| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Pharmacology, Case Western Reserve University School of Medicine, Cleveland, Ohio 44106
Address all correspondence and requests for reprints to: John H. Nilson, Ph.D., Department of Pharmacology, Case Western Reserve University School of Medicine, 2109 Adelbert Road, Cleveland, Ohio 44106-4965. E-Mail: jhn@po.cwru.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In part, this ambiguity is derived from the complex etiology of female reproductive disorders that are associated with elevated LH. For example, polycystic ovarian syndrome (PCOS) is characterized by elevated LH to FSH ratios, elevated androgen to estrogen ratios, and cystic ovaries. PCOS is also the most common cause of anovulation (5, 6). This profile is often accompanied by insulin resistance, obesity, hirsuitism, and ultimately infertility primarily due to anovulation (7, 8, 9, 10). Approximately 75% of anovulatory infertility cases can be attributed to PCOS (11).
Women with elevated LH who are diagnosed with PCOS have difficulty conceiving, and those who do conceive experience a miscarriage rate of 30–64% (compared with 12% in women with normal levels of LH) (12, 4). As a result of the multiple clinical manifestations of disorders such as PCOS, dissecting the cause from the consequence is obviously challenging. Thus, the development and analysis of adequate animal models for the study of this etiologically complex disorder becomes important.
Recently, our laboratory developed a transgenic mouse using the
proximal promoter from the bovine gonadotropin-
subunit gene to
direct expression of a chimeric bovine LH ß subunit fused to the
carboxyl terminal peptide (CTP) of hCG-ß. The resulting transgene is
expressed exclusively in gonadotropes of the anterior pituitary (13).
Serum LH levels are elevated due to robust transgene expression and an
extended half-life of the chimeric LH as a result of the CTP fusion
(13). While male transgenics do not hyper-secrete LH and are
phenotypically normal, females develop extensive pathology. This
pathology includes hormonal alterations such as elevated LH, with an
elevated LH/FSH ratio, and elevated testosterone and estradiol, with an
elevated T/E ratio. These hormonal changes lead to precocious puberty,
ovarian cysts and anovulation by three weeks of age (14).
The development of transgenic, ovarian pathology at puberty led us to investigate the quality of oocytes maturing within the context of the altered hormonal milieu. One approach that we have used in collaboration with Hirshfield and colleagues was to determine the impact of chronic LH hypersecretion on follicular pools. We have recently shown that primordial follicles are depleted by 45% in transgenics by 5 weeks of age (15). This loss contributes to infertility observed in these mice.
Elevation of the LH/FSH ratio can lead to excessive ovarian androgen synthesis (11). Abnormally elevated levels of androgens are thought to induce follicular atresia and oocyte degeneration (16, 17). In addition to androgen-induced follicular demise, elevated levels of LH may have a detrimental impact on oocyte development directly (18). Indeed, increased LH concentrations during the follicular phase may result in the inappropriate activation of meiotic prophase I arrested oocytes (19). Because the increased rate of miscarriage in women diagnosed with PCOS is often attributed to poor oocyte quality (2, 20), we assessed the quality of oocytes developing within the transgenic ovary. We also investigated the impact of chronic LH hypersecretion on embryo and maternal reproductive health.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
LHßCTP gene was previously described (13). Mice were genotyped by
PCR using primers specific to the bovine subunit promoter (5'-AAG
GGC TGA AAC AAG ATA AGA TAA A-3') and the LHß subunit reporter gene
(5'-CTG GAA CAT CTC CAT CCT TG-3'). All mice originated from one
founder line that was bred 6–10+ generations into the CF1 strain for
these studies. Female, age-matched, nontransgenic littermates were used
as controls (-) for all experiments. Mice were between the ages of
4–10 weeks, unless otherwise indicated. The Institutional Animal Care
and Use Committee (IACUC) of Case Western Reserve University
approved all animal studies.
Oocyte studies
Ovaries were collected from 15- and 21-day-old transgenic and
nontransgenic littermates. Oocytes were released into media via
mechanical puncture using 30-gauge needles (Becton Dickinson and Co.). This method results in the collection of growing oocytes
(small antral follicles) from 15-day-old mice, and grown ooyctes (large
antral follicles) from 21-day-old mice (21). Only oocytes with easily
removable cumulus cells and a germinal vesicle were selected for
incubation overnight at 37 C. They were subsequently scored for the
continued presence of the germinal vesicle (GV), germinal vesicle
breakdown (GVBD), polar body formation (PB), or death. Incubation was
performed in Waymouth’s Media (Life Technologies,
Gaithersburg, MD) supplemented with 10% FCS, 50 U/ml penicillin, 50
µg/ml streptomycin, and sodium pyruvate to a final concentration of
0.25 mM.
Embryo collection/breeding
For all pregnancy studies, embryos were generated by
superovulating mice with a standard PMSG (pregnant mare serum
gonadotropin, Calbiochem, La Jolla, CA)/hCG (human CG,
Wyeth-Ayerst Laboratories Inc., Philadelphia, PA) regimen
(22), followed by mating with proven stud males. Copulation plugs were
identified the morning following mating when embryo age was referred to
as "1 dpc" (one day post coitus). Embryo transfer experiments
involved the collection of 1 dpc embryos from superovulated donor
females, selection of embryos with two identifiable pronuclei, and
transfer into pseudopregnant host females. Pseudopregnant females were
generated by mating mice with proven vasectomized males, and
identifying copulation plugs the following morning. Pseudopregnant
recipients received no hormonal treatments. Embryos were handled in FHM
media and incubated in KSOM (both from Specialty Media, Lavallette, NJ)
at 37 C, 5% CO2, until transfer into the oviducts of host
females. All transfer surgeries were performed on the afternoon of the
copulation plug under avertin anesthesia. Term gestation was defined as
20 dpc, and all animals killed at 20 dpc had not yet undergone
parturition.
Hormone measurements
Blood samples were obtained by either retro-orbital sinus
sampling, or at the time of death via cardiac puncture. Sera were
prepared by clotting, centrifugation, and then collection of
supernatants. Sera were stored at -20 C before RIA. LH was assayed as
previously described (23). The limit of detection for LH was 0.89
ng/ml. Progesterone was assayed using a kit from Pantex
(Santa Monica, CA). The limit of detection was 0.2 ng/ml. Estradiol and
testosterone concentrations were determined using kits from Diagnostics
Biochem Canada Inc. (London, Ontario, Canada). The limit of detection
for 17ß-estradiol was 5 pg/ml and 0.01 ng/ml for testosterone. All
steroid hormone analysis kits were validated in our laboratory for use
with mouse serum. All hormone measurements were performed with single
aliquots due to the small amounts of sera available. At least three
samples were included in each data point. Estradiol concentrations
during pregnancy reflect an average of two separate assays on the same
samples.
Uterine receptivity
Decidualization was induced by injecting 100-µl safflower oil
into one uterine horn (at the oviduct/uterus junction) of
pseudopregnant mice on 4 dpc. Control mice received sham surgery (no
injection). At 6 dpc, mice were killed, uteri were collected, ovaries
and oviducts were removed, and uterine wet weights were obtained.
Decidualization was also scored by gross morphology.
Pregnancy rescue
Embryo transfers, using only nontransgenic embryos, were
performed as described except ovariectomy was performed immediately
following the transfer. Progesterone (10 mg/ml) in sesame seed oil
(Barron Pharmacy, Beachwood, OH) was administered at 2 mg/day sc
starting on 3 dpc. 17ß-estradiol (Sigma Chemical Co.,
St. Louis, MO) dissolved in safflower oil to 250 ng/ml, was
administered at 25 ng sc on 5 dpc then lowered to 12.5 ng/day on
6 dpc.
Statistical analysis
Statistical differences were assessed by single factor ANOVA,
and reported as the mean ± the SEM. Differences where
P < 0.05 were considered statistically
significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). The
gross appearance of this phenotype correlates with the onset of
precocious puberty in transgenic mice at 21 ± 0.15 days of age
compared with 30 ± 0.86 days in nontransgenics (14). Oocytes
developing within this pathological ovary are exposed to elevated LH,
androgens, and estrogens, which prompted us to evaluate their quality.
We first determined the age at which oocyte meiotic competency was
acquired in transgenics compared with nontransgenic littermate controls
(Fig. 2).
|
|
), or in 21-day-old pubertal transgenics
compared with their prepubertal nontransgenic littermates (Fig. 2B). At
day 15, fewer oocytes were collected from transgenic mice
(transgenic = 33.2 ± 5.0, n = 9, nontransgenic =
50.5 ± 5.8, n = 12, P < 0.05). However, at
day 21 there were no significant differences in number of oocytes
retrieved (transgenic = 35.5 ± 5.6, n = 6,
nontransgenic = 29.8 ± 5.6, n = 5). Normally, most
oocytes from day 15 mice are not competent to complete the first
meiotic division in vitro (24). At day 15, approximately
half of the oocytes in transgenic and nontransgenic mice resumed
meiosis spontaneously with germinal vesicle breakdown (GVBD) upon
release into culture. The absence of the germinal vesicle and the lack
of a polar body indicates these oocytes were arrested in meiosis I,
most likely at metaphase. The majority of the remaining oocytes did not
mature as reflected by the maintenance of the germinal vesicle (GV) or
death, after overnight incubation. Less than 10% of oocytes produced a
polar body (PB) to complete meiosis I, a reflection of their
developmental immaturity. In contrast, by 21 days of age, in both
transgenics and nontransgenics approximately half of the oocytes were
arrested in metaphase II with formation of a polar body. Very few
oocytes from the 21-day-old mice were unable to resume meiosis. These
data demonstrate that transgenic oocytes acquire meiotic competency
normally despite their development within a pathological ovary and
early exposure to an altered endocrine milieu.
Anovulation can be reversed by administering a pharmacological LH
surge, but a transgenic pregnancy cannot be maintained due to maternal
defects
Transgenic females are infertile primarily due to anovulation.
Despite repetitive mating, females have not become spontaneously
pregnant, and oviducts did not contain ova following observation of a
vaginal plug (data not shown). Adult transgenics exhibit persistent
leukocytic vaginal smears (14), however we have observed multiple
consecutive copulation plugs in transgenics (up to four consecutively),
suggesting an altered behavior of chronic estrus-like receptivity. To
determine if anovulation was a result of constant elevated LH exposure
with no LH surge, we treated nonstaged, randomly cycling mice with a
pharmacological bolus of hCG, creating a peak in LH-like activity.
Ovulation was induced by administration of hCG as determined by the
presence of ova in the oviducts the morning after treatment. There was
no difference in number of ova collected or likelihood to ovulate
(transgenic = 50% ovulated average of 2.2 ±1.1 ova, n = 8,
nontransgenic = 33% ovulated average of 1.8 ± 0.65 ova,
n = 15 P > 0.1).
Superovulation, induced by treatment with PMSG followed by hCG 48
h later, also resulted in ovulation (Fig. 3). Ovaries collected the morning after
hCG administration, were sectioned and evaluated for the presence of
corpora lutea (CL). We found, in sections containing oviducts, ova
still associated with cumulus cells. Figure 3 contains a transgenic
ovary section with hemorrhagic follicles adjacent to healthy follicles
and luteinized structures resembling CL (+). The oviduct contains at
least seven ova, as indicated by the arrows. The
nontransgenic counterpart has normal follicles along with CL, and two
ova in the adjoining oviduct (-). This study indicates that transgenic
anovulation can be rescued by administration of a pharmacological LH
surge.
|
displays a typical transgenic (+)
pregnancy at 15 dpc compared with a nontransgenic (-) pregnancy.
Transgenic pregnancy was achieved with hormonal manipulation but could
not be maintained, resulting in fetal resorption. No differences were
seen between the number of fertilized ova obtained at 2 dpc
(transgenic = 30.2 ± 7.9, nontransgenic = 27.5 ±
9.6). However, average litter size per pregnancy (average from 2–20
dpc) was slightly, but not statistically (P > 0.05),
reduced in transgenics (12.4 ± 2.3 pups/litter) compared with
nontransgenics (17.3 ± 1.6 pups/litter). This may suggest a
population of oocytes from transgenics that are abnormal but do not
comprise the majority of the ovulated oocyte pool. Alternatively, this
may imply a uterine receptivity defect in transgenics even after
superovulation. To determine at what developmental stage transgenic
pregnancy began to fail, we evaluated in more detail the time course of
transgenic pregnancy resorption.
|
). The average
percent resorbed pups per pregnancy in transgenics (30%) (+) was
elevated compared with nontransgenic (5%) (-) at 10 dpc. By 12 dpc,
transgenic pregnancy resorption approaches 100%, whereas nontransgenic
resorption increases to 21%. From 12 to 20 dpc, transgenics maintained
an average of 98 ± 1% resorption, whereas nontransgenic
resorption rate averaged 33 ± 7%, P < 0.001.
Transgenic pregnancy resorption occurring at midgestation implies
either a late embryonic defect, or more likely, a defect in maternal
pregnancy support.
|
, row 1). Four of the five transfers
resulted in a normal pregnancy to term (20 dpc) when recipients were
killed and the pregnancy scored. Live, healthy pups from the recipients
were recovered, and the pregnancy was determined by PCR to contain
approximately 50% transgenic pups (as expected). One transfer surgery
resulted in no observable pregnancy, either due to surgical error, or
inadequate pseudopregnancy in the recipient. These results indicate
that embryos from transgenic mice are healthy and are capable of normal
development to term when provided a nontransgenic maternal
environment.
|
, row 2). Surprisingly, we never observed any
pregnancy. These data are in contrast to our earlier study showing that
transgenics were capable of achieving pregnancy but were unable to
maintain the pregnancy beyond midgestation. One possible explanation
for the difference between these studies, is that previous transgenics
were treated with superovulating hormones, whereas in the embryo
transfer studies, recipients were made pseudopregnant simply by mating
with a proven vasectomized male. The apparent requirement for exogenous
hormonal priming in transgenics to achieve pregnancy, suggested that
the transgenic mice might have a uterine receptivity defect.
Transgenic pregnancy defects include lack of uterine receptivity
due to inappropriate priming, and alterations in gestation hormonal
profiles
To test transgenic uterine receptivity, we challenged
pseudopregnant transgenics with a decidualization stimulus and scored
their response in comparison to nontransgenics. When oil was injected
into one uterine horn of pseudopregnant mice on 4 dpc, uteri collected
2 days later from nontransgenics exhibited a typical decidualization
response (Fig. 6). In nontransgenics
(-), the injected left uterine horn ballooned from the tip of the
horn, to the cervix, whereas the uninjected right horn remained
unresponsive. In contrast, transgenic uteri (+) never exhibited a
decidualization response. Both uterine horns, regardless of oil
injection, remained small and unresponsive.
|
). In nontransgenics
(-), all four oil stimulated animals exhibited an increased uterine
wet weight to 0.63 ± 0.096 g from an average of 0.17
± 0.028 g in six sham surgeries, a 3.7-fold increase. All
five oil treated transgenics (+) however, did not increase uterine wet
weight (0.14 ± 0.010 g) compared with their five sham treated
counterparts (0.14 ± 0.023 g). Oil-treated transgenic uteri were
statistically different (P < 0.001) from nontransgenic
oil-treated uteri. The lack of decidualization following oil stimulus
indicates that transgenic pseudopregnancy fails to appropriately prime
the uterus to be receptive for implantation. This could reflect a
hormone imbalance because treatment with PMSG/hCG rescues this
defect.
|
).
In nontransgenics, a gradual rise in serum progesterone concentrations
was observed from a basal value of 4.4 ± 1.5 ng/ml to a peak at 6
dpc of 20 ± 5.6 ng/ml. Pseudopregnancy progesterone levels
resolved, as expected, to close to basal levels by 12 dpc (6.9 ±
2.5 ng/ml). Both basal and 12 dpc progesterone levels in nontransgenics
were statistically different from the peak at 6 dpc (basal
vs. 6 dpc P < 0.05, 6 vs.12 dpc
P < 0.05). In contrast, transgenic serum progesterone
levels were higher than nontransgenics before pseudopregnancy
(11.5 ± 4.1 ng/ml) and rose to 26.5 ± 6.7 ng/ml by 6 dpc
(basal vs. 6 dpc p 
0.05). This rise reflects only a
2.3-fold increase in progesterone concentrations compared with 4.6-fold
increase in nontransgenics over the same time period. Finally, unlike
the nontransgenic profile, transgenic progesterone production did not
return to basal levels by 12 dpc (26 ± 8.8 ng/ml). Indeed,
progesterone levels remain unchanged from their 6 dpc peak. It is
possible that this altered progesterone profile following
pseudopregnancy induction may contribute to the transgenic uterine
receptivity defect.
|
). As expected, serum LH
concentrations are elevated in transgenics (7.5 ± 0.89 ng/ml)
throughout gestation compared with nontransgenics (0.96 ± 0.028
ng/ml); P < 0.001 (Fig. 9A). On average, transgenic LH
concentrations remained 7.8-fold elevated over nontransgenics. These
levels are somewhat lower than, but still consistent with prepregnancy
values of 17.8 ± 5.0 ng/ml in transgenics vs. 1.5
± 0.33 ng/ml in nontransgenics, a difference of 11.9-fold.
|
). Estradiol concentrations from
8–14 dpc, a critical window of development during which transgenic
pregnancy fails, are elevated on average 4 ± 1-fold over
nontransgenic levels. Estradiol values continue to rise throughout
gestation, although animal variability in transgenics prevents
conclusive statistical analysis. In addition to estradiol, testosterone
levels are elevated in pregnant transgenics (Fig. 9C). Testosterone
levels become elevated (and remain so) in transgenics starting at 14
dpc. On average, transgenic testosterone levels are 11.2 ±
2.5-fold elevated over nontransgenics from 14–20 dpc. In contrast to
estradiol and testosterone, progesterone levels were unchanged in
transgenics. In response to superovulation, average progesterone levels
increase in nontransgenics dramatically (87 ± 16 ng/ml, Fig. 9D, 2 dpc) compared with baseline levels (4.4 ± 1.5 ng/ml, Fig. 8, basal), and compared with progesterone levels at 2 dpc following
pseudopregnancy induction (11 ± 3 ng/ml, Fig. 8, 2dpc).
Transgenic progesterone levels at 2 dpc following superovulation still
approximate those of 2 dpc following pseudopregnancy (33 ± 19
ng/ml vs. 25 ± 9.0 ng/ml). However, by 4 dpc, there
appears to be no statistically significant difference between
progesterone levels in pregnant transgenics compared with
nontransgenics (Fig. 9D). These data suggest that corpus luteum failure
is not responsible for the midgestation pregnancy resorption in
transgenics.
Transgenic pregnancy defects can be overcome by restoring an
appropriate hormonal environment
While ovulation in transgenics could be rescued, we questioned
whether the uterine receptivity defect, and midgestation pregnancy
failure was reversible, or if a permanent physiological change had
occurred in the transgenics. We hypothesized that hormonal factors from
the transgenic ovary were responsible for these defects. If transgenic
pathophysiology is reversible, removal of the ovary and replacement
with hormones at normal levels should rescue pregnancy. Pseudopregnant
mice were ovariectomized and given progesterone starting on 3 dpc, to
mimic the level that would have been produced by the corpora lutea (25, 26). An implantation-inducing dose of estradiol was also administered
on 5 dpc (27, 28). From 6–20 dpc, additional estradiol was
administered to maintain sufficient levels of uterine progesterone
receptors (29, 30). In nontransgenics, 3 of 10 animals had normal
pregnancies with live pups, an additional 5 animals were pregnant but
resorbed the embryos (Table 2, row 1).
Only 2 animals showed no sign of pregnancy upon examination. This
experiment, when performed on 18 transgenics, also resulted in 3 normal
pregnancies with live pups (Table 2, row 2). An additional four animals
were pregnant but resorbed the embryos, while 11 animals had no sign of
pregnancy. These findings demonstrate that transgenics are capable of a
normal pregnancy producing live pups when ovariectomized and displaying
a normal hormonal environment. In addition, under these conditions,
uterine priming can occur normally in nonsuperovulated transgenics,
suggesting that the lack of uterine receptivity found in the previous
studies is due to inappropriate ovarian signaling. Finally, this study
also indicates that, while transgenic pregnancy defects can be rescued,
this paradigm is not sufficient to restore transgenic physiology to
that of completely normal, as only 20% (2/10) of the nontransgenics
failed to establish pregnancy compared to 61% (11/18) of the
transgenics (Table 2). Although these studies demonstrate that
transgenic gestation defects are due to the reversible effects of
ovarian factors, they do not reveal the primary ovarian agent
responsible for transgenic midgestation failure.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The transgenic ovarian phenotype becomes apparent by 3 weeks of age. While multiple studies have shown that oocyte development is compromised when exposed to elevated LH and androgens (1, 2, 17, 19, 20), we have found that oocytes developing in transgenics are remarkably healthy. Although primordial follicles are 45% depleted in transgenics by 5 weeks of age (15), we found that GV containing oocytes were equally represented at three weeks of age.
While transgenics develop elevated androgens by 2 weeks of age (14), we found that there were no meiotic maturation abnormalities in oocytes from transgenics. In addition, the onset of precocious puberty in transgenics at 3 weeks of age does not coincide with advanced oocyte meiotic maturity. This outcome is surprising in light of data indicating that exposure to elevated androgens can cause premature oocyte meiotic activation (19). These findings are additionally interesting because inappropriate meiotic activation is thought to be one cause of early pregnancy loss in PCOS patients with elevated androgens (20).
We further evaluated oocytes from transgenics by determining if they could undergo normal fertilization and embryonic development. We showed first that transgenic anovulation could be reversed by administration of a LH-like surge (hCG bolus). This suggests that infertility due to anovulation is a result of high LH with no LH surge.
Using a superovulation regimen, we observed that transgenics could become pregnant, although pregnancy failed at midgestation. While early pregnancy and embryonic development occurred normally until 8 dpc, by 10 dpc there was a trend toward pregnancy failure. By 12 dpc, transgenic pregnancy failure approached 100%. The developmental timing of this is strongly suggestive of a maternal vs. an embryonic defect because by 10 dpc, murine embryo organogenesis is almost entirely complete. After this point, most development is growth-related (31). It is possible that a pregnancy defect occurred earlier in gestation but that it was not manifested until midgestation (for example, a defect in placentation). In nontransgenics, resorption rates increase from 10–16 dpc as well, with an average from 12–20 dpc of 33% compared with 98% in transgenics (P < 0.001). The nontransgenic resorption rate may reflect a superovulation- induced increased pregnancy size that is beyond the normal gestation capacity of mice (average 16 pups at 16 dpc in both groups).
Using embryo transfer procedures, we demonstrated that transgenic midgestation pregnancy failure is maternal in origin. Embryos transferred from transgenics to nontransgenic recipients, developed normally to term. This indicates that oocytes developing within the transgenic ovary are capable of normal meiotic activation, fertilization, and embryonic development to term.
Reciprocal embryo transfers, where nontransgenic embryos were transferred into pseudopregnant transgenics, did not result in any pregnancies, despite 14 attempts. This outcome was surprising, considering previous superovulation studies where transgenics became pregnant but resorbed the pregnancy at midgestation. This suggested that pseudopregnancy, initiated by the mating stimulus, was insufficient to induce uterine receptivity in transgenics. When pseudopregnant mice were injected at 4 dpc with oil in one uterine horn, at 6 dpc transgenics showed no sign of decidualization, whereas nontransgenics did. This procedure has been used in the past as a measure of uterine receptivity (32). Unaltered uterine wet weights in transgenics also reflected this lack of uterine receptivity. These findings indicate that transgenics were indeed unable to respond appropriately to the mating stimulus to induce uterine receptivity.
One outcome of the mating stimulus resulting in pseudopregnancy is the rescue of the corpus luteum, leading to increased progesterone production. We measured serum progesterone levels before and after the induction of pseudopregnancy. Nontransgenics developed a 4.6-fold rise in progesterone, resolving to near baseline by 12 dpc, the anticipated length of pseudopregnancy. Transgenics, however, started with elevated progesterone levels that rose only 2.3-fold by 6 dpc and failed to drop by 12 dpc. These data indicate that the normal progesterone profile, induced by the mating stimulus, is altered in transgenics. This may contribute, along with other endocrine changes (such as elevated estradiol levels) to defective uterine receptivity in transgenics. The apparent normal uterine receptivity in transgenics following superovulation can be explained by the temporary alteration in their endocrine profile when subjected to superovulating hormones.
To evaluate the role of the maternal hormonal environment on transgenic
pregnancy failure, we measured LH, estradiol, testosterone, and
progesterone during pregnancy. All but progesterone were elevated
compared with nontransgenics. Elevated LH, due to the transgene, causes
elevated testosterone and estradiol. Estradiol is elevated during the
critical window from 8–14 dpc when pregnancy failure occurs.
Testosterone, however becomes elevated in transgenics from 14–20 dpc,
after the onset of pregnancy failure. While elevated androgens have
been speculated to be associated with early pregnancy loss (20), and
poor oocyte quality (33), testosterone itself is not thought to be an
embryonic toxin (34). In contrast, elevated estradiol during
midgestation has been shown to be toxic to embryos (35). Elevated
estradiol results in midgestation pregnancy resorption, as observed in
mice lacking the 5-reductase type I gene (34). Although these mice
have estradiol levels at least 2-fold higher than their wild-type
counterparts (from 6–14 dpc), LH-hypersecreting transgenics exhibit
approximately 4-fold higher estradiol compared with nontransgenics over
a similar time period (8–14 dpc). In light of this data, it is
possible that estradiol toxicity is responsible for pregnancy loss in
transgenics.
To test if pregnancy failure in transgenics is reversible and due to ovarian factors, we performed embryo transfers accompanied by ovariectomy and hormone replacement. It has been shown previously that uterine receptivity can be induced following ovariectomy by administering progesterone, followed by an implantation dose of estradiol (27, 28, 31). In addition, other studies have determined that progesterone and low levels of estradiol are required to maintain a pregnancy (29). Estradiol has also been shown to be necessary to induce and maintain the expression of progesterone receptors (30). This experiment produced three pregnancies with normal pups in both transgenics and nontransgenics. While informative, this procedure was inefficient; we also found approximately equal number of pregnancies had resorbed in each group. The difference between transgenics and nontransgenics became apparent in the number of "not pregnant" outcomes observed. While we found two (20%) nontransgenics who were not pregnant, 11 (61%) transgenics were not pregnant. This difference may reflect the transgenic uterine receptivity defect, which is more difficult to overcome using this procedure as opposed to superovulation. It is also possible that ovariectomy at 1 dpc does not provide sufficient clearance time for transgenics to return to a normal hormonal profile during implantation. Regardless of efficacy, the finding of healthy pregnancies in transgenics using this procedure suggests that midgestation pregnancy failure is a reversible maternal defect induced by ovarian factors.
To determine if estrogen toxicity caused the pregnancy failure, we
attempted to block the effects of estrogen using tamoxifen (data not
shown). In contrast to the pregnancy rescue observed in mice lacking
5-reductase type I (34), we were unable to rescue pregnancy using
tamoxifen. This may be due to partial agonist activity of tamoxifen in
the uterus (36, 37, 38). LH-hypersecreting mice have estradiol levels
approximately 2-fold higher than those observed in 5-reductase type
I deficient mice; thus the amount of tamoxifen required to block the
effects of estradiol might have induced pregnancy failure due to
agonist activity (39). We therefore elected to reduce estradiol by
inhibiting P450-aromatase, the enzyme responsible for
converting androgens to estrogens, using an injection paradigm of
4-androsten-4-ol-3,17-dione, an aromatase inhibitor. Unfortunately,
this approach was inadequate to reduce estradiol levels (data not
shown), and hence could not be used to study the role of elevated
estradiol on pregnancy in these mice.
The mechanism through which elevated estradiol during gestation induces pregnancy loss in mice is not clear. It has been speculated that estradiol may exert its toxic effects via alteration in vascular permeability leading to hemorrhage events (34). It is known that estradiol can regulate nitric oxide production in the endothelium, probably by inducing nitric oxide synthases (40, 41), and that nitric oxide itself can induce vasodilatation (42). This hypothesis, however, remains to be tested.
In summary, these studies further elucidate the physiological impact of chronic LH hypersecretion on oocyte, embryo, and maternal reproductive health. We conclude that chronic elevation of LH results in anovulation due to lack of an LH surge, lack of uterine receptivity due to inappropriate uterine priming following the mating stimulus, and mid- gestation pregnancy failure possibly due to estradiol toxicity. Surprisingly, chronic LH exposure and resulting chronic androgen exposure, does not prevent the development of meiotically normal oocytes that are capable of normal fertilization and development to term.
This study suggests that pregnancy failure in women diagnosed with PCOS may be attributed to a hostile maternal environment contributing to inappropriate uterine priming and/or pregnancy loss. Thus, these findings may enhance our understanding of infertility disorders (such as PCOS) in women involving elevated LH and androgen/estrogen ratios. Further analysis of LH hypersecreting transgenic mice and identification of the molecular mechanism(s) involved in their reproductive pathophysiology should provide an avenue for the future development of therapeutic agents.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received September 29, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-reductase type I caused by estrogen
excess. Mol Endocrinol 11:917–927This article has been cited by other articles:
 |
|
 |
|
  A. A. Murray, A. K.E. Swales, R. E. Smith, M. D. Molinek, S. G. Hillier, and N. Spears Follicular growth and oocyte competence in the in vitro cultured mouse follicle: effects of gonadotrophins and steroids Mol. Hum. Reprod., February 1, 2008; 14(2): 75 - 83. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Inzunza, A. Morani, G. Cheng, M. Warner, J. Hreinsson, J.-A. Gustafsson, and O. Hovatta Ovarian wedge resection restores fertility in estrogen receptor beta knockout (ERbeta-/-) mice PNAS, January 9, 2007; 104(2): 600 - 605. [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]
|
|
|
|
 |
|
 O. W.H. van der Heijden, Y. P.G. Essers, M. E.A. Spaanderman, J. G.R. De Mey, G. J.J.M. van Eys, and L. L.H. Peeters Uterine Artery Remodeling in Pseudopregnancy Is Comparable to That in Early Pregnancy Biol Reprod, December 1, 2005; 73(6): 1289 - 1293. [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]
|
|
|
|
 |
|
 J. S. Jorgensen, C. C. Quirk, and J. H. Nilson Multiple and Overlapping Combinatorial Codes Orchestrate Hormonal Responsiveness and Dictate Cell-Specific Expression of the Genes Encoding Luteinizing Hormone Endocr. Rev., August 1, 2004; 25(4): 521 - 542. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. P. Mohammad, R. A. Abbud, A. F. Parlow, J. S. Lewin, and J. H. Nilson Targeted Overexpression of Luteinizing Hormone Causes Ovary-Dependent Functional Adenomas Restricted to Cells of the Pit-1 Lineage Endocrinology, October 1, 2003; 144(10): 4626 - 4636. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. J. Mann, R. A. Keri, and J. H. Nilson Consequences of Elevated Luteinizing Hormone on Diverse Physiological Systems: Use of the LH{beta}CTP Transgenic Mouse as a Model of Ovarian Hyperstimulation-induced Pathophysiology Recent Prog. Horm. Res., January 1, 2003; 58(1): 343 - 375. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. B. Rulli, A. Kuorelahti, O. Karaer, L. J. Pelliniemi, M. Poutanen, and I. Huhtaniemi Reproductive Disturbances, Pituitary Lactotrope Adenomas, and Mammary Gland Tumors in Transgenic Female Mice Producing High Levels of Human Chorionic Gonadotropin Endocrinology, October 1, 2002; 143(10): 4084 - 4095. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Danilovich, I. Roy, and M. R. Sairam Emergence of Uterine Pathology during Accelerated Biological Aging in FSH Receptor-Haploinsufficient Mice Endocrinology, September 1, 2002; 143(9): 3618 - 3627. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. L. Milliken, R. K. Ameduri, M. D. Landis, A. Behrooz, F. W. Abdul-Karim, and R. A. Keri Ovarian Hyperstimulation by LH Leads to Mammary Gland Hyperplasia and Cancer Predisposition in Transgenic Mice Endocrinology, September 1, 2002; 143(9): 3671 - 3680. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. E. Owens, R. A. Keri, and J. H. Nilson Ovulatory Surges of Human CG Prevent Hormone-Induced Granulosa Cell Tumor Formation Leading to the Identification of Tumor-Associated Changes in the Transcriptome Mol. Endocrinol., June 1, 2002; 16(6): 1230 - 1242. [Abstract] [Full Text] [PDF] |
