| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Pathology (T.R.K., G.P., P.W., M.M.M) Department
of Cell Biology (M.M.M.) Department of Molecular and Human Genetics
(M.M.M.) Baylor College of Medicine Houston, Texas 77030
Department of Medicine and Neurobiology and Physiology
(T.K.W.) Northwestern University Chicago, Illinois 60611
Department of Molecular Biology and Pharmacology (I.B.)
Washington University School of Medicine St. Louis, Missouri
63110
Genetics Institute (M.C.B.) Cambridge,
Massachusetts 02140
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-subunit that is
noncovalently linked to a hormone-specific ß-subunit (1). The
pituitary gonadotropins LH and FSH bind to structurally related but
distinct receptors in the gonads and control gonadal growth,
differentiation, and steroidogenesis. FSH receptors (FSHRs) are
localized to Sertoli cells in the testis and granulosa cells in the
ovary (2). Expression of the glycoprotein -subunit and the
hormone-specific FSHß-subunit is regulated by the hypothalamic
peptide GnRH, steroids, and the gonadal and pituitary peptides,
activins, and inhibins (3, 4). To generate animal models for human diseases involving the gonadotropin signal transduction pathway, we recently produced loss-of-function mice deficient in the FSHß-subunit using embryonic stem cell technology (5). FSH-deficient female mice are infertile and demonstrate small ovaries resulting from a block in folliculogenesis at the preantral stage. In contrast, male mice deficient in FSH are fertile despite having small testes with reduced sperm number and motility (5). This loss-of-function model phenocopies human primary amenorrhea due to defective FSHR signaling in the ovary (6). Although loss-of-function mutations in the FSHß gene or FSHR gene could explain some forms of female infertility, including ovarian dysgenesis and hypogonadism, it is unclear whether ovarian hyperstimulation syndromes and ovarian cancer in women are due to elevated FSH levels or altered FSH signaling in the ovary.
Inhibins are members of the transforming growth factor-ß superfamily that includes important proteins such as activins, growth differentiation factor-9, and Müllerian-inhibiting substance (7). Inhibins were originally discovered as gonadal peptides that suppress pituitary FSH synthesis and secretion (8). The major sites of inhibin production in the gonads are Sertoli cells in the testis and granulosa cells in the ovary (9). To study gonadal growth and differentiation, we earlier generated an animal model in which mice deficient in inhibin develop multiple sex cord-stromal tumors (i.e. granulosa/Sertoli cell tumors) as early as 4 weeks of age with 100% penetrance (10). These tumors are usually multifocal and often hemorrhagic and secrete large amounts of estradiol and activins into circulation (4). The gonadal tumor-prone inhibin-deficient mice display characteristic hunchback and sunken eye appearance and eventually die as a result of a severe wasting (cancer cachexia-like) syndrome accompanied by hepatocellular necrosis around the central vein and a block in differentiation of several gastric cell lineages (11). Thus, inhibin was identified as a novel secreted tumor suppressor with gonadal specificity. Consistent with the known role of inhibin to negatively regulate FSH, inhibin-deficient mice demonstrate elevated levels of serum FSH. Although these types of gonadal tumors are rare in humans, elevated levels of serum FSH have been associated with some forms of ovarian epithelial cancers in elderly women (12). However, to date, there is no direct in vivo evidence to support the involvement of elevated levels of FSH in gonadal tumorigenesis. Earlier we showed that mice deficient in inhibin and GnRH (and therefore have suppressed levels of FSH and LH) survive for more than 1 yr and do not develop cachexia. These mutant male mice do not develop testicular cancers, but females show only premalignant lesions in the ovary (13). Although these studies indicated that gonadotropins (FSH and LH) are essential modifier factors for gonadal sex cord-stromal tumor development, we could not delineate the individual roles of FSH and LH in this pathway.
In this manuscript, we have addressed the biological role of FSH in gonadal growth and tumorigenesis. Using genetic approaches, including the production of gain-of-function transgenic mice overexpressing human FSH (hFSH) [expressed from a mouse metallothionein-1 (mMT-1) promoter], we have studied gonadal development. In an independent set of experiments, using a loss-of-function approach and genetic intercrosses, we generated double-homozygous mutant mice that are deficient in both inhibin and FSH to examine the role of FSH in gonadal tumor development/progression in these mice.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
and MT-hFSHß Transgenic Mice and Analysis of
Fertility
or MT-hFSHß transgenes. Using an hCG-specific probe fragment, we
identified two MT- male founder mice that had approximately 50–60
copies of the MT- transgene. The 700-bp human -probe fragment,
which did not hybridize to the endogenous mouse- subunit gene
sequences, permitted us to unequivocally identify the MT-
transgene-positive mice. In an independent set of pronuclear
microinjection experiments, four MT-hFSHß founder mice (two male and
two female) were identified using an hFSHß-specific 3'-untranslated
region (UTR) probe. Southern blot analysis and breeding experiments
confirmed that one male founder (line 2) had less than 5 copies of the
transgene, and the other (line 1) had 2 chromosomal integrations of the
hFSHß transgene (see below). At one site, less than 5 copies of the
transgene had integrated, and at the other site, approximately 50
copies were independently segregated and were transmitted to progeny
successfully. Both of the female founder mice had approximately 50
copies of the transgene.
To determine the tissue sites of expression of the transgenes, we
prepared total RNA from different tissues of the hFSH transgenic mice,
and duplicate RNA blots were separately hybridized with either an
hCG probe or an hFSHß 3'-UTR probe and subsequently stripped and
reprobed with an 18S rRNA probe. As shown in Fig. 1
, both of the transgenes (hCG in Fig. 1A and hFSHß in Fig. 1B) are expressed in multiple tissues with the
highest level of expression in the liver.
|
founders were fertile and stably transmitted the
transgene to subsequent generations. Similarly, the low copy-bearing
MT-hFSHß male founders, the female progeny derived from these lines,
and one male founder bearing high-copy MT-hFSHß transgene were all
fertile. These mice stably transmitted the transgene for several
subsequent generations. In sharp contrast, both female founders that
contained 50 copies of the MT-hFSHß transgene were infertile and
never mated to proven fertile male mice over a 6-month period.
Histological analyses performed on the ovaries from these founders and
female progeny mice obtained from the high-copy male founder confirmed
that these mice did not undergo estrous cycles. In contrast to ovaries
from control female mice (Fig. 2E), there
were no obvious corpora lutea, and many of the sections obtained from
these ovaries showed accumulation of a periodic acid Schiff
(PAS)-positive substance in the interstitial cells (Fig. 2F).
Immunohistochemical analysis using a hFSHß-specific monoclonal
antibody confirmed that the PAS-positive material in the interstitial
cells was hFSHß (data not shown). In ovaries from many of the older
female transgenic mice containing multiple copies of the hFSHß
transgene, there were often visible fluid-filled cysts, and
morphologically the ovaries looked pale yellow in contrast to the
highly vascularized ovaries in wild-type littermate mice. There were no
obvious defects in oocytes or granulosa and thecal cells. Thus, the
high-copy hFSHß transgene carrying female mice demonstrate
intraovarian defects of unknown etiology leading to infertility. It is
unclear whether the high levels of free FSHß subunit in the ovarian
interstitial cells are interfering with some key function.
|
and MT-hFSHß lines of mice. The progeny mice
were screened by Southern blot analysis, and mice positive for both
transgenes were identified. Reciprocal crosses were made between both
sexes of MT- and low-copy MT-hFSHß mice. This line of mice is
referred to as weak hFSH expressors. Both male and female weak hFSH
(i.e. dimer) expressors were fertile and indistinguishable
from the control wild-type littermates. These mice did not show any
gross phenotypic abnormalities or any pathology upon detailed surgical
and histological analysis when analyzed up to 1 yr of age. The
description and use of the weak hFSH expressors (i.e. 48.0
and 115.9 mIU/ml hFSH for males and females, respectively) have been
published previously (14).
Since high-copy MT-hFSHß females were infertile, male mice with the
high-copy number hFSHß transgene were bred to female mice that carry
the MT- transgene. The resulting mice that carry both transgenes are
referred to as high-copy hFSH (dimer) expressors. All further analyses
described in this manuscript were carried out on the high-copy hFSH
expressors, and hereafter these mice will be referred to as hFSH
(i.e. dimer) transgenic mice.
To determine the level of expression of the transgene mRNA driven by
the mMT-1 promoter under basal conditions, we analyzed the levels of
hFSH in the mouse serum using a specific fluoroimmunoassay that did not
cross-react with endogenous mouse FSH and did not detect the hFSHß
subunit. In addition, we measured the steroid hormone levels by
specific RIAs. Both of these studies demonstrated high levels of these
hormones in adult transgenic mice compared with age-matched wild-type
littermates (Table 1 and see below).
These results confirm that the hFSH heterodimer could be efficiently
assembled, processed, and secreted in large quantities from multiple
tissues of these transgenic mice. In addition, these data also suggest
that this ectopically produced hFSH is biologically active, leading to
enhanced gonadal steroid output into the serum.
|
, A and B). Likewise, there were no differences in
the weights of the epididymides. However, the seminal vesicles from the
transgenic mice were enlarged, appeared highly translucent, and were
greater than 2-fold larger (Fig. 2B) compared with the age-matched
wild-type male mice (Fig. 2A; 223.1 ± 10.9 mg, transgenic
vs. 105.3 ± 10.2 mg, wild-type, n = 6;
P < 0.05). The seminal vesicles were enlarged as early
as 3 weeks of age in the male transgenic mice consistent with elevated
testosterone levels in the serum (36.8 ± 6.7 ng/ml, transgenic
vs. 1.9 ± 1.5 ng/ml, wild-type, n = 5;
P < 0.05). In addition, histological analysis of the
testes did not show any obvious defects. The tubules appeared healthy
and intact and contained abundant spermatoza in the lumen, and normal
numbers of Leydig cells were observed in the interstitial spaces (Fig. 2C). To examine whether there were any quantitative differences in
sperm parameters, epididymal sperm from 6-week-old wild-type and
transgenic male mice were analyzed. There was a significant increase in
sperm number in the transgenic male mice (1.4 ± 0.2 x
107, transgenic vs. 0.8 ± 0.1 x
107, wild-type, n = 5; P < 0.05), but
no significant differences were observed in motility or viability when
wild-type and transgenic mice were compared (data not shown).
Castration of 42-day-old male transgenic mice resulted in regression of
the seminal vesicles similar to castrated age-matched wild-type male
mice (23.9 ± 5.9 mg, transgenic vs. 19.4 ± 1.8
mg, wild-type, n = 6; P > 0.05). This experiment
suggested that the increased size of the seminal vesicle was due to the
elevated testosterone and not a direct action of the hFSH on the
seminal vesicles. Thus, these hFSH-expressing male mice are essentially
infertile, demonstrate high levels of serum testosterone, and have
enlarged seminal vesicles. Based on our experiments, the infertility in
these hFSH-expressing transgenic male mice could be due to some
reproductive behavioral defects because of either the high hFSH
or the high testosterone levels in the serum.
hFSH-Overexpressing Female Mice Are Infertile and Develop
Hemorrhagic and Cystic Ovaries
More than 95% of the female transgenic mice appeared weak
and died between 6–9 weeks. To study the gain-of-function effects of
hFSH in female mice, reproductive tracts were examined morphologically
and histologically. At 6 weeks of age, ovaries from the transgenic
female mice were completely hemorrhagic and enlarged and appeared
cystic (Fig. 2D). The uteri were fluid filled and enlarged and appeared
translucent (Fig. 2D). These phenotypic characteristics were obvious as
early as 2 weeks of age. In contrast to control female mice (Fig. 2E),
histological analysis of the ovaries of these adult hFSH-transgenic
female mice demonstrated minimally intact follicular architecture and
no progression of follicles beyond the preantral follicle stage with
massive hemorrhagic and cystic islands within the ovaries (Fig. 2, H
and I). However, those follicles that remained intact showed
normal-appearing oocytes and granulosa and thecal layers. The ovarian
defects and hemorrhage appeared less severe at earlier time points, and
histological analysis showed the initiation of hemorrhage and cyst
formation as discrete foci. Furthermore, as early as 2 weeks, the
ovarian histology appeared normal with many immature follicles
including primary and early antral follicles (Fig. 2G). In two of the
female transgenic mice that survived to 13 weeks of age, the ovaries
were massively hemorrhagic and cystic with no apparent signs of
folliculogenesis. In addition, there were no obvious signs of any
tumors (data not shown). Consistent with these morphological and
histological findings, all of the female transgenic mice were
infertile. Serum IGF-I and LH levels, which are known to be elevated in
human polycystic ovarian disease, did not show any differences between
wild-type and hFSH-overexpressing female mice (Table 1). There were no
apparent defects in other tissues examined except for the kidneys and
bladders (described below). These findings in the hFSH-overexpressing
female mice resemble some of the features of gonadotropin-induced
ovarian hyperstimulation in human patients (see
Discussion).
Urinary Tract Abnormalities in hFSH-Overexpressing Female Mice
The immediately obvious defects secondary to excessive
stimulation of the ovaries were enlarged kidneys and urinary bladders
in the majority of the hFSH-transgenic female mice. Morphologically,
the bladders of the transgenic mice appeared thick and filled with
deposition of a white proteinaceous stone-like material (Fig. 2J).
Although the bladder enlargement was apparent as early as 3–4 weeks,
only mice that were more than 6 weeks old demonstrated this
accumulation inside the bladder. This material could be extracted into
0.1 N HCl and was found to be immunoreactive in an FSH RIA
(data not shown). As the deposition of this immunoreactive FSH-like
material progressed in the bladder, the urinary output declined, and in
two 13-week-old mice, there was no obvious sign of urine in the
bladder. Consistent with these observations, the kidneys were enlarged
and sometimes even cystic. Histological analysis of the kidneys from
these animals demonstrated that the architecture of many of the
glomeruli was damaged (Fig. 2L), the tubules were enlarged, and there
was a significant infiltration of macrophage-like cells at different
sites within the tubules (not shown). These results suggest that the
majority of the hFSH-overexpressing female mice developed urinary tract
obstruction leading to death.
The Majority of the
inham1/inham1,
fshbm1/fshbm1
Double-Homozygous Mutant Mice Fail to Develop a Wasting Syndrome
To determine the role of FSH in gonadal tumor development, we
generated double-homozygous mutant mice that lack both inhibin and FSH
by a genetic cross. The first overt sign of gonadal tumor development
in inhibin-deficient mice is severe weight loss caused by an
activin-related cachexia-like syndrome that eventually results in the
death of these mice (11). Therefore, the double-homozygous mutant mice
were weighed weekly, the weights were compared with those of mice
deficient only in inhibin (inham1/inham1), and
the percentage of survivors was calculated. As seen in Fig. 3A, 95% of the inhibin-deficient male
mice die by 12 weeks of age. In sharp contrast, the majority of the
double-homozygous mutant male mice survived to 1 yr and did not show
any dramatic weight loss. Alternatively, about 70% of the female
double-homozygous mutant mice survived past 17 weeks. However,
100% of them eventually lost weight and died by 39 weeks, in contrast
to 95% of the female mice deficient in inhibin alone that die by 17
weeks (Fig. 3B). These results suggest that absence of FSH was
affecting the gonadal tumor development/progression in
inhibin-deficient mice.
|
, A and B). Similarly, 9-
to 12-week-old female mice deficient in inhibin demonstrated hemorrhage
and disruption of the normal follicular architecture of the invasive
tumors. These tumors often contained masses of granulosa cells,
undifferentiated gonadal stromal derivatives, or clusters of
mitotically active stromal cells reminiscent of seminiferous tubules
(i.e. mixed granulosa/Sertoli cell tumors) (Fig. 5D).
|
|
and 5B). These tumors,
which initially appeared less hemorrhagic compared with the
inham1/inham1 mice, were very small in a few of
the 12-week-old double-mutant male mice (Fig. 4, C and D). Some of the
double-homozygous mutant male mice (6 of 11) did not develop any tumors
beyond 1 yr of age, and there were no signs of hemorrhage in the testes
of these mice. Histological analysis performed on the testes from these
double-mutant mice showed normal tubules that contained abundant sperm
in the lumen and normal numbers of Leydig cells (Fig. 4, E and F).
Testicular tumors that developed in some of the younger
double-mutant male mice often failed to disrupt the gross tubular
architecture; however, these tubules were often filled with
proliferating tumor cells that appeared to be less aggressive (Fig. 4D), unlike those in the testes of
inham1/inham1 male mice (10). Thus, absence of
FSH in the inhibin knockout male slows tumor development and, in some
cases, prevents the formation of gonadal tumors.
Although at 12 weeks, the ovaries of some of the double-mutant
female mice appeared normal morphologically (Fig. 5B), histological
analysis on ovarian tumors obtained from double-homozygous mutant mice
showed signs of hemorrhage and cyst formation as early as 12 weeks of
age (Fig. 5G). At this and later stages, the ovaries contained obvious
cysts with many tubule-like structures that resemble those in the
ovaries of inhibin-deficient female mice (Fig. 5, E–H). Ovaries from
some of the double-homozygous mutant female mice demonstrated bilateral
mixed granulosa/Sertoli cell tumors that were less invasive (Fig. 5E).
The spectrum of the histological features of these slow-growing and
less invasive gonadal tumors in the double-homozygous mutant male and
female mice is summarized in Table 2 and
can be compared directly to mice lacking inhibin alone (10).
|
). Consistent with these
results, the livers and glandular stomachs from these double-mutant
mice did not show any morphological or histological abnormalities (data
not shown). Whereas inhibin-deficient mice showed elevated levels of
serum estradiol, serum levels were suppressed in the double-homozygous
mutant mice (Table 3). In accordance with this, when total RNA prepared
from inham1/inham1 or
inham1/inham1,
fshbm1/fshbm1 female mouse ovaries was
hybridized with an aromatase probe, the levels were found to be
significantly down-regulated in the double-homozygous mutant mice
compared with inhibin-deficient mice (data not shown). Additionally,
comparison of the large-scale gene expression profiles in the ovaries
of wild-type and fshbm1/fshbm1 female mice
using an oligonucleotide-based gene chip assay revealed that activin
ßA and activin ßB mRNAs were present in wild-type ovaries at 23 and
31 copies per million total copies of mRNA, respectively, but the mRNA
for both subunits was undetectable in the knockout ovaries (M. C.
Byrne, P. Wang, T. R. Kumar, and M. M. Matzuk,
unpublished results). Furthermore, four of four double-homozygous
mutant male mice up to 14–16 months of age were fertile when
mated to wild-type females and produced viable double-heterozygous
mutant mice. The epididymal sperm number in these mice was comparable
to that in age-matched control wild-type mice (data not shown).
Together, these results confirm that absence of FSH leads to important
functional alterations in gonadal tumors that develop in
inhibin-deficient mice.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-subunit minigene or the
hormone-specific hFSHß subunit gene and then intercrossed these mice
to obtain hFSH-expressing mice. Although the mMT-1 promoter is active
in pituitary as confirmed by a RT-PCR performed on total RNA from
MT-hFSHß mouse pituitaries (data not shown), we have confirmed the
lack of expression of hFSHß protein in pituitary gonadotropes in
these mice by a simultaneous double-immunofluorescence technique using
rat LHß and hFSHß-specific antibodies (data not shown). Thus, these
data support our hypothesis that the infertility in the female hFSHß
(subunit) transgenic mice is caused somehow by the local accumulation
of the hFSHß subunit in the ovarian interstitium. Thus, the
mMT-1-driven hFSHß transgene was mainly ectopically expressed in
sites other than the normal site of FSH synthesis (i.e.
pituitary gonadotropes). In vitro biochemical data have demonstrated that the hFSHß subunit alone is not or inefficiently secreted out of the cell (17). In another assay, the free ß-subunit of glycoprotein hormones has been shown to compete with the heterodimeric hormones to bind their corresponding cognate gonadal receptors and inhibit hormone signaling (18). Recently, Markkula et al. (19) have shown the presence and/or expression of gonadotropin subunits in the ovary; however, the significance of this expression is unknown. Although we did not detect free hFSHß subunit in the serum of MT-hFSHß transgenic female mice, we noticed an accumulation of large amounts of hFSHß polypeptide in the ovaries of these mice (where mMT-1 promoter is also active). It is not clear whether this deposition interfered in an autocrine or paracrine way with the folliculogenesis and resulted in the complete infertility of these mice. However, the MT-hFSHß male mice were fertile, and therefore we could successfully generate the hFSH-transgenic mice and study the phenotypes of these mice.
Phenotypic Characteristics of hFSH-Overexpressing Mice
We have generated an hFSH-overexpressing mouse model to
study the consequences of elevated levels of FSH on reproductive
function and gonadal tumorigenesis. Our results clearly suggest that
ectopic production of hFSH in large quantities does not affect
testicular growth and differentiation including spermatogenesis. Male
MT-FSH transgenic mice were infertile and demonstrated increased
epididymal sperm number and enlarged seminal vesicles resulting from
elevated testosterone levels. Our data suggest that the elevated hFSH
levels, signaling possibly through LH receptors in the Leydig cells,
resulted in increased testosterone output.
Female transgenic mice expressing high levels of hFSH were infertile and developed hemorrhagic and cystic ovaries. They have kidney and urinary tract abnormalities secondary to elevated testosterone, estradiol, and progesterone levels in serum. These female mice died by 13 weeks of age as a result of urinary tract obstruction and had no signs of tumors. We do not know whether ovarian tumors would have eventually developed in these mice. However, our previous studies (14) with the weak hFSH expressor mice, in which serum hFSH levels (116 mIU/ml) are comparable to those in postmenopausal women, suggest that prolonged exposure to elevated FSH levels for more than 1 yr do not directly cause ovarian tumorigenesis. Female-specific characteristics of hFSH overexpression are kidney abnormalities and enlarged bladder with hFSH deposition. Although the exact mechanism for this is not clear, one possibility could be estrogen-stimulated aberrant glycosylation of liver-derived hFSH, which may give rise to insoluble acidic forms of hFSH. The majority of male transgenic mice (>95%) and all the female transgenic mice overexpressing hFSH were infertile. In males, the infertility could result from a failure of the functional competence of sperm or it could be caused by aberrant seminal vesicle secretions. The infertility in female transgenic mice is caused by disruption of normal folliculogenesis and the development of cysts in the ovary. As early as 2 weeks of age we could notice distinct gross morphological differences in the ovaries between wild-type controls and female transgenic mice. The ovarian and kidney phenotypes observed in the adult female transgenic mice are similar to those seen in transgenic mice in which expression of an LH analog is targeted to the pituitary (20). The elevated levels of serum testosterone in male and estradiol in female hFSH-transgenic mice could be caused by cross-talk of hFSH in large excess with the LH receptors. This suggests that elevated levels of either of the gonadotropins (i.e. LH or FSH) can result in pathological defects in the urogenital system.
Two independent lines of FSH-transgenic mice that differ from our present model were developed earlier. In one strain, bovine FSH expression was targeted to the mammary gland (21), using a modified rat ß-casein gene-based expression system. These mice expressed bioactive hFSH up to 60 IU/ml, vectorially in milk, but not in serum, at a much lower level compared with that in our hFSH-overexpressing transgenic mice. However, no physiological/pathological consequences of this ectopic expression in milk were reported. The second strain was developed, using 10 kb of hFSHß gene sequences (22). This hFSHß transgene was appropriately targeted to pituitary gonadotropes and contained the necessary GnRH and steroid-responsive elements (23, 24). Whereas there was a marginal increase in serum testosterone levels and testis weight in male transgenic mice, no differences were observed in female transgenic mice. The hFSHß mRNA in the pituitaries of this line of mice appears to be expressed at 3- to 4-fold higher levels than the endogenous mouse FSHß mRNA (M. J. Low, personal communication). Both male and female transgenic mice of this line were fertile with no additional abnormalities in other tissues (22).
MT-hFSH Transgenic Mice as Models of Human Reproductive
Disorders
The phenotypic characteristics in our hFSH-overexpressing
transgenic mice resemble, to some extent, known human reproductive
disorders. The infertility of hFSH-transgenic male mice could be the
result of a physical obstruction by the enlarged seminal vesicles at
the junction where the vas deferens opens into the urethra, thus
preventing the epididymal sperm ejaculation. However, the infertility
of hFSH-transgenic male mice could be caused by a reproductive
behavioral defect. Further analyses are required to confirm this. In
male patients who have pituitary adenomas secreting large amounts of
bioactive hFSH, no testicular phenotypes were observed (25). It is not
known whether or not these men were infertile. It is relevant to note
that inactivating mutations in either the FSHß subunit (5) or the
FSHR in mice (26) and men (27) do not affect male fertility.
In marked contrast to the male, the results obtained with our hFSH-overexpressing female mice more closely resemble clinical features in human patients. Cyst formation and hemorrhage are often associated with ovarian cancers in postmenopausal women who also have elevated serum hFSH levels (28). More striking similarity is seen when compared with patients who suffer from an ovarian hyperstimulation syndrome. These patients have high estradiol levels in serum, renal abnormalities including pyelonephritis in the kidney and oligouria, and massive hemorrhage and formation of cysts in the ovary (29). In addition, a premenopausal woman patient who had pituitary adenoma secreting high levels of bioactive hFSH has also been clinically documented (30). Her serum estradiol levels were found to be high, and the ultrasound scan revealed the presence of multiple ovarian cysts in the ovary (30). Although, hFSH levels are several fold higher in our transgenic mice compared with human patients, we believe hFSH-overexpressing mice are useful to study the molecular pathobiology of some of these clinical disorders. In addition, these mice may be useful for pharmacological testing of drugs that can block hemorrhage and cyst formation in the ovary.
Genetic Dissection of Gonadal Tumor Development in
Inhibin-Deficient Mice and the Role of FSH in Gonadal Sex
Cord-Stromal Tumor Development
To study the complex process of gonadal growth and
differentiation, we previously generated inhibin-deficient mice
using a gene-targeting strategy in embryonic stem cells. These mice
develop focally invasive gonadal sex cord-stromal tumors of granulosa
or Sertoli cell origin with 100% penetrance and eventually die due to
a severe wasting syndrome (4). Thus, inhibin is identified as a novel
secreted tumor suppressor specific to the gonads. The incidence of this
type of gonadal tumor in humans is rare (4), and it is not known how or
if mutations in the inhibin -subunit gene or in the inhibin signal
transduction pathway cause such cancers in humans. Similar to many
types of human cancers, the gonadal tumors in inhibin-deficient
mice arise as focal lesions, and not all of the granulosa or Sertoli
cells of the gonads demonstrate a malignant transformation and form
tumors. This indicates that other secondary event(s) are
necessary for malignant growth. To identify these modifier foci/factors
and to dissect out the individual components involved in the cascade of
events that lead to the formation of gonadal tumors in
inhibin-deficient mice, we have taken a genetic approach. We have
generated mice with multiple genetic lesions by selected crosses and
studied how the development/progression of gonadal sex cord-stromal
tumor is affected. For example, double-homozygous mutant male mice
deficient in inhibin and Müllerian-inhibiting substance
demonstrate synergistic effects of these two proteins in accelerating
Leydig cell neoplasia (31). Double-homozygous mutant mice that lack
inhibin and an activin receptor II showed continued growth of the
gonadal tumors, elevated levels of serum activin, and normal livers but
no characteristic cachexia (15). Therefore, activins, secreted from the
gonadal tumors and signaling through type II activin receptors, have
been implicated as mediators of cachexia symptoms in inhibin-deficient
mice (15). In addition, based on observations from Beamer’s group
(32), who reported an important role of androgens in granulosa cell
tumorigenesis, we generated male mice deficient in inhibin and a
functionally inactive androgen receptor. Analysis of these
double-mutant mice demonstrated that androgens do not influence gonadal
tumor development in inhibin-deficient male mice (33).
Contrary to what was observed in our inhibin/GnRH-deficient mouse
model, double-homozygous inhibin and FSH-deficient mice developed
gonadal tumors. However, there was a significant delay in the tumor
development/progression compared with mice deficient in inhibin alone.
Interestingly, there was little or no cachexia in these mice, and the
serum activin levels as well as the estradiol levels were suppressed.
This suggests that the tumors in these mice are functionally different
as they lack the trophic stimulation by FSH. The gender differences in
mechanism of FSH action on the gonads were reported previously (34).
This is reflected in the fact that female mice that lack both inhibin
and FSH did exhibit some differences compared with male mice that lack
these proteins. Male mice, but not female mice, deficient in both
inhibin and FSH were fertile and lived longer, and the gonadal tumors
histologically appeared to be less locally invasive. Inhibin/FSH
double-mutant mice demonstrate different phenotypes compared with our
previously characterized double-mutant mice, which lack both inhibin
and GnRH. This could be explained because absence of GnRH leads to
complete suppression of both LH and FSH, whereas LH is still present in
double-mutant mice that lack both inhibin and FSH. Granulosa and
stromal cell tumors were also observed in a proportion of transgenic
female, but not male, mice that overexpress either a bovine LHß
transgene or a bovine LHß-CTP-analog (fused to the bovine
-glycoprotein hormone subunit) in the pituitary (20). These
studies, along with our present genetic analyses, suggest that altered
gonadotropin ratios (i.e. LH/FSH) in the serum may be
important in gonadal tumorigenesis. We will generate mice deficient in
inhibin and LH to further distinguish the roles of LH and FSH in
gonadal tumorigenesis.
Mechanisms of cell cycle regulation by gonadotropins in gonadal cells are not clearly understood. Cyclin D2, an FSH-responsive cell cycle-regulatory gene, has been shown to be up-regulated in many human ovarian granulosa cell tumors (35), and cyclin D2-deficient female mice are infertile and display hypoplastic ovaries (36). The ovarian granulosa cells from these mutant mice do not proliferate both in vivo and in vitro in response to FSH, suggesting that the FSH signal transduction pathway is impaired. It will be interesting in the future to examine how specific cell cycle events in gonadal sex cord-stromal tumors are influenced by gonadotropins.
In conclusion, we have generated two different strains of transgenic mice: one in which hFSH expression from multiple tissues is directed by a mMT-1 promoter, and the other, a double-homozygous mutant that lacks both inhibin and FSH. These studies provide in vivo evidence to suggest that FSH is not directly involved in gonadal tumor formation but significantly influences the tumor progression in inhibin-deficient mice. These and previously generated gain-of-function and loss-of-function mice are important models for studying hemorrhagic cyst formation and sex cord-stromal tumor development. In addition, these mouse models will be useful in the future in formulating and testing a generalized mechanism of gonadal growth and differentiation.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-minigene (37). The same promoter was also fused to 5.2 kb of the
hFSHß (17) gene sequences. The hFSHß sequences, which were
engineered to start at -100 bp (after the HindIII site) in
the 5'-flanking sequences, contain all of the hFSHß exons and introns
and 1 kb 3'-flanking hFSHß sequences. The transgene fragments were
released from the vector backbone with appropriate restriction enzyme
digestions, purified, and microinjectioned into fertilized eggs to
produce transgenic mice (14).
Generation of Transgenic Mice
Independent lines of mice harboring either the MT- transgene
or MT-hFSHß transgene were separately generated by standard
pronuclear injections into fertilized eggs from C57BL/6/C3H x ICR
hybrid mice. Stable pedigrees of transgenic mice were obtained by
crossing Southern blot-positive founder (FO) mice to control wild-type
littermates. Mice expressing hFSH heterodimers were generated by
crossing MT- and MT-hFSHß lines of mice to produce double
transgene-positive mice. All animal studies were conducted in
accordance with the guidelines for Care and Use of Experimental
Animals.
Generation of
inham1/inham1,
fshbm1/fshbm1
Double-Mutant Mice
Generation of inham1/inham1 (10) and
fshbm1/fshbm1 (5) mice were as described.
Initially, fshbm1/fshbm1 male mice were bred to
inham1/+ female mice to obtain inham1/+,
fshbm1/+ double-heterozygous mice. These mice were later
intercrossed to obtain inham1/inham1,
fshbm1/fshbm1 double-homozygous mutant mice
at a 1:16 frequency. In addition, inham1/+,
fshbm1/fshbm1 male mice were also bred to
double-heterozygous mutant female mice to increase the frequency
of generating double-homozygous mutant mice to 1:8.
Southern Blot Analysis
For genotype analysis of the offspring, Southern blot analyses
were performed on tail DNA samples using 32P-labeled probes
as previously described (5, 10). Tail DNA samples from MT-
transgenic mice were screened with a 700-bp
HindIII-HindIII probe fragment, and MT-hFSHß
mice were screened with a 450-bp PstI-BamHI
hFSHß 3'-UTR (hFSHß-specific) probe. The identification of the
inham1 and fshbm1 mutant alleles in mice was as
described (5, 10).
Northern Blot Analysis
Total RNA was extracted from different tissues of wild-type,
MT-hFSH transgenic mice and from gonadal tumors of
inham1/inham1,
fshbm1/fshbm1 mice by the TRI-Reagent method
(38). RNA was denatured, separated on 1.4% agarose-formaldehyde gels,
and transferred to nylon membranes. The membranes were hybridized at 63
C with hCG, hFSHß, activin ßA, or activin ßB probes, washed,
and exposed to autoradiographic film as described (38). The blots were
stripped and rehybridized with an 18S probe as an internal control
(38).
RT-PCR
Total RNA was extracted from individual pituitaries of 10 kb
hFSHß transgene MT-hFSHß transgenic, and wild-type adult male mice
by the TRI-Reagent method (38). After isopropanol precipitation and air
drying of the RNA pellet, the RNA was solubilized in 6 µl diethyl
pyrocarbonate-treated water, and 2 µl of an aliquot from each
sample was subjected to a RT-PCR reaction using the ONE TUBE PCR kit
(Boehringer Mannheim, Indianapolis, IN) according to
the manufacturer’s instructions. The hFSHß-specific 3'-UTR primers
used in the reactions were: 5'-AAACACAACAATGGCTTCTT-3' (forward),
5'-ATTCCAAAGAAGTGGATCCT-3' (reverse). The amplified 450-bp fragment was
separated on a 2% agarose gel and visualized by ethidium bromide
staining.
Immunohistochemistry
Adult female wild-type or MT-hFSHß transgenic mice were
transcardially perfused with 4% paraformaldehyde in PBS (pH 7.2), and
the pituitaries and ovaries were collected and postfixed overnight at 4
C in the same fixative containing 10% sucrose, embedded in OCT medium,
and frozen on dry ice, and 16-µm sections were cut using a cryostat.
Simultaneous dual immunofluorescence was performed according to
previously published procedures (21) using a hFSHß-specific
monoclonal antibody (Medix, 1:500) and a guinea pig polyclonal
antiserum to rat LHß (NIDDK, 1:1000). The antigen-bound primary
antibodies were visualized by appropriate secondary antibodies
conjugated to either fluorescein isothiocyanate or rhodamine
isothiocyanate dyes.
Hormone Assays
Mice were Metofane anesthetized and exsanguinated by closed
cardiac puncture. Sera were collected and stored frozen at -20 C until
further use. Rat FSH and LH were iodinated by iodogen and chloramine-T
methods (21), respectively, and RIA was performed using NIDDK kits as
described (21, 39). hFSH (holoprotein) was measured by
fluoroimmunoassay using a Baxter automated fluoroimmunoassay detection
system according to the manufacturer’s instructions. Serum
testosterone (sensitivity = 0.1 ng/ml) and progesterone
(sensitivity = 0.3 ng/ml) were measured using solid-phase RIA
kits, and estradiol (sensitivity = 5 pg/ml) was quantitated using
an ultrasensitive liquid phase double-antibody assay kit according to
the manufacturer’s instructions. Serum IGF-I levels were measured
using a rat IGF-I RIA kit after acid-ethanol extraction as per the
instructions provided by the manufacturer. The enzyme-linked
immunosorbent assay for activin A was performed according to previously
published methods (40).
Evaluation of the Sperm Parameters
Epididymal sperm from adult male mice (6–7 weeks) were
collected into 1 ml M-2 medium by incubating at 37 C for 20 min. The
sperm number and motility were calculated using a hemocytometer at a
1:20 dilution. The viability of the sperm was determined by an eosin-Y
method at 1:40 dilution as described (5).
Histological Analysis
Testes and epididymides were either formalin or Bouin’s fixed
overnight and rinsed several times in LiCO3-saturated 70%
ethanol. Seminal vesicles, kidneys, bladders, and ovaries were fixed in
formalin overnight. The tissues were processed and paraffin embedded,
and 4-µm sections were cut and stained with PAS/hematoxylin reagents
as described (10).
Superovulation Experiment
Immature ICR strain female 24-day-old mice were injected with
PMSG (5 IU ip/mouse) and 48 h later with hCG (5 IU ip/mouse) and
mated with MT-hFSH-transgenic male mice as described (5). Vaginal plugs
were monitored the next morning to confirm matings.
Statistical Analysis
Statistical analysis was done by Student’s t test
using a Microsoft Corp. (Redford, WA) Excel
(version 6.0) software program. A P value <0.05 was
considered significant.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
These studies were supported in part by NIH Grants CA-60651 and HD-07495 to M.M.M.
Received for publication February 1, 1999. Revision received March 18, 1999. Accepted for publication March 19, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-Inhibin is a tumour-suppressor gene with gonadal specificity in
mice. Nature 360:313–319[CrossRef][Medline]
This article has been cited by other articles:
 |
|
 |
|
  S. A. Williams and P. Stanley Mouse fertility is enhanced by oocyte-specific loss of core 1-derived O-glycans FASEB J, July 1, 2008; 22(7): 2273 - 2284. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. P. Purdue, B. I. Graubard, S. J. Chanock, M. V. Rubertone, R. L. Erickson, and K. A. McGlynn Genetic Variation in the Inhibin Pathway and Risk of Testicular Germ Cell Tumors Cancer Res., April 15, 2008; 68(8): 3043 - 3048. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Q. Li, J. M. Graff, A. E. O'Connor, K. L. Loveland, and M. M. Matzuk SMAD3 Regulates Gonadal Tumorigenesis Mol. Endocrinol., October 1, 2007; 21(10): 2472 - 2486. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. A. Pangas, C. J. Jorgez, M. Tran, J. Agno, X. Li, C. W. Brown, T. R. Kumar, and M. M. Matzuk Intraovarian Activins Are Required for Female Fertility Mol. Endocrinol., October 1, 2007; 21(10): 2458 - 2471. [Abstract] [Full Text] [PDF]
|
|
