The Journal of Clinical Endocrinology & Metabolism Vol. 83, No. 5 1797-1800
Copyright © 1998 by The Endocrine Society
Molecular Basis of Severe Gynecomastia Associated with Aromatase Expression in a Fibrolamellar Hepatocellular Carcinoma1
Veena R. Agarwal,
Kazuto Takayama,
Judson J. Van Wyk,
Hironobu Sasano,
Evan R. Simpson and
Serdar E. Bulun
Cecil H. and Ida Green Center for Reproductive Biology Sciences and
Departments of Obstetrics-Gynecology and Biochemistry (V.R.A., K.T.,
E.R.S., S.E.B.) University of Texas Southwestern Medical Center,
Dallas, Texas 75235-9051; Department of Pediatrics (J.J.V.W.),
University of North Carolina, Chapel Hill, North Carolina 27599-0001;
and Department of Pathology (H.S.), Tohoku University of Medicine,
Sendai 980, Japan
Address all correspondence and requests for reprints to: Serdar E. Bulun, M.D., Green Center for Reproductive Biology Sciences, Department of Obstetrics-Gynecology, University of Texas Southwestern Medical School, 5323 Harry Hines Boulevard, Dallas, Texas 75235-9051. E-mail: bulun{at}grnctr.swmed.edu
 |
Abstract
|
|---|
This report represents the first study in the literature linking
development of severe gynecomastia, in a 171/2-yr-old boy, to
high levels of aromatase expression in a large fibrolamellar
hepatocellular carcinoma, which gave rise to extremely elevated serum
levels of estrone (1200 pg/mL) and estradiol-17ß (312 pg/mL) that
suppressed FSH and LH (1.3 and 2.8 IU/L, respectively), and
consequently testosterone (1.53 ng/mL). After removal of a 1.5-kg
hepatocellular carcinoma, gynecomastia partially regressed, and
essentially, normal hormone levels were restored (estradiol-17ß, <50
pg/mL; estrone, 74 pg/mL; testosterone, 6.85 ng/mL; and FSH/LH, 6.3/3.7
mIU/mL). Conversion of C19 steroids to estrogens occurs in
a number of human tissues and is catalyzed by aromatase P450
(P450arom), the product of the CYP19 gene in a number of
human tissues. Tissue-specific promoters are used to regulate P450arom
gene transcription in adult human tissues, e.g.
promoters I.4 and I.3 in adipose fibroblasts, and promoter II in the
gonads. Human fetal liver uses promoter I.4 to express markedly high
levels of P450arom, whereas hepatic P450arom expression normally
becomes undetectable in postnatal life. Using immunohistochemistry,
diffuse intracytoplasmic aromatase expression was detected in the liver
cancer cells from this severely feminized boy. Northern analysis
indicated the presence of P450arom transcripts in total RNA from the
hepatocellular cancer but not in the adjacent liver nor in disease-free
adult liver samples. Promoter use for aromatase expression was
determined by a specific RT-PCR method. Promoters I.3 and II were used
for P450arom gene expression in the hepatocellular cancer tissue.
Because aromatase is not expressed in the disease-free adult liver, the
presence of extremely high levels of aromatase expression in this
fibrolamellar hepatocellular carcinoma tissue is intriguing,
particularly because there is preferential use of the proximally
located P450arom promoters I.3 and II by the tumor, instead of the much
more distally located fetal liver-type promoter I.4.
|
Introduction
|
|---|
MILD GYNECOMASTIA at puberty is a common
but self-limiting condition, which ordinarily disappears by the end of
pubertal development. On the other hand, the development of severe
gynecomastia in prepubertal and adolescent boys and young men prompts
the physician to immediately rule out an estrogen-secreting tumor.
Although testicular Sertoli cell tumors and the syndrome of excessive
peripheral aromatization are some of the best studied causes of
gynecomastia of prepubertal onset, other testicular and adrenal tumors
have been reported to secrete sufficient quantities of estrogen to give
rise to rapid development of severe gynecomastia at any age (1, 2). It
should be pointed out here that gynecomastia in young boys may also be
caused by inadvertent application of exogenous estrogens such as
ointments that contain diethylstilbestrol (3). Several cases of
gynecomastia also have been reported in association with the presence
of a liver tumor (4, 5, 6). In none of these reports, however, was
aromatase expression in the tumor tissue studied. This is the first
report in which aromatase expression in a liver carcinoma was
demonstrated, which gave rise to extremely high plasma levels of
estradiol-17ß, gynecomastia, and hypogonadotropic hypogonadism. In
the steroidogenic pathway, aromatase is the enzyme that converts
C19 steroids to estrogens in a number of human cells, such
as the placental syncytiotrophoblast, ovarian granulosa cells, and
adipose and skin fibroblasts (7). Aromatase expression also is detected
at very high levels in the fetal liver (the second highest, only next
to placenta, among fetal tissues) (8). Aromatase expression in liver,
however, disappears in the postnatal period. Aromatase expression in
human tissues is regulated by use of several alternative promoters,
giving rise to transcripts with promoter-specific untranslated 5'-ends
but with an identical coding region (7). For example, whereas the
distally located promoter I.1 is exclusively used in the placenta,
giving rise to transcripts containing exon I.1; promoters I.3 and I.4
are responsible for aromatase expression in the adipose tissue in a
hormone-dependent fashion, giving rise to transcripts containing exons
I.3 and I.4 (7). In the ovary and testis, on the other hand, only the
proximally located promoter II is used (7, 9). Aromatase expression in
the fetal liver is under the control of promoter I.4 (8, 10). We have
demonstrated that promoters I.3 and II are preferentially used for
inappropriate aromatase expression in a number of human tumors, such as
endometrial carcinoma (11), uterine leiomyomas (12), adipose
fibroblasts surrounding breast cancer (13), testicular Sertoli cell
tumors, and feminizing adrenal carcinoma (2, 9). In the present report,
we demonstrated the use of promoters I.3 and II in this fibrolamellar
liver carcinoma.
|
Subject and Methods
|
|---|
This 171/2-yr-old boy was evaluated for development of
severe gynecomastia. He was born after a normal pregnancy with normal
birth weight and height. He had no major illnesses or surgeries during
childhood or early adolescence. He developed mild gynecomastia at age
12 yr. Gynecomastia never regressed but did not progress for 4 yr. From
age 161/2 yr, breast size rapidly increased to reach a severe
degree of gynecomastia. He had pain in his right flank for the past
6–9 months and mentioned recent wt loss of 5 pounds. Aside from
gynecomastia, he had a normal pubertal development and had erections
and nighttime ejaculations. His family history is negative for
gynecomastia.
On physical examination, his height was 171.2 cm (just under the 25th
percentile) and his wt was 50.6 kg (<5th percentile). He had bilateral
severe gynecomastia with mammary tissue, on the right side, of 6
x 6 x 6-cm size and, on the left side, of 8 x 8 x
7-cm size. He had a fair amount of axillary hair and apocrine secretory
activity. His liver was enlarged (vertical span of 12 cm) with an
ill-defined lower margin. Tanner stage 4 pubic hair development was
noted. His phallus was 13 cm stretched; and his testes, which measured
4.5 x 2.3 x 1.9 and 5.1 x 2.8 x 2 cm, were soft
in consistency without any evidence of palpable masses. His serum
estradiol-17ß level was extremely high, whereas gonadotropins and
testosterone were low. Hormone levels are reported in Table 1
. The rest of the blood work was
essentially normal, except for mildly elevated liver enzymes. Bone age
was consistent with chronological age. Ultrasonic examination of the
testes and adrenals indicated no abnormalities. Abdominal ultrasound
revealed a large mass, with a mean diameter of 15 cm, in the right
liver lobe and an adjacent 3.6-cm mass in the quadrate lobe. An
magnetic resonance imaging scan of the abdomen confirmed these
findings. A chest x-ray was strongly suggestive of left pulmonary
metastases.
The hepatic tumor in the right lobe, the quadrate lobe, and a portion
of the left lobe was removed surgically. The specimen weighed 1590
g. Metastatic tumors were also resected from the left lung.
Pathological diagnosis was fibrolamellar hepatocellular carcinoma.
After tumor resection, hormone levels were restored to essentially
normal for an adult man (Table 1
). Subsequently, he was given
chemotherapy and underwent bilateral mastectomies within 6 months.
Within 1 yr, a recurrence was noted in the liver, and the patient
underwent a second tumor resection.
Tissue acquisition
All tissue samples were obtained at the time of surgery and were
immediately frozen in liquid nitrogen until RNA isolation, or processed
and paraffin embedded for immunohistochemistry. Total RNA was isolated
using the guanidinium thiocyanate-cesium chloride method, as previously
described (14). For the following studies, we obtained written informed
consent, which was approved by the Institutional Review Committee of
the University of North Carolina at Chapel Hill.
Immunohistochemistry
The anti-aromatase P450 (anti-P450arom) antibody used in this
study was a rabbit IgG fraction of an antiserum raised against the
enzyme purified from human placenta (15). The immunohistochemical
procedures were performed, as previously described, on 2.5-µm-thick
sections mounted on poly-L-lysine-coated slides using the
biotin-streptavidin amplified technique with a Histone immunostaining
kit (Nichirei, Tokyo, Japan) (16). Briefly, the staining procedure was
performed as follows: 1) routine deparaffinization; 2) inactivation of
endogenous peroxidase activity with 0.3% H2O2
in methyl alcohol for 30 min at 23 C; 3) blocking with 1% goat serum
for 45 min at 23 C; 4) incubation with the primary antibody at 4 C for
18 h; 5) incubation with biotinylated goat antirabbit antibody for
30 min at 23 C; 6) incubation with peroxidase-conjugated streptavidin
for 30 min at 23 C; 7) colorimetric reaction with a solution containing
0.05% Tris-HCl (pH 7.6), 0.66 mol/L 3,3'-diaminobenzidine, and 2 mol/L
H2O2; and 8) counterstaining with 1% methyl
green.
Northern blot analysis and quantitative RT-PCR
Northern blot analysis was performed as described previously
(9). Total RNA samples from all indicated tissues and cells were
size-fractionated by electrophoresis on formaldehyde-agarose (1%) gel
and transferred to a nylon membrane by capillary elution. Hybridization
was conducted for 16 h at 42 C using a human complementary DNA
probe radiolabeled with [32P]-deoxycytidine
triphosphate.
Exon-specific competitive RT-PCR was performed according to a recently
standardized method (17). For total P450arom transcript levels, coding
region was amplified using specific oligonucleotides, as previously
described (17). We also used sense oligonucleotides specific for
untranslated first exons to amplify promoter-specific transcripts. The
antisense primer was designed from a sequence in the coding exon III.
To check the integrity and comparative quantity of RNA used in
amplification of CYP19 gene transcripts, transcripts of
glyceraldehyde-3-phosphate dehydrogenase (a housekeeping gene) were
amplified by the RT-PCR method, as described previously (17). A trace
amount of [32P]-deoxycytidine triphosphate was added to
each of the PCR samples. The reaction products were analyzed on 4%
nondenaturing polyacrylamide gels. Gels were either autoradiographed
with an x-ray film or scanned on a PhosphorImager (Molecular Dynamics,
Sunnyvale, CA) and quantitatively analyzed using ImageQuant software.
PCR amplified products obtained were of expected sizes.
|
Results and Discussion
|
|---|
Regulation of inappropriate aromatase expression in a liver cancer
tissue, which gave rise to the development of severe gynecomastia, was
studied. To date, only two reports have been published describing a
link between hepatocellular carcinoma and gynecomastia. In the first
case (a 38-yr-old man), serum estrogen levels did not change after the
removal of the tumor (4), and in the second case (a 15-yr-old boy), the
patient died before the tumor could be resected (5). In neither case
was aromatase expression in the tumor studied. In the present study, we
hypothesized that the massive amounts of estrone and estradiol-17ß
were secreted directly by the liver cancer. This pertains because
levels of estrogens, testosterone, and gonadotropins promptly returned
to normal after removal of the carcinoma (Table 1). Low serum
testosterone before surgery was a result of low LH that was suppressed
by extremely high levels of circulating estradiol-17ß, giving rise to
hypogonadotropic hypogonadism.
The next set of questions pertains to the sources of testosterone and
androstenedione, the substrates for aromatase in this boy’s liver
cancer. Were these C19 steroids synthesized within the
liver tumor, or did the rates of testosterone and androstenedione
production, as calculated from their plasma levels and metabolic
clearance rates, account for the estrogen secretion by the tumor? Both
of these mechanisms might have provided the C19 precursors
for aromatase activity in the tumor. We, however, do not have an
answer, because variables such as the metabolic clearance rates of
these C19 steroids and estrogens and the rates of
aromatization by the liver tumor and hepatic blood flow were not
determined.
To study aromatase expression in the liver cancer tissue, we used
immunohistochemistry, northern blot analysis, and exon-specific RT-PCR
techniques. Upon hematoxylin and eosin staining of the liver tumor, we
observed large polygonal cells with abundant cytoplasm and sharply
demarcated borders (Fig. 1A). These
sheets of neoplastic cells were separated by collagen-rich stroma
containing fibroblasts (fibrolamellar variant). Using a polyclonal
antibody, abundant immunoreactive aromatase expression was detected in
the cytoplasm of large neoplastic cells of the hepatocellular carcinoma
(Fig. 1B). Immunoreactive aromatase was also detected in the stromal
fibroblasts. Higher staining intensity in overwhelming numbers of the
neoplastic hepatocellular cancer cells, however, suggested that
aromatase expression in these cells accounted for the largest portion
of estrogen production in the tumor tissue. No immunoreactive aromatase
was detected in the breast tissue of this patient, indicating that
peripheral aromatization in the adipose tissue did not significantly
contribute to estrogen excess in this boy (Fig. 1C). Northern blot
analysis (Fig. 2) showed readily
detectable P450arom mRNA in total RNA samples from the liver tumor and
fetal liver, whereas no transcripts were detected in total RNA from the
adjacent liver, recurrent tumor, or normal liver tissues (obtained from
persons who died in traffic accidents). Placenta and adipose stromal
cells, treated with dibutyryl AMP, were used as positive controls for
northern analysis. Similar results were obtained using RT-PCR as a more
sensitive technique, and we were also able to observe aromatase
expression, to a lesser degree, in the recurrent tumor (Fig. 3).
View larger version (74K):
[in this window]
[in a new window]
|
Figure 1. Immunohistochemical detection of
P450arom (brown stain) in the histological sections of
the fibrolamellar hepatocellular carcinoma. A, Hematoxylin and eosin
staining of the liver tumor show large polygonal cells with abundant
cytoplasm and sharply demarcated borders (arrows). These
neoplastic cells are separated into sheets and nests by relatively
acellular collagen-rich stroma containing fibroblasts
(arrowheads). B, Abundant immunoreactive aromatase
(light brown stain) is detected primarily in the
cytoplasm of neoplastic cells of the hepatocellular carcinoma
(arrows). C, No immunoreactive aromatase is detected in
glandular epithelial cells (arrowheads) or stromal
fibroblasts (arrows) of the breast tissue of the
patient.
|
|
View larger version (59K):
[in this window]
[in a new window]
|
Figure 2. Northern blot analysis showing expression of
P450arom. An equal amount of total RNA (20 µg) from each tissue was
used. Adj. Liver, Normal liver tissue adjacent to the tumor; Rec.
tumor, recurrent tumor; ASC (Bt2cAMP), adipose stromal
cells treated with dibutyryl cAMP; Normal liver, disease-free control
liver samples from two adults who died in traffic accidents.
|
|
View larger version (53K):
[in this window]
[in a new window]
|
Figure 3. I, RT-PCR of promoter-specific P450arom
transcripts in total RNA from liver tumor (A), adjacent normal liver
tissue (B), recurrent tumor (C), normal liver from a disease-free adult
(D), and fetal liver (E). Coding region, Coding exons II and III that
are common to all P450arom transcripts (size: 194 bp); Promoter
II-specific, promoter II-specific transcripts (size: 305 bp); Exon I.3,
promoter I.3-specific transcripts (size: 289 bp); Exon I.4, promoter
I.4-specific transcripts (size: 294 bp). PCR-amplified products were of
expected size (17). II, Amplification of the ubiquitous marker,
glyceraldehyde-3-phosphate dehydrogenase, from total RNA of above
mentioned tissues.
|
|
To characterize the regulation of aromatase expression in the
hepatocellular carcinoma, it was essential to determine the alternative
promoter(s) responsible for gene transcription in this tissue.
Therefore, we used an exon-specific RT-PCR technique to determine the
promoter-specific 5'-ends of P450arom transcripts in the liver tumor
(17). We observed that promoters I.3 and II were used in the liver
cancer, whereas promoter I.4-specific sequences were found in the fetal
liver tissue, as expected (Fig. 3). This was somewhat surprising,
because regression to use the fetal-type promoter I.4 might be expected
in hepatic cells after neoplastic transformation. Instead, the
cAMP-inducible promoters, I.3 and II, were used in this liver cancer
tissue. This is similar to our recent findings in breast cancer, where
promoter switching takes place from promoter I.4 in the disease-free
breast fat to promoters I.3 and II in the tumor-bearing breast adipose
tissue (13). Interestingly, in an adrenal cancer too, promoter II was
used for aromatase expression, as was the case also in Sertoli cell
tumors (2, 9). High expression of aromatase transcripts and protein in
these tumors and the use of promoters II and I.3 indicate some common
mechanism of regulation of P450arom gene expression in various
endocrine tumors. Because in nonneoplastic tissues, use of promoter II
is dependent on steroidogenic factor 1 and cAMP response element
binding protein, it will be of interest to investigate the role of
these factors in aromatase expression in these neoplastic tissues,
including liver cancer (18).
|
Footnotes
|
|---|
1 This work was supported, in part, by research Grants CA-67167 and
DAMD17–94-J-4188 from the National Cancer Institute and the U.S. Army
Medical Research and Development Command (to S.E.B.) and AG-08174 from
the National Institute on Aging (to E.R.S.). 
Received October 2, 1997.
Revised January 14, 1998.
Accepted January 22, 1998.
|
References
|
|---|
-
Bulun SE, Albini CH, Brodie AM, McGillivray MH, Yu
W, Simpson ER. 1994 Quantification of aromatase cytochrome P450
transcripts in a testicular tumor and breast tissue samples from a
prepubertal boy with gynecomastia and Peutz-Jeghers syndrome using
competitive RT-PCR technology. Proc Endocr Soc. 76:498 (Abstract).
-
Young J, Bulun SE, Agarwal V, et al. 1996 Clinical
case seminar. Aromatase expression in a feminizing adrenocortical
tumor. J Clin Endocrinol Metab. 81:3173–3176.[CrossRef][Medline]
-
Halperin DS, Sizonenko PC. 1983 Prepubertal
gynecomastia following topical inunction of estrogen containing
ointment. Helv Paediatr Acta. 38:361–366.[Medline]
-
Salles G, Vital-Durand D, Mackiewicz R, Pugeat M,
Levrat R. 1987 Hepatocellular carcinoma and hyperestrogenemia in a
male. Gastroenterol Clin Biol. 11:607–609.[Medline]
-
Hany MA, Betts DR, Schmugge M, et al. 1997 A
childhood fibrolamellar hepatocellular carcinoma with increased
aromatase activity and a near triploid karyotype. Med Pediatr Oncol. 28:136–138.[Medline]
-
Nappi O, Ferrara G, Ianniello G, Wick MR. 1992 Metastatic hepatocellular carcinoma of the breast, simulating
gynecomastia: diagnosis by fine-needle aspiration biopsy. Diagn
Cytopathol. 8:588–592.[Medline]
-
Simpson ER, Mahendroo MS, Means GD, et al. 1994 Aromatase cytochrome P450, the enzyme responsible for estrogen
biosynthesis. Endocr Rev. 15:342–355.[CrossRef][Medline]
-
Toda K, Simpson ER, Mendelson CR, Shizuta Y, Kilgore
MW. 1994 Expression of the gene encoding aromatase cytochrome P450
(CYP19) in fetal tissues. Mol Endocrinol. 8:210–217.[Abstract]
-
Bulun SE, Rosenthal IM, Brodie AMH, et al. 1993 Use of tissue-specific promoters in the regulation of aromatase
cytochrome P450 gene expression in human testicular and ovarian sex
cord tumors, as well as in normal fetal and adult gonads. J Clin
Endocrinol Metab. 77:1616–1621.[Abstract]
-
Zhao Y, Mendelson CR, Simpson ER. 1995 Characterization of the sequences of the human CYP19 (aromatase) gene
that mediate regulation by glucocorticoids in adipose stromal cells and
fetal hepatocytes. Mol Endocrinol. 9:340–349.[Abstract]
-
Bulun SE, Economos K, Miller D, Simpson ER. 1994 CYP19 (aromatase cytochrome P450) gene expression in human malignant
endometrial tumors. J Clin Endocrinol Metab. 79:1831–1834.[Abstract]
-
Bulun SE, Simpson ER, Word RA. 1994 Expression of
the CYP19 gene and its product aromatase cytochrome P450 in human
leiomyoma tissues and cells in culture. J Clin Endocrinol Metab. 78:736–743.[Abstract]
-
Agarwal VR, Bulun SE, Leitch M, Rohrich R, Simpson
ER. 1996 Use of alternative promoters to express the aromatase
cytochrome P450 (CYP19) gene in breast adipose tissues of
cancer-free and breast cancer patients. J Clin Endocrinol Metab. 81:3843–3849.[Abstract/Free Full Text]
-
Chirgwin JM, Przybyla AE, MacDonald RJ, Rutter WJ. 1979 Isolation of biologically active ribonucleic acid from sources
enriched in ribonuclease. Biochemistry. 18:5294–5299.[CrossRef][Medline]
-
Harada N. 1988 Cloning of a comparative cDNA
encoding human aromatase: immunochemical identification and sequence
analysis. Biochem Biophys Res Commun. 156:725–732.[CrossRef][Medline]
-
Sasano H, Nagura H, Harada N, Goukon Y, Kimura M. 1994 Immunolocalization of aromatase and other steroidogenic enzymes in
human breast disorders. Hum Pathol. 25:530–535.[CrossRef][Medline]
-
Agarwal VR, Bulun SE, Simpson ER. 1995 Quantitative
detection of alternatively spliced transcripts of the aromatase
cytochrome P450 (CYP19) gene in aromatase-expressing human cells by
competitive RT-PCR. Mol Cell Probes. 9:453–464.[CrossRef][Medline]
-
Michael MD, Michael LF, Simpson ER. 1997 A CRE-like
sequence that binds CREB and contributes to cAMP-dependent regulation
of the proximal promoter of the human aromatase P450 (CYP19) gene. Mol
Cell Endocrinol. 134:147–156.[CrossRef][Medline]
This article has been cited by other articles:
|
|
|
|
 
M. T. Smith, E. R. Blatt, P. Jedlicka, J. D. Strain, and L. Z. Fenton
Best Cases from the AFIP: Fibrolamellar Hepatocellular Carcinoma
RadioGraphics,
March 1, 2008;
28(2):
609 - 613.
[Full Text]
[PDF]
|
|
|
|
|
|
|
L. A. M. Castagnetta, B. Agostara, G. Montalto, L. Polito, I. Campisi, A. Saetta, T. Itoh, B. Yu, S. Chen, and G. Carruba
Local Estrogen Formation by Nontumoral, Cirrhotic, and Malignant Human Liver Tissues and Cells
Cancer Res.,
August 15, 2003;
63(16):
5041 - 5045.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. K. McLarney, P. T. Rucker, G. N. Bender, Z. D. Goodman, N. Kashitani, and P. R. Ros
From the Archives of the AFIP: Fibrolamellar Carcinoma of the Liver: Radiologic-Pathologic Correlation
RadioGraphics,
March 1, 1999;
19(2):
453 - 471.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
