| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Gene in Female Mice: Characterization of Ovarian Responses and Phenotype in the Adult1
Departments of Obstetrics and Gynecology and Cell Biology (D.W.S.), Duke University Medical Center, Durham, North Carolina 27710; the Receptor Biology Section (J.F.C., K.S.K.), Laboratory of Reproductive and Developmental Toxicology, National Institute of Environmental Health Sciences, National Institutes of Health, Research Triangle Park, North Carolina 27709-2233; Departments of Biochemistry and Child Health (D.B.L.), University of Missouri, Columbia, Missouri 65211; the Departments of Biochemistry, Molecular and Cell Biology (A.M., K.E.M.), Northwestern University, Evanston, Illinois 60208-0835; and the Chemical Industry Institute of Toxicology (M.S.), Research Triangle Park, North Carolina 27709-2137
Address all correspondence and requests for reprints to: Dr. Kenneth Korach, NIH-NIEHS, P.O. Box 12233/111 Alexander Drive, Research Triangle Park, North Carolina 27709-2233. E-mail: korach{at}niehs.nih.gov
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
gene (estrogen
receptor- knockout; ERKO) results in a highly novel ovarian
phenotype in the adult. The ERKO mouse model was used to characterize
ER-dependent processes in the ovary. Visualization of the ovaries of
10-, 20-, and 50-day-old wild-type (WT) and ERKO mice showed that the
ERKO phenotype developed between 20 and 50 days of age. Developmental
progression through the primordial, primary, and antral follicle stages
appeared normal, but functional maturation of preovulatory follicles
was arrested resulting in atresia or in anovulatory follicles, which in
many cases formed large, hemorrhagic cysts. Corpora lutea were absent,
which also indicates that the normal biochemical and mechanical
processes that accomplish ovulation were compromised.
Northern and ribonuclease protection analyses indicated that ERKO ovary
FSH receptor (FSHR) messenger RNA (mRNA) expression was approximately
4-fold greater than in WT controls. Ovarian LH receptor (LHR) mRNA
expression was also higher in the ERKO animals. Cellular localization
studies by in situ hybridization analysis of ERKO
ovaries showed a high level of LHR mRNA expression in the granulosa and
thecal layers of virtually all the antral follicles. Ribonuclease
protection analyses showed that ovarian progesterone receptor and
androgen receptor mRNA expression were similar in the two groups. These
results indicated that ER action was not a prerequisite for LHR mRNA
expression by thecal or granulosa cells or for ovarian expression of
progesterone receptor mRNA.
Ovarian estrogen receptor ß (ERß) was detected
immunohistochemically, was sharply compartmentalized to the granulosa
cells, and was expressed approximately equally in the ERKO animals and
the WT controls. In contrast, ER staining was present in the thecal
cells but not the granulosa cells of the WT animals.
The summary findings indicate that in the adult the major cause of the
ERKO phenotype is high circulating LH interacting with functional LHR
of the theca and granulosa cells. These features result in a failure of
the normal maturational events leading to successful ovulation and
luteinization and presumably involve both hypothalamic-pituitary and
intraovarian mechanisms dependent upon ER action. The presence of
ERß in the granulosa cells did not rescue the phenotype of the ovary.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The principal extra-ovarian sites of estrogen action critical to
ovarian function involve cells of the hypothalamus and anterior
pituitary that regulate gonadotropin secretion. ER-mediated regulation
at this level was demonstrated in adult mice in which ER gene
disruption (estrogen receptor- knockout; ERKOs) resulted in
increased steady-state FSHß and LHß mRNA expression and elevated
circulating LH levels (9, 10), indicating a disruption in the negative
feedback response. In adult ERKOs, preovulatory follicle development is
altered, and, rather than progressing to ovulation, culminates in many
cases in the formation of large, cystic, and hemorrhagic follicles
(11). Extended treatments with the estrogen antagonists such as ZM
189,154 and EM-800 (12, 13, 14) also produce changes in
hypothalamo-pituitary and ovarian function similar to those obtained
with ER gene disruption. However, their effects can vary
significantly with dose and do not specifically designate the actions
of the ER or of ERß, a new, recently cloned separate ER form (15).
Also, as pointed out above, it is difficult to know the extent to which
local follicular estrogen production modulates ovarian cellular
responses to these agents.
The ERKO mouse model previously described provides a more direct
experimental approach by which to characterize the intraovarian
processes dependent upon ER and, simultaneously, to indicate
mechanisms that might be dependent upon ERß. ERß mRNA is highly
expressed in the granulosa cells of the rat ovarian follicle
(16) and is normally expressed in the ovaries of ERKO mice (17). The
working hypotheses we tested with this model in this study were: 1)
that LHR expression in granulosa cells requires ER action, 2)
whether expression of progesterone receptor (PR) in the ovary requires
ER action, and 3) whether, in addition to ERß mRNA, ERß protein
is actually expressed in the ovary.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. The breeding, genotyping, and daily care of the mice
have been described (11, 18). All procedures involving animals were
approved under an NIEHS Animal Care and Use protocol and were performed
in accordance with USPHS guidelines. Five or six 15- to 16-week-old
animals of each genotype were injected sc with a single dose of 2.2 IU
PMSG (Sigma Chemical Co.) at 1300–1400 h followed by 3.2
IU human CG (hCG) (Sigma Chemical Co.) 48–52 h later.
This dose of gonadotropin is the routine protocol used at NIEHS.
Animals were killed 20–24 h after the hCG injection, and the ovaries
trimmed and processed histologically.
Histological preparation
Ovaries for routine histological analysis were placed into 10%
buffered formalin at 4 C for a period of 6–18 h, followed by
transfer to 70% ethanol. The tissue was then imbedded in paraffin,
sectioned, and stained with hematoxylin and eosin according to standard
histological procedures or used for immunohistochemical analysis as
described below.
Gonadotropin and steroid receptor mRNA analysis
Generation of riboprobes. All riboprobes used were generated
from linearized templates using Maxiscript reagents with the
appropriate RNA polymerase (Ambion, Inc., Austin, TX), and
the incorporation of [32P]-CTP (Amersham Life Sciences, Arlington Heights, IL). The antisense riboprobe for
the mouse PR mRNA corresponded to bp 2417–2781 of the mouse PR
complementary DNA (cDNA) (GenBank accession no. M68915). This was
generated from a subclone (bp 1672–2781) of the mouse PR cDNA in
pBluescript KS- (Stratagene Cloning Systems, La Jolla,
CA), kindly provided by Dr. Vicki Davis, that was first linearized with
the internal EcoRI site (at bp 2417) and then transcribed
with T3 RNA polymerase. The riboprobe for the mouse androgen
receptor (AR) mRNA was generated from a subclone of bp 2379–2817 of
the mouse AR cDNA (GenBank accession no. X53779) in pBluescript SK-
(Stratagene), kindly provided by Dr. Jonathan Lindzey. The
mAR antisense riboprobe was generated from a template linearized with
EcoRI and then transcribed with T3 RNA polymerase. To allow
for normalization of the amounts of total RNA loaded per lane on a
Northern blot, an antisense riboprobe for the mRNA of the mouse
ribosomal protein L-7 (bp 371–639; GenBank accession no. M29016) was
used. To allow for normalization between samples in the RNase
protection assay (RPA), an antisense riboprobe for the mouse
cyclophilin mRNA was generated from the template pTRI-Cyc
(Ambion, Inc.) and used.
The generation of labeled antisense riboprobes for the mouse FSHR and LHR first required cloning of the coding sequences into a suitable vector. This was carried out by RT-PCR amplification of partial cDNA sequences of the FSHR and LHR from wild-type (WT) mouse ovarian RNA. The following primers targeting the extracellular domains of each receptor were used: for FSHR, 525-bp fragment (bp 128–652 of rat cDNA; GenBank accession no. LO2842): forward 5'-CACTGGCTGTGTCATTGCTCT-3' to reverse 5'-CTGAGTTCCGTTGAATGCACA-3'; for LHR, 505-bp fragment (bp 256–760 of rat cDNA; GenBank accession no. M81310) forward 5'-TCTCTCAGAGTGATTCCCTG-3' to reverse 5'-AGCGTCTGAATGGACTCCAG-3'. Reverse transcriptase generation of cDNA was performed on 0.4 µg of poly A RNA using the GeneAmp RNA PCR kit (Perkin Elmer Corp., Norwalk, CT) with random hexamers according to the manufacturer’s protocol except that all reagents were scaled up to 50 µl per reaction. PCR was carried out using 3.0 µl of the reverse-transcriptase-generated cDNA per reaction with the appropriate primers at 100 pmol each, dinucleotide triphosphates at 0.2 mM each, 2.5 U AmpliTaq DNA polymerase (Perkin Elmer Corp.) and the corresponding optimal buffer (LHR, Buffer N; FSHR, Buffer A) (Invitrogen Corp., San Diego, CA) at 1x in a total volume of 50 µl. Thermal cycling was carried out at 95 C/30 sec; 58 C/1.0 min; 72 C/1.0 min for 35 cycles in a GeneAmp 9600 (Perkin Elmer Corp.). The amplified sequences were then cloned into the Srf 1 site of PCR-Script (SK(+) (Stratagene, La Jolla, CA) according to the manufacturer’s protocol. Antisense riboprobes were generated from linearized templates using the appropriate RNA polymerase, T7 for LHR and T3 for FSHR.
Northern blot analysis
Total RNA was extracted from pooled ovarian tissue (1.7 g) from
adult WT adult females and from pooled ovarian tissue (2.9 g) from
adult ERKO females using TRIZOL reagent (Gibco BRL,
Gaithersburg, MD) according to the manufacturer’s protocol. Final
yield was quantified by UV spectrophotometry, and the RNA was checked
for integrity on a 1% agarose gel. Duplicate 20 µg fractions of
total RNA from each genotype were electrophoresed on a 1.5%
agarose/1x MESA (MOPS-EDTA-sodium acetate) buffer/6.7%
formaldehyde gel system. The gel was blotted to Hybond N nylon
(Amersham Life Science, Arlington Heights, IL) by salt
capillary transfer according to the manufacturer’s instructions. The
resulting membrane was crosslinked using a UV cross-linker
(Stratagene Cloning Systems, La Jolla, CA). The
blot was then halved and each half probed for either the FSHR or LHR
mRNAs. Prehybridization consisted of 6–8 h in a Hybaid oven at 65 C in
50% formamide, 3 x SSC, 5 x Denhardt’s, 0.02
M NaPO4, 10% dextran solution, and 1% SDS.
For each antisense riboprobe, 3 x 106 cpm/ml was
added and allowed to hybridize overnight at 65 C. The membranes were
then washed 2 x 15 min in 2x SSC/0.1% SDS at room temperature
followed by 2 x 14 min in 0.1x SSC/0.1% SDS at room temperature
followed by 2 x 15 min in 0.1x SSC/0.1% SDS at 68 C. The
resulting bands were visualized and quantified with the Phosphorimager
425 and accompanying ImageQuant Software (Molecular Dynamics, Inc., Sunnyvale, CA) followed by exposure to x-ray film.
RNase protection assay
Assays were carried out on 2.5 µg total ovarian RNA from
individual adult WT and ERKO females using the Hybspeed RPA kit
(Ambion, Inc.) according to the manufacturer’s protocol.
Assays for each particular mRNA were carried out in separate tubes
(except for those for the FSHR and LHR mRNAs) containing target RNA and
1 x 105 cpm each of the respective antisense
riboprobe and the antisense mouse cyclophilin riboprobe. Protected
fragments were separated on a 6% bis-acrylamide/8.3 M
urea/1 x TBE gel (National Diagnostics Systems, Atlanta,
GA). Each antisense riboprobe produced the following size-protected
fragments: AR = 439 nt, Cyc = 103 nt, FSHR = 355 nt,
LHR = 236 nt, and PR = 365 nt. Note that the antisense
riboprobes for the FSHR and LHR mRNAs were shortened to produce
protected fragments of sizes that were more optimum to the RPA. The
gels were then fixed and dried using the Hoeffer Easy Breeze gel drying
system. The resulting bands were visualized and quantified with the
Storm 850 Phosphorimager and accompanying ImageQuant software
(Molecular Dynamics, Inc., Sunnyvale, CA) followed by
exposure to x-ray film.
Antibodies and immunocytochemistry
A rabbit polyclonal antibody (PAI-310) raised against a
synthetic peptide corresponding to the C-terminal amino acid residues
467–485 of rat ERß was purchased from Affinity BioReagents, Inc. (Golden, CO). The characterization of this antibody by
Western blot and gel supershift was accomplished using rat ERß
overexpressed by COS-7 cells. ER monoclonal antibody (clone 1D5;
DAKO Corp., Carpinteria, CA) binds to estrogen receptors
and localizes ER in target tissues both by immunofluorescence and
immunoperoxidase (19).
The paraffin sections were first deparaffinized and then treated with
3% H2O2 in PBS (pH 7.6) for 5 min. These steps
were followed by heating the sections in a microwave oven (3–4 min
each) for antigen retrieval using a citrate buffer pH 5.5–5.7
(HIER buffer, 1:10 dilution, Ventana Medical Systems, Inc., Santa
Barbara, CA) and processed for immunostaining by the avidin-biotin
peroxidase method as previously described (20). The sections were
incubated overnight at 4 C with ERß antibody, preadsorbed ERß
antibody or ER monoclonal antibody, (ID5, DAKO Corp.).
ERß antibody was used at a concentration of 5 µg/ml and ID5
monoclonal antibody at a concentration of 0.1–0.2 µg/ml. Sections
were washed in 1.0 mM PBS (pH 7.6) followed by incubation
with the secondary antibody, goat-antirabbit IgG or horse-antimouse
IgG, and Elite avidin-biotin peroxidase at a concentration of either
1:100 or 1:200 for 30 min each at room temperature. After a 5-min wash,
the sections were treated with liquid diaminobenzadine (DAB)
(BioGenex Laboratories, Inc., San Ramon, CA) followed by a
10-min wash in PBS and then counterstained with hematoxylin.
Specificity of the ERß antibody was established by incubating the
sections either with normal rabbit serum or preadsorbed ERß antibody,
prepared by incubating 5.0 µg of ERß antibody with 20 µg peptide
in 1.0 ml for 24 h at 4 C. The specificity of ER immunostaining
was established by incubating sections of ovary and oviduct with the
normal mouse IgG, which did not show specific immunoreaction.
In situ hybridization
Twenty-micrometer sections of fresh-frozen ERKO and WT ovaries
were prepared using a Reichert 820 cryostat (Buffalo, NY) and mounted
onto gelatin and poly-L-lysine-coated glass slides for
in situ hybridization as described previously (21).
Hybridization probes used were [35S]-UTP-labeled
riboprobes derived from a rat LH-receptor cDNA subclone. Hybridization
was continued for 12–18 h at 47 C in a humidified chamber. Sense
riboprobes were used as controls. Subsequently, the slides were washed
to a final stringency of 0.2x SSC at 55 C after a 1-h treatment with
20 µg/ml RNase at 37 C. Slides were then processed for emulsion
autoradiography (NTB-2, Eastman Kodak Co., Rochester, NY).
Exposure time on emulsion was 2 weeks. After development, slides were
stained with hematoxylin to visualize nuclei. The sections were then
examined and photographed using a macroscope (Wild M420, Leica Corp., Heerbrugg, Switzerland) or a microscope
(Nikon Optiphot, Nippon Kogaku [USA] Inc., Garden City,
NY).
Histochemical detection of apoptosis
Ovaries from adult WT and ERKO females were removed and fixed in
4% paraformaldehyde (pH 7.0) for 4–6 h at 4 C. The tissues were then
transferred to 70% ethanol at 4 C for 24 h, embedded in paraffin,
and sectioned at 5 microns. Sections were mounted on silanized slides
(Oncor, Gaithersburg, MD) and processed through the MEBSTAIN Apoptosis
Kit (Medical & Biological Laboratories, Watertown, MA) according to the
manufacturer’s protocol. This kit utilizes a fluorescent label and is
based on the TUNEL method. A single modification was made where the
incubation time during the proteinase K step was increased from 30
m to 1 h. Stained sections were coverslipped using Antifade
reagent (Oncor) as a mounting medium to maintain the fluorescent signal
as long as possible. Sections were evaluated and photographed using a
Zeiss Photomicroscope and Fugi Provia 1600 film at ASA 800 (Carl Zeiss, Thornwood, NY).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
gene disruption were restricted to adult animals. We evaluated whether
the adult phenotype observed earlier (11) was developmentally
determined or a result of aging. Neonatal ERKO females have ovaries
which are indistinguishable from those of WT females from a gross
morphological standpoint, and which at 10 days of age contain
histologically normal primordial and primary follicles (Fig. 1
, A and B, WT ovaries; E and F, ERKO
ovaries). At 20 days, the ovarian phenotype of the ERKO animals changes
slightly. The ovaries contain some follicles that have developed to the
antral stage (Fig. 1, C and D, WT ovaries; G and H, ERKO ovaries),
perhaps secondary to an elevation in gonadotropin levels, first
noticeable in ERKO animals at about this time (10). The ovarian
phenotype in 50-day, postpubescent, ERKO animals is very similar to the
initial description in older (12 week), acyclic adult animals (11, 22).
Follicle development in the ERKO animals culminates in atresia or as
hemorrhagic cysts, whereas corpora lutea, representative of successful
ovulation in response to appropriately timed gonadotropin surges, are
present in the WT ovary and not in the ERKO (compare Fig. 2A and 2B). The absence of corpora lutea
is not conclusive evidence of anovulation but is highly correlative. To
date, we have not observed corpora lutea in any of the sections
analyzed. Figure 2C (arrow) shows a high-power view of a
hemorrhagic, cystic follicle in which only a single granulosa cell
layer remains. Adjacent antral follicles contain several granulosa cell
layers, indicative of an initial, functional follicular growth response
to gonadotropin. Figure 2D shows a thecal blood vessel of a
PMSG-stimulated ovary hemorrhaging into the follicular antrum.
Hemorrhage into the follicles can occur at multiple points and also
occurs before death of all the granulosa cell layers (not shown). We
attempted to induce ovulation in adult ERKO mice in the absence of a
GnRH antagonist. Exogenous PMSG followed by hCG treatment of adult
animals did not result in ovulation as indicated by the absence of
corpora lutea. The thecal layer of these follicles was hypertrophied.
The granulosa cells tended to dissociate from each other and from the
lamina propria suggesting a lack of cell-to-cell association in this
compartment (not shown).
|
|
) corresponds to the
most abundant transcript previously described in both rat testis and
granulosa cells (23, 24). Similarly, several LHR mRNA transcripts
correspond to those previously described for rat ovary, mouse testis,
and mouse MA-10 Leydig cells (25). These results validate the accuracy
and specificity of the probes. When RNA pooled from several trimmed
ovaries of each genotype were compared by Northern analysis, the FSHR
and LHR transcripts representing the ERKO were appreciably elevated
compared with the WT ovary. Figure 4A
demonstrates expression of the most abundant FSHR and LHR mRNA
transcripts when analyzed in individual WT and ERKO ovarian samples by
RNase protection assay. In agreement with the Northern, ERKO ovary FSHR
mRNA expression averaged approximately 4-fold higher than WT controls.
Average LHR mRNA expression was slightly higher in the ERKO ovaries
(Fig. 4B). Because factors such as the number of preovulatory follicles
and corpora lutea can introduce variability and complicate the
interpretation of LHR mRNA expression on the basis of whole-ovary mRNA
levels in these type comparisons, we performed in situ
hybridization analyses of LHR mRNA expression.
|
|
–mediated
actions. In situ FSHR expression was marginally above
background and was not sufficiently definitive (not shown). Figure 5 (top panel) indicates the
specificity of the LHR probe used in all of the studies: the sense
probe produced only a very low background. The antisense probe detected
LHR mRNA in specific ovarian cell types in PMSG-stimulated WT and
ERKO ovaries. The representative WT ovary shows LHR mRNA expression by
the granulosa and thecal layers of the follicle as well as the
interstitium and the corpus luteum (Fig. 5; antisense; middle
panels). The representative ERKO ovary shows high-level LHR mRNA
expression in the granulosa and thecal compartments of virtually all of
the large antral follicles present (Fig. 5; antisense; lower
panels). These results indicate that ER action is not required
for LHR mRNA expression by interstitial, thecal, or granulosa
cells.
|
, A and C,
shows low and high power views, respectively, of ovarian ERß
localization in WT ovaries. The principal feature is pronounced
granulosa cell nuclear staining. Some staining is also evident in
precorona radiata and precumulus cells immediately surrounding the
oocyte. Theca interna cells did not express detectable ERß protein.
All ovarian cell types stained at background levels in sections
preabsorbed with peptide (low and high power views, Fig. 6, B and D,
respectively). Ovarian ERß staining in the ERKO animals was
essentially the same as that of WT controls, showing pronounced
expression by granulosa cells and the cell types immediately
surrounding the ovum (Fig. 6, E and F, low and high power,
respectively). In contrast, ER protein was absent in the granulosa
compartment of comparable follicles in the WT animals but was expressed
by thecal and interstitial cells (Fig. 6G). Nuclear staining was absent
in sections processed without primary antibody, indicating specificity
of the antibody (Fig. 6H).
|
is a representative example of a
mature antral follicle with several granulosa layers, a morphology
compatible with that of a preovulatory follicle. However, many
granulosa cells are undergoing apoptosis as indicated by nuclear
pycnosis, those sloughed into the follicular antrum have undergone
lysis, and the oocyte is degenerating. Beyond this stage, the surviving
follicles become cystic. Figure 7B shows by another procedure,
fluorescence TUNEL methodology, that the granulosa cells of ERKO
follicles undergo apoptosis. The incidence of granulosa cell apoptosis
was comparable in both genotypes (data not shown), suggesting that the
process is not principally dependent upon ER.
|
).
Therefore, the anovulatory ERKO phenotype is not associated with marked
differences in steady-state ovarian PR and/or AR gene expression.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
disruption results in altered developmental ovarian
morphology, enhanced androgen, estrogen, and LH secretion, elevated
FSHR mRNA expression, and the formation of cystic, hemorrhagic
follicles in association with anovulation. ER disruption did not
inhibit formation of the ovary itself or of the oocyte, did not alter
circulating serum FSH levels, and did not prevent LHR, AR, PR mRNA
expression, or ERß protein expression.
Because ER plays such a crucial role in female reproductive
development, it was not unreasonable to expect that ER disruption
might produce a phenotype lacking ovaries or one severely depleted of
oocytes at an early developmental stage. This was not the case. Oocytes
were present and biochemically functional in the sense that they
contributed to granulosa cell and follicle formation. This implies that
ER action is not essential for the functional activity of at least
one oocyte-derived growth factor because disruption of the oocyte
factor GDF-9 gene leads to premature demise of ovarian follicles at the
one-layer granulosa cell stage of primary follicle development (27).
ER disruption was compatible with apparently normal follicle
development up to 10 days after birth, with differences beginning to be
evident at 20 days. The accelerated follicle development is probably
secondary to elevated gonadotropin levels, first observable at about
this time. In postpubertal-age animals, antral follicles developed and
were steroidogenically active in terms of androgen and estrogen
secretion (26) but did not progress to ovulation. The mechanisms
contributing to anovulation presumably are responsible in part for the
dramatic finding of large, cystic, hemorrhagic follicles in the ERKO
animals (11).
Other genetic models, such as the gene disruptions of FSHß, insulin-like growth factor I, cyclin D2, COX-2, PR, and vitamin D exemplify follicular arrest at various developmental stages (28, 29, 30, 31, 32). Follicular development in the FSHß and insulin-like growth factor I knockout models progressed only to the early antral stage. In the majority of cases, follicle development in mice overexpressing Follistatin arrests between the primary and antral follicle stages (33). Development in the cyclin D2-knockout animals progressed only to follicles with four granulosa cell layers, and COX-2 knockout animals developed ovulatory follicles that luteinized and formed corpora lutea but failed to release the ovum. Ovaries of aromatase knockouts develop follicles with numerous granulosa cells but fail to ovulate; those of Connexin 37 knockouts do not develop Graafian follicles and also fail to ovulate (34, 35). PR knockout animals formed preovulatory follicles that failed to ovulate (36). These models emphasize the complexity and interdependence of various systems having regulatory input to the processes of follicular development and ovulation. The now characteristic ERKO phenotype of cystic, hemorrhagic follicles did not develop in any of the above-described models.
A major contributory factor to the anovulatory and cystic, hemorrhagic
follicular state(s) in the ERKO females seems to involve the
physiological regulation of LH and its receptor. LH levels were
significantly higher than in WT controls, but its secretory pattern was
presumably acyclic because an estrogen-initiated response at the
hypothalamic-pituitary level is necessary to elicit the ovulatory
gonadotropin surge (37). The mouse pituitary expresses ER, but
levels of ERß mRNA are low to absent (16, 38). Ovaries of the ERKO
animals show biochemical and histological evidence of elevated LH
stimulation: 1) circulating androgen levels are very high, and 2)
thecal and interstitial cells are hypertrophied. The ovarian
histological picture is quite similar to those of transgenic mice
overexpressing bLHß-CTP (39, 40). Both models exhibit normal
follicular morphology in the early stages, but development then
culminates in the formation of anovulatory, hemorrhagic, cystic
follicles or granulosa cell tumors. Some corpora lutea develop in the
bLHß-CTP-overexpressing animals but are absent in the ERKO. This is
one slight difference in ovarian morphology between the two models.
In situ analysis of LHR mRNA expression corroborated the
histological picture. LHR mRNA expression was at a high level
throughout the ovary and was more consistent among ERKO animals because
of a relatively greater number of large antral follicles. There was
considerable variability in LHR mRNA expression in the WT ovaries,
which contained fewer large follicles of comparable size to those of
the ERKOs, presumably because the WT animals better reflected the
distribution of follicular development over the normal stages of the
estrous cycle. A large proportion of signal was localized in the
granulosa compartment. This is important because previous work
established convincingly that estrogen action along with that of FSH
was necessary to induce granulosa cell LHR binding and mRNA expression
in the rat granulosa cell (8). Our combined RPA and in situ
results show that ER action is not necessary to achieve granulosa
cell LHR mRNA expression in either the granulosa or the thecal
compartment. ERß mRNA is highly expressed in granulosa cells and it
is a likely candidate to explain the earlier studies that indicated
estrogen action was necessary, but not sufficient, for LHR induction
(8). LH itself is another candidate effector. Although ovulatory levels
of LH transiently down regulate LHR, LH has also been shown to maintain
granulosa cell LHR binding (41, 42).
PR knockout mice also fail to ovulate (36). Thus, the action of this receptor is important and potentially very relevant to the ovulatory failure in the ERKOs because ER activity is prerequisite for the induction of at least some forms of PR mRNA expression in the uterus (18). PR mRNA was low but comparable in both genotypes. If PR mRNA expression were ERß-regulated, one would expect levels to be high in the ERKO because circulating estrogen is elevated in these animals. It is more likely that ovarian PR mRNA expression is regulated by cAMP (43, 44). Our results would be consistent with this observation. Although the lack of PR expression in a specific cell type of the ovary associated with ovulation is theoretically possible, these results also indicate that the anovulatory ERKO phenotype cannot be ascribed to an absence of PR mRNA expression.
Failure of the ovulatory mechanism in the ERKO animals is not due to
impairment of steroidogenesis per se. The increased androgen
and estrogen levels are strong evidence that the rate-limiting steps of
StAR, P450scc, and the other P450-dependent steps, C17
lyase and aromatase, are not significantly impaired by disruption of
the ER gene (26).
Ovarian ERß expression was detected at the protein level and it was
sharply compartmentalized to the granulosa cells. Granulosa cells of
the preantral follicles thus express ERß as well as FSHR well before
expression of LHR binding and mRNA expression (8). ER gene
disruption caused a significant increase in FSHR mRNA levels. This
leads to the hypothesis that this receptor may be regulated in part by
a combination of ER and ERß action or by an ER/ERß
heterodimer.
The summary findings indicate that the major cause of the adult ERKO
phenotype is the disruption in negative feedback at neuroendocrine
centers (9), resulting in high circulating LH interacting with
functional ovarian LHR. These processes presumably involve both
hypothalamic-pituitary and intraovarian mechanisms dependent upon
ER. Also, ERß may be preserving roles of estrogen action
previously associated with some, but not all of the actions of ER
within the ovary, and ERß may be uniquely regulating several
additional biochemical endpoints that remain to be identified. The ERKO
model should therefore be invaluable in helping to dissect the
respective roles of these two ERs in ovarian function.
|
Footnotes |
|---|
Received November 16, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
in the
anterior pituitary gland. Mol Endocrinol 11:674–681 (ERKO) mice. Program of the 80th Annual
Meeting of The Endocrine Society, New Orleans, LA, 1998 (Abstract
OR46-2), p 112
(ER) and estrogen receptor-ß (ERß) messenger
ribonucleic acid in the wild-type and ER-knockout mouse.
Endocrinology 138:4613–4621This article has been cited by other articles:
 |
|
 |
|
  S. Ikeda, K. Nakamura, K. Kogure, Y. Omori, S. Yamashita, K. Kubota, T. Mizutani, K. Miyamoto, and T. Minegishi Effect of Estrogen on the Expression of Luteinizing Hormone-Human Chorionic Gonadotropin Receptor Messenger Ribonucleic Acid in Cultured Rat Granulosa Cells Endocrinology, April 1, 2008; 149(4): 1524 - 1533. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. C. Gieske, H. J. Kim, S. J. Legan, Y. Koo, A. Krust, P. Chambon, and C. Ko Pituitary Gonadotroph Estrogen Receptor-{alpha} Is Necessary for Fertility in Females Endocrinology, January 1, 2008; 149(1): 20 - 27. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K.R. Barnett, C. Schilling, C.R. Greenfeld, D. Tomic, and J.A. Flaws Ovarian follicle development and transgenic mouse models Hum. Reprod. Update, September 1, 2006; 12(5): 537 - 555. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. F. Couse, M. M. Yates, K. F. Rodriguez, J. A. Johnson, D. Poirier, and K. S. Korach The Intraovarian Actions of Estrogen Receptor-{alpha} Are Necessary to Repress the Formation of Morphological and Functional Leydig-Like Cells in the Female Gonad Endocrinology, August 1, 2006; 147(8): 3666 - 3678. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. B Rulli and I. Huhtaniemi What have gonadotrophin overexpressing transgenic mice taught us about gonadal function? Reproduction, September 1, 2005; 130(3): 283 - 291. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. K. Woodruff and K. E. Mayo To {beta} or Not To {beta}: Estrogen Receptors and Ovarian Function Endocrinology, August 1, 2005; 146(8): 3244 - 3246. [Full Text] [PDF]
|
|
|
|
|
|
Y. Omori, K. Nakamura, S. Yamashita, H. Matsuda, T. Mizutani, K. Miyamoto, and T. Minegishi Effect of Follicle-Stimulating Hormone and Estrogen on the Expression of Betaglycan Messenger Ribonucleic Acid Levels in Cultured Rat Granulosa Cells Endocrinology, August 1, 2005; 146(8): 3379 - 3386. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. F. Couse, M. M. Yates, B. J. Deroo, and K. S. Korach Estrogen Receptor-{beta} Is Critical to Granulosa Cell Differentiation and the Ovulatory Response to Gonadotropins Endocrinology, August 1, 2005; 146(8): 3247 - 3262. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. E. Vaskivuo, M. Maentausta, S. Torn, O. Oduwole, A. Lonnberg, R. Herva, V. Isomaa, and J. S. Tapanainen Estrogen Receptors and Estrogen-Metabolizing Enzymes in Human Ovaries during Fetal Development J. Clin. Endocrinol. Metab., June 1, 2005; 90(6): 3752 - 3756. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. A. Emmen, J. F. Couse, S. A. Elmore, M. M. Yates, G. E. Kissling, and K. S. Korach In Vitro Growth and Ovulation of Follicles from Ovaries of Estrogen Receptor (ER){alpha} and ER{beta} Null Mice Indicate a Role for ER{beta} in Follicular Maturation Endocrinology, June 1, 2005; 146(6): 2817 - 2826. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. H. Herynk and S. A. W. Fuqua Estrogen Receptor Mutations in Human Disease Endocr. Rev., December 1, 2004; 25(6): 869 - 898. [Abstract] [Full Text] [PDF]
|
|
|
|
