| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Pharmacology, Case Western Reserve University School of Medicine, Cleveland, Ohio 44106
 |
ABSTRACT |
|---|
 TOP ABSTRACT I. IntroductionII. Pathophysiology Induced by Elevated...III. ConclusionsACKNOWLEDGEMENTSREFERENCES |
|---|
)-subunit promoter to direct transgene expression specifically to gonadotropes in the anterior pituitary. Previously, it was shown that female LHßCTP mice are infertile due to anovulation, develop granulosa cell tumors, and undergo precocious puberty from elevated LH and steroid hormones that fail to completely repress the -subunit promoter. This chapter will discuss recent studies that further elucidate the impact of chronically elevated LH on diverse physiological systems. Granulosa cell tumors induced by elevated LH are strain dependent and prevented when transgenics are treated with human chorionic gonadotropin (hCG) surges. A granulosa cell tumor-associated transcriptome is generated, revealing several possible gene candidates for ovarian granulosa cell tumorigenesis. Primordial follicles in LHßCTP transgenics become depleted and oocytes exhibit increased rates of meiotic segregation defects, although meiotic competency is acquired normally. Anovulation can be rescued in transgenics by superovulation, though pregnancy fails at midgestation due to maternal factors. Uterine receptivity defects prevent implantation of normal embryos following induction of pseuodpregnancy. Transgenics develop Cushing-like adrenocortical hyperfunction with increased corticosterone production following induction of adrenal LH receptor expression. Elevated LH acts as a tumor promoter in the gonads and the adrenal gland, when expressed in conjunction with the inhibin- SV40 transgene. Finally, chronic elevated LH promotes mammary tumorigenesis. The understanding of multiple clinical pathologies — including ovarian cancer, perimenopausal reproductive aging, premature ovarian failure, polycystic ovarian syndrome, Cushing’s syndrome, and breast cancer — may be enhanced through further study of this useful transgenic mouse model. |
I. Introduction |
|---|
|
TOPABSTRACTI. IntroductionII. Pathophysiology Induced by Elevated...III. ConclusionsACKNOWLEDGEMENTSREFERENCES |
|---|
subunit and a unique ß subunit, which confers hormone specificity (Bousfield et al., 1994). LH is involved primarily in gametogenesis and steroid hormone production in both males and females.
To fully understand the function of LH physiologically, one must understand the context in which it exists. This context includes where it comes from, what drives its production, and what regulates its repression. The major system that controls reproductive physiology in mammals is the hypothalamic-pituitary-gonadal (HPG) axis. Gonadotropin-releasing hormone (GnRH), a decapeptide, is secreted in a pulsatile fashion from neurons located in the hypothalamus into the hypothalamic-hypophyseal portal system (Page, 1994). It then exerts its effect through GnRH receptors located on the surface of gonadotrope cells in the anterior pituitary. The seven transmembrane GnRH receptor is a G protein-coupled receptor that stimulates multiple signaling pathways, including modulation of intracellular calcium concentrations and activation of protein kinase C and other protein kinases (e.g., mitogen-activating protein kinase, MAPK) (Kraus et al., 2001). To what extent these paths cross-talk, or differentially regulate LHß subunit versus the gonadotropin subunit, is still under investigation (Gajewska and Kochman, 2001). GnRH signaling stimulates both the synthesis and secretion of the gonadotropins LH and FSH from the anterior pituitary.
LH exerts its effect by binding the G protein-coupled LH receptor, initiating signaling primarily through the second messenger, cyclic AMP (cAMP) (Ascoli et al., 2002). In females, LH stimulates ovarian thecal cell production of androstenedione, which diffuses to granulosa cells and undergoes conversion to estrogen by aromatase. FSH controls aromatase activity and also induces the appearance of LH receptors in the granulosa cells of preovulatory follicles, further augmenting estrogen production (Mathews et al., 1987). During the preovulatory surge, LH binds to LH receptors on granulosa cells, stimulating meiotic maturation of the oocyte and release of the oocyte from the follicle. Thus, granulosa cells with both FSH and LH receptors are capable of responding to LH directly, causing luteinization and release of progesterone (Gore-Langton and Armstrong, 1994). These steroids exert regulatory feedback onto the HPG axis by acting on the hypothalamus and gonadotrope cells, primarily in a negative (repressive) manner (Haisenleder et al., 1994). In males, the role of LH in spermatogenesis via stimulation of androgen production has been well established. Activation of LH receptors on testicular Leydig cells stimulates cAMP release, inducing synthesis of testosterone from cholesterol. In adults, LH-stimulated testosterone is the primary determinant of spermatogenesis (Sharpe, 1994). Due to the LHßCTP sexually dimorphic phenotype, the remainder of this review will focus primarily on the impact of LH hypersecretion in females.
A highly regulated system has evolved to ensure that LH is produced at appropriate concentrations, which, in females, change with time in a cyclic nature. Abnormalities leading to over- or underactivation of the HPG axis have extreme consequences on reproductive health (de Roux and Milgrom, 2001). Too little LH results in underdevelopment of the gonads and subsequent decline (or cessation) of gametogenesis. Too much LH causes precocious puberty and excessive steroid hormone production, which can lead to cell immortalization and tumorigenesis. In humans, inappropriate LH regulation can result in inability to produce a genetically related family, a hardship that is not only physiologically difficult to overcome but also potentially psychologically debilitating. Thus, it is important that we understand the physiologic role of LH and its ability to induce reproductive pathology upon dysregulation.
B. GENERATING LH HYPERSECRETION IN LHßCTP TRANSGENIC MICE
Many approaches can be taken to generate mice with elevated serum levels of LH. However, most of these methods rely on pharmacological means to elevate LH to nonphysiological levels. In addition, it is difficult to devise protocols for chronic administration of exogenous LH that mimic endogenous LH pulse patterns. By using a transgenic approach, our laboratory not only achieved a physiological means of elevating LH but also limited its synthesis and secretion strictly to gonadotrope cells in the anterior pituitary. This approach leaves intact the HPG axis and permits studies that depend on the responsiveness of this axis. Studying the impact of chronically elevated LH in the whole animal permits a physiological method of disease modeling in multiple, diverse systems, as described in this review.
The LHßCTP transgenic mouse was generated by designing a transgene (depicted in Figure 1) that incorporates the C-terminal peptide (CTP) of the hCGß subunit fused to the C-terminus of the bovine LHß cDNA and its first intron. The 87-bp CTP fragment extends the half-life of the resulting LHßCTP fusion product by 2- to 3-fold (Risma et al., 1995). An SV40 polyadenylation signal was added at the C-terminus of the transgene. Transgene expression is driven by the 360-bp (-315 to +45) bovine gonadotropin -subunit promoter, which has been previously characterized and targets expression specifically to gonadotrope cells in the anterior pituitary (Kendall et al., 1991; Hamernik et al., 1992). This promoter retains responsiveness to GnRH, estrogens, and androgens in transgenic mice (Keri et al., 1991; Hamernik et al., 1992; Clay et al., 1993).
|
5-fold over control females by 2 weeks of age; estradiol also is elevated by 3 weeks of age by 2- to 5-fold (Risma et al., 1995,1997).
Despite pathologically elevated levels of steroid hormones, the bovine gonadotropin -subunit promoter driving expression of the LHßCTP transgene does not become repressed. In contrast, the endogenous mouse LHß-subunit gene becomes completely silent during the transition from the neonatal period to adulthood (Abbud et al., 1999). Interestingly, the endogenous mouse gonadotropin -subunit gene remains active in addition to the transgene. Abbud et al. showed that -subunit promoter activity in the presence of elevated steroid hormones is due to independence from GnRH regulation and, as a result, a lack of responsiveness to estradiol-negative feedback. Since this bovine -subunit promoter previously was shown to be responsive to GnRH and estradiol in transgenic studies using a chloramphenicol acetyl transferase (CAT) reporter gene (Keri et al., 1991; Hamernik et al., 1992), loss of -subunit promoter responsiveness to GnRH and estradiol must be mediated by the pathological effects of exposure to high levels of LH early in life, possibly through the reprogramming of key signal transduction cascades.
Female LHßCTP transgenic mice are infertile due to chronic anovulation, while males appear to have normal fertility. However, male founder transgenics appeared to be slightly subfertile due to breeding delays and had significantly smaller testes than controls (Risma et al., 1995). Transgenic females enter puberty precociously at 21 days of age vs. 30 days in controls (Risma et al., 1997). By 3 weeks of age, transgenic females have accelerated folliculogenesis and multifollicular ovaries. By 5–6 weeks of age, transgenic ovaries become significantly enlarged and contain multiple fluid- or blood-filled cysts (Risma et al., 1997). Prolonged luteal life span with elevated serum progesterone is induced in transgenics following hemiovariectomy and mating with vasectomized males. Morphological analyses of ovaries from older animals reveal ovarian granulosa cell tumors or theca-interstitial cell tumors by 4 months of age (Risma et al., 1995).
Transgenic females develop renal abnormalities — including enlarged bladders, dilated ureters, and hydronephrosis — sometimes associated with acute pyelonephritis that can lead to death (Risma et al., 1995). This renal pathology probably is due to exposure to chronically elevated steroids, since hydronephrosis has been observed in rats chronically administered estradiol (Corriere and Murphy, 1968) and is well documented during pregnancy in women (Andriole, 1975). Interestingly, this renal pathology appears to occur most frequently in CF-1 female transgenics but only rarely in [C57BL/6xCF-1] F1 hybrids (Nilson et al., 2000). Thus, female transgenics on C57BL/6 background live longer and go on to develop pituitary hyperplasia by 4 months of age, with pituitary adenoma formation by 10–12 months of age (Nilson et al., 2000).
|
II. Pathophysiology Induced by Elevated LH |
|---|
|
TOPABSTRACTI. IntroductionII. Pathophysiology Induced by Elevated...III. ConclusionsACKNOWLEDGEMENTSREFERENCES |
|---|
|
To estimate the number of recessive genes involved in granulosa cell tumor strain dependency, backcrosses between the C57BL/6xCF-1 F1 transgenic hybrids (
) and CF-1 nontransgenics (
) were performed. Of the 138 transgenic females generated from this study, 19 harbored at least one granulosa cell tumor. The fit between the predicted and observed (13.8%) frequency of tumor appearance supports a three-gene model (Keri et al., 2000). The expected (predicted) frequency of tumor formation for two-, three-, and four-gene models are, respectively, 25%, 12.5%, and 6.25% (predicting 34.5, 17.3, and 8.6 tumors out of 138 evaluated). Subsequent to this publication, more transgenic females have been evaluated, altering the final observed frequency to 30/295 (10.2%), further supporting that three unlinked, recessive genes may discriminate tumor susceptibility in response to elevated LH (R.A. Keri, personal communication).
Ongoing studies to determine the identity of the genes required for granulosa cell tumors are being performed using microsatellite markers and bulk segregant analysis to identify contributing loci. Recent preliminary data indicate phenotype linkage to mouse chromosomes 10 and 11 (R.A. Keri, personal communication).
2. LH Surges Prevent Granulosa Cell Tumor Formation
While granulosa cell tumors comprise only 10% of ovarian tumors, they exhibit the potential for malignancy and recurrence, making them clinically significant (Wynder et al., 1969, Fontanelli et al., 1998; Lee et al., 1999). Efforts to identify the molecular mechanisms leading to development of these tumors are important, as they may reveal possible new therapeutic targets.
Clinical studies have implicated elevated gonadotropins levels in tumorigenesis. The strongest correlative evidence comes from postmenopausal women, who represent the largest cohort of patients with granulosa cell tumors (Amsterdam and Selvaraj, 1997).
Studies using other types of transgenic mice support this view as well. Mice deficient in inhibin develop granulosa cell tumors at an early age (Matzuk et al., 1992). The lack of inhibin leads to an increase in LH, FSH, and estrogens (Kumar et al., 1999). When these animals also are made genetically deficient for GnRH (hpg/hpg), tumors fail to form, further underscoring the importance of LH and FSH (Kumar et al., 1996). In contrast, mice deficient in FSH and inhibin continue to develop tumors, although with increased latency and decreased penetrance (Kumar et al., 1999). These data suggest that elevated LH, independent of FSH, can induce granulosa cell tumor formation when mice also are deficient for inhibin , while FSH may accelerate this process (Owens et al., 2002).
In Owens et al. (2002), subtractive gene expression profiling was employed, comparing normal ovaries, LHßCTP (CF-1) granulosa cell tumors, and LHßCTP (F1 hybrids) luteomas. This subtractive method (diagrammed in Figure 3) was employed to evaluate the differential gene expression profile that distinguishes a granulosa cell tumor from a normal or luteoma-bearing ovary. Equal amounts of total RNA were pooled from at least four mice for each experimental group, to minimize changes due to interindividual variation. Global changes in gene expression profiles were assessed with Affymetrix Mu11K oligodeoxynucleotide microarrays containing approximately 11,000 genes and expressed sequence tags (ESTs). Samples were normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and analyzed using self-organizing maps (SOMs) to reveal informative patterns of gene expression (Tamayo et al., 1999). The data were filtered using Affymetrix parameters to exhibit changes that were at least 3-fold, since these are highly reproducible in replicate samples (Cho et al., 1998). Clusters were generated of 47 genes whose expression dramatically decreased and 75 whose expression increased significantly.
|
|
|
Owens et al. further evaluated the role of elevated LH in granulosa cell tumor development by determining whether the lack of ovulatory surges of LH in genetically predisposed LHßCTP (CF-1) mice might contribute to the granulosa cell tumor phenotype. LHßCTP CF-1 mice were treated with ovulatory doses of hCG every fourth day for 5 months, beginning at 2 weeks of age. Transgenic female littermates in control groups received saline injections. All transgenic animals (n = 4) receiving hCG injections failed to develop granulosa cell tumors while, as expected, all littermate controls (n = 3) receiving saline injections developed tumors by 5 months of age (Owens et al., 2002). Ovaries from all four hCG-treated animals developed luteomas, a histological phenotype indistinguishable from those seen in F1 hybrid transgenics (C57BL/6xCF-1). This result indicates that long-term restoration of ovulatory surges of hCG is capable of antagonizing the genetic predisposition for granulosa cell tumor formation in CF-1 transgenic mice. Interestingly, the ovarian transcriptome of CF-1 transgenic mice treated with hCG closely resembled that of the F1 hybrid transgenics (C57BL/6xCF-1). These results are shown in the right column of Tables I and II.
Figure 4 depicts the finding that chronically elevated LH initiates a molecular pathway that, in conjunction with a genetic predisposition, leads to the development of granulosa cell tumors. However, in the absence of this genetic predisposition, or in the presence of ovulatory-like surges of hCG, chronically elevated levels of LH result in the formation of a luteoma rather than a granulosa cell tumor. Further molecular clarification of the signaling pathways involved in these two different outcomes may help refine our understanding of human granulosa cell tumors as well as the molecular pathways regulated by LH.
|
Accelerated primordial follicle depletion also is seen in the disorder known as premature ovarian failure (POF). Although of unknown origin, POF clearly is associated with early and accelerated follicle atresia (Rebar, 1982; Cohen and Speroff, 1991). POF is characterized by primary or secondary amenorrhea, with elevated levels of serum gonadotropins and early menopause (Bione et al., 1998). Women with POF are infertile due to anovulation, experience depletion of their follicular reserve, and show hormonal profiles similar to older women who are postmenopausal (Russell, 1997).
In Flaws et al. (1997), ovaries from LHßCTP mice were evaluated for follicular depletion. By scoring serial sections of ovaries obtained from transgenics and control littermates, a 45% decline in primordial follicles was observed in transgenics by 5 weeks of age. By 3 months, ovaries from transgenic mice contained 68% fewer primordial follicles, when compared to controls (Figure 5). As the primordial follicle pool becomes depleted, primary follicle numbers begin to decline as well. By 3 months, primary follicles are 45% depleted in transgenics vs. controls (Flaws et al., 1997). These data suggest that elevated LH leads to depletion of the primordial and primary follicle pool and thus alters the timing of reproductive senescence. Further evaluation of the molecular mechanism of this depletion may increase our understanding of primordial pool depletion observed during perimenopause and POF.
|
Meiotic competency, or the ability to resume meiosis after release from the ovarian follicle, is indicated by germinal vesicle breakdown (GVBD), followed by polar body (PB) extrusion after overnight culture. No differences could be observed between transgenics or control littermates: at 15 days of age, oocytes arrest at GVBD, while at 21 days, meiotic competency has been acquired and oocytes extrude a polar body, presumably arrested at metaphase of meiosis I (Mann et al., 1999). Thus, despite development within the context of elevated LH, estrogen, and testosterone, and the depletion of primordial follicles, acquisition of meiotic competency in oocytes from transgenics overtly appears to remain unaltered.
3. Elevated LH Increases the Rate of Oocyte Meiotic Segregation Defects
While acquisition of meiotic competency occurs normally in transgenics, elevated LH does significantly impact the organization of chromosomes on the meiotic I spindle during oocyte maturation (Hodges et al., 2002). This study suggests that the altered endocrine milieu of LHßCTP transgenics results in an increase in meiotic nondisjunction, similar to that observed during human reproductive aging.
Errors in chromosome segregation during human meiotic divisions result in the loss of a significant number of conceptions. The high incidence of meiotic errors in humans has a strong correlation to maternal age (Hassold and Chiu, 1985), though the mechanisms underlying these errors remain unclear. Data show that the age-related increase in oocyte meiotic errors is characteristic of all racial groups (Hassold and Chiu, 1985) and that most meiotic errors have their genesis at meiosis I (Hassold and Hunt, 2001). Studies on human trisomies have further elucidated the impact of age on meiotic nondisjunction but have not revealed the molecular mechanisms involved (Hodges et al., 2002). Recently, immunofluorescent studies have suggested that early events in meiosis I may contribute to subsequent nondisjunction (Battaglia et al., 1996; Volarcik et al., 1998).
Congression failure, or the inability of chromosomes to move to the equator of the meiosis I (MI) spindle due to gross aberrations in spindle morphology and/or chromosome alignment, is another age-related phenomenon (Volarcik et al., 1998). Since female meiosis (unlike mitosis or male meiosis) is not subject to chromosomal monitoring checkpoint mechanisms of the metaphase/anaphase transition, this alignment failure may be linked to eventual nondisjunction (LeMaire-Adkins et al., 1997; Woods et al., 1999; Burke, 2000; Shah and Cleveland, 2000).
Because the altered endocrine environment is known to disrupt folliculogenesis in the LHßCTP female, Hodges et al. evaluated oocytes from LHßCTP transgenics (and other mouse models) for evidence of congression failure and subsequent indications of nondisjunction. On both a heterogeneous background and the C57BL/6 inbred strain, oocytes from the first follicular wave in LHßCTP were analyzed for chromosome alignment at metaphase I (6 hours in culture) and metaphase II (16–18 hours in culture). Table III shows the results of these studies (Hodges et al., 2002). A high level of congression failure was observed at both MI (18.2% vs. 0–1% in controls) and meiosis II (MII) metaphase (38.6% vs. 0% in controls) in LHßCTP mice. In addition, meiotically immature oocytes from nontransgenics (partially competent 18- to 20-day-old mice) did not exhibit congression failure. These studies show that congression failure is not a symptom of meiotic immaturity but, rather, a reflection of oocyte growth in an altered environment. It also was observed that the degree of meiotic disturbance appeared to correlate to the extent of ovarian pathology, another indicator of the abnormal endocrine environment.
|
Hodges et al. also asked if congression failure causes an increase in oocyte aneuploidy through nondisjunction and premature sister chromatid segregation (PSCS). Table IV shows that LHßCTP-derived oocytes do exhibit a significant increase in total hyperploidy and/or PSCS (18% vs. 5% controls). These data, along with data from another congression failure mutant (not shown), support the conclusion that increased frequency of congression failure at MI results in increased impairment of segregation of homologous chromosomes at anaphase I. These results suggest that changes in the endocrine milieu that occur during human reproductive aging and the perimenopausal period may contribute significantly to the age-related increase in oocyte meiotic nondisjunction, possibly due to errors in chromosomal alignment at MI, such as those seen during congression failure.
|
Women with elevated LH who are diagnosed with PCOS have difficulty conceiving, frequently requiring the use of fertility medications to induce ovulation. Unfortunately, those who do conceive experience an elevated miscarriage rate of 30–64% (compared to a rate of 12% in women with normal LH) (Hamilton-Fairley and Franks, 1990; Regan et al., 1990). Because this increased miscarriage rate in women with PCOS often is attributed to poor oocyte quality (Brzyski et al., 1995), we evaluated oocyte, embryo, and maternal reproductive health in LHßCTP transgenics (Mann et al., 1999).
As discussed previously, oocytes from LHßCTP transgenic mice acquire meiotic competency normally, despite early exposure to elevated LH and androgens. Subsequent studies revealed abnormalities in LHßCTP meiotic chromosomal segregation (Hodges et al., 2002). It is possible that a pool of these oocytes would develop abnormally and result in pregnancy failure if evaluated. However, when superovulation was used to recruit a large group of oocytes, which then were fertilized and transferred to pseudopregnant, nontransgenic recipients, they developed into normal, live-born pups.
LHßCTP females are infertile primarily because of anovulation, without ovulating spontaneously, though they will mate, as exhibited by the presence of repetitive vaginal plugs. However, both a single dose of hCG and a superovulatory regimen of pregnant mare serum gonadotropin (PMSG), followed by hCG 48 hours later, induces LHßCTP ovulation. This suggests that chronically elevated LH causes a loss of normal LH ovulatory surges. Following superovulation and mating, LHßCTP transgenics will become pregnant but they experience pregnancy failure at midgestation. By 12 days post coitus (dpc), transgenic pregnancy resorption approaches 100%, whereas nontransgenic resorption reaches only 21% (elevated slightly from normal, presumably due to large litters from superovulation). The rate of pregnancy loss is plotted in Figure 6.
|
Levels of LH, estradiol, testosterone, and progesterone were measured throughout superovulation-induced pregnancy and compared to pregnant nontransgenics. Progesterone levels were not significantly different from nontransgenics following superovulation, suggesting that corpus luteum failure was not responsible for the midgestation pregnancy resorption phenotype. Testosterone does not become elevated until 14 dpc, after the onset of pregnancy failure. Interestingly, estradiol becomes elevated 4-fold over nontransgenics from 8–14 days gestation, during the critical window of pregnancy failure. Elevated estradiol during midgestation has been shown to be toxic to embryos (Huggett and Pritchard, 1945). Elevated estradiol causes midgestation pregnancy failure in mice lacking the 5-reductase type I gene (Mahendroo et al., 1997). Thus, it is possible that excess estradiol is the cause of LHßCTP midgestation pregnancy failure. Interestingly, estrogen toxicity has been used to explain some human miscarriages (Trout and Seifer, 2000). LHßCTP pregnancy defects could be rescued, with live pups born when animals were ovariectomized following embryo transfer and treated with estrogen and progesterone during pregnancy. While inefficient, the success of this regimen demonstrated that the reproductive abnormalities experienced by LHßCTP females are reversible (Mann et al., 1999).
This study suggests that pregnancy failure in women with elevated LH, such as those diagnosed with PCOS, may be attributed to a hostile maternal environment contributing to lack of uterine receptivity and/or pregnancy loss. Future studies to determine the molecular mechanisms involved in LHßCTP pregnancy failure may help us to further understand the complexities behind increased miscarriage rates in women with PCOS.
D. ADRENAL DISTURBANCES: MODELING CUSHING’S SYNDROME
Although most of the androgens in PCOS are secreted from the ovaries, in approximately 50% of cases, excessive production of adrenal androgens also occurs (Azziz, 1996). The underlying cause of this adrenocortical disturbance remains unclear.
The primary regulator of glucocorticoid production in the adrenal cortex is adrenocorticotropin (ACTH) (Orth and Kovacs, 1998). In humans, ACTH drives the adrenal gland to produce the glucocorticoids cortisol and corticosterone. However, corticosterone is the only glucocorticoid produced in the mouse (Spackman and Riley, 1978).
In addition to ACTH, several other hormones (e.g., LH, prolactin, insulin-like growth factor-1 (IGF-1)) have been implicated in the regulation of adrenal androgen production (Orth and Kovacs, 1998). There is some evidence that LH acts directly at the adrenal gland, although this is controversial (Parker and Odell, 1980). A case of postmenopausal Cushing’s syndrome has been described in which the adrenocortical hyperfunction was found to be LH dependent and responsive to treatment with GnRH agonist (Lacroix et al., 1999). In addition, a significant proportion of women with chronic anovulation have elevated serum levels of LH and adrenal-derived androgen precursors such as dehydroepiandrosterone-sulfate (DHEA-S) but normal ACTH levels (Hoffman et al., 1984).
In Kero et al. (2000), adrenal gland function and LH receptor expression were evaluated in LHßCTP transgenics. They presented evidence for a novel mechanism that might explain the adrenal-derived androgens in PCOS and some LH-dependent Cushing’s syndrome cases.
Adrenal glands from 5-month-old transgenic females were evaluated and found to be morphologically altered, compared to nontransgenic littermates. Adrenal glands from transgenics weighed significantly more than controls (6.0 ± 0.86 mg vs. 3.3 ± 0.3 mg, respectively). Histological signs of cortical stimulation included increased cortical width and centripetal extension of lipid-depleted cells, widening the zona reticularis. There were also foci of acute and chronic inflammatory cells (Kero et al., 2000). Notably, lymphocyte infiltration of the adrenal gland has been observed in some patients with Cushing’s syndrome (Willenberg et al., 1998).
While at 1 month of age, both transgenics and controls had a typical area of the andrenal X-zone visible, by 5 months of age, the X-zone had disappeared in transgenics. A poorly characterized area, the X-zone is thought to be involved in steroid hormone production (Hu et al., 1999). Interestingly, in normal male mice, the X-zone disappears after puberty, possibly due to testosterone (Asari et al., 1979). However, in normal females, the X-zone remains until after the first pregnancy (Holmes and Dickson, 1971).
To determine whether the altered adrenal morphology reflected a change in adrenal steroidogenesis, serum corticosterone levels were measured and determined to be elevated 14-fold in female transgenics, when compared to nontransgenic controls. This elevation decreased to nearly undetectable levels following adrenalectomy, indicating that the primary source of the corticosterone was the adrenal gland. In addition, transgenic ovariectomy resulted in decreased corticosterone levels to that of nontransgenic controls. Interestingly, gonadectomy, which induces elevated LH in both males and females, did not increase corticosterone levels in nontransgenics. This suggests that the polycystic ovaries found in transgenics play a significant role in adrenal hyperfunction and that chronically elevated LH alone cannot account for the increased adrenal steroidogenesis (Kero et al., 2000).
The steroidogenic capacity of andrenal glands from 3-month-old trangenics was assessed further through culturing dispersed adrenal cells. Unstimulated cells from transgenic females produced elevated levels of cAMP, progesterone, and corticosterone, relative to nontransgenics. In addition, transgenic-derived adrenal cells responded to both hCG (in a dose-dependent manner) and ACTH treatment. However, cells from nontransgenics responded only to ACTH. These data indicate that, in addition to ACTH, an LH analog stimulates adrenal corticosterone production in transgenics. This increased steroidogenic capacity is associated with increased LH receptor expression in the transgenic adrenal gland. Significant and specific LH receptor binding in adrenal homogenates was observed in transgenic females. RT-PCR was used to demonstrate that long-term elevation of serum LH concentrations is associated with the presence of LH receptor mRNA in adrenal glands. LH receptor expression was observed in female transgenics by 3 months of age and in gonadectomized nontransgenics after 5 months (although at lower levels) (Kero et al., 2000). In situ hybridization localized LH receptor expression in transgenic adrenal glands across the whole cortex, including the zona glomerulosa, zona fasciculata, and zona reticularis. Thus, the presence of the cystic ovary, in conjunction with elevated LH, seems to be a requirement for achieving significant adrenal LH receptor expression along with functional responsiveness of adrenal cells to hCG.
Besides serving as a model for PCOS, these findings of increased corticosterone production in LHßCTP transgenics suggest that these mice may provide a useful model for Cushing-like adrenocortical hyperfunction (Kero et al., 2000). Other manifestations of Cushing’s syndrome (e.g., obesity) have been observed in these mice. LHßCTP mice become obese, weighing 30% more than their nontransgenic littermates (R.A. Keri, J. Kero, I.T. Huhtaniemi, J.H. Nilson, unpublished data). Interestingly, in rare cases of ACTH-independent Cushing’s syndrome, the disease becomes more prominent during pregnancy and may improve or remit spontaneously after delivery (Da Motta et al., 1991; Buescher et al., 1992; Close et al., 1993).
The mechanism involved in the development of this Cushing-like adrenocortical hyperfunction is unclear. It is known that LHßCTP mice have 2- to 3-fold elevated circulating estradiol (Risma et al., 1995). Estradiol is known to elevate prolactin, which is upregulated in transgenics by 2.5-fold (Kero et al., 2000). Prolactin is an important regulator of LH receptor expression at both the level of transcription and translation (Huhtaniemi and Catt, 1981; Gafvels et al., 1992; Pakarinen et al., 1994). This may explain the ovarian dependence for induction of adrenal LH receptor expression and subsequent stimulation of corticosterone production by elevated LH seen in transgenics. This represents a novel mechanism behind the altered adrenocortical function observed in transgenics, which results in elevated glucocorticoid production. As these mice appear to model some aspects of both PCOS and Cushing’s syndrome, further studies on adrenocortical function are warranted in humans with chronically elevated gonadotropin levels (Kero et al., 2000).
E. LH ACTS AS A TUMOR PROMOTER IN LHßCTP/INHTAG DOUBLE TRANSGENICS
Previous studies on mice transgenic for the inhibin- promoter directing expression of the SV40 T-antigen showed that 100% of these mice developed granulosa cell or Leydig cell tumors by 5–8 months of age (Kananen et al., 1995,1996a). When gonadectomized, these mice developed adrenal gland tumors with 100% penetrance by 6–8 months (Kananen et al., 1996b). While the adrenal tumor dependency on gonadectomy suggested a role for increased levels of gonadotropins — and, indeed, was shown to be gonadotropin dependent (Kananen et al., 1997) — these studies did not differentiate between the actions of FSH and LH. Thus, a new line of double transgenics was created to introduce the LHßCTP transgene onto the inhTag line (Kero et al., 2002).
In the double-transgenic mice, gonadal tumorigenesis starts earlier and progresses faster, compared to inhTag transgenics. At 3 months of age, LHßCTP female ovaries are enlarged and multicystic, while ovaries from double transgenics are even larger and present with highly proliferating tumors of granulosa cell origin. Ovaries from nontransgenics and inhTag transgenics are normal at this time point. The presence of elevated LH in female double transgenics enhanced the aggressiveness of their ovarian tumors. While gonadal tumors in inhTag transgenics never metastasized to other tissues, double transgenics were observed to generate metastases to the lungs and liver.
Interestingly, these studies revealed some previously unseen phenotypes in male LHßCTP transgenics. While serum LH levels originally were not observed to be elevated in males (Risma et al., 1995), this study found slight, but significant, elevation of LH at 9–12 weeks of age (5.18 ± 0.45 µg/mL vs. 1.35 ± 0.34 µg/mL). Additionally, some Leydig cell hyperplasia was observed. As previously identified, testicles from LHßCTP males were slightly, but significantly, reduced in size (Risma et al., 1995). This can be explained by the known effects of elevated LH with concomitantly low FSH before puberty, leading to early differentiation of testicular cells and reduced testicular size (Gaytan et al., 1994). Three-month-old male double transgenics had increased testicular weight and showed massive hemorrhagic and invasive tumors originating from Leydig cells, while inhTag transgenic males showed only microscopic tumor foci at this age.
Histological signs of adrenal gland tumorigenesis in all double-transgenic females were observed in the presence of the gonad. These tumors clearly originated from the adrenal cortex. In contrast, adrenal tumors were observed in inhTag transgenics only after gonadectomy. Not surprisingly, no evidence of adrenal tumorigenesis was seen in male double transgenics, presumably due to their only slightly elevated LH levels. These findings suggest that chronically elevated LH, in the presence of normal to decreased FSH, can act as a tumor promoter on the inhTag transgenic background.
The progression of ovarian tumorigenesis in double transgenics is associated with elevated LH receptor and inhibin- gene expression. Interestingly, high serum levels of LH fail to induce downregulation of LH receptor in both LHßCTP and double transgenics. Indeed, LH receptor is upregulated, indicating that the normal desensitization of LH receptor by LH is altered in these animals (Dufau, 1998). Chronic activation of the LH receptor function could act as a protooncogene, as shown with some other G protein-coupled receptors (Allen et al., 1991; Parma et al., 1993). It is possible that the enhanced tumorigenesis seen in double transgenics could be explained by LH receptor stimulation of increased cAMP production and subsequent activation of the cAMP response element (CRE) contained within the inhibin- transgene promoter (Pei et al., 1991).
F. MAMMARY TUMORIGENESIS: MODELING BREAST CANCER
Mammary gland development is a hormonally regulated process. Progression of the rudimentary mammary ductal system present at birth to an extensive network following puberty and finally into a differentiated, milk-producing organ during pregnancy and lactation is mediated by estrogen, progesterone, and prolactin (Hennighausen and Robinson, 1998). These hormones also regulate the occurrence and timing of reproductive events that influence a woman’s risk of developing breast cancer. While many mouse models of mammary cancer exist, few address the role that hormones play in spontaneous tumor formation.
Known risk factors for development of breast cancer include early age at menarche, late age of menopause, late age at first full-term pregnancy, or nulliparity (Kvale, 1992; Stoll et al., 1994; Armstrong et al., 2000). Interestingly, parity has a dual effect on breast cancer. While full-term pregnancy actually reduces long-term tumor incidence in humans and rodents (MacMahon et al., 1982; Grubbs et al, 1986), a short-term increased risk in tumor formation has been observed immediately following pregnancy (Hsieh et al., 1994; Lambe et al., 1994).
In Milliken et al. (2002), the impact of LH-mediated alterations in the hormonal milieu of LHßCTP transgenics on the development of the mouse mammary gland was investigated. Initial studies revealed that mammary gland development is accelerated in LHßCTP transgenics. Precocious puberty occurs in transgenics around postnatal days 21–22, at least 5 days earlier than nontransgenic littermates (Risma et al., 1997; Keri et al., 2000). To assess the impact of early ovarian activity on mammary gland morphogenesis, Milliken et al. examined glands at 3 and 5 weeks of age. By 3 weeks of age, accelerated development of the mammary gland was observed in transgenic mice (Figures 7A and B). This accelerated growth of the transgenic mammary gland was more apparent by 5 weeks of age (Figures 7C and D). While the ductal network in nontransgenics progressed through only half of the mammary gland fat pad by 5 weeks of age, the entire fat pad of the transgenic mouse was filled with ducts that displayed abundant alveoli and a loss of terminal end buds (Milliken et al., 2002). This pattern of development closely resembles that seen at mid- to late pregnancy in normal mice (Hennighausen and Robinson, 1998).
|
Ovariectomy was performed to determine whether the mammary hyperplasia observed in transgenics was due to direct action of LH or whether it required LH-induced ovarian hyperstimulation. Mammary glands collected after 21 days postovariectomy showed complete regression of the gland, compared to sham surgery controls. These results indicate that hyperplasia is reversible and requires ovary-derived factors.
While transgenics develop spontaneous mammary tumors (mostly mammary intraepithelial neoplasias (MINs)) with 50% penetrance by 41 weeks (Figure 8), treatment with the mammary carcinogen, 7,12-dimethylbenz(a)anthracene (DMBA) accelerates tumor formation. Transgenics treated with DMBA develop invasive mammary carcinomas with squamous metaplasia beginning at 13.5 weeks posttreatment (with 100% penetrance by 20 weeks), while only 20% of nontransgenic controls treated with DMBA developed tumors by 56 weeks postexposure. These data demonstrate that LHßCTP transgenics are predisposed to mammary carcinogenesis, when compared to their nontransgenic littermates.
|
It is known that growth of mammary tumors can be classified as either hormone dependent or independent. Analysis of human breast tumors has revealed a significant correlation between expression of estrogen and progesterone receptors and hormone growth dependency (DePlacido et al., 1990; Ormandy et al., 1997). Immunohistochemical analysis of the spontaneous mammary tumors formed in transgenics for progesterone receptor (PR) indicated that while normal tissue adjacent to tumor cells did express nuclear PR, tumor cells did not. Unfortunately, detection of estrogen receptor (ER) expression was not possible at that time.
These studies have demonstrated that chronic exposure to elevated LH leads to precocious mammary gland development and ovary-dependent mammary hyperplasia. Hyperplasia is due to an increase in epithelial cell proliferation. It will be important to determine whether precocious puberty in LHßCTP transgenics significantly contributes to their development of mammary tumors, considering that early puberty is a risk factor for breast cancer in humans (Apter et al., 1989).
Despite histological and molecular markers suggesting the mammary glands are in a midpregnancy-like state, LHßCTP transgenics have accelerated tumor formation. This finding contrasts with studies indicating a protective effect of pregnancy against mammary tumor formation. The apparent discrepancy may suggest that the protection achieved through pregnancy is regulated by the extent and timing of hormone elevation.
Further studies are being performed to dissect the molecular mechanisms that govern the hormone-induced mammary tumorigenesis seen in this model. Gene-expression profiling has revealed many genes that behave similarly in transgenic and pregnant mammary glands, as compared to virgin nontransgenics. However, there are also many genes that behave differentially between the transgenic and pregnant glands. Understanding these similarities and differences may reveal novel mechanisms by which to develop therapies to treat and prevent human breast cancer.
|
III. Conclusions |
|---|
|
TOPABSTRACTI. IntroductionII. Pathophysiology Induced by Elevated...III. ConclusionsACKNOWLEDGEMENTSREFERENCES |
|---|
The diversity of physiological systems ultimately impacted by LH hypersecretion is revealed through the LHßCTP transgenic mouse model. Future studies investigating the downstream signaling pathways and molecular mechanisms involved in the generation of these LH-induced pathologies should support the notion that the LHßCTP transgenic mouse model may be a useful resource for both transcriptome and proteome profiling.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTI. IntroductionII. Pathophysiology Induced by Elevated...III. ConclusionsACKNOWLEDGEMENTSREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTI. IntroductionII. Pathophysiology Induced by Elevated...III. ConclusionsACKNOWLEDGEMENTSREFERENCES |
|---|
-subunit gene to gonadotropin-releasing hormone and estrogens. Mol Endocrinology
13
: 1449
–1459 subunit gene by androgen may involve direct binding of androgen receptor to the proximal promoter. J Biol Chem
268
: 13556
–13564-subunit genes is regulated by distinct cis-acting elements. Mol Endocrinol
6
: 1745
–1755-subunit gene despite the absence of a high affinity binding site for estrogen receptor. Mol Endocrinol
5
: 725
–733-subunit promoter/Simian virus 40 T-antigen fusion gene. Oncogene
, in press-reductase type I caused by estrogen excess. Mol Endocrinol
11
: 917
–927This article has been cited by other articles:
 |
|
 |
|
  M. Bielinska, S. Kiiveri, H. Parviainen, S. Mannisto, M. Heikinheimo, and D. B. Wilson Gonadectomy-induced Adrenocortical Neoplasia in the Domestic Ferret (Mustela putorius furo) and Laboratory Mouse. Vet. Pathol., February 1, 2006; 43(2): 97 - 117. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Bielinska, E. Genova, I. Boime, H. Parviainen, S. Kiiveri, J. Leppaluoto, N. Rahman, M. Heikinheimo, and D. B. Wilson Gonadotropin-Induced Adrenocortical Neoplasia in NU/J Nude Mice Endocrinology, September 1, 2005; 146(9): 3975 - 3984. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. L. Britt, P. K. Saunders, S. J. McPherson, M. L. Misso, E. R. Simpson, and J. K. Findlay Estrogen Actions on Follicle Formation and Early Follicle Development Biol Reprod, November 1, 2004; 71(5): 1712 - 1723. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. F. Couse, M. M. Yates, R. Sanford, A. Nyska, J. H. Nilson, and K. S. Korach Formation of Cystic Ovarian Follicles Associated with Elevated Luteinizing Hormone Requires Estrogen Receptor-{beta} Endocrinology, October 1, 2004; 145(10): 4693 - 4702. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
