| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Departments of Medicine (M.C.-T., K.Y.R., J.C.), Pathology, Immunology and Laboratory Medicine (G.Y.L.), and Pharmacology and Therapeutics (K.T.S.), College of Medicine, University of Florida, Gainesville, Florida 32610
Address all correspondence and requests for reprints to: Martha Campbell-Thompson, Box 100214 JHMHC, Department of Medicine, College of Medicine, University of Florida, Gainesville, Florida 32610. E-mail: thompmc{at}medicine.ufl.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Environmental factors are important etiological factors in human gastric carcinogenesis with N-nitrosamino compounds the most commonly investigated agents. These compounds are found in cured meats, some drinking water, and certain vegetables. They can be converted by certain bacteria such as Helicobacter pylori into nitrosamines that have been found to cause gastric cancer in animals. Chronic administration of N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) in drinking water to rats produces adenocarcinomas in the distal glandular stomach (antrum) and upper small intestines (duodenum) (11, 12, 13, 14, 15). The MNNG model has been used extensively to study mechanisms of human chemical gastric carcinogenesis because the antral adenocarcinomas share gender-related incidences, histological features, and distribution patterns with human upper gastrointestinal (GI) adenocarcinomas (11, 16). MNNG acts directly on upper GI epithelial cells, causing methylation of DNA and proteins. It is rapidly converted to N-methyl-N'-nitroguanidine that is not carcinogenic. Adenocarcinomas develop at the gastric lesser curvature wherein higher cell proliferative rates and thiol concentrations may increase local susceptibility (17, 18).
Large differences have been observed in the sensitivity of different rat strains to MNNG-induced tumors (reviewed in Ref. 19). These differences were observed following identical treatment with MNNG in the drinking water. Male Wistar and ACI rats were very sensitive to MNNG, whereas male Sprague Dawley rats had intermediate sensitivity and Buffalo rats were resistant. Several parameters have been studied to account for genetic differences in sensitivity to MNNG that include MNNG metabolism by N-denitrosation, tissue glutathione levels, O-methylguanine demethylation (DNA repair), and cell proliferation rates (19, 20). Of all these factors, cell proliferation rate (number and range of distribution) was the only factor that predicted strain sensitivity (20). Thus, the proliferative response of the gastric mucosa was proposed to be a key factor in the difference of susceptibility to gastric carcinogens between male rats of various strains.
Within these strains, gender differences in tumor incidence have also been determined (21). Rats were administered MNNG (83 µg/ml) in their drinking water for 8 months and were killed at 18 months. The incidence of gastric adenocarcinoma in ACI rats was 80% in males and 47% in females. In contrast, the incidence was 18% in males and 0% in females in Buffalo rats. The F1 progeny of sensitive (ACI) and resistant (Buffalo) stains had a dominant resistant phenotype in males and sensitive phenotype in females (17% male vs. 8% female). These incidences indicate complex interactions between MNNG susceptibility and gender. Sprague Dawley rats are reported to have intermediate sensitivity (19, 22), but gender incidences have been not reported. In another study, administration of MNNG for 4 months with tumor incidences determined at 12 months resulted in gastric carcinomas in 80% of Wistar male rats compared with an incidence of < 6% of ovariectomized and intact female rats (8). Castration decreased the incidence to 30%. Another report from the same laboratory showed a slight reduction in tumor incidences in male rats receiving estradiol treatment (88% vs. 68% (23). Estradiol was administered as twice weekly injections (400 µg/kg) for 4 months starting 2 months after beginning MNNG treatment. This evidence suggests that female hormones suppress, and male hormones promote, gastric cancer risk. Although male rats are known to be more susceptible than females to MNNG-induced gastric tumors, the gender-specific mechanism has not been investigated.
Estrogens regulate cell growth and differentiation in numerous tissues
in both females and males and have critical roles in breast and uterine
carcinogenesis (24). Estrogens act on cells through intracellular
estrogen receptors (ER) that regulate target gene expression although
nongenomic effects have also been reported (25). Recently, two ER
subtypes, ER
and ERß, have been characterized with differences in
tissue distribution, ligand binding, transcriptional activity, and
dimerization specificity (26, 27, 28). Expression of both ER and ERß
have been reported in epithelial cells of the upper GI tract with a
predominance of ERß over ER expression (29, 30, 31). Additional
complexity would be expected in those cell types that express both
receptors as ER homodimer (ER/ER; ERß/ERß) or heterodimer
(ER/ERß) formation can occur (32).
Mechanisms for protective effects of estrogens in gastric carcinogenesis are not known and could involve direct effects on mucosal cells, indirect effects mediated through other agents, or both. As a first step in identifying the roles of estrogens in GI carcinogenesis, a rat model for human gastric cancer was used to examine the influence of systemic estrogen exposure on morphological progression of preneoplastic lesions and to define the cell populations at which estrogens may act. Systemic and local effects of 17ß-estradiol administration were examined in male rats that received sc implants of 17ß-estradiol (E2) or oil vehicle before and during the initiation phase of gastroduodenal carcinogenesis induced by MNNG.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Animals and experimental protocol
Young adult male Sprague Dawley rats (6 weeks old, weight 
150
g, n = 48; Harlan Sprague Dawley, Inc.,
Indianapolis, IN) were used. All experiments were approved by the
Institutional Animal Care and Use Committee using protocols according
to the NIH Guide for the Care and Use of Laboratory Animals and by the
Environmental Health and Safety office at the University of Florida.
Rats were randomly assigned in groups of 2–3/cage and housed on wood
chips (14-h light, 10-h dark cycle). Rats were acclimated to a powdered
diet (AIN 76A, Tekland, Madison, WI) containing casein protein rather
than soy protein, a major source of genistein and other phytochemicals
with potential estrogenic activity (33). Following 1 week on the AIN
diet, rats were anesthetized and SILASTIC brand silicon tubing
(Dow Corning, Midland, MI) containing either
17ß-estradiol (E2; 300 µg/ml in sesame oil;
Sigma) or oil vehicle were placed sc as previously
described (29). All implants were replaced after 2.5 weeks.
One week following the first implant placement, rats were given MNNG (CAS: 70–25-7)(100 µg/ml, (12)) or plain tap water. Stock solutions of MNNG (1 mg/ml in tap water) were prepared fresh weekly and stored at 4 C. The stock solution was diluted with tap water just before use and placed in a tin-foil covered bottle to prevent photolysis of MNNG. Food and water intakes were recorded and drinking solutions replenished three times a week. Weekly body weights were determined.
After 4 weeks of MNNG treatment and 2 h before they were killed, rats received an injection of the thymidine analog bromo-deoxyuridine (BrdU; 50 mg/kg ip) to label cells in the S phase of the cell cycle. Two hours later, rats were anesthetized with sodium pentobarbital (60 mg/kg) and the abdominal aorta was cannulated for serum collection and in vivo perfusion with 100 ml phosphate buffer (PBS, 10 mM KPO4, 150 mM NaCl, pH 7.4) followed by 200 ml 2% paraformaldehyde-lysine-periodate (PLP) (34). During perfusion, the stomach was distended with 2 ml PLP. The stomach and duodenum (proximal 5 cm) were removed and opened and ingesta rinsed off with PBS. Tissues were postfixed overnight in PLP at 4 C. Sections of fundus, antrum (midline on lesser curvature) and duodenum were embedded in paraffin for light microscopy. Sections from antrum were embedded in Lowicryl resin for immunogold labeling as previously described (35) or in plastic resin (TAAB 812, Merivac, Halifax, Nova Scotia). Blood samples were collected before perfusion, stored on ice for 30 min, and spun at 3000 x g for 20 min. Serum was collected and stored at -75 C.
Histopathology
Alterations in mucosal morphology were determined on sections
immersion-fixed in 10% formalin and stained with hematoxylin-eosin for
nuclear detail and periodic acid-Schiff (PAS) histochemistry for mucin
staining. Histological classification of lesions was made using
established criteria from human pathology for grade of dysplasia (0–3
scale, none to severe) (36). Sections were coded so that the observer
(GYL) was unaware of the treatment group.
Image analysis
Antral gland heights and duodenal villus height and crypt depth
were measured in regions in which full-length glands were oriented
perpendicular to the submucosa. Height of antral and duodenal glands
was measured using a Carl Zeiss Axiophot microscope fitted
with a 10x objective and 1.25 optivar setting. Images were captured
with a video camera (Hamamatsu C2400, Hamamatsu City, Japan) linked to
a personal computer with frame grabber and image analysis software
(Vidas 2.1, Carl Zeiss, Thornwood, NY). The image analysis
system was calibrated for each objective and optivar combination by
tracing a reference distance on a stage micrometer. For each rat, 5–8
antral glands were measured in proximal, middle and distal regions.
Duodenal villous and crypt heights were measured in four to eight
glands from proximal and distal regions in each rat. Total duodenal
gland height was calculated as the sum of villous and crypt height for
each gland.
Gastrin immunogold analysis
Gastrin protein levels within storage granules were quantified
by immunogold analysis. Ultrathin sections were placed on
formvar-coated nickel grids and incubated with 0.1 M
NH4Cl for 1 h. Grids were washed in Tris buffer (TBS;
10 mM Tris HCl, pH 7.5, 150 mM NaCl) followed
by blocking buffer (10 mM Tris HCl, pH 7.5, 500
mM NaCl, 1% BSA). Grids were incubated with rabbit
polyclonal antibodies to gastrin (1:500) in blocking buffer overnight
at 4 C followed by washing in TBS. Grids were incubated with goat
antirabbit antiserum conjugated with 15 nm gold particles (1:200,
Amersham Pharmacia Biotech) followed by washing with TBS.
Sections were counterstained with uranyl acetate and lead citrate
before examination using a Carl Zeiss transmission
electron microscope. Gastrin cells were randomly selected and cells
containing a nucleus were photographed. Storage granules were selected
from photographs (16,000 final magnification) and defined as
symmetrical gray densities 0.3–0.7 mm in diameter. Gold particles
within four to eight storage granules were counted and data expressed
as average number per storage granule/cell. This average was then used
to calculate a mean value for each rat (4–7 gastrin cells/rat).
Controls for immunogold staining included replacement of the primary
antisera with normal rabbit serum or PBS.
Immunohistochemistry for BrdU incorporation and gastrin cell
numbers
Cells in the S phase of the cell cycle and gastrin-producing
neuroendocrine cells were counted following immunohistochemistry.
Following rehydration to PBS, tissue sections were placed in citrate
buffer (10 mM, pH 6.0) and heated in a microwave at full
power (750 W) for 10 min with buffer replenishment every 2.5 min.
Sections were cooled for 20 min at room temperature and washed in PBS
for 5 min. Sections were incubated in 10% normal goat serum for 10 min
following by the primary antibody overnight at 4 C (mouse monoclonal
anti-BrdU peroxidase conjugate, Roche Molecular Biochemicals; rabbit polyclonal anti-gastrin). Sections were
washed in PBS and incubated in 3% H2O2/PBS for
10 min to inactivate endogenous peroxidase. Sections were incubated for
10 min with peroxidase conjugated antirabbit antibodies for gastrin
cell detection. Immunoreactivity was localized by incubation with
diaminobenzidine tetrahydrochloride substrate. Cell counting was
performed in regions with glands oriented perpendicular to the
submucosa. Numbers of positive BrdU nuclei and gastrin cells in the
antral glands were counted in five to eight rectangular reference
frames with the 40x objective and 1.6 optivar setting. Number of
positive BrdU nuclei in duodenal crypts were counted as above using the
20x objective and 2.0 optivar setting. The reference frame from the
imaging system was overlaid on the section with the shortest side
aligned at the muscularis mucosa to encompass four to five glands. The
viewer was unaware of treatment conditions.
Apoptosis rates in antral glands
Several methods are available for determining rates of apoptosis
including morphological assessment on stained sections and in
situ DNA end labeling methods (37, 38). Cell necrosis following
inadequate fixation has been reported to effect apoptosis rates
determined by the end labeling method in the rat intestinal tract (38).
The accuracy of apoptosis detection was enhanced by using semithin
plastic resin sections with toluidine blue staining rather than using
H&E sections (39). Cells undergoing apoptosis were estimated by
counting numbers of apoptotic cell bodies/500 cells. Apoptotic bodies
were identified as small, round cells with dense nuclear staining
surrounded by a distinct clear cytoplasm. Cells in the entire antral
gland unit were counted from superficial epithelium to the base of
glands in regions where glands were oriented perpendicular to the
submucosa.
Data analysis
Three independent experimental trials were performed with 2–6
rats/treatment group in each trial. Data are presented as mean ±
SEM (N = number of rats/treatment group) and were
analyzed by one-way ANOVA with posthoc t-tests for
significant differences (P < 0.05).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
show that MNNG
induced striking alterations in antral and duodenal gland morphology.
In the antrum, MNNG induced mucosal hyperplasia and surface epithelial
dysplasia (Fig. 1B). Nuclear alterations were observed in surface cells
following MNNG. Surface epithelial cells in the proximal glandular
region (gastric fundus) did not show hyperplasia or dysplasia (not
shown). Staining for mucin, determined by PAS histochemistry, was
reduced throughout the antral glands particularly in superficial and
neck mucous cells. Grade of antral dysplasia was significantly
increased compared with control (Table 1)
and dysplastic foci or adenocarcinoma were detected in two animals. In
contrast, MNNG induced duodenal villus shortening and widening and
mucosal hyperplasia and epithelial dysplasia (Fig. 1, D–F). Systemic
17ß-estradiol treatment modulated glandular alterations induced by
MNNG in both antrum and duodenum. Mucin staining was increased to
control levels throughout the antral glands following 17ß-estradiol
treatment (Fig. 1C). As well, antral dysplastic foci or adenocarcinomas
were not observed and grade of dysplasia was improved by
17ß-estradiol treatment (Table 1). Duodenal gland heights were
increased in 17ß-estradiol treated rats (Fig. 1F). These data show
that MNNG induced differential patterns of changes in antral and
duodenal glands and that these alterations could be modulated by
17ß-estradiol administration.
|
|
). These effects
represented an increase of 164% in antral gland height and a decrease
of 47% in duodenal villous height compared with control rats. Duodenal
crypt height was not changed by MNNG treatment (Fig. 2).
17ß-Estradiol administration significantly reversed MNNG-induced
alterations in duodenal total gland and villous heights. Increases in
antral gland heights induced by MNNG were slightly decreased by
17ß-estradiol cotreatment (P = 0.068) (Fig. 2). Thus,
17ß-estradiol treatment primarily had a protective effect on duodenal
gland height in this model.
|
). The effect of MNNG on intracellular
gastrin protein levels was quantified using immunogold analysis.
Immunoreactivity for gastrin was found in the small, electron dense
storage granules, over intracellular membranes and occasionally within
the nucleus (Fig. 4). 17ß-estradiol
treatment significantly increased gastrin levels in storage granules
compared with MNNG treatment alone (Fig. 4).
|
|
Cell proliferation and apoptosis rates
We next used BrdU immunohistochemistry to evaluate the effects of
MNNG on antral and duodenal crypt cells undergoing DNA synthesis in the
S phase of the cell cycle. In antral glands following MNNG
administration, the number of proliferating cells was increased
approximately 2.5-fold and their position within the gland shifted from
the base upward toward the superficial epithelium (Fig. 5). Proliferating cells were detected in
the superficial epithelium of an antral adenocarcinoma that developed
in an MNNG-treated rat (Fig. 5). Quantitation of the antral labeling
index showed that pretreatment with 17ß-estradiol did not decrease
antral cell proliferation induced by MNNG (Fig. 5). Similarly in the
duodenum, MNNG-induced increases in the labeling index (1.6-fold) were
not significantly altered by 17ß-estradiol cotreatment (Fig. 5).
|
). 17ß-estradiol cotreatment did
not alter this endpoint.
|
). Food intake tended
to be lower in the MNNG-treated groups particularly in the rats with
17ß-estradiol implants (Table 2). Testicular weights, when normalized
to body weight, were not altered by 17ß-estradiol treatment and serum
testosterone levels were not significantly different between groups
(Table 2). Serum estrogen levels measured 18 days following placement
of 17ß-estradiol implants were similar to those in normal cycling
female Sprague Dawley rats from our lab (12.3 ± 3 vs.
12.0 ± 3 pg/ml, respectively, n = 6/group). These serum
levels are similar to other published levels for physiological ranges
of estradiol fluctuations during normal estrus cycles (40). The control
and MNNG males had serum estrogen concentrations of 7.7 ± 2 and
6.2 ± 1 pg/ml, respectively (n = 5/group). These values are
similar to previously reported serum estrogen levels in male rats
(41).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The present study confirms earlier reports that preneoplastic proliferative lesions are observed as early as 4 weeks following MNNG treatment and that these early changes correlate with tumor development (14, 45). A greater susceptibility to the carcinogenic effects of MNNG has been reported for the upper small intestine (duodenum) compared with the glandular stomach in rats with preneoplastic alterations consisting of villus blunting and crypt elongation as shown in this study (15, 47). We present new data showing that the upper intestine is a particularly sensitive site for estrogenic protection. These data are compelling in light of the probable protective effect of estrogen replacement therapy on reducing large intestinal, i.e. colon, cancer risk in postmenopausal women (reviewed in Ref. 5). Altered mucus secretion is also a premalignant alteration observed following MNNG treatment (48, 49). This study shows that mucin levels were similar to control conditions in rats receiving MNNG and 17ß-estradiol treatments.
MNNG treatment resulted in antral cell hyperplasia with displacement of stem cells toward the superficial epithelium. These changes are considered obligatory preneoplastic lesions that increase the potential for direct stem cell exposure to MNNG (49). One potential mechanism of action of estrogens on mucosal cells could involve modulation of stem cells leading to reductions in cell proliferation. The present study shows a significant increase in number of cells undergoing DNA synthesis in antral glands and duodenal crypts following MNNG treatment. These alterations in labeling indices paralleled changes in antral gland and duodenal crypt heights. 17ß-estradiol administration did not alter proliferation rates under these conditions. A compensatory mechanism for excessive cell proliferation is an increased rate of programmed cell death or apoptosis (50). Apoptosis rates were increased following MNNG treatment but were not altered by 17ß-estradiol treatment. More sensitive detection methods, such as ultrastructural analysis by electron microscopy, may be required to detect significant differences between treatment groups, if they exist, as basal rates of apoptosis in the antrum are very low (37, 50). Apoptosis rates in human gastric carcinomas are closely associated with cell proliferation rates and apparently do not have biological significance as a prognostic factor (51).
Gastrin is a well-characterized hormonal regulator of proximal gastric and duodenal cell growth and mucosal barrier production (52, 53). Gastrin is synthesized as a prohormone, and following processing in the Golgi apparatus, is concentrated in small electron dense storage granules that contain several processing enzymes (54). This study reports for the first time that gastrin protein levels in storage granules are significantly increased following MNNG and 17ß-estradiol treatments. The differences in intragranular gastrin protein levels are not expected to yield significant alterations in serum gastrin levels as gastrin cell numbers and messenger RNA steady-state levels (data not shown) are not different between groups. The effects of sex steroid hormones on serum gastrin levels have been studied. In Sprague Dawley rats and guinea pigs, males have higher levels of serum gastrin levels than females and serum levels correlate with antral gastrin concentrations or gastrin cell numbers (55, 56, 57). However, 17ß-estradiol treatment in either castrated or ovariectomized rats does not significantly lower serum gastrin levels implying that estrogens alone may not have a role in gender differences (55, 56). In contrast, 17ß-estradiol administration in ovariectomized mice results in significant increases in serum gastrin levels (58).
The effects of systemic estrogens on differentiation of mucous and gastrin cells could result from improved mucosal barrier function as both cell types have apical membranes raising the potential for direct exposure to luminal MNNG. This proposal is supported by studies showing protective roles of estrogens in acute rat gastric injury models (59). Interestingly, estrogens were prescribed at one time for duodenal ulcers in humans (60). Mechanisms for improved mucosal barrier function could result from alterations in MNNG metabolism or luminal pH. Gender differences in glutathione levels and antioxidant enzymes are known for different regions of the rat GI tract (61). Glutathione levels are significantly higher in gastric and duodenal mucosa of female Sprague Dawley rats compared with males (61). However, higher glutathione levels would probably increase MNNG susceptibility (18, 19, 62). Glutathione peroxidase activity is also significantly higher in gastric mucosa of females compared with males (61). 17ß-estradiol treatment decreases gastric acid secretion in ovariectomized rats however MNNG is more efficiently absorbed at higher pH levels (63, 64). Taken together, these data suggest a mechanism distinct from known effects of estrogens in the upper GI tract for mediating protective effects during MNNG carcinogenesis.
Our data confirm decreased body weight gain following MNNG treatment (14) and further show that decreased body weight gain was associated with small reductions in food intake. Reduction of appetite by estrogens has been previously reported (65). As expected, rats treated with both 17ß-estradiol and MNNG had decreased weight gain and food intake, with testicular weights reduced in parallel with body weights. These changes occurred with serum testosterone levels remaining similar between treatment groups.
Despite the severity of lesions induced by MNNG, systemic estrogen
treatment in male rats resulted in significant protective effects.
Direct effects in mucosal cells, secondary effects mediated by other
agents, or both, could mediate protective effects. The potential for a
direct effect is suggested by studies showing that ER, particularly the
recently identified ERß subtype, are expressed in rat antral and
duodenal epithelium (29, 66). In another study using male rats,
expression of antral ER, determined by RIA, remained constant while
androgen and epidermal growth factor receptors declined throughout the
course of MNNG treatment (67). ER expression in GI epithelial
cells suggests that direct hormonal manipulation of cell function is
possible. Further studies in conjunction with in vitro cell
culture models and ER gene knockout animal models (68), particularly
those with deletions for ERß, are needed to further test this
proposal. Such studies would have relevance for understanding potential
mechanisms by which female hormones reduce GI cancer risk.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received February 23, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and estrogen receptor ß to partial estrogen agonists/antagonists. Mol
Pharmacol 54:105–112 and ERß at AP1 sites. Science 277:1508–1510 and
ß expression in rat upper gastrointestinal tract with regulation of
trefoil family factor 2 mRNA levels in ovariectomized rats. Biochem
Biophy Res Commun 240:247–518[CrossRef][Medline]
) and
estrogen receptor ß (ER ß) messenger RNA in the midgestational
human fetus. J Clin Endocrinol Metab 82:3509–3512This article has been cited by other articles:
 |
|
 |
|
  M. Ohtani, A. Garcia, A. B. Rogers, Z. Ge, N. S. Taylor, S. Xu, K. Watanabe, R. P. Marini, M. T. Whary, T. C. Wang, et al. Protective role of 17 -estradiol against the development of Helicobacter pylori-induced gastric cancer in INS-GAS mice Carcinogenesis, December 1, 2007; 28(12): 2597 - 2604. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Lindblad, W. Ye, C. Rubio, and J. Lagergren Estrogen and Risk of Gastric Cancer: A Protective Effect in a Nationwide Cohort Study of Patients with Prostate Cancer in Sweden Cancer Epidemiol. Biomarkers Prev., December 1, 2004; 13(12): 2203 - 2207. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Campbell-Thompson, I. J. Lynch, and B. Bhardwaj Expression of Estrogen Receptor (ER) Subtypes and ER{beta} Isoforms in Colon Cancer Cancer Res., January 1, 2001; 61(2): 632 - 640. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
