| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Division of Molecular Genetics (N.H., N.Y., Y.T.), Institute for Comprehensive Medical Science, Fujita Health University, Toyoake, Aichi 470, Japan; and the Department of Laboratory Medicine (H.O., T.K.), Shinshu University School of Medicine, Matsumoto, Nagano 390, Japan
Address all correspondence and requests for reprints to: Nobuhiro Harada, Division of Molecular Genetics, Institute for Comprehensive Medical Science, Fujita Health University, Toyoake, Aichi 470–11, Japan. nharada@fujita-hu.ac.jp.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Aromatase is known to participate in the regulation of reproductive functions in gonadal tissues. Aromatase has been further found in human adult livers (1) and hepatocellular carcinomas (2), as well as other extragonadal tissues or tumors. The estrogen receptor also has been shown to be present not only in livers of experimental animals (3) and human (4), but also in hepatic adenomas (5) and hepatocellular carcinomas (6, 7). It can therefore be postulated that estrogens participate in neoplastic transformation as paracrine or autocrine promoters of hepatocarcinogenesis, as suggested in experimental animals (8, 9). In fact, liver lesions may be dependent on estrogen for proliferation (10), and there have been reports of complete regression of hepatocellular adenoma after withdrawal of oral contraceptives (11, 12) and antiestrogenic therapy for hepatomas (6). However, the role of estrogens in the development of hepatic tumors is still controversial.
Recently, aromatase was immunocytochemically observed in human breast tumor epithelial cells (13, 14) or stromal cells adjacent to tumors (15, 16). This observation was further supported by in situ hybridization and quantitative analyses of aromatase messenger RNA (mRNA) in the breast tumors (14, 17, 18), suggesting that hormone-dependent breast tumors may proliferate in response to autocrine or paracrine stimuli from estrogen locally produced by aromatase. In this study, we showed localized aberrant expression of aromatase in regions adjacent to malignant liver tumors. RT-PCR analysis of aromatase mRNA further indicated that aberrant increase of aromatase in regions closely associated with tumors is caused by locally increased aromatase mRNA.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The tissues were obtained from 14 patients with primary or
metastatic malignant liver tumors, undergoing surgery at Shinshu
University Hospital. Histological classification was based on the World
Health Organization Histological Typing. Clinical details of the 14
patients are given in Table 1
. The liver
tissues obtained by surgical excision were separated into tumor and
surrounding liver tissues. The latter were further divided into regions
distal and proximal to the tumors. The tissue specimens were frozen
immediately and stored at -80 C until used to prepare microsomal
fractions and total RNA.
|
The ICC staining of aromatase in the sections was performed with a standard streptavidin-biotin complex procedure using a polyclonal antibody raised in a rabbit against human placental aromatase (19). This antibody has been confirmed to be monospecific (19) and has been validated for detection of aromatase by ICC staining in several laboratories (15, 16). The specificity of the ICC reaction was further controlled by omission of primary or secondary antibodies and use of primary antibody preabsorbed with purified aromatase. After immunoreaction with biotinylated goat anti-rabbit Ig and then with peroxidase-conjugated streptavidin complex, ICC staining was performed using 0.05% 3,3-diaminobenzidine and 0.01% hydrogen peroxide for 5–10 min. The sections were counterstained with hematoxylin-eosin.
Assay of aromatase and NADPH-cytochrome P450 reductase activities and Western blotting analysis
Microsomal fractions were prepared by successive centrifugations (19). Aromatase activity was determined by a radiometric method using [19-14C] androst-4-ene-3,17-dione (DuPont/NEN, Boston, MA) as a substrate (20). Microsomal NADPH-cytochrome P450 reductase activity was assayed as described previously (21). Western blotting analysis was performed using aromatase antibody (19) and horseradish peroxidase-conjugated anti-rabbit Ig antibody (Bio-Rad, Richmond, CA), as previously described (22).
RT-PCR analysis
Total RNA fractions were prepared from frozen tissues according to the method of Chirgwin et al. (23). Aromatase mRNA in the total RNA fraction was fluorometrically determined by RT-PCR using a fluorescent dye, FAM (Perkin Elmer Co., Foster City, CA)-labeled primer in the presence of an internal standard RNA, as previously described (18, 24, 25). Oligonucleotides of an antisense primer (5'-AACCACGATAGCAC TTTCGT-3') for reverse-transcription of aromatase mRNA, and antisense (5'-TGTTAGAGGTGTCCAGCATG-3') and sense (5'-TACTACAACCGGG TATATGG-3') primers for PCR of aromatase cDNA were synthesized (26). Total RNA (5 µg) and an internal standard RNA (0.05 attomoles) were subjected to reverse-transcription at 42 C for 40 min and PCR amplification using a FAM-labeled primer for 26 cycles. The fluorescent PCR products were analyzed with a Gene Scanner 362 (Perkin-Elmer Corp.). The amount of aromatase mRNA was calculated from peak areas of fluorescent PCR products by an internal standard method. The quantitativeness of this RT-PCR analysis was previously confirmed by proportionality between amounts of aromatase mRNA and fluorescent peak areas over a wide range from 0.001–10 attomoles (27). To check quantity and integrity of RNA samples, ß-actin mRNA, as a control, was also analyzed by RT-PCR as described previously (25).
Statistical analysis
Statistical analysis was performed using one-way ANOVA, followed by the Scheffe’s test. A P value < 0.05 was considered to be significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The aromatase immunolocalization in the metastatic colon carcinoma
of the liver was shown in Fig. 1, A and
B. Hematoxylin-eosin staining indicated colon carcinoma cells
proliferating invasively and forming anaplastic glandular structures.
ICC staining of aromatase in serial sections revealed relatively strong
immunoreactivity in nonneoplastic hepatocytes adjacent to metastatic
tumor tissue, as shown in Fig. 1, C and D. The hepatocytes around the
tumor showed strong cytoplasmic immunoreactivity for aromatase.
Relatively weak immunoreactivity of aromatase also was observed in
inflammatory cells. However, aromatase immunoreactivity was hardly
detected in either tumor cells or distal hepatocytes. Similar aromatase
immunoreactivity also was observed around other types of malignant
liver tumors (data not shown).
|
To confirm localized aberrant expression of immunoreactive
aromatase in malignant liver tumors, a hepatocellular carcinoma
(patient 1), a cholangiocellular carcinoma (patient 3), and metastatic
hepatic tumor tissues from a sigmoid colon carcinoma (patient 9) and a
rectal carcinoma (patient 10) were separated into tumor, proximal, and
distal regions; and Western blotting analysis of aromatase in their
microsomal fractions was carried out (Fig. 2). In all cases, each fraction showed a
single major band of an immunoreactive protein (51,000 Da)
corresponding to human aromatase. The intensity of immunochemical
staining was obviously weaker in the tumor regions than in the
surrounding regions. Furthermore, immunoreactive bands of aromatase in
the tissues adjacent to tumors were significantly stronger than those
in the distal regions. These results were essentially consistent with
the ICC observation.
|
The data for activities of aromatase and NADPH-cytochrome P450
reductase, an electron donor for aromatase, in liver tumors are
summarized in Table 2. Activities of both
enzymes were obviously lower in the tumor regions than in the adjacent
tissues (P < 0.0001). There was no statistically
significant difference between proximal and distal regions, regarding
aromatase activity, but the proximal regions of all patients (except
for patient 12) demonstrated higher values than the distal regions,
whereas NADPH-cytochrome P450 reductase activities were significantly
greater in the liver tissues distal to tumors (P =
0.0035).
|
Quantitative analysis of aromatase mRNA was performed by RT-PCR.
Representative results are shown in Fig. 3. The almost equal peak heights and
areas of PCR products of the internal standard RNA indicate that the
efficiency of RT-PCR was essentially constant. High levels of aromatase
mRNA were observed in all tissues adjacent to tumors. The levels of
aromatase mRNA in the tumors were calculated to be only 15–75% of
those in the proximal regions. In the distal regions, only trace levels
of aromatase mRNA were observed. Contents of ß-actin mRNA in the
total RNAs, simultaneously quantitated as controls, demonstrated no
significant differences among the tissue regions.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
to be
possible local factors exerting regulatory effects on aromatase in
adipose stromal cells (28). Breast tumor extracts and cyst fluids also
were reported to influence aromatase in adipose tissue (29). Such local
factors, synthesized by liver primary or metastatic tumors, may be
secreted into the external local environment, inducing expression of
aromatase in the adjacent liver tissues. Alternatively, such factors
may be synthesized in some other cell elements, in response to the
presence of malignancy, and act in an autocrine fashion. Changed
physiological conditions in the local environment adjacent to tumors
may induce aromatase-inducing factors in the surrounding tissue. In
fact, expression of aromatase in cultured adipose stromal cells is
markedly increased by removal of serum from the culture medium (18, 30), suggesting involvement of some inhibitory substances. It would be
of interest to examine the distribution of other hepatic cytochrome
P450 species and drug (steroid)-metabolizing enzymes and compare them
with that of aromatase. Recently, hepatic P450 1A, P450 3A, epoxide
hydrolase, and glutathione S-transferase were shown to be present in
hepatocellular carcinomas, as well as normal hepatocytes, indicating
that expression of these enzymes was regulated differently from that of
aromatase. These enzymes may partially contribute to inactivation of
estrogens locally produced by aromatase.
Recently, we determined the expression of aromatase mRNA in the livers
of disease-free individuals (31). The levels were essentially
comparable with those in the regions distal to liver tumors in the
present study. It has been shown that human aromatase is highly
expressed in fetal livers, compared with adult livers (31, 32).
Aberrant expression in tumors also has been observed for
-fetoprotein, a fetal liver-specific protein, and often found in
livers carrying tumors as a carcinoembryonic protein. This suggests
that aberrant expression of aromatase in liver tumors may be caused by
a similar mechanism to that responsible for -fetoprotein production
during carcinogenesis.
The ICC localization of aromatase in the livers adjacent to tumors was in line with the results of Western blotting and activity analyses, consistent with the quantitative data for aromatase mRNA. However, there were several inconsistent points. Aromatase was immunocytochemically limited to the boundary region of liver tissue adjacent to tumors and was hardly detectable in either tumors or the distal liver tissues, whereas the other analytical methods revealed significant levels of expression by these populations. Several possible explanations can be formulated. First, differences in the detection thresholds among these analytical methods could be responsible. Second, differences in analytical targets may have played a role. ICC localization is based on in situ observation, whereas the others are based on biochemical or molecular analysis of tissues. It is impossible to preclude the possibility of mutual contamination in tissue preparations. Third, the ratios of apo-aromatase to holo-aromatase or the degradation rates of aromatase protein and mRNA may be different from each other in the tissue preparations. Finally, the variable ratios of aromatase to NADPH-cytochrome P450 reductase observed may explain a part of the inconsistency. Aromatase activity depends not only on its own content, but also on that of NADPH-cytochrome P450 reductase. The finding of the lowest reductase activities in tumors and the highest activities in the distal regions might therefore have complicated the results.
In this study, localized aberrant expression of aromatase in the livers adjacent to tumors was not limited to primary hepatocellular and cholangiocellular carcinomas but also was found for metastatic hepatic tumors from gastric, colonic, and rectal carcinomas. In preliminary experiments, we observed estrogen receptors in primary and metastatic liver tumors, as well as hepatocytes, in accordance with previous reports (4, 6, 7), although we could not find any significant correlation with expression levels of aromatase. Estrogen receptors in tumors were very variable and, in many cases, lower than those in nontumorous livers. Their presence in tumors may support possible participation of locally produced estrogens as paracrine or autocrine mitogenic promoters, although clear evidence of participation of estrogens in proliferation of metastatic liver tumors is lacking. To block the possible proliferative effects of estrogens on tumors, application of antiestrogenic or aromatase inhibitor therapy has been described for hepatocellular (6) and pancreatic cancers (33, 34), as well as breast cancer. Further investigations are needed to clarify the influence of hyperestrogenic stimuli during carcinogenesis.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received December 13, 1996.
Revised July 8, 1997.
Revised October 24, 1997.
Accepted November 3, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  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]
|
|
|
|
 |
|
 H. Sumitani, M. Shozu, T. Segawa, K. Murakami, H.-J. Yang, K. Shimada, and M. Inoue In Situ Estrogen Synthesized by Aromatase P450 in Uterine Leiomyoma Cells Promotes Cell Growth Probably Via an Autocrine/Intracrine Mechanism Endocrinology, October 1, 2000; 141(10): 3852 - 3861. [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 |
