| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
-Subunit in High Grade Prostate Cancer1
Institute of Reproduction and Development (S.L.M., M.G.R., G.P.R.) and Prince Henry’s Institute of Medical Research (D.M.R.), Monash Medical Center, Clayton, Victoria 3168; and Melbourne Pathology (J.S.P.), Collingwood, Victoria 3066, Australia
Address all correspondence and requests for reprints to: Dr. G. P. Risbridger, Institute of Reproduction and Development, Monash University, Level 3, Block E, Monash Medical Center, 246 Clayton Road, Clayton, Victoria 3168, Australia. E-mail: gail.risbridger{at}med.monash.edu.au
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
gene in male and female mice
results in the development of primary gonadal granulosa/Sertoli cell
tumors. The aim of this study was to determine whether inhibin
-subunit gene and protein expression are altered in prostate cancer.
Messenger ribonucleic acid expression was studied by in
situ hybridization, and protein localization was studied by
immunohistochemistry. Inhibin -subunit messenger ribonucleic acid
expression and protein localization were observed in the epithelium of
tissues from men with benign prostatic hyperplasia, in regions of basal
cell hyperplasia, and in nonmalignant regions of tissue from men with
high grade prostate cancer. In the malignant regions of tissue from men
with high grade prostate cancer, the expression of the inhibin
-subunit gene was suppressed and was not detectable in poorly
differentiated tumor cells. These results demonstrate that in contrast
to ovarian granulosa cell tumors, inhibin gene expression is
down-regulated in poorly differentiated prostate cancer. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-subunit (inhibin
-subunit) linked by disulfide bonds to form inhibin A and inhibin B,
respectively. These proteins were first identified by their ability to
regulate the secretion of pituitary FSH (2). Inhibins are the negative
feedback regulators of FSH secretion and are produced by Sertoli cells
of the testis (3) and granulosa cells of the ovary (4). In contrast to
inhibins, activins stimulate pituitary FSH secretion (5).
Both - and ß-subunits have been found as free forms;
however, they do not have the ability to suppress FSH secretion
unless dimerization with the corresponding or ß occurs (6). The
inhibin -subunit is synthesized in precursor forms consisting of
pre, pro, N, and C components. The
precursor - and ß-subunits link to form a 105-kDa inhibin, which
forms bioactive 31- to 34-kDa inhibin C-ß after
posttranslational modification (cleavage of the pre, pro, and
N regions from the -subunit and the pro region from
the ß-subunit) (7). The N region has no inhibin-like
immunoreactivity or bioactivity, and its functional role remains
unclear, but it has been suggested that N may have a
role in promoting ovulation, because immunization of ewes against
N peptide decreased the ovulation rate and subsequent
fertility (8).
The tumor suppressor activity of the -subunit of inhibin
( inhibin) was first identified after functional deletion of the
inhibin gene in male and female mice that resulted in the development
of primary gonadal sex cord-stromal tumors, i.e.
granulosa/Sertoli cell tumors (9). After gonadectomy, these mice
developed adrenal cortical tumors, and therefore, it was suggested that
inhibin may be a tumor-suppressor gene for the gonads and adrenal
glands (10). If the -subunit gene is overexpressed in male mice
expressing the inhibin -subunit promoter-simian virus 40 Tag fusion
gene, then testicular tumors arise from Leydig cells, rather than
Sertoli cells (11). In contrast, postmenopausal women with granulosa
and mucinous epithelial cell tumors of the ovary displayed elevated
levels of serum inhibin (12), and the inhibin -subunit is considered
valuable as a marker for ovarian cancer (13, 14). These latter studies
suggest that dimeric inhibin and -subunit gene expression is
increased in ovarian cancers. The question remains as to whether the
inhibin gene is tumor suppressive in other human tissues.
Using human prostate cancer cell lines, ßA and ßB
(but not ) inhibin subunit messenger ribonucleic acid (mRNA) were
detected by RT-PCR, and it was postulated that the absence of inhibin
gene expression was important in the development of prostate cancer
(15, 16, 17). In the human prostate, we have previously reported the
expression and localization of activin (ßA and ßB) mRNA and protein
in tissue specimens from patients with benign prostatic hyperplasia
(BPH) (18). Inhibin -subunit gene expression was detected by RT-PCR,
indicating that the human prostate has the potential to synthesize
inhibin A or B.
The aim of this investigation was to determine whether
there was a change in inhibin gene expression in the
development of prostate cancer. The expression and localization of
inhibin in tissue sections from men with BPH were compared to those
from men with high grade carcinoma of the prostate. The latter tissues
provided a further comparison between malignant tumor tissue,
i.e. Gleason grade 4 and 5, and the adjacent nonmalignant
tissue regions of the biopsies. Inhibin -subunit gene expression was
performed by in situ hybridization, and localization of the
subunit proteins was examined with specific antibodies raised to the
carboxyl- and amino-terminal (C and N)
regions of the -subunit. The results demonstrate the down-regulation
of inhibin gene expression in tissues from men with high grade
prostate cancer.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Prostate tissues were obtained from a total of 29 patients, which were grouped according to diagnosis into 3 groups, i.e. those with BPH, basal cell hyperplasia, or prostate cancer. Transrectal ultrasound needle biopsies were obtained from 15 patients with BPH, 2 patients with basal cell hyperplasia, and 12 patients with prostate cancer, each having a Gleason score of 7–10, containing regions of histological Gleason grade 4 or 5 cancer. None of the patients had received androgen therapy. Two patients with basal cell hyperplasia were identified by histological diagnosis. The tissues were fixed in 10% buffered formalin and processed in paraffin. Three-micron sections were cut from the specimen and used for immunohistochemistry and/or in situ hybridization as described below.
Antibodies
Polyclonal antibody C41 was produced in
sheep by initial immunization with recombinant bovine inhibin
C fusion protein and boosted with human recombinant
inhibin A. The antisera was immunopurified by absorption to a column of
immobilized bovine C-subunit fusion protein before use.
This antibody has been used to measure inhibin and inhibin
-subunit levels in serum from normal and postmenopausal women using
an immunofluorometric assay (19). Polyclonal antibody N320 was
raised in sheep by immunization with a fusion protein consisting of a
peptide (amino acid 1–26) of the N region of bovine
inhibin -subunit (20). The antiserum was immunopurified by
absorption to a column of immobilized bovine N fusion
protein. Immunostaining for cytokeratin was performed using the
monoclonal antibody NCL-LP34 obtained from Novocastra Laboratories
(Newcastle upon Tyne, UK).
Immunohistochemistry
Sections were dewaxed, rehydrated, and placed in Target Retrieval Solution (Dako, Carpenteria, CA). Antigenic sites were exposed using an antigen retrieval process; slides were immersed in the solution, pH 6.1, and exposed to microwaves at 2.25 watts/ml·min for 2 min, followed by 0.3 watts/ml·min for 8 min (Akai MW-420 microwave, Akai Electric Co., Korea). After washing in 0.01 mol/L phosphate-buffered saline (PBS; 10 mmol/L phosphate buffer and 154 mmol/L NaCl, pH 7.4), endogenous peroxidase was blocked with 3% H2O2 for 30 min. Sections were incubated with 0.2% Triton X-100 (Sigma Chemical Co., St. Louis, MO) for 10 min and then blocked with a 1:1 mixture of CAS block (Zymed, San Francisco, CA) and 10% normal rabbit serum at room temperature for 20 min.
Inhibin was localized using the C41 polyclonal antibody
(IgG concentration, 1.6 µg/ml in PBS) and the N320 polyclonal
antibody (IgG concentration, 1.9 µg/ml in PBS). Basal cells were
localized using cytokeratin monoclonal antibody (1:100 in PBS). All
antibodies were incubated at 4 C overnight. Control sections were
incubated with sheep (inhibin) or mouse (cytokeratin) IgG at a matched
dilution or protein concentration. After overnight incubation, the
sections were washed in PBS and incubated with biotinylated rabbit
antisheep IgG (Vector Laboratories, Burlingame, CA; inhibin) or
biotinylated rabbit antimouse IgG (Dako; cytokeratin) for 60 min. The
sections were washed with PBS and incubated with the Vectastain Elite
ABC Kit (Vector Laboratories) for 60 min. After additional washes with
PBS, peroxidase activity was detected using a liquid
3,3'-diaminobenzidine tetrahydrochloride substrate kit (Zymed). The
reaction was terminated by immersion in distilled water, and the
sections counterstained with Mayers’ hemotoxylin (Sigma Diagnostics,
St. Louis, MO) and Scott’s tap water (Sigma), dehydrated, and
permanently mounted with DPX (BDH, Poole, UK).
In situ hybridization: probe synthesis
Digoxigenin (Dig)-labeled riboprobes
were prepared using the method outlined in the Boehringer Mannheim
riboprobe labeling kit (Indianapolis, IN). Rat and human
inhibin -subunit share 82% homology, and riboprobes to both rat and
human sequences were used in this study.
Dig antisense and sense complementary RNA probes (gift from
Dr. Moira O’Bryan, Institute of Reproduction and Development, Monash
University, Melbourne, Australia) were synthesized from a 
400-bp
partial rat -inhibin subunit complementary DNA (cDNA) cloned into
pGEM 4Z (21). Antisense probes were transcribed from
EcoRI-linearized plasmids with T7 RNA polymerase, and sense
complementary RNA was generated from HindIII-linearized
plasmids with SP6 RNA polymerase. The amount of Dig-labeled RNA was
determined by comparison to a Dig-labeled RNA control using dot blot
analysis.
An approximately 400-bp PstI/PvuII
fragment of the human inhibin -subunit cDNA (gift from Biotech
Australia, Roseville, Australia) was subcloned into pGEM 4z. The cDNA
corresponds to positions 702-1115 of the published human inhibin
-subunit nucleic acid sequence (22). Antisense probes were
synthesized by linearizing the plasmids with HindIII and
transcribed with SP6 RNA polymerase. Sense probes were obtained after
linearizing with EcoRI and were transcribed with T7.
After dewaxing, sections were washed in 1 x PBS (twice, 5 min each time) and treated with proteinase K (20 µg/ml) for 30 min at 37 C. After digestion, sections were washed in PBS containing 0.2% glycine for 5 min, followed by 5 min fixation in 4% paraformaldehyde. Sections were then washed in PBS (twice, 5 min each time), equilibrated for 2 min in 0.1 mol/L triethanolamine, and acetylated in 0.25% acetic anhydride in triethanolamine for 5 min. After rinsing in PBS, prehybridization was conducted at 42 C for 60 min in hybridization buffer, which contained 50% formamide, 10% dextran sulfate, 1 x Denhardt’s solution, 5 x SSC (1 x SSC = 0.15 mol/L NaCl and 0.015 mol/L sodium citrate), 45 mmol/L phosphate buffer, herring sperm DNA (200 µg/ml; Promega, Madison, WI), and transfer RNA (500 µg/ml; Sigma). Riboprobe was diluted in hybridization buffer to a concentration of 200-1000 ng/ml and denatured at 65 C for 10 min to remove secondary structures. Slides were then incubated at 80 C for 10 min, and hybridization was performed under coverslips in a humidified box at 42 C overnight.
After hybridization, coverslips were removed in 4 x SSC, and slides were washed twice (5 min each time) in 2 x SSC. An ribonuclease A digestion (20 µg/ml) was performed at 37 C for 30 min, followed by SSC washes of increasing stringency (twice, 5 min each time) in 1 x SSC and once (20 min) in 0.5 x SSC at 42 C. The tissues were briefly rinsed in 0.1 mol/L maleic acid-0.15 mol/L NaCl, pH 7.5, and nonspecific binding was removed with a blocking buffer containing 1% skim milk powder in 0.1 mol/L maleic acid-0.15 mol/L NaCl, pH 7.5, for 30 min at room temperature. Slides were then incubated in Cas-block (Zymed) for 20 min at room temperature. Antidigoxigenin alkaline phosphatase conjugate antibody (Boehringer) was diluted 1:1000 in blocking buffer, and sections were incubated overnight at 4 C. After washing (three times, 10 min each time) in 0.1 mol/L maleic acid-0.15 mol/L NaCl, immunoreactivity was detected with 4-nitro blue tetrazolium chloride (NBT)-5-bromo-4-chloro-3-indolyl-phosphate (BCIP) substrate (NBT/BCIP one step, Pierce Chemical Co., Rockford, IL). After appropriate color development (1–20 h), the reaction was halted by immersion in water.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In the glandular epithelial tissue from
patients with BPH, basal cells were localized using a specific
cytokeratin monoclonal antibody, as shown in Fig. 1A
. No cytokeratin immunoreactivity was
observed in the control sections (Fig. 1B). Immunoreactivity was
identified in epithelial cells in tissue sections from 15 patients with
BPH using both C41 and N320 antibodies. As shown in Fig. 1C, inhibin C-subunit was detected in basal cells (six of
eight patient tissues), with variable immunoreactivity in the secretory
cells (two of eight patient tissues). No immunoreactivity was present
in the control section (Fig. 1D). Positive immunoreactivity for the
N inhibin subunit was localized to both the secretory
and basal cells (six of six patient tissues; Fig. 1E). No
immunoreactivity was detected in the control section (Fig. 1F).
|
-subunit was localized to the epithelial basal cells (Fig. 1G
) in five patients. In one of five patients, -subunit mRNA was
localized to both basal and secretory epithelial cells (Fig. 1I). No
signal was detected using the sense control riboprobe (Fig. 1, H and
J). A summary of these results is presented in Table 1A.
|
Tissue sections obtained from two patients
with basal cell hyperplasia were used to detect inhibin -subunit
gene expression and protein localization. Identification of regions of
basal cell hyperplasia were confirmed using cytokeratin antibody as
shown in Fig. 2A. No cytokeratin
immunoreactivity was localized in the control section (Fig. 2B).
Inhibin C and N immunoreactivities were
also localized to these regions of the tissue sections and confirmed
inhibin protein localization to basal cells (Fig. 2, C and E). No
immunoreactivity was detected in the control sections (Fig. 2, D and
F). The presence of inhibin -subunit mRNA in basal cell hyperplasia
was confirmed in one patient using in situ hybridization
(Fig. 2G); no localization was detected using the corresponding sense
labeled riboprobe (Fig. 2H).
|
In 12 patients with poorly differentiated
prostate cancer, the localization of C inhibin protein
(11 of 12 patient tissues), N inhibin protein (6 of 12
patient tissues), and inhibin -subunit mRNA (8 of 12 patient
tissues) was determined and compared in malignant (histological Gleason
grade 4/5) and adjacent nonmalignant regions of the tissues. As
observed in tissue from patients with BPH, the C-subunit
protein was predominantly localized to the basal cells of nonmalignant
regions of tissue sections in 8 of 11 patient tissues (Fig. 3A) and to basal and secretory cells in 3
of 11 patient tissues. In the adjacent, poorly differentiated tumor
tissue, no positive immunoreactivity was observed (Fig. 3B). Similarly,
the pattern of staining of N inhibin was predominantly
localized to the basal and secretory epithelial cells in the
nonmalignant regions of tissue sections from 6 men with high grade
stage cancer of the prostate (Fig. 3C); no immunoreactivity was
observed in the adjacent tumor tissue (Fig. 3D). The control sections
for both the malignant and nonmalignant regions displayed no positive
staining (Fig. 3, E and F, respectively).
|
protein localization observed
above. Inhibin -subunit gene expression was detected in basal cells
of seven of eight patients in nonmalignant regions; however, in some
patients, signal was detected in both basal and secretory cells (Fig. 3G). Malignant tumor cells in adjacent regions of the same patient
biopsies did not display any -subunit gene expression (Fig. 3H). No
signal was observed with the inhibin -subunit sense riboprobe (Fig. 3, I and J). The results for the nonmalignant regions of the patient
tissues are summarized in Table 1C. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-subunit gene is not expressed and the protein is not detected in
malignant tissue from men with poorly differentiated prostate cancer.
In contrast, the expression and localization of inhibin mRNA and
protein are observed in tissues from adjacent nonmalignant regions of
the same patients or in men with BPH. Therefore, there is loss of the
expression and localization of inhibin -subunit in high grade
prostate cancer. These results contrast with the observation that
inhibin -subunit is elevated in postmenopausal women with granulosa
and mucinous epithelial cell tumors (12).
In our previous study of BPH tissue we were able to
detect inhibin gene expression by RT-PCR (18); the current study
extends those observations to demonstrate that the basal epithelial
cells are a predominant site of synthesis of the inhibin gene. Both
the N and C regions of the inhibin
protein were also localized to these cells within the prostatic
epithelium. Basal cell hyperplasia is a relatively rare and incidental
finding in the hyperplastic prostate (23), and in two patients with
basal cell hyperplasia the expression and localization of inhibin mRNA
and protein provide additional evidence that the basal cells are a
predominant site of inhibin gene expression in benign prostate
tissue. The proliferative compartment of the hyperplastic human
prostate is located in the basal cell layer, where about 70% of
proliferating cells are of the basal cell phenotype (24). As
proliferative and antiproliferative effects of inhibins have been
described in other cells and tissues (25, 26, 27, 28, 29), it is possible that
inhibin may have a role in the regulation of basal cell proliferation.
Bonkhoff and co-workers have proposed that there is a stem cell
population (located in the basal cell layer) that gives rise to all
epithelial cell lineages (neuroendocrine, secretory cells, and basal
cells) in benign and malignant prostate disease; hence, the role of
inhibin may be important in BPH and prostate cancer (30). Although
basal cells are the predominant site of localization of the inhibin
C-subunit protein, the N protein is
localized to both basal and secretory epithelial cells. It is possible
that the loss of inhibin expression is related to the loss of basal
cells during the progression to malignancy. However, the demonstration
that N is present in secretory epithelial cells in
nonmalignant tissue regions, but not in poorly differentiated tumor
cells would support the idea that inhibin -subunit expression is
turned off in the progression to malignancy.
The synthesis and production of the inhibin -subunit in
the prostatic epithelium correlate with our previous demonstration that
the inhibin ßA- and ßB-subunit proteins are localized to tissue
specimens from patients with BPH (18). If both inhibin - and
ß-subunits are expressed in the same cells, then these tissues have
the ability to produce dimeric inhibin proteins such as inhibin A and
inhibin B. Inhibin B is considered to be the physiologically important
form of inhibin in men in the regulation of pituitary FSH (31, 32). The
differential effects and production of inhibin A or inhibin B in
the prostate gland remain to be determined. The presence of novel
inhibin dimers is implicated by detection of the
ßC-subunit (33), which may give rise to a putative
ligand, inhibin C. It is noteworthy that the N protein
was also localized to the epithelium and may be a secretory product of
the prostate, but its role in this tissue remains to be
established.
In malignancy, this study reports that inhibin -subunit
gene expression is down-regulated, and no mRNA or protein is detectable
in the epithelial cells of poorly differentiated tumors. These findings
are similar to those in which mice bearing a functional deletion of the
inhibin gene develop gonadal and adrenal tumors. It is noted that
no inhibin is expressed in any cell or tissue from these mice,
whereas in men with high grade prostate cancer, it is presumed that
inhibin will be synthesized in other tissues and may be able to act on
malignant prostate tumor cells that lack inhibin -subunit
expression. However, inhibin gene expression is not present in the
human prostate tumor cell lines LNCaP, DU145, and PC3 (15, 16, 17), and we
recently reported that exogenously added inhibin A failed to antagonize
the effects of activin A on DU145 and LNCaP cells (34).
Alternatively, the failure to observe an effect of inhibin A
on the tumor cell lines may be similar to the resistance to TGFß that
is observed in cancer. Even though TGFß is up-regulated in
malignancy, many cancers are resistant to the effects of TGFß. In
human colon cancers, somatic mutations inactivate TGFß receptor type
II, and it is suggested that resistance to the action of this ligand is
a significant event in the progression to malignancy (35, 36, 37, 38, 39, 40). As for
TGFß, the tumor-suppressing action of inhibins may be mediated
through specific receptor complexes similar to those described
for TGFß and activin. However, although there is a body of indirect
evidence to suggest that there are inhibin receptors, no receptors for
inhibin have yet been identified (41, 42). Whether the regulation of
inhibin activity in tumor cells occurs during ligand synthesis or
through the regulation of receptor or postreceptor events remains to be
established. Finally, it is important to note that these observations
directly contrast with the previous demonstration that the inhibin
gene is up-regulated in human ovarian cancers (13, 14). The reason why
the present results contrast with those from patients with ovarian
cancer is unknown, but may relate to tissue- and cell-specific actions
of the inhibin -subunit gene.
Finally, the loss of inhibin -subunit expression in poorly
differentiated tumors contrasts with the continued expression of the
ß-subunits and the production of activins (43). The growth inhibitory
effects of activins contrast with the malignant characteristics
of the tumors, and another mechanism of conferring resistance to the
activins must be considered. Although inhibin A failed to block the
effect of activins on the tumor cells, follistatins (which are binding
proteins for activins), were able to completely antagonize the effect
of activins on the LNCaP and DU145 human prostate tumor cells (34). In
the absence of inhibin production, the interplay between activins and
follistatins appears to be more important in the regulation of growth
responses to activins by prostate tumor cells.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received August 7, 1997.
Revised November 13, 1997.
Accepted November 17, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
N) of the alpha 43
kDa subunit of inhibin (43) on fertility of ewes. Reprod Fertil Dev. 6:265–267.[CrossRef][Medline]
-subunit promoter directs SV40
T-antigen to Leydig cells in transgenic mice. Mol Cell Endocrinol. 119:135–146.[CrossRef][Medline]
N) of
the alpha 43 subunit of inhibin impairs fertility in sheep. Endocrinology. 124:3122–3124.[Abstract]
This article has been cited by other articles:
 |
|
 |
|
  M. P. Purdue, B. I. Graubard, S. J. Chanock, M. V. Rubertone, R. L. Erickson, and K. A. McGlynn Genetic Variation in the Inhibin Pathway and Risk of Testicular Germ Cell Tumors Cancer Res., April 15, 2008; 68(8): 3043 - 3048. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Florio, P. Ciarmela, F. M Reis, P. Toti, L. Galleri, R. Santopietro, E Tiso, P. Tosi, and F. Petraglia Inhibin {alpha}-subunit and the inhibin coreceptor betaglycan are downregulated in endometrial carcinoma Eur. J. Endocrinol., February 1, 2005; 152(2): 277 - 284. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. F. Schmitt, D. S. Millar, J. S. Pedersen, S. L. Clark, D. J. Venter, M. Frydenberg, P. L. Molloy, and G. P. Risbridger Hypermethylation of the Inhibin {alpha}-Subunit Gene in Prostate Carcinoma Mol. Endocrinol., February 1, 2002; 16(2): 213 - 220. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. P. Risbridger, J. F. Schmitt, and D. M. Robertson Activins and Inhibins in Endocrine and Other Tumors Endocr. Rev., December 1, 2001; 22(6): 836 - 858. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. L. Mellor, M. Cranfield, R. Ries, J. Pedersen, B. Cancilla, D. de Kretser, N. P. Groome, A. J. Mason, and G. P. Risbridger Localization of Activin {beta}A-, {beta}B-, and {beta}C-Subunits in Human Prostate and Evidence for Formation of New Activin Heterodimers of {beta}C-Subunit J. Clin. Endocrinol. Metab., December 1, 2000; 85(12): 4851 - 4858. [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 |
