| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Articles: Hormones and Reproductive Health |
Pediatric and Reproductive Endocrinology Branch, National Institute of Child Health and Human Development (N.H., C.P.C., S.B., S.R.B.), and Laboratory of Pathology, National Cancer Institute (M.A.-A.), National Institutes of Health, Bethesda, Maryland 20892; Department of Pediatrics, Monash University (A.L.), Clayton, Victoria 3168, Australia; Division of Pediatric Endocrinology, New York University Medical Center (B.K.), New York, New York 10016; Department of Medicine and School of Biological Sciences, University of Manchester (A.W.), Manchester M13 9PT, United Kingdom; and Department of Endocrinology, University of Düsseldorf, (S.R.B.), 40225 Düsseldorf, Germany
Address all correspondence and requests for reprints to: S. R. Bornstein, M.D., Department of Endocrinology, University of Düsseldorf, Mooren Str 5, 40225 Düsseldorf, Germany. E-mail: STEFAN.BORNSTEIN{at}UNI-DUSSELDORF.DE
Abstract
We describe the first case of an adrenocortical-pituitary hybrid tumor causing Cushing’s syndrome in a 17-yr-old boy. Adrenal vein sampling confirmed elevated secretion of both cortisol and ACTH precursors from a right adrenal mass, whereas pituitary ACTH levels, as determined by bilateral inferior petrosal sinus samples (IPSS), were unresponsive to CRH and equal to peripheral levels. There was no biochemical or histological evidence for a pheochromocytoma, but, rather, the tumor demonstrated lipid-rich clear cells characteristic of an adrenocortical adenoma. Immunohistochemical analysis revealed ACTH immunoreactivity and synaptophysin proteins in the tumor. Isolation of tumor cells by the novel technique of laser capture microdissection and subsequent RT-PCR showed expression of POMC messenger ribonucleic acid and cytochrome p450 enzyme messenger ribonucleic acid within the same cells. Finally, ultrastructural analysis provided ultimate proof for adrenocortical-pituitary hybrid cells exhibiting the characteristic vesicular mitochondria and abundant smooth endoplasmic reticulum of steroid cells and the typical secretory granules of corticotrophs within the cytoplasm of the same cells. The adrenocortical tumor expressed the pituitary transcription factor pituitary homeobox factor 1 and the steroidogenic factor 1. The intermingling of the centrally located ectodermally derived pituitary tissue with the mesodermally derived adrenocortical tissue in this adenoma suggests a hitherto unrecognized genetic and phenotypic plasticity within the hypothalamic-pituitary-adrenal axis.
CUSHING’S SYNDROME can be ACTH dependent,
most commonly due to a pituitary corticotroph tumor or ectopic ACTH
production, or ACTH independent, due to adrenocortical,
cortisol-producing tumors (1, 2, 3). Here we present an
unusual case that challenges this classification and our views on tumor
development in the hypothalamicpituitary-adrenal (HPA) axis. We
report an adrenal tumor containing both pituitary and adrenal elements
that produced Cushing’s syndrome. Adrenal vein sampling suggested the
secretion of both cortisol and ACTH from a right adrenal mass, whereas
ACTH levels in the inferior petrosal sinuses were equal to peripheral
levels. There was no imaging, biochemical, or histological evidence of
a pheochromocytoma; in fact, the tumor was comprised of lipid-rich
clear cells, characteristic of an adrenocortical adenoma.
Immunohistochemical analysis revealed the presence of ACTH
immunoreactivity and synaptophysin in the tumor. Isolation of tumor
cells by laser capture microdissection and subsequent RT-PCR showed
expression of POMC, 17
-hydroxylase, and pituitary homeobox factor-1
messenger ribonucleic acid (mRNA) within the same cells. Finally,
ultrastructural analysis provided ultimate proof that there were
adrenocortical-pituitary hybrid cells exhibiting both the
characteristic vesicular mitochondria and abundant smooth endoplasmic
reticulum (SER) of steroid-secreting cells in addition to the typical
secretory granules of corticotrophs.
Case Report
A 17-yr-old boy presented with a 1-yr history of hypertension, swelling of the extremities, increasing obesity, facial plethora and fullness, easy bruising, and abdominal enlargement with development of large purple striae. Physical examination revealed weight of 70 kg, height of 162.7 cm, and blood pressure of 151/99 mm Hg. Cushingoid features included facial plethora, bilateral supraclavicular fat pads and a dorsocervical fat pad, truncal obesity, and multiple violaceous striae on his abdomen, thighs, and axillae. The patient did not present with signs of hyperpigmentation.
Plasma cortisol concentrations were elevated, with loss of circadian rhythm [19.5 µg/dL (538 nmol/L) at midnight; 21.3 µg/dL (587 nmol/L) at 0730 h]. Mean 24-h urinary free cortisol excretion was elevated at 399 µg. Plasma ACTH values were persistently elevated, ranging from 29.0 pg/mL (6.42 pmol/L) to 63.5 pg/mL [14.1 pmol/L; normal range, 0–23 pg/mL (0–5.1 pmol/L)]. There was failure of suppression of morning cortisol after administration of 8 mg dexamethasone at midnight and a lack of response for ACTH after iv administration of 1 mg/kg ovine CRH. Coronal and sagittal sections (3 mm) through the pituitary using Tesla magnetic resonance imaging (MRI) did not show any abnormality. An MRI scan of the neck, chest, and abdomen and a computed tomography scan of the chest and pelvis failed to reveal a source of ectopic ACTH.
An adrenal CT scan revealed a 3-cm round mass with calcification and central necrosis and/or fat attached to the medial limb of the right adrenal gland. The remainder of the right gland and the left gland appeared atrophic, consistent with an ACTH-independent process. T2-weighted MRI of the adrenals was not suggestive of a pheochromocytoma. Plasma and urinary catecholamine and metanephrine levels were normal. There was no response to a glucagon stimulation test.
Bilateral adrenal venous sampling revealed cortisol production from the right adrenal gland (553 µg/dL; 15,257 nmol/L), which responded to exogenous ACTH (3,150 µg/dL; 86,908 nmol/L). Cortisol levels in the left adrenal vein (29.9 µg/dL; 824 nmol/L) were similar to peripheral levels (15 µg/dL; 413 nmol/L) and did not respond to ACTH (21 µg/dL; 579 nmol/L). Right adrenal vein samples contained ACTH (24 ng/L; 5.3 pmol/L) and large amounts of ACTH precursors (1,091 pmol/L).
The patient underwent laparoscopic right adrenalectomy. Postoperative cortisol and ACTH values were undetectable, necessitating hydrocortisone replacement. The clinical stigmata of Cushing’s syndrome resolved gradually.
Materials and Methods
Molecular and histological evaluation
Tissue from the patient’s tumor and from four cortisol-producing adenomas from other patients with ACTH-independent Cushing’s syndrome was transferred to tubes containing prechilled PBS (pH 7.6) and kept on ice until immunohistochemical and molecular studies were performed. Three tumors from patients with adrenocortical carcinoma, three pheochromocytomas, and three pituitary corticotroph adenomas were studied in parallel along with normal human adrenal glands and pituitaries from four patients who underwent nephrectomy for renal carcinoma or at autopsy. Approval of the use of human tissues for research was obtained by the corresponding committee on the protection of the rights of human subjects from each participating institute.
Assays
For all measurements, 8 mL blood were taken into a chilled
syringe with ethylenediamine tetraacetate and immediately centrifuged
at 1600 x g for 15 min at 4 C. Plasma was stored at
-70 C until analysis. ACTH [normal range, 0–23 pg/mL (0–5.1
pmol/L)] and cortisol levels [normal range, 5–25 µg/dL (138–689
nmol/L)] were measured by specific assays (ACTH, immunocytometric
assay, Mayo Clinic, Rochester, MN; cortisol, RIA,
Diagnostic Products, Los Angeles, CA). Controls were used
in the low and high sections of the standard curve. Samples were run in
duplicate, and standards were run in triplicate. The intra- and
interassay coefficients of variation were both less than 5%. ACTH
precursors were measured by a two-site immunoradiometric assay using a
pair of antibodies, one that recognizes ACTH and one that recognizes
the 
MSH sequence in N-proopiocortin. Both antibodies are
required to produce a signal, and the assay only recognizes POMC and
pro-ACTH, but not any of the derived peptides, such as ACTH,
N-proopiocortin, ß-lipotropin, etc. The
two-site immunoradiometric assay for ACTH precursors enables a direct
quantitation of their levels (4, 5).
Laser capture microdissection (LCM)
All adrenocortical tissues were fixed in 70% alcohol and stained with hematoxylin and eosin. Sections were dehydrated in graded alcohol, then in xylene, and were air-dried for 5 min before LCM. The optically transparent thin film caps were placed on top of tissue sections, and the tissue-film sandwich was viewed in an inverted microscope (model CK2, Olympus Corp., Tokyo, Japan). Using a phosphotide laser beam (Arcturus, Mountain View, CA), medullary and cortical cells were separated by melting film and target tissue and removing both from the surrounding area. Total RNA was then extracted from cells adherent to the film cap, in accordance with the RNA microisolation protocol (Stratagene, La Jolla, CA).
RT-PCR
Complementary DNA was synthesized by RT using transfer RNA from
the LCM and human pituitary gland mRNA (CLONTECH Laboratories, Inc., Palo Alto, CA) after treatment with deoxyribonuclease I
(GenHunter Corp, Nashville, TN). PCR amplification was performed using
primers for the following pituitary genes: CYP17, POMC,
pituitary homeobox factor-1 (Ptx1), leukemia inhibitory factor (LIF),
steroidogenic factor-1 (SF-1), pituitary specific transcription
factor-1 (Pit1), prophet of Pit1 (Prop1), prohormone
convertase-1 (PC1), and PC2 (Table 1
).
PCR was performed as previously described (6). All PCR
products were verified by sequencing. The negative control was a blank
prepared using all reagents with substitution of an equal volume of
distilled water for transfer RNA. The negative samples revealed no
amplification products.
|
To quantify the expression of ACTH receptor and CYP17, we
applied the TaqMan PCR using the 7700 Sequence Detector
(Perkin-Elmer Corp. PE Applied Biosystems,
Foster City, CA). Reactions contained 1 x TaqMan Universal PCR
Master Mix, 900 nmol/L forward and reverse primers for ACTH receptor
(sense, 5'-CGATCCCACACCAGGAAGAT-3'; antisense,
5'-TCAGTGTGATGGCCCCTTTC-3') and CYP17 (sense, 5'-CCAGC
CGGATCAGTTCATG-3'; antisense, 5'-AAATAGCTTACTGACGGTGAGATGAG-3'), and
200 nmol/L TaqMan probes (ACTH receptor, 5'-TCCACCCTCCCCAGAGCCAACA-3';
CYP17, 5'-CCCGCTGGATTCAAGAAACGCTCA-3'). 18S primers and probe were
added at 50 nmol/L. Thermal cycling proceeded with 50 cycles of 95 C
for 15 s and 60 C for 1 min. Input RNA amounts were calculated
with relative standard curves for both mRNAs of interest (ACTH receptor
and CYP17) and 18S.
Immunohistochemical analyses
Adrenocortical, chromaffin, and immune cells were characterized immunohistochemically in paraffin-embedded sections of adrenal tissue using antibodies against chromogranin A and synaptophysin (clone DAK-A3, DAKO Corp., Hamburg, Germany), CD45 (clones 2B11 and PD7/26, DAKO Corp.), and CD68 (clone KP1, DAKO Corp.), as previously described (6). Two antibodies against ACTH were used to determine the presence and distribution of ACTH-positive cells. The presence of ACTH was evaluated by immunostaining with two antibodies to ACTH [NCL-ACTH (clone 56, Novocastra Laboratories, Ltd., Burlingame, LA) and ACTH rabbit IgG antihuman (Sigma, St. Louis, MO)]. Antimouse SF-1 rabbit polyclonal IgG were purchased from Upstate Biotechnology, Inc. (Lake Placid, NY). Bound antibodies were detected by the linked streptavidin-biotin-peroxidase method (DAKO Corp.), and the enzyme reaction was visualized with 3-amino-9-ethylcarbazole (Dianova, Hamburg, Germany) chromogen solution containing 0.05% H2O2. All slides were counterstained with hematoxylin for 1 min, rinsed, and mounted. Monoclonal mouse Ig was used as a negative control.
Electron microscopy
For the ultrastructural investigations, small pieces of tissue were fixed in 4% paraformaldehyde-1% glutaraldehyde in 0.1 mol/L phosphate buffer (pH 7.3) for 3 h, postfixed for 90 min in 2% OsO4 in 0.1 mol/L cacodylate (pH 7.3), dehydrated in ethanol, and embedded in epoxy resin. Ultrathin sections (70 nm) were mounted on 200-mesh uncoated nickel grids. Sections were then stained with uranyl acetate and lead citrate, examined, and photographed at 80 kV under a Phillips electron microscope CM10 (Phillips Electronics, Mahway, NJ).
Results
Histological examination and immunohistochemical characterization
Microscopic examination of the patient’s adrenal tumor revealed a well vascularized nodule that contained characteristic adrenocortical adenoma cells. The cells were pale staining and lipid rich, with a cytoplasm that appeared optically clear. There was no diffuse growth pattern, cellular atypical mitotic figure pleomorphism, or other signs of malignancy. The remnant adrenal adenoma cells stained negative for chromogranin A, which was positive in normal adrenal medulla and three specimens of pheochromocytoma. Tumor cells from the patient, however, stained positively for synaptophysin a neuroendocrine tumor marker.
Expression of ACTH protein
The patient’s adenoma contained many cells that stained with
antibodies to ACTH (Fig. 1D). The
staining was similar to that found in normal corticotroph cells of the
pituitary (Fig. 1A). In contrast, the normal adrenal cortex and the
three other specimens of adrenal adenoma did not stain with an ACTH
antibody (Fig. 1, B and C).
|
Adenoma cells were selectively procured by laser capture
microdissection with visualization under the microscope, allowing
isolation of single cells and cell groups (Fig. 2, A–C). RT-PCR assay for both POMC mRNA
and CYP17- hydroxylase mRNA was performed on the RNA isolated from
these cells (Fig. 2D). POMC mRNA and CYP17- hydroxylase mRNA were
expressed in the patient’s tumor cells. In contrast, there was no POMC
mRNA expression in the normal human adrenocortical cells and the three
other cortisol-secreting adrenal adenomas, whereas these tissues
demonstrated positive expression for CYP17 hydroxylase mRNA.
Conversely, pituitary tissue demonstrated no expression of CYP17
hydroxylase mRNA, whereas there was a strong signal for POMC mRNA (Fig. 2D). The expression of CYP17 mRNA and ACTH receptor mRNA was determined
by TaqMan PCR. ACTH receptor expression was 115.7 ± 29.4% in the
tumor tissue compared with normal adrenal tissue 101.5 ± 19.5%.
CYP17 expression was 56.3 ± 3.5% in the tumor tissue compared
with 100.4 ± 14.2% in the normal adrenal tissue.
|
Ultrastructurally, the adrenocortical adenoma cells from controls
demonstrated ample SER, lysosomes, and mitochondria characterized by
round tubulovesicular cristae (Fig. 3, A
and E). Chromaffin cells from normal adrenals and pheochromocytoma
cells contained a large number of membrane-bound secretory granules and
dense core vesicles of approximately 60–300 nm in diameter (Fig. 3B).
These vesicles were either round or elongated. Medium density vesicles
and particulate substructure contained epinephrine. Vesicles with
electron-dense granules within a large lucent vacuole contained
norepinephrine. The cytoplasmic feature of pituitary corticotrophs
demonstrated vesicles with highly variable electron density, numerous
free ribosomes, and a moderately abundant rough endoplasmic reticulum.
The mitochondria were spherical or oval, with laminar or tubular
cristae. The secretory granules were numerous in normal pituitaries,
but less frequent in a patient with a sparsely granulated corticotroph
adenoma (Fig. 3C). The secretory granules demonstrated characteristic
variability in size, shape, and electron density, measuring from
150–700 nm. Secretory granules were frequently found in a linear
submembrane configuration. In the patient’s tumor cells, there were a
large number of granules with the characteristic shape and size of
pituitary corticotrophs (Fig. 3, D and F). Likewise, the granules lined
up in a similar manner to that seen in pituitary corticotrophs. The
tumor cells exhibited the round tubulovesicular mitochondria and ample
vesicular SER that characterize adrenocortical cells together with the
characteristic secretory granules of corticotrophs, all within the
cytoplasm of the same cells (Fig. 3F).
|
The patient’s adrenocortical tumor cells and the normal pituitary
expressed the mRNA for the Ptx1 (Fig. 4),
and SF-1. In addition, both the patient’s tumor and the normal
pituitary corticotrophs expressed LIF. SF-1 was present in normal
adrenocortical tissue, adrenocortical tumors, and normal pituitary.
Pit1 and Prop1 transcription factors, important for the differentiation
of PRL-, GH-, and TSH-producing cells, were expressed in the
pituitary tissue, but not in the ACTH-producing adrenocortical tumor
(Fig. 4A). The tumor did not express the enzymes PC1 and PC2 that are
responsible for processing POMC to pro-ACTH. Both enzymes were
expressed in human pituitary tissue used as a control (Fig. 4B).
|
Our patient’s adrenocortical adenoma was an unprecedented clinical challenge. As there was no partial suppression of ACTH levels with dexamethasone, and no response to the CRH stimulator test, ectopic ACTH secretion, rather than Cushing’s disease, was suspected. Yet, imaging studies, bilateral inferior petrosal sinus sampling, and adrenal vein sampling suggested an adrenocortical adenoma secreting ACTH and cortisol. Cellular and molecular analyses of the tumor cells revealed the astonishing case of a pituitary-adrenocortical hybrid tumor.
Embryologically, there are a limited number of patterns in the development of the HPA axis. The adrenocortical primordium is derived from mesenchymal tissue contiguous with the coelomic epithelium of the suprarenal cavity, near the urogenital ridge (7, 8). On the other hand, adrenomedullary cells appear in the adrenals of fetuses by migration of sympathogonia from the primitive sympathetic ganglia adjacent to the adrenals (7, 8). This embryonic development explains the morphogenesis of adrenal rest tissue in the gonads and the intermingling of cortical and chromaffin tissue within normal adrenal glands (9, 10). Interestingly, even normal human adrenocortical cells may exhibit properties of neuroendocrine cells (11), whereas various adrenocortical tumors aberrantly express neuroendocrine markers and/or receptors for neurotransmitters, neuropeptides, and cytokines (6, 12, 13, 14). In patients with primary pigmented nodular adrenal disease, we found expression of synaptophysin, and on the ultrastructural level, these cells had vesicle-like secretory granules (15). Our patient’s adrenal tumor, in addition to synaptophysin expression, produced large amounts of ACTH precursors that cross-reacted in the ACTH assay, and the tumor also contained secretory granules usually found in pituitary corticotrophs (16). Similar to a pituitary corticotroph tumor and in contrast to pheochromocytomas or other neuroendocrine tumors, the cells did not express the neuroendocrine tumor marker, chromogranin A (17). The ACTH released into the circulation consisted of less active ACTH precursors (4, 5), which apparently did not provide adequate stimulation to the contralateral side, explaining the atrophy of the left adrenal and the remnant right adrenal cortex in our patient. Quantitative TaqMan PCR of CYP17 mRNA and ACTH receptor mRNA did not reveal any significant changes compared with normal adrenal tissue. The expression of PC1 and PC2, the enzymes processing POMC peptides to ACTH, was negative in the patient’s tumor, providing further evidence that only ACTH precursors were released from the adrenal adenoma. Therefore, these data suggest that the ACTH precursor produced locally in this tumor did not exhibit a significant autocrine/paracrine action and do not explain the autonomous hypersecretion of cortisol.
The pituitary gland is derived from the oral ectoderm of the dorsomedial region of Rathke’s pouch and from an evagination of the infundibular portion of the diencephalon of the brain (16, 18, 19). Therefore, an intermingling of the centrally located ectodermally derived anterior pituitary tissue with the mesodermally derived adrenocortical tissue in the periphery should not be expected. Yet here we demonstrate that this can occur in humans, possibly revealing a hitherto unrecognized plasticity within the human HPA axis. Pituitary development and differentiation are regulated by the interaction of multiple transcription factors (19, 20, 21). These factors regulate pituitary lineage and commitment with their expression pattern, allowing recognition of progenitor cells from which particular tumors are derived (22). Recent evidence suggests that a similar code of transcription factors and trans-activators may regulate differentiation and gene expression in the adrenal and pituitary (23, 24, 25, 26). Thus, SF-1 is both a key transcription factor for the development of adrenal, gonadal steroidogenic tissue, and the pituitary (24, 25, 26). Likewise, Nur77, which belongs to the nuclear receptor family, is expressed at all three levels of the HPA axis (23). It activates transcription of the POMC gene in the pituitary and mediates ACTH-induced expression of steroidogenic enzymes in the adrenal cortex. Ptx1, a bicoid-related vertebrate homeobox gene, is a strong activator of several pituitary-specific promoters and is considered a pan-pituitary transcription factor. This is the first report describing the presence of this pituitary transcription factor in adrenal tissue. Ptx1 achieves cell- and promoter-specific transcriptional activation by cooperation with lineage-restricted transcription factors and growth factors (24). Thus, the combined expression of Ptx1 and LIF, a cytokine required for the differentiation of corticotrophs, has been implicated in the formation of ACTH-producing pituitary tumors (27, 28). Similarly, in our patient’s tumor both factors were expressed, whereas other pituitary transcription factors, such as Pit1 and Prop1, which are involved in the formation of lactotrophs, thyrotrophs, and somatotrophs, were absent (22). ACTH is the earliest intact human hormone to be secreted (29). Furthermore, the commitment of the POMC-secreting cell appears to occur in a cell lineage independent of the other pituitary trophic hormone-secreting cells and may reflect the possibility that an independent POMC-expressing cell may arise in the periphery. This intriguing possibility is also supported by the observation that POMC-expressing cells are most resistant to pituitary radiation or compression. The clinical, morphological, and molecular features of this case suggest a hitherto unrecognized genetic and phenotypic plasticity within the human HPA axis.
Acknowledgments
We thank Drs. Maria Tsokos and Heiner Westphal for helpful discussions, and Gregory T. Frey, Ninet Sinaii, Holger Willenberg, and Zahra Rassouli for technical assistance.
Received November 2, 2000.
Revised February 7, 2001.
Accepted March 2, 2001.
References
-Hydroxylase and chromogranin A in 6th week human
fetal adrenals. Horm Metab Res. 29:30–32.[Medline]
This article has been cited by other articles:
 |
|
 |
|
  D. Cartier, S. Jegou, F. Parmentier, I. Lihrmann, E. Louiset, J.-M. Kuhn, C. Bastard, P.-F. Plouin, M. Godin, H. Vaudry, et al. Expression profile of serotonin4 (5-HT4) receptors in adrenocortical aldosterone-producing adenomas Eur. J. Endocrinol., December 1, 2005; 153(6): 939 - 947. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Bertherat, V. Contesse, E. Louiset, G. Barrande, C. Duparc, L. Groussin, P. Emy, X. Bertagna, J.-M. Kuhn, H. Vaudry, et al. In Vivo and in Vitro Screening for Illegitimate Receptors in Adrenocorticotropin-Independent Macronodular Adrenal Hyperplasia Causing Cushing's Syndrome: Identification of Two Cases of Gonadotropin/Gastric Inhibitory Polypeptide-Dependent Hypercortisolism J. Clin. Endocrinol. Metab., March 1, 2005; 90(3): 1302 - 1310. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Lefebvre, C. Duparc, N. Chartrel, S. Jegou, A. Pellerin, A. Laquerriere, R. Ivell, H. Vaudry, and J.-M. Kuhn Intraadrenal Adrenocorticotropin Production in a Case of Bilateral Macronodular Adrenal Hyperplasia Causing Cushing's Syndrome J. Clin. Endocrinol. Metab., July 1, 2003; 88(7): 3035 - 3042. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
