronidase, extraction, and colorimetric quantitation employing the Porter-Silber for 17-hydroxycorticosteroids (17OHCS) and Zimmerman for 17-ketosteroids (17-KS) reactions. In patients with primary aldosteronism significantly higher lev-
Table 1
els of serum aldosterone and suppressed PRA were found. In the patients with Cushing's syndrome the levels of serum cortisol were high, and none of the five patients showed either a diurinal rhythm of serum cortisol level, or suppression of serum cortisol and urinary 17-OHCS by administration of 2 or 8 mg of dexamethasone. A 55-year-old woman with a left DOC-producing adrenal tumor had hypertension and a hypermineralocorticoid state, i.e., high levels of peripheral DOC (1.0 mg/L; normal value <0.13 mg/L) and left adrenal venous DOC (19.5 mg/L versus 1.05 mg/L on the right side), suppressed PRA (0.06 ng/L/s), normal plasma aldosterone (98 ng/L) concentration, and a low level of serum potassium (2.8 mEq/L).
Mitochondrial and microsomal fractions were obtained from the tumor and the adjacent portions of the adrenals of all patients and from the control adrenals by differential centrifugation. 7
P450scc activity assay. The mitochondrial fraction (30 p,g of protein) was disrupted by freezing and thawing. In a reconstituted enzyme system including 111 p.mol/L 20ec-hydroxycholesterol as substrate, and 14 p,mol/L adrenodoxin and 0.3 p,mol/L NADPHadrenodoxin reductase as an electron transfer system, 5 mmol/L isocitrate, 0.4 units isocitrate dehydrogenase per mL, and 5 mmol/L MgC12 as an NADPH-generating system were dissolved in 20 mmol/L potassium phosphate buffer (pH 7.4) and 0.3%
Clinical features of patients with primary aidosteronism, Cushing's syndrome, DOC-producing adenoma, and non-
hyperfunctioning tumor Blood
Number Primary aldosteronism Cushing's syndrome DOC-producing tumor Non-hyperfunctioning tumor
pressure
PRA
PAC
COR
ACTH
17-OHCS
17-OS
5 5
180/100 +- 10/10 165/95 ± 12/6
0.06 ± 0.02 0.82 ± 0.05
458 ± 50 a 72 ± 25
310 -+ 36 625 ± 52 a
2.4 ± 0.5 <5
18 ± 3 54 ± 5*
28.5 -+ 5.4 35.4 ± 6.2
1
170/100
0.06
98
180
26
12
12.4
0.17 ± 0.03
305 ± 57
290 ± 21
3.77 +- 0.92
19 ± 2
30.6 ± 3.1
10
140/80 -+ 5/3
Blood pressure (mm Hg); serum K (mEq/L; 3.8-4.5); PRA, plasma renin activity (ng/mL/s) (0.27-0.48); PAC, plasma aldosterone concentration (pmol/L) (137-685); COR, plasma cortisol concentration (nmol/L) (138-414); ACTH (pmol/L; <22.02); 17-OHCS (l~mol/ day; males:8.3-35.8, females:5.5-19.3); 17-OS (p.mol/day; males: 14-45.5, females:7-28.0); Normal values are shown in parentheses, a p < 0.05 versus control values.
Steroids, 1995, vol. 60, J a n u a r y
43
Papers Tween 20. The reaction was initiated by adding NADPH, and was terminated by adding 10% cholic acid. The product, pregnenolone, was converted to progesterone by adding cholesterol oxidase, and the final product was extracted with dichloromethane and analyzed by HPLC on a TSK gel silica 150 column (4.6 × 250 mm; Tosoh, Tokyo, Japan) with n-hexane: isopropanol (100:2) as the mobile phase. 8 The HPLC system consisted of a Trirotor 11 pump and a Uvidec-100-III detector (Jasco, Tokyo, Japan). We used 5 nmol of DOCA as the internal standard. P4501113 activity assay. The mitochondrial fraction (30 p.g) was disrupted by freezing and thawing. In a reconstituted enzyme system including t00 p~mol/L DOC as a substrate, and 14 ixmol/L adrenodoxin, 0.3 txmol/L NADPH-adrenodoxin reductase as an electron transport system, 5 mmol/L isocitrate, 0.4 units of isocitrate dehydrogenase per mL, and 5 mmol/L MgCl 2 as an NADPH-generating system were dissolved in 100 mmol/L potassium phosphate buffer (pH 7.4) and 0.1 mmol/L EDTA. The reaction was initiated by adding NADPH, and was terminated by adding dichloromethane. The reaction product was extracted with dichloromethane and analyzed by HPLC on a TSK gel silica 150 column with dichloromethane:ethanol:water (96:3.6:0.4) as the mobile phase, 9'1° P45017ol and P450C21 activity assay. The microsomal fraction (30 Ixg of protein) was reacted with 100 ixmol/L progesterone as a substrate, 5 mmol/L isocitrate, 0.4 units of isocitrate dehydrogenase per mL, and 5 mmol/L MgCI 2 as an NADPH-generating system in 1 mL of 100 mmol/L potassium phosphate buffer (pH 7.4) and 0.1 mmol/L EDTA. The reaction was initiated by adding NADPH and was terminated by adding dichloromethane. The reaction product was extracted with dichloromethane and analyzed by HPLC on a TSK gel silica 150 column with n-hexane:isopropanol:acetate (93:7:1) as the mobile phase, i~ We used 10 nmol of spironolactone as the internal standard.
Immunoblot analysis Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of mitochondrial and microsomal lysates obtained from human adrenal tumors was carried out on 7.5% gel according to the method of Laemmli. t2 lmmunoblotting onto polyvinylidene difluoride membranes was performed as described previously, t3 The membrane for each tumor was blocked with skimmed milk, and incubated with anti-bovine P450scc, anti-bovine P4501113, anti-porcine P45017a, or anti-bovine P450C21 IgG (the anti-bovine P4501113 IgG was a generous gift from Dr. F. Mitani, and the other antibodies were purchased from OXYgene Dallas, TX USA). The materials were then reacted with anti-rabbit IgG-biotin complex and visualized by reaction with diaminobenzidine. ~4 Table 2
Statistical analysis The results are expressed as the means --+ SEM. Statistical analysis was performed by Wilcoxon Signed-Rank for paired t-tests and Mann-Whitney U for unpaired t-tests. Significance was set at P < 0.05. Results
Steroid content of adrenals in primary aldosteronism, Cushing's syndrome, and the normal control adrenals The steroid contents of two kinds of hyperfunctioning adrenal tumor are shown in Table 2. In aldosterone producing adenomas the aldosterone content was much higher (14fold) than in the normal control adrenals. In cortisolproducing adenomas the cortisol content was significantly e l e v a t e d , by 2 . 3 - f o l d , w h i l e t h e a m o u n t s o f 18hydroxydeoxycorticosterone and aldosterone were significantly decreased, compared with control adrenals, to one fourth and one twenty-fifth, respectively. The steroid contents of the adrenal tumors from 5 of the l0 patients with non-hyperfunctioning adrenal adenomas were examined. No abnormal or distinctive values were found for non-hyperfunctioning adrenal adenomas, except for aldosterone which was lower than in the apparently normal adrenals.
Steroidogenic P450 activities and amounts in primao' aldosteronism, Cushing's syndrome, and normal control adrenals In primary aldosteronism, the 48.5 kDa protein was expressed only in the tumor portion of aldosterone-producing adenoma, as shown previously I (Figure 1). The expressed amounts and activities of P450scc, P45011 [3, P45017c~, and P450C21 in the tumor portion of aldosterone-producing adenoma were shown to be similar to those in the nontumor portion of the adenoma and the normal control adrenals (Figure 1 and Tables 2 and 3). In Cushing's syndrome, both P45017o~ and P450C21 activities were elevated in the tumor portion o f cortisolproducing adenoma, by approximately 3-fold compared to those in the control adrenals. In concert with these changes, amounts of these enzymes expressed were elevated in the tumor portion (Figure 1). Neither the activities nor amounts
Steroid content of adrenal from patients with primary aldosteronism, Cushing's syndrome, and non-hyperfunctioning t u m o r
Primary aldosteronism Cushing's syndrome Non-hyperfunctioning tumor
ALDO
11-DOC
18-DOC
B
COR
DHEA
PREG
PROG
810 -~ 150 a
98 ± 30
528 ± 25
25.5 -+ 12.5
50 -+ 2
660 -+ 250
7.8 ± 0.8 a
2600 -+ 300
5 + 1
60 -+ 35
145 _+ 18
18.5 ± 7.5
135 ± 25
1800 ÷ 1150
4.2 -+ 1.5
2300 ~ 300
76 +- 22
84 ± 23
350 -+ 60
29.8 ± 3.5
60 ± 22
900 ± 150
3.9 ± 0.9
5500 ± 750
ALDO, aldosterone (pmol; 153.4 -+ 10.9); 11-DOC, 11-deoxycorticosterone (pmol; 212 _+ 107); 18-DOC, 18-hydroxydeoxycorticosterone (pmol; 347 ~- 33); B, corticosterone (nmol; 18.6 ± 9.8); COR, cortisol (nmol; 49.5 ± 5.4); DHEA, dehydroxyoepiandrosterone (pmol; 1051 -+ 239); PREG, pregnenolone, (nmol; 3.85 -+ 0.56), PROG, progesterone (pmol; 13979 -- 4092). Normal values were obtained from apparently normal adrenals of patients with renal cell carcinoma (n = 5). a p < 0.05 versus control values.
44
Steroids,
1995, vol. 60, January
Steroidogenic enzymes in hyper- and non-hyperfunctioning adenoma: Suzuki et aL
P-450scc • 51K 12
T N 1
T N 2
T N 3
T N T N 1 2
i;i
T N 3
i ii
P-450~
50K 1 2
T
N 1
T N 2
T N 3
T N T N T N 1 2 3
T N
P-45017c~
55K 1 2
TN 1
T N 2
T N 3
T N T N 1 2
T N 3
T N
47K
P-450c21 1 2
NL
TN 1
T N 2
T N 3
T N T N 1 2
T N 3
T N
12
DOC
NL
12345 Non-
PA
CS
hyperfunetioning
Figure I
Immunoblot analysis of enzyme levels of P450scc, P4501113,P-45017~, and P450c21 from adrenals of patients with primary aldosteronism (PA), Cushing's syndrome (CS), DOC-producing adenoma (DOC), non-hyperfunctioning adrenal cortical tumor and normal control adrenals (NL). T and N represent the tumor and adjacent portions of the adrenal adenoma. Numbers 1 to 5 refer to individual cases.
Table 3 Enzyme activities in tumor portion of adrenals from patients with primary aldosteronism, Cushing's syndrome, DOC producing adenoma, and non-hyperfunctioning tumor
Primary aldosteronism Cushing's syndrome DOC-producing tumor Non-hyperfunctioning tumor
P450scc
P4501113
P450aldo
P45017a
P450C21
4.8 -+ 0.8 5.6 +- 1.5 1.0 6.3 +- 0.6
3.5 .+ 1.3 2.4 +- 0.3 0.9 1.8 --- 0.4
80 .+ 25a <3 <3 <3
4.2 .+ 0.2 8.6 -+ 2.0a <1 5.6 -+ 1.1
4.8 +- 0.6 9.1 .+ 1.8 3 3.4 ~ 1.0
P450scc (nmol/min/mg protein; 6.03 -+ 2.03); P4501113 (nmol/min/mg protein; 1.94 -+ 0.59); P450aldo (pmol/min/mg protein; <3); P-45017e (nmol/min/mg protein; 4.25 -+ 0.87); P450C21 (nmol/min/mg protein; 4.32 -+ 0.57). Normal values obtained from the apparently normal adrenals of patients with renal cell carcinoma (n = 5) are shown in parentheses. a p < 0.05 versus control values.
of P450scc and P4501113 in cortisol-producing adenoma were different from those in control adrenals. Both activities and amounts of P450scc, P45017ot, and P450C21 in the non-tumor portion of the adenoma were significantly lower than in the tumor portion of the adenoma. In the DOC-producing adenoma, both activities and amounts of cytochromes P45017a, P4501 ll~, and P450scc in the tumor portion of the adenoma were one third, one third, and one sixth of those in the non-tumor portion of the adenoma and the normal control adrenals, while those of other P450s in the tumor portion were not significantly different from normal control adrenals. The amounts and activities of P450scc, P45011~, P450aldo, P45017a, and P450C21 in the tumor and adja-
cent portions of the non-hyperfunctioning tumors from 5 of the 10 patients were similar to those in the normal control adrenals (Figure 1 and Tables 2 and 4). However, the amounts and activities of both P450scc and P45017a in the tumor portion in two cases (lines 1 and 5 in Figure 1) were moderately increased compared to the other 3 cases.
Histology In all 10 patients, grossly well-demarcated yellowish-brown tumors of the adrenal gland were seen. Microscopically, these tumors were surrounded by a thin, broken capsule. The surrounding cortex was not atrophic and apparently normal. The cells of both the tumor and adjacent portions
S t e r o i d s , 1995, v o l . 60, J a n u a r y
45
Papers Table 4
Enzyme activities in adjacent portion of adrenals from patients with primary aldosteronism, Cushing's syndrome, DOCproducing adenoma, and non-hyperfunctioning tumor
Primary aldosteronism Cushing's syndrome DOC-producing tumor Non-hyperfunctioning tumor
P450scc
P45011 [3
P450aldo
P45017e
P450C21
6.5 ± 1.2 1.8 -+ 0.9 4.0 4.8 -+ 0.4
2.6 ± 0.6 1.8 ± 0.3 2.0 1.6 ± 0.2
<3 <3 <3 <3
4.6 _+ 0.5 2.2 _+ 0.8 <1 2.9 ± 0.6
4.0 -+ 0.6 1.8 -+ 0.3 3 2.9 ± 1.0
P450scc (nmol/min/mg protein) (6.03 -+ 2.03); P4501113 (nmol/min/mg protein) (1.94 -+ 0.59); P450aldo (pmol/min/mg protein) (<3); P45017e (nmol/min/mg protein) (4.25 -+ 0.87); P450C21 (nmol/min/mg protein) (4.32 -+ 0.57). Units and normal values in parentheses were obtained from the apparently normal adrenals of patients with renal cell carcinoma (n = 5), respectively.
were arranged in nets and cords. No nuclear atypia or mitosis was observed,
Discussion In the tumor portion of aldosterone-producing adenoma, aldosterone synthase cytochrome P450 (P450aldo) was found to be specifically overexpressed. Our findings on the steroid content of aldosterone producing adenoma and cortisol producing adenoma are compatible with those reported by Kaplan, 15 Biglieri et al., 16 Ogo et al., 2'4 and other investigators. In Cushing's syndrome, both enzyme activities and amounts of P45017et and P450C21 were found to be significantly greater in the tumor portion of cortisol-producing adenoma than in the non-tumor portion of the adrenal and in normal control adrenals, while those of P450scc and P4501113 were unchanged. Evidence that both enzyme activities and amounts of P450scc, P45017et, and P450C21 were significantly more suppressed in the non-tumor portion than in the tumor portion may be the result of suppression of the plasma ACTH level by negative feedback inhibition of cortisol overproduction. These results suggest that overexpression of both P45017et and P450C21 may be involved in the pathogenesis of Cushing's syndrome. In DOC-producing adenoma, both activities and amounts of cytochromes P45017ct and P4501113 in the tumor portion of the adenoma were one third of those in the non-tumor portion of the adenoma and the normal control adrenals. These data suggest that underexpression as well as overexpression of steroidogenic enzymes may play a role in the pathogenesis of hyperfunctioning adrenal tumor. In the present study, apparently non-hyperfunctioning adrenal cortical adenomas produced adrenal steroids. Further, in the adrenal tissue containing the non-hyperfunctioning adrenal cortical adenoma, the adjacent portion was not atrophic and also bad steroidogenic enzymes. From these data, it is suggested that non-hyperfunctioning adrenal cortical adenomas are completely different from hyperfunctioning adenomas. There have been few reports on the steroid contents and steroidogenic enzymes in cases of non-hyperfunctioning adrenal cortical adenoma. 17'is One major reason is probably due to their heterogeneity. In the present study, we excluded metastasis, cyst, lipoma, and ganglioneuroma using CT scan and MR images, and also ruled out adrenal
46
S t e r o i d s , 1995, v o l . 60, J a n u a r y
cortical carcinoma histologically. Fukushima and Gallagher ~9 first reported that "non functional" adrenocortical adenoma was able to produce pregnenolone and progesterone. Kaplan 15 examined the steroid contents of adrenal adenomas from patients with essential hypertension and demonstrated that the contents of aldosterone and corticosterone in these patients were similar to those of normal adrenal tissue and much less than that of adenoma from patients with PA. Recently, Gr6ndel et al.~7 examined the basal and ACTH-stimulated cortisol and aldosterone release from adrenocortical adenoma of patients with primary aldosteronism, Cushing's syndrome and non-hyperfunctioning adenoma. Their study was not conclusive, given the varied histologic appearance of non-hyperfunctioning adrenal cortical adenoma. Two cases which may be similar to our data showed that ACTH increased tumor cortisol release to the same extent as in the surrounding cortex. In the present study, since all steroidogenic enzymes are found in the tumor and adjacent portion, it is possible that even nonhyperfunctioning adrenal cortical adenoma may respond to the stimuli. Recently, Ogo et al. is have documented the expression of adrenal steroidogenic P450 mRNAs in nonhyperfunctioning adrenal cortical adenoma, and found them to be not different from those in the normal adrenal gland. These data are essentially identical to the present findings on protein levels of steroidogenic P450s. In the present study, although we did not measure the steroid content of adjacent portion of non-hyperfunctioning tumor, there were some differences in enzyme activities between tumor and adjacent portion of non-hyperfunctioning tumor. In particular, both P450scc and P45017ot levels were greater in the tumor than in adjacent tissue. In our previous results of adrenal tumor from patients with Cushing's syndrome the activity of P45017a was found to be increased. 3 Moreover, the aldosterone content of the tumor portion was significantly lower. Taking these data together, we suggest that a shift of steroidogenesis from the 17-deoxypathway to 17hydroxypathway is occurring in some non-hyperfunctioning tumors. Alternatively, some of these tumors may develop pre-Cushing's syndrome. These results demonstrate that overexpression of P450aldo in primary aldosteronism, overexpression of P45017a and P450C21 in Cushing's syndrome, and underexpression of P45017o~ and P4501113 in DOC-producing adenoma may play some role in the pathogenesis of these three kinds of hyperfunctioning adrenal tumors. Con-
Steroidogenic enzymes in hyper- and non-hyperfunctioning adenoma: Suzuki et aL versely, incidentally discovered non-hyperfunctioning tumors possess a normal synthetic system for adrenocortical steroids, and mechanisms other than abnormal expression of steroidogenic enzymes may be related to tumorigenesis.
8.
9.
Acknowledgment
10.
Bovine adrenodoxin and bovine NADPH-adrenodoxin reductase were generous gifts from Dr. Fumiko Mitani, Department of Biochemistry, School of Medicine, Keio University.
11.
References
12.
1.
2.
3. 4.
5. 6. 7.
Ogishima T, Shibata H, Shimada H, Mitani H, Suzuki H, Saruta T, Ishimura Y (1991). Aldosterone synthase cytochrome P450 expressed in the adrenals of patients with primary aldosteronism. J Biol Chem 266:10731-10734. Ogo A, Haji M, Ohashi M, Nawata H (1991). Expression of cytochrome P-450 mRNAs in steroidogenesis of adrenocortical adenomas from patients with primary aldosteronism. Mol Cell Endocrinol 76:7-12. Shibata H, Suzuki H, Ogishima T, Ishimura Y, Saruta T (1993). Significance of steroidogenic enzymes in the pathogenesis of adrenal tumour. Acta Endocrinol 128:235-242. Ogo A, Haji M, Ohashi M, Nawata H (1991). Markedly increased expression of cytochrome P-45017 ct-hydroxylase (P-450c17) mRNA in adrenocortical adenomas from patients with cushing's syndrome. Mol Cell Endocrinol 80:83-89. Saruta T, Suzuki H, Handa M, Igarashi Y, Kondo K, Senba S (1986). Multiple factors contribute to the pathogenesis of hypertension in Cushing's syndrome. J Clin Endocrinol Metab 62:275-279. Kawabe H, Itaya Y, Suzuki H, Kondo K, Saruta T (1985). Metoclopramide in the diagnosis of pheochromocytoma. Jpn Heart J 26:557-566. Johnson D, Lardy H (1967). Isolation of liver or kidney mitochondria. Methods Enzymol 10:94-101.
13.
14. 15. 16. 17. 18.
19.
Sugano S, Morishima N, lkeda H, Horie S (1989). Sensitive assay of cytochrome P-450scc activity by high-performance liquid chromatography. Anal Biochera 182:327-333. Ogishima T, Mitani F, Ishimura Y (1989). Isolation of aldosterone synthase cytochrome P-450 from zona glomerulosa mitochondria of rat adrenal cortex. J Biol Chera 264:10935-10938. Momoi K, Okamoto M, Fujii S, Kim C, Miyake Y, Yamano T (1983). 19-hydroxylation of 18-hydroxy-ll-deoxycorticosterone catalyzed by cytochrome P-450 1113of bovine adrenocortex. J Biol Chem 258:8855-8860. Kominami S, Inoue S, Higuchi A, Takemori S (1989). Steroidogenesis in liposomal system containing adrenal microsomal cytochrome P-450 electron transfer components. Biochim Biophys Acta 98:293-299. Laemmli U (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage 1"4. Nature 227:680-685. Erickson P, Minter L, Lasher R (1982). Quantitative electrophoretic transfer of polypeptide from SDS-polyacrylamide gels to nitrocellulose sheets: a method for their re-use in immunoautotradiographic detection of antigens. J lmmunol Methods 51:241-249. Dellis R, Sternberger L, Mann R, Banks P (1979). Immunoperoxidase technics in diagnostic pathology. Am J Clin Pathol 71:483488. Kaplan NM (1967). The steroid content of adrenal adenomas and measurements of aldosterone production in patients with essential hypertension and primary aldosteronism. J Clin Invest 46:728-734. Biglieri EG, Hane S, Slaton PEJ, Forsham PH (1963). In vivo and in vitro studies of adrenal secretions in Cushing's syndrome and primary aldosteronism. J Clin Invest 42:516-524. Gr6ndel S, Grimelius L, Tboren M, Hamberger B (1991). Basal and ACTH-stimulated cortisol and aldosterone release from adrenocortical adenomas in vitro. Eur J Surg 157:179-183. Ogo A, Haji M, Ohashi M, Takayanagi R, Yanase T, Nawata H (1992). Disordered expression of adrenal steroidogenic P450 mRNAs in incidentally discovered nonfunctioning adrenal adenoma. Endocrinol Japon 39:73-80. Fukushima DK, Gallagher TF (1963). Steroid production in "nonfunctioning" adrenal cortical tumor. J Clin Endocrinol Metab 23: 923-927.
Steroids, 1995, vol. 60, January
47