| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Vollum Institute (T.R.K., K.E.G., M.J.L.), and Division of Endocrinology (K.E.G.), Oregon Health Sciences University, Portland, Oregon 97201; and Department of Pathology, Mt. Sinai Hospital (S.L.A.), Toronto, Ontario, Canada M5G 1X5
Address all correspondence and requests for reprints to: Malcolm J. Low, M.D., Ph.D., Vollum Institute, L-474, Oregon Health Sciences University, 3181 S.W. Sam Jackson Park Road, Portland, Oregon 97201. E-mail: low{at}ohsu.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
- and
ß-subunits in vitro (5, 6, 7). These features all support the
hypothesis that null cell adenomas are derived from gonadotrophs.
However, in vitro these tumors respond to multiple
hypothalamic releasing factors, including CRH, TRH, and GH-releasing
hormone, in addition to GnRH (6). This finding indicates the expression
of multiple receptors and suggests that null cell adenomas may be
derived from a pluripotent null cell still capable of differentiation
toward one of multiple pituitary cell lineages. Further, some null cell
adenomas have been shown to express dopamine receptors (8) and respond
to bromocriptine (9), whereas others express vasoactive intestinal
peptide and vasopressin (10). Many of these features have also been
observed in human gonadotroph adenomas, leaving the cell of origin of
null cell adenomas in question. We have produced transgenic mice expressing a temperature-sensitive mutant of simian virus 40 (SV40) T antigen targeted to gonadotrophs by human (h) FSHß genomic sequences. These animals develop gonadotroph hyperplasia that progresses to adenomas with immunohistochemical and ultrastructural features that provide a unique model of human null cell adenomas.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
), an EcoRI-NheI
fragment of the hFSHß gene encoding 4.3 kb of 5'-flanking sequences
was ligated to a PCR-generated NheI-EcoRV
fragment containing exon 1, the first intron, and a new
EcoRV site that was added immediately upstream of the
initiator methionine codon in exon 2. Sequence identity of the PCR
fragment was verified by dideoxy chain termination sequencing. A 2.7-kb
StuI-BamHI fragment of the large T and small t
coding regions of an SV40 temperature-sensitive T antigen A58 (a gift
from Dr. Chou, NIH, NCI) (12) was ligated to the EcoRV site
by adding XbaI linkers (Promega, Madison, WI) to the
EcoRV and StuI ends. The final construct was
completed with a three-way ligation of the 5'-hFSHß sequences from
EcoRI to EcoRV/XbaI linker,
XbaI linker/StuI-BamHI SV40tsTA58
fragment, and a BamHI-EcoRI fragment containing
2.5 kb of hFSHß 3'-flanking sequences in a pBSIISK (Stratagene, La
Jolla, CA) vector. From this, a 9.5-kb EcoRI-SphI
linearized fragment containing the hFSHß and T antigen sequences was
isolated and purified using Geneclean II (BIO 100, Vista, CA) and
Elu-tip ion exchange chromatography columns (Schleicher and Schuell,
Keene, NH) for microinjection.
|
Histology/immunohistochemistry
For routine light microscopy analyses, pituitaries were fixed in
10% neutral buffered formalin and paraffin embedded, and 3- to 5-µm
sections were stained with hematoxylin and eosin (H&E) and with the
Gordon-Sweet silver method. Immunocytochemical stains to localize
adenohypophysial hormones were performed using the streptavidin
peroxidase technique on paraffin-embedded sections. Primary antisera
directed against pituitary hormones were obtained through the National
Hormone and Pituitary Program, NIDDK, NICHHD, USDA, and were used at
the following dilutions: rat (r) GH, 1:2500; rPRL, 1:2500; rTSHß,
1:3000; rFSHß, 1:600; and rLHß 1:2500. A prediluted polyclonal
antiserum against ACTH (Dako Corp., Carpinteria, CA) was used at an
additional 1:15 dilution. A polyclonal antiserum against SF-1 (Upstate
Biotechnology, Waltham, MA) was used at a dilution of 1:1000 after
pretreatment with microwave antigen retrieval in a citrate buffer.
For immunofluorescence experiments, age-matched control and transgenic mice were anesthetized and perfused transcardially with 4% paraformaldehyde in PBS, pH 7.2. Pituitary glands were postfixed and freeze-protected in the same fixative with 10% sucrose overnight at 4 C. Twenty-micrometer cryostat sections were used for colocalization of FSHß and LHß, TSHß, ACTH, PRL, GH, and SV40 T antigen. For colocalization of mouse (m) FSHß and LHß, a mouse monoclonal anti-hFSHß (Medix Biotech, Foster City, CA) that cross-reacts with mFSHß, but not LHß (14), was used at a dilution of 1:200. NIDDK rabbit polyclonal primary antisera were used in the following dilutions: rFSHß, 1:200; rLHß, 1:5000; rTSHß, 1:5000; ACTH IC-1, 1:1000, rPRL S9, 1:25, and hGH S2, 1:1000. A rabbit polyclonal SV40 T antigen antiserum was a gift from D. Hanahan (15), and a mouse monoclonal antibody was purchased from Oncogene Science (Cambridge, MA). Binding was detected with either goat antirabbit IgG-fluorescein or -rhodamine isothiocyanate (1:50; Tago, Burlingame, CA) or with goat antimouse IgG-rhodamine isothiocyanate (1:50; American-Qualex, LaMirada, CA).
For all immunohistochemistry, appropriate positive and negative controls were performed to demonstrate the sensitivity and specificity of each reaction.
Electron microscopy
For transmission electron microscopy, pituitary tissue was fixed
in 2.5% glutaraldehyde, postfixed in 1% osmium tetroxide, dehydrated
in graded ethanols and propylene oxide, and embedded in epoxy-resin.
Semithin sections were stained with toluidine blue; ultrathin sections
of selected areas were stained with uranyl acetate and lead citrate and
examined with a Philips 301 electron microscope (Philips, Mahway,
NJ).
Surgical manipulation
Eight- to 10-week-old male and female mice were castrated or
ovariectomized under anesthesia with Avertin (2% tribromoethanol, 15
ml/kg, ip; Aldrich, Rouses Point, NY) for evaluation of hormonal
responsiveness and were observed for up to 2 yr.
To evaluate transformation, cells from pituitary tumors were dispersed in 0.2% trypsin, treated with 0.2% deoxyribonuclease I with 10 mM MgCl2 and 0.05% soybean trypsin inhibitor (Life Technologies, Grand Island, NY), resuspended in sterile Hanks’ buffered salt solution (all except inhibitor from Sigma Chemical Co., St. Louis, MO), and injected sc into the flank of intact and castrate male nude mice.
RIA
Serum FSH levels and FSH concentrations in medium samples were
quantitated using NIDDK RP-2 kits by a double antibody method as
described previously (14). RIA grade FSH was iodinated by either
Iodogen (Sigma Chemical Co., St. Louis, MO) (16) or a chloramine-T
method (17), and FSH antisera were used at a dilution of 1:21,000. The
sensitivity of the assay was 4 ng/ml.
Cell culture
Pituitary glands were harvested from male transgenic animals and
minced with sterile instruments. They were enzymatically dispersed in
0.2% trypsin, treated with 0.2% deoxyribonuclease I with 10
mM MgCl2 and 0.05% soybean trypsin inhibitor,
and resuspended in sterile Hanks’ buffered salt solution. They were
plated at densities ranging from 100,000–500,000 cells/cm2
on either plain, poly-L-lysine-coated (Sigma),
extracellular matrix-coated (Oregon Regional Primate Center, Beaverton,
OR), or Primaria (Falcon, Franklin Lakes, NJ) plates in DMEM containing
15% FCS, 0.45% glucose, 1% penicillin-streptomycin, and 1%
nonessential amino acids or 10% horse serum, 2.5% FCS, 1%
L-glutamine, 1% nonessential amino acids, and 1%
penicillin-streptomycin and maintained at 33 C until temperature shift
to 40 C. Transfection of the full 10-kb hFSHß gene was performed with
calcium phosphate precipitation (18). GnRH pulses were obtained by
adding 10 nM GnRH (Sigma) to the medium for 1 h twice
daily. Cells were evaluated for FSH expression by RIA of medium
samples, immunocytochemistry as described above, and ribonuclease
protection assay using [32P]riboUTP-labeled riboprobes
for mFSHß and hFSHß and ß-actin (19). GnRH receptor numbers were
determined by binding to [125I]buserelin (20).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). All lines had normal embryonic viability, growth,
sexual development, and reproduction.
Tumor formation
Over 80 male and 35 female mice including the 16 founders and
progeny from the 4 established lines were examined at various ages for
pituitary pathology. In male transgenic mice, a consistent pattern of
pituitary enlargement starting at 6–8 months of age was observed,
demonstrating pituitary weights of up to 5–8 mg compared with normal
pituitary weight of 1–2 mg (Fig. 2).
After 9 months of age, male mice developed multifocal nodular tumors
weighing 30 mg or more. These tumors demonstrated a characteristic
shape predominant in the anterior portion with central bulging and
often had associated pars tuberalis cysts. With advanced age (24
months), male mice developed massive, displacing noninvasive tumors
weighing up to 300 mg. Gonadectomy had only minimal effect on tumor
development. After 6 months of age, male mice demonstrated pituitary
tumors with a penetrance greater than 90%. Female mice failed to
demonstrate significant gross pituitary pathology even after
gonadectomy; however, at advanced age (24 months), mild diffuse
enlargement was seen (average pituitary weight, 5 mg).
|
Transplantation of primary tumor cells sc into nude mice resulted in slow growing secondary tumors in both intact and castrate male mice with characteristics of adenohypophyseal cells, consistent with a transformed phenotype.
Histological and immunohistological findings in pituitaries
The pituitaries of male mice exhibited histological and
immunohistochemical evidence of gonadotroph hyperplasia progressing to
multifocal adenomatous nodules (Figs. 3-5). The earliest change was the
presence of individual gonadotrophs with marked nuclear enlargement and
hyperchromasia (Fig. 3A). At 6–9 months of age, immunofluorescence
showed diffuse FSHß and LHß immunoreactivity (Figs. 4, B and D, and 5A). After 8 months of
age, this progressed to frank adenomas with loss of the reticulin fiber
network (Fig. 3B). Although the smaller nodules contained
immunoreactivity for the gonadotropins, as the tumors progressed in
size and host age, hormone positivity diminished in intensity and
distribution (Fig. 3C); however, SF-1 nuclear reactivity persisted
(Fig. 3D). The tumors compressed the surrounding normal tissue, and in
the 2-yr-old males with massive tumors, only a thin rim of nontumorous
tissue was identified at the edge of the gland (Fig. 5C).
|
|
|
) and LHß staining;
however, scattered gonadotrophs had enlarged, pleomorphic nuclei (Fig. 3F). Even at 2 yr of age, female mice continued to have increased
numbers of scattered gonadotrophs with immunoreactive FSHß (Fig. 5D)
and LHß, but no evidence of tumor formation. T Antigen was not demonstrable by immunofluorescence or Western blot, but messenger RNA for T antigen was detected by in situ hybridization (not shown).
Electron microscopy
The pituitaries of young male transgenic mice contained scattered
cells with atypical large nuclei harboring clumped chromatin and scant
cytoplasm with few subcellular organelles (Fig. 6A); these corresponded to the atypical
gonadotrophs identified by light microscopy (cf. Fig. 3A).
In animals that had been castrated, gonadotrophs exhibited the usual
features of gonadectomy or castration cells: the cytoplasm was filled
with dilated rough endoplasmic reticulum containing flocculent contents
(Fig. 6B). The ultrastructure of tumor cells from older male transgenic
mice was that of poorly differentiated cells with irregularly shaped
nuclei and abundant cytoplasm but poorly developed subcellular
organelles resembling those seen in null cell adenomas of the human
pituitary (Fig. 6C). The presence of short profiles of rough
endoplasmic reticulum, Golgi complexes, and occasional secretory
granules indicated the potential for hormone synthesis and secretion by
these cells. Most male mice with tumors also had pars tuberalis cysts.
Histologically, these cysts had prominent vascular channels and were
lined with fibrous tissue with a few clusters of epithelial cells with
ultrastructural features characteristic of adenohypophysial cells with
marked dilation of the endoplasmic reticulum, numerous mitochondria,
and secretory granules (Fig. 6D).
|
). Normal male
mice had serum FSH levels averaging about 19 ng/ml; levels in
transgenic animals showed a large degree of variability, but in all
cases were much higher and ranged from 40 to over 750 ng/ml. Levels
tended to increase with age related to increasing tumor bulk;
consistent with the finding of decreasing immunoreactivity in tumors,
there was not a close correlation of serum FSH levels with tumor size,
and serum FSH levels increased only 5- to 10-fold despite tumor size
increasing up to 150-fold. Serum FSH levels did not increase
significantly after castration. Female mice had lower serum FSH levels
than males, but transgenic females had higher levels than wild-type
females both before and after ovariectomy.
|
In vitro characteristics of gonadotroph tumor cells
Twelve tumors removed from animals from five different lines were
enzymatically dissociated and grown in culture. Three of these tumors
yielded slowly dividing pituitary cells that survived crisis, were
passaged for many months, and survived freeze-thawing. Although FSH
immunoreactivity was observed initially in all cultures examined, FSH
expression and secretion evaluated by immunocytochemistry, RIA of
conditioned medium, and ribonuclease protection assay showed an early
and progressive decline to undetectable levels in all lines
established. One cell line expressed only low numbers of GnRH
receptors, and pulsatile GnRH stimulation was unable to restore FSH
expression. These cells did not express endogenous mFSHß or
transfected hFSHß subunit messenger RNAs. In addition, there were no
apparent effects in vitro on the morphology of cells or on
FSH expression with temperature shift from 33 to 40 C. In an attempt to
culture cells before potential dedifferentiating events, cultures were
made from fetal mouse pituitary cells, but no cell lines were
established using this method.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
It is interesting to speculate whether targeting of T antigen
expression to other pituitary cell types would result in similar
pathology, indicating a non-cell-specific dedifferentiation process.
Schechter et al. reported that pituitary glands from mice of
the -T7 line, which develop pituitary tumors induced by an
-subunit-driven T antigen, demonstrated cells of the
gonadotroph/thyrotroph lineage with very small, poorly abundant
granules with short irregular endoplasmic reticulum and small Golgi
(24). Ultrastructural features of LßT2 cells have not been described
in detail, but they are known to have secretory granules, little rough
endoplasmic reticulum, and Golgi apparatus; ultrastructural features of
the -T3 cell line, another glycoprotein hormone cell line, have not
been reported. However, the ultrastructural appearance of a melanotroph
intermediate cell line immortalized using the same
temperature-sensitive T antigen sequences is quite different, with
occasional intercellular junctions, although also with a few short
profiles of rough endoplasmic reticulum and small, atypical secretory
granules (25). These observations all suggest that the features we
observed are specific for gonadotrophs. In addition, analysis of other
neuroendocrine tumors generated by tissue-specific transgenic
expression of SV40 T antigen indicate persistent features of
differentiation. These include hormone expression and secretory
capacity with numerous secretory granules and well developed rough
endoplasmic reticulum in glucagon-SV40 T antigen-induced neuorendocrine
colonic tumors (26) and typical ß-cell type secretory granules with
insulin production, processing, and secretion in a ß-cell tumor line
developed from an insulin SV40 T antigen-induced insulinoma (27).
Moreover, the proglucagon SV40 T antigen-derived enteroendocrine cell
line, GLUTag, showed expression of proglucagon and cholecystokinin and
numerous small secretory granules, although there were only short
profiles of rough endoplasmic reticulum, and proglucagon processing was
abnormal (28).
As FSHß is the last hormone to be expressed ontogenically on
embryonic day 17.5 in the mouse (29), we chose to transform pituitary
gonadotrophs with FSHß promoter elements to produce cells with the
greatest potential for fully differentiated gonadotroph function.
Tissue-specific T antigen immortalization has been used in a number of
tissues to generate cell lines (21, 22) as well as to study mechanisms
of oncogenesis (30) and senescence (31). In particular, targeting of T
antigen expression to pituitary gonadotrophs using -subunit or LHß
promoter elements generated the -T3 and LßT2 cell lines,
respectively (21, 32). However, -T3 cells do not express LH or FSH
ß-subunits, and LßT2 cells do not express FSHß. This has been
ascribed to the late expression of FSHß in the ontogeny of the
pituitary, with the argument that because of immortalization at the
time of expression of T antigen [embryonic day 11.5 for -subunit or
embryonic day 16.5 for LHß (29), respectively], the later hormones
are never expressed, as opposed to a dedifferentiation event in the
course of tumor development (21).
FSH production and secretion and SF-1 expression, markers of
differentiated gonadotroph function, were retained in tumors
transformed by hFSHß-driven T antigen expression. This is in contrast
to what has been observed in tumors in which T antigen has been
targeted with -subunit or LHß promoter sequences and FSHß
expression was not observed, even though SF-1 immunoreactivity was
retained in the -T3 cell line (33). Cells from the gonadotroph
tumors have been immortalized in culture and maintained for more than 2
yr. However, they demonstrated FSH secretion in vitro for
only a short time, and when evaluated after an extended time in
culture, they did not express GnRH receptors or respond to pulses of
GnRH (19). This supports the theory that these cells do not
dedifferentiate at the time of T antigen expression, but rather slowly
lose differentiated function, as documented by decreasing hormone
immunoreactivity with increased tumor size.
Use of the A58 temperature-sensitive variant of T antigen results in a
protein that is stable at 33 C, partially active at 37 C, and inactive
at 40 C due to loss of interaction with DNA polymerase , which is
important for its transforming activity (34). In other cell systems,
expression of the temperature-sensitive T antigen results in
transformed cells at the permissive temperature, with loss of T antigen
action and recovery of differentiated function of the transformed
tissue as evidenced by morphological changes in fibroblasts and thyroid
epithelial cells and expression of tissue-specific proteins in thyroid
epithelial cells (30) at restrictive temperatures. However, temperature
shift in cells cultured from the gonadotroph tumors did not change cell
morphology or recover hormone production, but did result in decreased
cell viability. Further attempts to culture these tumors to develop a
gonadotropin-producing cell line are underway, exploring various
conditions for growth and hormone secretion.
Gonadotroph adenomas in humans have been reported to demonstrate a sexual dimorphism with regard to ultrastructural features. Specifically, a minority of gonadotroph adenomas in women demonstrate a characteristic honeycomb appearance of the Golgi apparatus (35). There is some debate as to whether these cells are actually of gonadotroph origin. As tumors were not observed in female transgenic animals, we were unable to evaluate this feature in T antigen-induced tumors. The absence of tumors in any female mice was striking, but unexplained. It is known that FSH gene expression is higher in male mice than in females (36). The reason for this is not clear, but it is postulated to be secondary to higher inhibin tone in female mice. Theoretically, FSHß promoter-driven T antigen expression would therefore be greater in male animals. However, tumors were not seen in female mice even after removal of negative feedback of gonadal steroid and peptides by ovariectomy.
One unusual feature of these T antigen-induced pituitary tumors was the finding of pars tuberalis cysts. The functional significance of these cysts is not clear, but is interesting given that gonadotrophs are the major constituent of the pars tuberalis (37). To our knowledge, cysts in this location have not been previously described in other T antigen-induced tumors, either with pituitary-targeted or ubiquitous T antigen expression. Neither are they found in spontaneous pituitary tumors in humans or rodents, although Rathke’s cysts are noted incidentally in a significant number of human pituitary glands at autopsy (38) and in transgenic mice expressing GH-driven leukemia inhibitory factor in the pituitary gland (39).
The tissue specificity of T antigen expression was tightly controlled by the hFSH regulatory elements. In over 120 mice killed for evaluation, only a few other tumors were incidentally noted in mice of advanced age, including two thymic tumors, a paravertebral mass, a calcific skull mass and a massive abdominal vascular tumor. Ubiquitous expression of T antigen results in aggressive choroid plexus tumors that result in death of the animals by 3 months of age, although thymic tumors are also seen (40). These other tumors are within the realm of what is normally observed in mice of advanced age and are probably incidental to the transgene expression.
Despite massive tumor growth, the pituitary tumors in our mice showed no evidence of either local invasion or metastasis. It is interesting that pituitary-specific expression of T antigen results in such a benign lesion. Use of the temperature-sensitive mutant of T antigen, which is fully active at 33 C, but only partially active at 37 C, may result in a weaker oncogenic stimulus. In humans and animal studies, the incidence of spontaneous pituitary carcinoma, however, is rare (41); this argues that a second oncogenic event is required for such a phenomenon. There has been one report of polyoma viral DNA sequences detected by PCR in normal and adenomatous human pituitary tissues (42) that speculated about the possibility of tumor virus DNA initiation of human pituitary adenomas. The finding of almost identical gross and ultrastructural pathology in our T antigen-induced tumors to that of human pituitary tumors is consistent with but certainly does not prove this hypothesis.
In conclusion, expression of a temperature-sensitive variant of T antigen in pituitary gonadotrophs by targeting with hFSH promoter and 3'-flanking sequences resulted in transformation of gonadotrophs and development of slow-growing gonadotroph adenomas in male transgenic mice. These tumors demonstrated gonadotropin production and secretion, but with growth exhibited reduced hormone expression. They had ultrastructural features remarkably similar to those of human null cell adenomas, supporting the hypothesis that null cell adenomas are derived from gonadotrophs. Although they have not yet yielded a gonadotropin-producing cell line, these tumors have significant potential as a source for immortalized gonadotrophs, given the evidence of their hormone production in vivo and short term production in vitro.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 These authors contributed equally to this work.
3 Current address: Department of Pathology, Baylor College of
Medicine, One Baylor Plaza, Houston, Texas 77030.
Received December 22, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
,6-diphenyl glycouril (Iodogen). Anal
Biochem 117:136–146[CrossRef][Medline]
-subunit gene in pituitary gonadotropes. Mol Endocrinol 8:878–885:
studies with various T-antigen mutants of SV40. Arch Virol 118:113–125[CrossRef][Medline]
This article has been cited by other articles:
 |
|
 |
|
  K.-H. Jeong, J. C. Gill, V. Nose, A. F. Parlow, R. S. Carroll, and U. B. Kaiser Expression of a Gonadotropin-Releasing Hormone Receptor-Simian Virus 40 T-Antigen Transgene Has Sex-Specific Effects on the Reproductive Axis Endocrinology, July 1, 2009; 150(7): 3383 - 3391. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Gejman, D. L. Batista, Y. Zhong, Y. Zhou, X. Zhang, B. Swearingen, C. A. Stratakis, E. T. Hedley-Whyte, and A. Klibanski Selective Loss of MEG3 Expression and Intergenic Differentially Methylated Region Hypermethylation in the MEG3/DLK1 Locus in Human Clinically Nonfunctioning Pituitary Adenomas J. Clin. Endocrinol. Metab., October 1, 2008; 93(10): 4119 - 4125. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. J. Cushman, D. E. Watkins-Chow, M. L. Brinkmeier, L. T. Raetzman, A. L. Radak, R. V. Lloyd, and S. A. Camper Persistent Prop1 expression delays gonadotrope differentiation and enhances pituitary tumor susceptibility Hum. Mol. Genet., May 1, 2001; 10(11): 1141 - 1153. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Zhai and L. Comai A Kinase Activity Associated with Simian Virus 40 Large T Antigen Phosphorylates Upstream Binding Factor (UBF) and Promotes Formation of a Stable Initiation Complex between UBF and SL1 Mol. Cell. Biol., April 1, 1999; 19(4): 2791 - 2802. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. L. Asa and S. Ezzat The Cytogenesis and Pathogenesis of Pituitary Adenomas Endocr. Rev., December 1, 1998; 19(6): 798 - 827. [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 |
