Adrenocortical Zonation in Humans under Normal and Pathological Conditions
Koshiro Nishimoto, Ken Nakagawa, Dan Li, Takeo Kosaka, Mototsugu Oya, Shuji Mikami, Hirotaka Shibata, Hiroshi Itoh, Fumiko Mitani, Takeshi Yamazaki, Tadashi Ogishima, Makoto Suematsu, and Kuniaki Mukai
Departments of Urology (K.Ni., K.Na., T.K., M.O.), Biochemistry and Integrative Medical Biology (D.L., F.M., M.S., K.M.), and Internal Medicine (H.S., H.I.), School of Medicine, Keio University, and Division of Diagnostic Pathology (S.M.), Keio University Hospital, Tokyo 160-8582, Japan; Laboratory of Molecular Brain Science (T.Y.), Graduate School of Integrated Arts and Sciences, Hiroshima University, Higashi-Hiroshima 739-8521, Japan; Department of Chemistry (T.O.), Faculty of Sciences, Kyushu University, Fukuoka 812-8581, Japan; and Department of Urology (K.Ni.), Federation of National Public Service Personnel Mutual Aid Associations, Tachikawa Hospital, Tokyo 190-8531, Japan
Context: Aldosterone synthase (CYP11B2) and steroid 11ß-hydroxylase (CYP11B1) catalyze the termi- nal steps for aldosterone and cortisol syntheses, respectively, thereby determining the functional dif- ferentiation of human adrenocortical cells. Little is known, however, about how the cells expressing the enzymes are actually distributed in the adrenals under normal and pathological conditions.
Objective: The objective of the study was to determine the localization of CYP11B2 and -B1 in human adrenal specimens by using developed antibodies capable of distinguishing the two en- zymes from each other.
Results: Under normal conditions, CYP11B2 was sporadically detected in the zona glomerulosa, whereas CYP11B1 was entirely detected in the zonae fasciculata-reticularis. Adrenocortical cells lacking both en- zymes were observed in the outer cortical regions. In addition to conventional zonation, we found a variegated zonation consisting of a subcapsular cell cluster expressing CYP11B2, which we termed aldo- sterone-producing cell cluster, and a CYP11B1-expressing area. Aldosterone-producing adenomas differed in cell populations expressing CYP11B2 from one another, whereas CYP11B1-expressing and double-neg- ative cells were also intermingled. Adenomas from patients with Cushing’s syndrome expressed CYP11B1 entirely but not CYP11B2, resulting in atrophic nontumor glands. The nontumor portions of both types of adenomas bore frequently one or more aldosterone-producing cell clusters, which sustained CYP11B2 expression markedly under the conditions of the suppressed renin-angiotensin system.
Conclusion: Immunohistochemistry of the human normal adrenal cortex for CYP11B2 and CYP11B1 revealed a variegated zonation with cell clusters constitutively expressing CYP11B2. This technique may provide a pathological confirmatory diagnosis of adrenocortical adenomas. (J Clin Endocrinol Metab 95: 2296-2305, 2010)
T he adrenal cortex of mammals consists of three major cell zones, which differ from each other in morphology and steroidogenesis. The most potent mineralocorticoid, aldo- sterone, is synthesized in the outermost zone, the zona glo-
merulosa (zG). Cortisol and/or corticosterone are formed in the midzone, the zona fasciculata (zF). In some species in- cluding humans, adrenal androgens are produced in the in- nermost zone, the zona reticularis (zR) (1). The classical view
Abbreviations: APA, Aldosterone-producing adenoma; APCC, aldosterone-producing cell cluster; CPA, cortisol-producing adenoma; CS, Cushing’s syndrome; CYP11B1, steroid 11ß-hydroxylase cytochrome P450; CYP11B2, aldosterone synthase cytochrome P450; CYP17, steroid 17a-hydroxylase and C17-20 lyase cytochrome P450; 3ßHSD, 3ß-hydrox- ysteroid dehydrogenase; PA, primary aldosteronism; RCC, renal cell carcinoma; TF, Trigger factor; UTUC, upper urinary tract urothelial carcinoma; zF, zona fasciculata; zFR, zonae fasciculata-reticularis; zG, zona glomerulosa; zR, zona reticularis.
of the functional zonation of the adrenal cortex was estab- lished based on histological examinations and biochemical studies on steroids produced from isolated tissue fragments (2-4).
The molecular basis of the functional zonation has been derived mainly from studies using experimental animals in- cluding rodents. The differential production of the steroid hormones in rodents, especially for either aldosterone or corticosterone, is attributed to the spatially restricted ex- pression of the two enzymes responsible for the terminal steps of their syntheses. These two enzymes are aldosterone synthase cytochrome P450 or P450aldo (CYP11B2) and steroid 11ß-hydroxylase cytochrome P450 or P45011B (CYP11B1). These enzymes catalyze the syntheses of aldo- sterone and corticosterone, respectively, from the common substrate 11-deoxycorticosterone (5-8), which is synthe- sized from cholesterol by successive actions of CYP11A, 3ß-hydroxysteroid dehydrogenase (3BHSD), and CYP21. These latter enzymes are expressed in both the zG and zF. The localization of the two CYP11B enzymes by means of immunohistochemical examination confirmed the classical view and further revealed the histological details of the func- tional zonation. Under normal dietary sodium conditions, only a small cell population in the zG expresses CYP11B2, whereas the zF cells express entirely CYP11B1 (9). Cells devoid of both CYP11B enzymes, namely functionally undifferentiated cells, exist in the outer cortical region including the zG (10). In adult rat adrenal cortex, the undifferentiated cells are comprised of the zG cells without CYP11B2, which is inducible when the renin- angiotensin system is activated (9, 11), and possible stem or progenitor cells (12-17).
In humans, CYP11B2 and -B1 also catalyze the terminal steps of aldosterone and cortisol syntheses, respectively (18-20). The production of cortisol, but not corticosterone, as the major glucocorticoid is due to 17a-hydroxylation catalyzed by CYP17 in humans. Thus, expression of either CYP11B2 without CYP17 or CYP11B1 with CYP17 deter- mines the corticosteroids synthesized in human adrenocor- tical cells. Based on immunohistochemistry, the localization of CYP17 has been specified to the zonae fasciculata- reticularis (zFR) (21). However, little is known about the localization of the two CYP11B proteins because immuno- histochemical techniques for the two enzymes have not been established for human specimens prepared for routine his- topathological examinations. Although in situ hybridiza- tion studies on mRNAs for the two enzymes have been performed by different research groups, consistent localiza- tion of the mRNAs has not been unambiguously defined for normal adrenal cortex as well as pathological glands in hu- mans (22-24).
The present study reveals the functional differentiation of adrenocortical cells in humans under normal and pathological conditions by demonstrating immunohisto- chemical localization of the two CYP11B enzymes. The successful application of the immunological detection of the two enzymes to formalin-fixed paraffin-embedded specimens of human samples provides a useful tool for a definitive diagnosis of adrenocortical tumors. The immu- nohistochemical localization leads to the novel finding of cell clusters markedly expressing CYP11B2 in the normal adrenal cortex and also in nontumor portions of adreno- cortical adenomas, even under the suppressed renin-an- giotensin system. This observation suggests that aldoste- rone is produced constitutively in such a cell cluster and inducibly in the zG of the conventional zonation under normal conditions.
Materials and Methods
Adrenal specimens
Adrenal glands from eight patients with renal cell carcinoma (RCC), a patient with upper urinary tract urothelial carcinoma (UTUC), 24 patients with primary aldosteronism (PA), and six patients with adrenal Cushing’s syndrome (CS) were examined. The patients with RCC had no clinical manifestation of hyper- tension, and the patient with UTUC had mild hypertension that was controlled by nicardipine. Diagnosis of PA and CS was con- firmed on the basis of The Endocrine Society clinical practice guidelines of PA (25) and CS (26), respectively (see supplemental data, published as supplemental data on The Endocrine Society’s Journals Online web site at http://jcem.endojournals.org). The present study was approved by the Medical Ethics Committee of the School of Medicine, Keio University.
Antibodies
Antibodies to human CYP11B2 or -B1 were raised in rabbits and affinity-purified as we described previously (19). Because the two enzymes have 93% identical residues in their amino acid sequences (18), the respective peptides corresponding to amino acid residues 80-90 (RYNLGGPRMVC for CYP11B2 and RYDLGGAGMVC for CYP11B1) were used for immuniza- tion. The sulfhydryl group of the carboxy terminal cysteine res- idue was used for conjugation with equine myoglobin (19) and with column media for affinity preparation of the antibodies (9). Anti-3ßHSD antiserum was raised in rabbits against purified bovine 3ßHSD (Yamazaki, T., unpublished result). Rabbit anti- guinea pig CYP17 antiserum was described previously (27).
Immunoblotting
An inner and outer layer of a normal adrenal (case 9), a nontumor and tumor portion of an aldosterone-producing adenoma (APA; case 27), and Escherichia coli cells expressing recombinant CYP11B2 or -B1 proteins were examined using anti-CYP11B2 and anti-CYP11B1 antibodies (1:40,000 dilu- tion) with an ECL Plus reagent pack (GE Healthcare, Buck- inghamshire, UK). Blocking peptides, which had the same amino acid sequences as those used for immunization, were
| Case no. | Clinical diagnosis | Adrenal cortex | Tumor | Figure | |||
|---|---|---|---|---|---|---|---|
| CYP11B2 | CYP11B1 | Size (mm) | CYP11B2 | CYP11B1 | |||
| 1 | RCC | zG | zFR(+++) | 2, A-G | |||
| 2 | RCC | zG, APCC | zFR(+++) | ||||
| 3 | RCC | zG, APCC | zFR(+++) | 2, H, I | |||
| 4 | RCC | zG, APCC | zFR(+++) | ||||
| 5 | RCC | zG, APCC | zFR(+++) | ||||
| 6 | RCC | APCC | zFR(+++) | 2, J-O | |||
| 7 | RCC | APCC | zFR(+++) | ||||
| 8 | RCC | APCC | zFR(+++) | ||||
| 9 | UTUC | APCC | zFR(+++) | 1A | |||
| 10 | PA | zG, APCC | zFR(++) | 2, 8ª | IV, III | II, II | Suppl A-D |
| 11 | PA | zG | zFR(+++) | 5 | IV | I | |
| 12 | PA | zG, APCC | zFR(+++) | 6 | IV | I | |
| 13 | PA | zG, APCC | zFR(++) | 7 | II | ||
| 14 | PA | — | zFR(++) | 7 | IV | — | |
| 15 | PA | APCC | zFR(+) | 7 | III | I | 3, O-R |
| 16 | PA | zG, APCC | zFR(++) | 8 | III | ||
| 17 | PA | zG | zFR(+++) | 9 | II | ||
| 18 | PA | — | zFR(++) | 9 | IV | I | |
| 19 | PA | zG, APCC | zFR(+++) | 9 | II | 3, F-N | |
| 20 | PA | — | zFR(++) | 9 | IV | I | 3, A-E |
| 21 | PA | — | zFR(+) | 11 | II | I | |
| 22 | PA | — | zFR(+++) | 12 | III | ||
| 23 | PA | — | zFR(++) | 12 | IV | I | |
| 24 | PA | APCC | zFR(++) | 14 | III | ||
| 25 | PA | — | zFR(+++) | 14 | III | I | |
| 26 | PA | zG | zFR(++) | 14 | III | ||
| 27 | PA | APCC | zFR(++) | 14 | I | 1A | |
| 28 | PA | — | zFR(+) | 15 | II | ||
| 29 | PA | APCC | zFR(++) | 15 | II | ||
| 30 | PA | APCC | zFR(++) | 15 | III | ||
| 31 | PA | — | zFR(+) | 16 | III | Suppl D | |
| 32 | PA | — | zFR(+++) | 20 | III | II | Suppl D |
| 33 | PA | — | Atrophic (+)b | 20 | I | III | |
| 34 | CS | APCC | Atrophic (+)b | 15 | — | IV | 4, D-O |
| 35 | CS | — | Atrophic (+)b | 25 | — | IV | |
| 36 | CS | APCC | Atrophic (±)b | 25 | — | IV | 4, A-C |
| 37 | CS | APCC | Atrophic (+)b | 25 | — | IV | |
| 38 | CS | APCC | Atrophic (±)b | 30 | — | IVC | |
| 39 | CS | — | Atrophic (±)b | 32 | — | IVC | |
Intensities of CYP11B1 staining in normal and nontumor glands: ±, 0-10%; +, 11-50%; ++, 51-90%; +++, 91-100% of the average intensity in zFR of normal adrenals. Cell population positive for CYP11Bs in tumors: - , not detected; I, up to 10%; II, 11-50%; III, 51-90%; IV, 91-100%.
a The smaller tumor in case 10 was subjected to laser capture microdissection for the experiments in supplemental data.
b zFR of nontumor glands were atrophic.
” Twenty and 60% of the tumor cells in cases 38 and 39 were weakly positive for CYP11B1, respectively.
used at 0.25 µM for verification of the specificities. The ex- perimental procedures are described in the supplemental data.
Immunohistochemistry
Serial sections (4 pm) from formalin-fixed paraffin-embed- ded specimens were deparaffinized and treated with Target re- trieval solutions (Dako, Glostrup, Denmark) at pH 9 for staining of CYP11B2 and -B1 and at pH 6 for staining of 3ßHSD and CYP17. Immunohistochemical staining was performed using a Dako autostainer universal system according to the manufac- ture’s protocol. Anti-CYP11B2 and anti-CYP11B1 antibodies were used at 1:10,000 and 1:5,000 dilutions, respectively. An- tisera specific to 3ßHSD and CYP17 were used at 1:10,000 and
1:4,000 dilutions, respectively. Secondary antibodies used were an Envision reagent coupled with horseradish peroxidase (Dako) or a simple stain reagent coupled with alkaline phosphatase (Nichirei, Tokyo, Japan). Peroxidase activity was visualized with 3,3’-diaminobenzidine tetrahydrochloride and hydrogen perox- ide (Dako). Alkaline phosphatase activity was visualized with 5-bromo-4-chloro-3-indoxyl phosphate and nitro blue tetrazo- lium chloride (Dako). To show CYP11B2 and -B1 with different colors on a single specimen, the sections were first immuno- stained for CYP11B2 with alkaline phosphatase reaction. Sub- sequently the sections were treated again with antigen retrieval reagents to remove the bound antibodies and were stained for CYP11B1 with peroxidase reaction.
Results
Antibodies specific to human CYP11B2 and -B1
Immunoblotting was performed to characterize the an- tibodies raised against human CYP11B2 and -B1 using a normal adrenal gland (case 9, Table 1) and an APA sample (case 27). An outer and inner layer from the normal ad- renal and a nontumor and tumor portion from the APA were examined. The antibody raised against CYP11B2 gave a weak signal with a molecular mass of 48.5 kDa from the outer layer (Fig. 1A, lane1) but no detectable signal from the inner layer or from the nontumor portion (lanes 2 and 3). The tumor portion of the APA showed a well-defined band (lane 4). The antibody raised against CYP11B1 gave an easily detectable signal with a molec- ular mass of 50 kDa from both layers from the normal gland (lanes 5 and 6); the signal intensity of the inner layer was higher than that of the outer layer. A strong signal was observed from the nontumor portion of the APA sample (lane 7). The tumor portion also showed a considerable signal, although an APA (lane 8). These results agreed well with our previous observations (19).
To confirm there was no cross-reactivity of the two antibodies between CYP11B2 and -B1, bacterially pro- duced recombinant fusion proteins (TF-hB2 and TF-hB1) were tested by immunoblot analysis. The results showed that anti-CYP11B2 antibody recognized only TF-hB2 (Fig. 1B, lanes 1 and 2), whereas anti-CYP11B1 antibody recognized only TF-hB1 (lanes 3 and 4).
In the experiments mentioned above, antigenic peptides were used to examine whether they could block the immu- noreactivity of the two antibodies. The antibodies were pre- incubated with the free peptides and then used for immuno- blotting of the tissue homogenates and the recombinant proteins. The signals obtained with either antibody were blocked specifically by addition of its own peptide but not by addition of the other (Fig. 1A, lanes 9-24, and B, lanes 5-12). Thus, these results indicated that the two antibody prepara- tions reacted specifically with either CYP11B2 or -B1.
Additionally, the expression levels of mRNAs for CYP11B2 and -B1 in tissue specimens including a nontu- mor portion of an APA and tumor portions of APAs were examined using laser capture microdissection and quan- titative RT-PCR analysis. The mRNAs results were con- sistent with the results from the immunoblotting (Fig. 1) and the immunohistochemistry described below (Figs. 2-4) (see supplemental data).
Normal adrenal cortices in humans
Immunohistochemical examination using the two an- tibodies in a successive manner enabled us to detect CYP11B2 and -B1 simultaneously in a single adrenal sec-
A
Tissue homogenate
APA
APA
Normal
Normal
- Outer
₪ Inner
& Non-tumor
+ Tumor
Un Outer
® Inner
~ Non-tumor
œ Tumor
hB2
None
hB1 50k
ns +
48.5k
ns +
45k
45k
Peptide
Ponceau
Ponceau
9
10
11
12
13
14
15
16
hB2
ns +
ns+
17
18
19
20
21
22
23
24
hB1
ns >
ns +
Anti-hB2
Antibody
Anti-hB1
B E. coli lysate
-TF-hB1
NTF-hB2
Vector
@ TF-hB1
^ TF-hB2
108k
None
TF-hB2
110k
TF-hB1
97k
Peptide hB2
5
6
7
8
9
10
11
12
hB1
Anti-hB2
Antibody
Anti-hB1
tion. We found that normal adrenocortical tissue exhib- ited two types of distribution of the two enzymes: the con- ventional distribution (Fig. 2, A-G) and the variegated distribution (Fig. 2, H-O), as discussed below.
093
f
r
A CYP11B2/B1
m
B 3BHSD
C CYP17
-
C
09.3 u
3
f
DHE
E CYP: 182/B1
F 3BHSD
G
CYP17
APCC
4 APCC
APCC
H HE
CYP11B2/B1
J HE
K CYP11B2/B1
APCC
LHE
M CYP11B2/B1
N
3BHSD
☐ CYP17
-
The normal adrenocortical tissues from the nine pa- tients with RCC or UTUC exhibited conventional histol- ogy consisting of the zG, zF, and zR (Fig. 2D for case 1, Fig. 2H for case 3, Fig. 2J for case 6; Table 1). Double immu- nostaining of sections from cases 1-5 for CYP11B2 and -B1 showed that CYP11B2 (blue) was sporadically
detected in the zG, whereas CYP11B1 (brown) was entirely detected in the zF and zR (Fig. 2, A and E, for case 1). It was noted that cells devoid of both CYP11B2 and -B1 were observed in the outer areas including the zG. Negative controls (nor- mal rabbit IgG or secondary antibody only) gave no signals, and preincuba- tion of the antibodies with appropriate amounts of each of their antigenic pep- tides blocked the signals (not shown), suggesting that the immunostainings were specific.
Localization of 3@HSD and CYP17 was examined for comparison with the local- izations of CYP11B2 and -B1. Cells ex- pressing 3ßHSD evidently corresponded to the zG cells expressing CYP11B2 and the zF cells expressing CYP11B1 (Fig. 2, Band F). 3ßHSD was hardly detected in the zR. Cells expressing CYP17 corresponded to those expressing CYP11B1 (Fig. 1, C and G). Thus, the spatially restricted expression of CYP11B2 and -B1 agreed well with the conventional zonation of the adrenal cortex.
Interestingly, eight of the nine normal adrenals (cases 2-9) exhibited a varie- gated pattern consisting of cell clusters markedly expressing CYP11B2 with the remaining areas expressing CYP11B1 (Fig. 2, H-O, for cases 3 and 6). On the basis of morphological criteria, these normal adrenal cortices exhibited the conventional zonation (Fig. 2, H and J; note that the medulla is not seen in panel H). However, double staining for CYP11B2 and -B1 revealed cell clusters showing signals of CYP11B2 with a width of 200-1300 um and a depth of 100-500 um beneath the capsule (Fig. 2, I, K, and M). Such cell clusters, termed aldosterone-producing cell clusters (APCCs) in the present study, were com- posed of morphological zG cells in contact with the capsule and inner columnar zF- like cells forming cords along sinusoids (Fig. 2, L and M). The cells in the remaining subcapsular areas other than APCCs were devoid of both CYP11B en- zymes (Fig. 2I) or expressed only CYP11B1 (Fig. 2K). Ex- pression of 3ßHSD but not CYP17 in APCCs was consistent with their ability to synthesize aldosterone (Fig. 2, N and O).
NT
NT
T
A CYP1 1B2/B1
F
CYP11B2/B1
BHE
C 3BHSD
G
HE
H 3BHSD
D CYP17
E CYP11B2/B1
CYP17
J CYP11B2/B1
C
g
NT
K HE
CYP11B2/B1
M 3BHSD
N CYP17
APCC
O HE
P CYP11B2/B1
Q 3BHSD
R CYP17
Based on the two patterns of the distribution of CYP11B2-expressing cells as described above, the func- tional zonation in humans was classified into the con-
ventional zonation with sporadic expression of CYP11B2 in the zG and the novel variegated zonation with an APCC.
APAS
Twenty-four APAs were immunohistochemically ex- amined for CYP11B2 and -B1. They were found to contain varied cell populations expressing CYP11B2 (Table 1). Figure 3 shows two cases: one was almost entirely positive for CYP11B2 (Fig. 3, A and E, for case 20); the other was approximately 40% positive for it (Fig. 3, F and J, for case 19). Similarly, cell populations expressing CYP11B1 were different from one another; a few percent in the former (Fig. 3, A and E) and 20% in the latter (Fig. 3, F and J). Taken together, the APAs essentially consisted of at least three cell types: CYP11B2-positive and -B1-negative cells; CYP11B2-negative and -B1-positive cells; and double- negative cells. Double-positive cells were not detectable under the experimental conditions. The cell populations expressing CYP11B2 appeared to be inversely correlated with the tumor sizes and with the cell populations express- ing CYP11B1 (Table 1). Expression of 3ßHSD was de- tected throughout the tumors, irrespective of tumor sizes and the cell populations expressing CYP11B2 or -B1 (Fig. 3, Cand H). The localization of CYP17-expressing cells in the tumors almost corresponded to that of CYP11B1-ex- pressing cells (Fig. 3, D and I). Thus, these results sug- gested that APAs produced both aldosterone and cortisol.
Nontumor portions of APAS
Two examples of nontumor portions from the 24 APAs are shown in Fig. 3; one exhibited the conventional zona- tion histologically and immunohistochemically (Fig. 3, K-N, for case 19); the other contained an APCC and ex- hibited no or weak expression of CYP11B1 in the non- APCC area (Fig. 3, O and P, for case 15). Ten of the 24 cases contained APCCs (Table 1), even under the sup- pressed renin-angiotensin system (<9.8 pg renin per mil- liliter plasma). Similar to the expression profile in the APCCs of normal adrenal cortex, the APCCs in the non- tumor portions expressed CYP11B2 and 3ßHSD but not CYP11B1 and CYP17.
Cortisol-producing adenomas (CPAs) and their nontumor portions
Two of six cases of CPAs are shown in Fig. 4 (Fig. 4, A-C, for case 36 and Fig. 4, D-O, for case 34). CYP11B1 was detected throughout the tumors, whereas CYP11B2 was not (Fig. 4, A and E). 36HSD and CYP17 were also detected throughout the tumors (Fig. 4, C, F, and G). These results suggested that the CPAs produced cortisol but not aldosterone.
APCC
1
APCC ,
NT
1
T
A CYP11B2/B1
BHE
C 3BHSD
T
D HE
E CYP11B2/B1
F 3BHSD
G CYP17
NT
H HE
| CYP11B2/B1
J 3BHSD
K CYP17
NT
APCC
LHE
M CYP11B2/B1
N 3BHSD
☐ CYP17
The nontumor portions of the CPAs were atrophic (Fig. 4, B, H, and L) and faintly immunostained for CYP11B1 (Fig. 4, A, I, and M; Table 1). CYP11B2 was hardly de- tectable (Fig. 4I), whereas 3ßHSD and CYP17 were ex- pressed weakly (Fig. 4, J and K). APCCs were found fre-
quently in four of six cases (Fig. 4, A and M; Table 1). It was again noted that the APCCs evidently expressed CYP11B2 under the suppressed renin- angiotensin system in the patients (<9.8 pg renin per milliliter plasma) due to ex- cess mineralocorticoid activity caused by overproduction of cortisol. 3BHSD was expressed in the APCCs (Fig. 4, Cand N). Although CYP17 was not expressed in most APCCs in the atrophic nontumor portions of CPAs, the APCC shown in Fig. 4O contained CYP17-expesssing cells. Thus, the nontumor portions of CPAs were inactive for production of corticosteroids except for APCCs.
Discussion
In the present study, immunohisto- chemical detection of human CYP11B2 and -B1 clearly visualized functional differentiation of adrenocortical cells. This was enabled by development of a pair of specific antibodies that distin- guished between the two enzymes and by the application of the antibodies to standard formalin-fixed paraffin-em- bedded specimens. The establishment of the immunohistochemical technique led us to first discover the variegated zonation of the gland and confirm the conventional zonation under normal and pathological conditions (Fig. 5).
The variegated zonation in the hu- man adrenal cortices under normal conditions is a novel finding obtained in the present study. The finding raises a question about physiological and de- velopmental roles of an APCC. The functional and morphological features of APCCs may provide important clues to these roles. First, the fact that APCCs frequently occur in normal adrenal cor- tices and nontumor portions of adeno- mas suggests that the human adrenal cortex consists of the variegated zona- tion and the conventional zonation. Second, the feature that cells in an APCC form columnar cell arrangements along sinusoids similar to those of the zF supports the view that an APCC retains the normal histology, at least in part. Third, compared with the conventional zG cells that ex-
| Cells | Expression of steroidogenic enzymess | Synthesis of aldosterone or cortisol | Zonation | |||
|---|---|---|---|---|---|---|
| CYP11B2 | CYP11B1 | 3BHSD2 | CYP17 | |||
| Normal | ||||||
| zG, induced | + | - | + | - | Aldosterone | Conventional |
| zG, uninduced ZU | } - | - | - | - | - | P |
| ZF | - | + | + | + | Cortisol | Variegated |
| ZR | - | + | - | + | a | |
| APCC | + | - | + | . b | Aldosterone | |
| Adenoma | ||||||
| APA cell type 1 | + | - | + | - | Aldosterone | |
| cell type 2 | - | + | + | + | Cortisol | |
| cell type 3 | - | - | + | - | - | |
| CPA | - | + | + | + | Cortisol | |
press CYP11B2 only sporadically under control of the renin-angiotensin system (reviewed in Refs. 1 and 28), APCCs express CYP11B2 remarkably, even under the pathological conditions of PA and CS in which the endo- crine system is suppressed. On the basis of these observa- tions, we propose that aldosterone production in the hu- man adrenal cortex is constitutive in APCCs of the variegated zonation and inducible in the zG of the con- ventional zonation.
An APCC is morphologically isolated with a defined boundary from the neighboring areas with CYP11B1 ex- pression, albeit without an apparent fibrous capsule. This feature contrasts with the conventional zonation in which steroidogenic functions of cells change radially inward with cell arrangements from the subcapsular zG to the inner zF. Previous studies using chimeric and transgenic animals (29, 30) suggested that adrenocortical cells ar- ranging centripetally are clonal cells. Therefore, the con- ventional zonation is based on two principles of cell ar- rangements, namely the centripetal arrangements of clonal cells and the concentric arrangements of function- ally differentiated cells (10, 31). In this context, an APCC could not occur on the basis of the two principles for the conventional zonation. Thus, development of an APCC in the human adrenal cortex may involve a cellular mechanism different from those for the conventional zonation. More- over, because the autonomous expression of CYP11B2 in
APCCs is shared with APAs, it is possible that an APCC develops into an APA.
Besides the finding of the variegated zo- nation, the primary observation of the present study is that CYP11B2 is expressed in the zG, whereas CYP11B1 is expressed in the zFR of the adrenal cortex in human (Fig. 5). The distribution of the two enzymes is consistent with the functional differentia- tion of the conventional zonation. The pres- ence of the functionally undifferentiated cells devoid of the two enzymes in the outer regions further indicates that the histology of the conventional zonation in humans is essentially the same as that in rodents (10). By analogy with rats (12-17), it is possible in humans that the cell populations devoid of the two enzymes consist of uninduced zG cells and stem-like or progenitor cells (Fig. 5). Direct evidence for the presence of stem- like or progenitor cells would provide more insights into development and maintenance of the adrenal cortex.
Another key steroidogenic enzyme re- sponsible for the functional differentiation of human adrenocortical cells is CYP17, which carries out two separate reactions: 17a-hydroxy- lation and C17-20 cleavage. Adrenocortical expression of CYP17 in human and some other mammals, but not in rodents, results in the production of cortisol as the major glucocorticoid and also androgen such as dehydroepi- androsterone. The present study, which showed the local- ization of CYP17 in the zFR, verified the production of cor- tisol in the zFR of humans. Furthermore, our data demonstrating the colocalization of CYP17 with CYP11B1 in APAs, as well as in CPAs, histochemically proved that the tumors were able to synthesize cortisol.
Previously a few research groups performed in situ hy- bridization examinations on mRNAs for the two CYP11B enzymes with human specimens (22-24). Their results that the zG of the nontumor portions expresses CYP11B2 mRNA agreed well with our present immunohistochem- ical result for CYP11B2 in the conventional zonation. As for CYP11B1 mRNA, Enberg et al. (23) reported expres- sion of CYP11B1 mRNA in the zF, consistent with our results, whereas the other groups reported conflicting ob- servations that CYP11B1 mRNA was coexpressed with CYP11B2 in the zG (22, 24). On the other hand, our re- sults obtained using highly specific antibodies indicated that the two enzymes are expressed in a mutually exclusive manner. Thus, the present immunohistochemical tech- nique provided to give more conclusive and clearer im-
ages than those collected from the in situ hybridization techniques.
In situ hybridization examinations also showed that nonpathological portions of APAs had cell clusters ex- pressing CYP11B2 mRNA (23, 24). By comparison of their results for mRNA expression profile with our results for the enzyme proteins, such cell clusters seem to be very similar to APCCs. As shown in the present study, however, the normal adrenal cortices frequently (eight of nine) con- tained APCCs, indicating that an APCC occurs normally in the human gland.
It is unknown about how adrenocortical adenomas develop. Besides the notion that an APA originates from an APCC, the undifferentiated cells with the putative plu- ripotency might be another candidate because APAs con- tain double-negative cells in addition to cells expressing CYP11B2 or -B1 (Fig. 5). On the other hand, because CPAs consist almost of cells expressing CYP11B1 but not those expressing CYP11B2 (Fig. 5), the zF cells might be an origin of a CPA. These putative candidates might undergo neoplas- tic changes to develop adenomas in combination with altered regulation of the steroidogenic gene expression.
Highly sensitive immunohistochemical detection of CYP11B2 as well as -B1 in surgically excised specimens is clinically of great significance in understanding and diag- nosing pathologic conditions of adrenal disorders based on the following considerations. First, because CYP11B2 is expressed in APAs but not CPAs whereas CYP11B1 is expressed in both, detection of CYP11B2 is of importance in diagnosing whether a tumor is an APA or a CPA. Sec- ond, patients with an APA with a low level of CYP11B2 mRNA expression tend to have a poor prognosis of blood pressure (32). The tendency supports the importance of highly sensitive detection of CYP11B2. Third, because much less is known about pathophysiological conditions of nontumor PAs, such as idiopathic hyperaldosteronism and unilateral adrenal hyperplasia, the immunohisto- chemical detection of CYP11B2 and -B1 would provide new insights into their pathophysiologies. Finally, compared with previous in situ hybridization techniques, which are generally complicated and result in low-resolution imaging, the present immunohistochemical technique is more effec- tive and suitable for a pathological confirmatory diag- nosis of adrenocortical adenomas. Our technique can be further used for other rare clinical conditions such as glucocorticoid-remediable aldosteronism, familial al- dosteronisms type II, and adrenocortical carcinomas.
Acknowledgments
We thank Shinya Sasai, Yuka Yokomichi, and Hitomi Shimizu (Department of Pathology, Tachikawa Hospital) for excellent
technical assistance with immunohistochemistry and the Col- laborative Research Resources, School of Medicine, Keio Uni- versity for assistance with laser capture microdissection. And we thank Dr. Ryon M. Bateman for reviewing the manuscript.
Address all correspondence and requests for reprints to: Kuniaki Mukai, Department of Biochemistry and Integrative Medical Biol- ogy, School of Medicine, Keio University, 35 Shinanomachi, Shin- juku-ku, Tokyo 160-8582, Japan. E-mail, k-mukai@a3.keio.jp.
This work was supported by Global Center of Excellence Program for Human Metabolomic Systems Biology the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan and in part by Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science and by grants from Keio University. D.L. is a research fellow supported by Global Center of Excellence Program from MEXT.
Disclosure Summary: The authors have nothing to disclose.
References
1. Stewart PM 2007 The adrenal cortex. In: Kronenberg HM, Melmed S, Polonsky KS, Larsen PR, eds. Williams textbook of endocrinol- ogy. 11th ed. Philadelphia: Saunders Elsevier; 445-504
2. Giroud CJ, Stachenko J, Venning EH 1956 Secretion of aldosterone by the zona glomerulosa of rat adrenal glands incubated in vitro. Proc Soc Exp Biol Med 92:154-158
3. Ayres PJ, Eichhorn J, Hechter O, Saba N, Tait JF, Tait SA 1960 Some studies on the biosynthesis of aldosterone and other adrenal steroids. Acta Endocrinol (Copenh) 33:27-58
4. Davis WW, Burwell LR, Casper AG, Bartter FC 1968 Sites of action of sodium depletion on aldosterone biosynthesis in the dog. J Clin Invest 47:1425-1434
5. Ogishima T, Mitani F, Ishimura Y 1989 Isolation of aldosterone synthase cytochrome P-450 from zona glomerulosa mitochondria of rat adrenal cortex. J Biol Chem 264:10935-10938
6. Domalik LJ, Chaplin DD, Kirkman MS, Wu RC, Liu WW, Howard TA, Seldin MF, Parker KL 1991 Different isozymes of mouse 11ß- hydroxylase produce mineralocorticoids and glucocorticoids. Mol Endocrinol 5:1853-1861
7. Mukai K, Imai M, Shimada H, Okada Y, Ogishima T, Ishimura Y 1991 Structural differences in 5’-flanking regions of rat cytochrome P-450aldo and P-450118 genes. Biochem Biophys Res Commun 180: 1187-1193
8. Mukai K, Imai M, Shimada H, Ishimura Y 1993 Isolation and char- acterization of rat CYP11B genes involved in late steps of mineralo- and glucocorticoid syntheses. J Biol Chem 268:9130-9137
9. Ogishima T, Suzuki H, Hata J, Mitani F, Ishimura Y 1992 Zone- specific expression of aldosterone synthase cytochrome P-450 and cytochrome P-450118 in rat adrenal cortex: histochemical basis for the functional zonation. Endocrinology 130:2971-2977
10. Mitani F, Suzuki H, Hata J, Ogishima T, Shimada H, Ishimura Y 1994 A novel cell layer without corticosteroid-synthesizing enzymes in rat adrenal cortex: histochemical detection and possible physio- logical role. Endocrinology 135:431-438
11. Shibata H, Ogishima T, Mitani F, Suzuki H, Murakami M, Saruta T, Ishimura Y 1991 Regulation of aldosterone synthase cytochrome P-450 in rat adrenals by angiotensin II and potassium. Endocrinol- ogy 128:2534-2539
12. Mitani F, Ogishima T, Miyamoto H, Ishimura Y 1995 Localization of P450aldo and P450118 in normal and regenerating rat adrenal cortex. Endocr Res 21:413-423
13. Mitani F, Miyamoto H, Mukai K, Ishimura Y 1996 Effects of long term stimulation of ACTH and angiotensin II-secretions on the rat adrenal cortex. Endocr Res 22:421-431
14. Mitani F, Mukai K, Miyamoto H, Ishimura Y 1998 Localization of replicating cells in rat adrenal cortex during the late gestational and early postnatal stages. Endocr Res 24:983-986
15. Mitani F, Mukai K, Miyamoto H, Suematsu M, Ishimura Y 1999 Development of functional zonation in the rat adrenal cortex. En- docrinology 140:3342-3353
16. Miyamoto H, Mitani F, Mukai K, Suematsu M, Ishimura Y 1999 Studies on cytogenesis in adult rat adrenal cortex: circadian and zonal variations and their modulation by adrenocorticotropic hor- mone. J Biochem 126:1175-1183
17. Mitani F, Mukai K, Miyamoto H, Suematsu M, Ishimura Y 2003 The undifferentiated cell zone is a stem cell zone in adult rat adrenal cortex. Biochim Biophys Acta 1619:317-324
18. Mornet E, Dupont J, Vitek A, White PC 1989 Characterization of two genes encoding human steroid 11ß-hydroxylase (P-450118). J Biol Chem 264:20961-20967
19. Ogishima T, Shibata H, Shimada H, Mitani F, Suzuki H, Saruta T, Ishimura Y 1991 Aldosterone synthase cytochrome P-450 expressed in the adrenals of patients with primary aldosteronism. J Biol Chem 266:10731-10734
20. Curnow KM, Tusie-Luna MT, Pascoe L, Natarajan R, Gu JL, Nadler JL, White PC 1991 The product of the CYP11B2 gene is required for aldosterone biosynthesis in the human adrenal cortex. Mol Endocrinol 5:1513-1522
21. Sasano H, Mason JI, Sasano N 1989 Immunohistochemical study of cytochrome P-45017@ in human adrenocortical disorders. Hum Pathol 20:113-117
22. Pascoe L, Jeunemaitre X, Lebrethon MC, Curnow KM, Gomez- Sanchez CE, Gasc JM, Saez JM, Corvol P 1995 Glucocorticoid- suppressible hyperaldosteronism and adrenal tumors occurring in a single French pedigree. J Clin Invest 96:2236-2246
23. Enberg U, Volpe C, Höög A, Wedell A, Farnebo LO, Thorén M, Hamberger B 2004 Postoperative differentiation between unilateral adrenal adenoma and bilateral adrenal hyperplasia in primary al- dosteronism by mRNA expression of the gene CYP11B2. Eur J En- docrinol 151:73-85
24. Shigematsu K, Kawai K, Irie J, Sakai H, Nakashima O, Iguchi A,
Shimamatsu J, Shimamatsu K, Kusaba Y, Takahara O 2006 Analysis of unilateral adrenal hyperplasia with primary aldosteronism from the aspect of messenger ribonucleic acid expression for steroidogenic en- zymes: a comparative study with adrenal cortices adhering to aldoste- rone-producing adenoma. Endocrinology 147:999-1006
25. Funder JW, Carey RM, Fardella C, Gomez-Sanchez CE, Mantero F, Stowasser M, Young Jr WF, Montori VM 2008 Case detection, diagnosis, and treatment of patients with primary aldosteronism: an endocrine society clinical practice guideline. J Clin Endocrinol Metab 93:3266-3281
26. Nieman LK, Biller BM, Findling JW, Newell-Price J, Savage MO, Stewart PM, Montori VM 2008 The diagnosis of Cushing’s syn- drome: an Endocrine Society Clinical Practice Guideline. J Clin En- docrinol Metab 93:1526-1540
27. Shinzawa K, Ishibashi S, Murakoshi M, Watanabe K, Kominami S, Kawahara A, Takemori S 1988 Relationship between zonal dis- tribution of microsomal cytochrome P-450s (P-45017a,lyase and P-450C21) and steroidogenic activities in guinea-pig adrenal cortex. J Endocrinol 119:191-200
28. Müller J 1988 Regulation of aldosterone biosynthesis: physiological and clinical aspects. Berlin: Springer-Verlag KG
29. Iannaccone PM, Weinberg WC 1987 The histogenesis of the rat adrenal cortex: a study based on histologic analysis of mosaic pat- tern in chimeras. J Exp Zool 243:217-223
30. Morley SD, Viard I, Chung BC, Ikeda Y, Parker KL, Mullins JJ 1996 Variegated expression of a mouse steroid 21-hydroxylase/ß-galac- tosidase transgene suggests centripetal migration of adrenocortical cells. Mol Endocrinol 10:585-598
31. Mukai K, Nagasawa H, Agake-Suzuki R, Mitani F, Totani K, Yanai N, Obinata M, Suematsu M, Ishimura Y 2002 Conditionally im- mortalized adrenocortical cell lines at undifferentiated states exhibit inducible expression of glucocorticoid-synthesizing genes. Eur J Bio- chem 269:69-81
32. Lenzini L, Seccia TM, Aldighieri E, Belloni AS, Bernante P, Giuliani L, Nussdorfer GG, Pessina AC, Rossi GP 2007 Heterogeneity of aldosterone-producing adenomas revealed by a whole transcrip- tome analysis. Hypertension 50:1106-1113