Human Molecular Genetics, 2001, Vol. 10, No. 11 1141-1153
© 2001 Oxford University Press
Persistent Prop1 expression delays gonadotrope differentiation and enhances pituitary tumor susceptibility
1Department of Human Genetics, University of Michigan Medical School, Ann Arbor, MI 48109-0638, USA and 2Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN 55905, USA
Received 13 December 2000; Revised and Accepted 23 March 2001.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
The ‘paired’-like homeodomain transcription factor Prop1 is essential for the expansion of the pituitary primordia and for the differentiation and/or function of the hormone-producing cells of the anterior pituitary gland. Prop1 expression is normally extinguished before transcription of most differentiation markers is initiated. We report that constitutive expression of Prop1 interferes with anterior pituitary cell differentiation and increases the susceptibility for pituitary tumors. The terminal differentiation of pituitary gonadotropes is delayed, resulting in transient hypogonadism and a delay in the onset of puberty. Thyrotrope differentiation occurs normally, but thyrotrope function is impaired resulting in mild hypothyroidism. Aged mice exhibit defects consistent with misregulation of pituitary cell proliferation, including adenomatous hyperplasia with the formation of Rathke’s cleft cysts and tumors. Thus, silencing Prop1 is important for normal pituitary development and function. These data suggest that gain-of-function mutations in PROP1 could contribute to the most common human pituitary endocrinopathies and tumors.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Pituitary gland dysfunction, manifested by reduced hormone production, results from both developmental defects and tumor formation. Alterations in anterior pituitary cell proliferation and differentiation are responsible for congenital hormone deficiencies in approximately 1 in 4000 live births (1) and for pituitary adenomas in up to 30% of the general population (2). Genetically engineered and spontaneous mouse mutants provide important model systems in which to examine the regulatory mechanisms that may be involved in these disorders. The molecular mechanisms that direct cell proliferation and differentiation in the pituitary gland include the secretion of inductive signals from neighboring structures as well as the expression of a cascade of homeodomain transcription factors (3,4). The combination of these events leads first to the development of the pituitary primordium, Rathke’s pouch. Subsequently, intense proliferation occurs in the presumptive anterior lobe. This period is followed by differentiation into the five cell types of the mature gland (5). By birth, the anterior pituitary is composed of functioning corticotropes, thyrotropes, somatotropes, gonadotropes, and lactotropes (6). These cells, which are defined by the hormones they secrete, produce adrenocorticotropic hormone (ACTH), thyroid-stimulating hormone (TSH), growth hormone (GH), gonadotropins follicle-stimulating hormone (FSH) and luteinizing hormone (LH), and prolactin (PRL), respectively.
One of the genes whose expression is essential for proper pituitary gland ontogeny is Prop1, which encodes a ‘paired’-like homeodomain transcription factor (7). The expression of the murine Prop1 gene is restricted to the developing pituitary gland. Its expression is detectable at embryonic day 10.5 (E10.5), peaks at E12.5, and decreases after E14.5, prior to the onset of most hormone gene transcription. No Prop1 expression is evident at or beyond birth. Mutations in the mouse Prop1 and human PROP1 genes reveal the necessity for appropriate Prop1 expression during pituitary gland organogenesis. Ames dwarf mice (Prop1df/df) exhibit growth insufficiency, hypothyroidism and infertility which result from a serine
proline substitution at residue 83 in the murine Prop1 gene (7). This conserved residue lies in the first 
-helix of the homeodomain and, when mutated, results in a marked decrease in the DNA binding ability of the PROP1 protein. This hypomorphic mutation leads to the development of an extremely hypocellular pituitary gland with a profound lack of thyrotropes, somatotropes and lactotropes, as well as reduced gonadotrope function (7–9). In humans, more than 11 different loss-of-function and null mutations have been identified in the PROP1 gene in patients with combined pituitary hormone deficiency (CPHD) (10–12). The majority of patients with PROP1 mutations lack GH, TSH, PRL, FSH and LH. Most of these cases are congenital, although late-onset hypopituitarism, hyperpituitarism, progessive hormone loss and acquired ACTH deficiency have been documented (13,14). The basis for the variability in age of onset and severity of CPHD in patients carrying PROP1 mutations is not understood and cannot be explained by the type of PROP1 mutation (15). However, loss of PROP1 function in humans and mice generally has the same effect, namely congenital deficiencies of GH, TSH, PRL, FSH and LH. This suggests that the mouse is an excellent mammalian model for human pituitary development and disease.
The control of gene expression during pituitary gland organogenesis is a dynamic process. Proper development requires both gene activation and gene repression at the appropriate times (11). Fetal-specific expression of the transcription factors Prop1 and Rpx (or Hesx1) exemplifies this temporal control of gene expression in the embryonic pituitary (7,16). Studies in mutant mice reveal that normal pituitary gland organogenesis requires the timely activation and repression of both Prop1 and Rpx during gestation. In addition, Prop1 is necessary for the extinction of Rpx, as well as the activation of Pit1, another transcription factor gene in this pathway (17,18).
The interactions between Rpx and Prop1 suggested that Pit1 expression might be necessary for Prop1 extinction and that Prop1 may be persistently expressed in Pit1-deficient animals (3). We disprove this hypothesis by demonstrating a normal developmental profile of Prop1 expression in Snell dwarf animals (Pit1dw/+ versus Pit1dw/dw). We demonstrate the importance of appropriate silencing of Prop1 expression using transgenic mice that express Prop1 constitutively throughout Rathke’s pouch and in the adult pituitary thyrotropes and gonadotropes. These animals have delayed gonadotrope differentiation, transient hypogonadism, mild hypothyroidism, and an enhanced susceptibility for pituitary adenomas resulted. Thus, Prop1 is not only essential for early developmental events including the expansion of Rathke’s pouch, but its repression is important for differentiation, function and controlled proliferation of specialized pituitary cells.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Prop1 is extinguished in Pit1dw/dw mutants
PROP1 loss-of-function mutants have prolonged expression of the ‘paired’-like homeodomain gene Rpx (or Hesx1) and failure to activate the POU homeodomain transcription factor Pit1 (16,18). Thus, Prop1 expression is essential to repress Rpx and activate Pit1. To test whether Pit1 expression is necessary for the appropriate repression of Prop1 gene expression in late gestation, we performed in situ hybridization with a Prop1 riboprobe in wild-type and Snell dwarf (Pit1dw/dw) mice. Snell dwarf mice have a loss-of-function mutation in the POU homeodomain of the Pit1 gene that eliminates DNA binding (19). If PIT1 function is required for the appropriate extinction of Prop1, then Prop1 expression would be extended in the Snell dwarf animals compared with wild-type controls. Peak expression of Prop1 at E12.5 is equivalent in Pit1dw/+ and Pit1dw/dw mice (Fig. 1, top panels). The decline in expression evident at E14.5 and E16.5 is also indistinguishable between these two groups of animals (Fig. 1, bottom panels and data not shown), indicating that Pit1 is not necessary for Prop1 repression.
|
The
GSU-Prop1 transgene is expressed at high levels in the anterior pituitaryTo assess the effects of persistent Prop1 expression, 10 lines of
GSU-Prop1 transgenic mice were generated. The transgene consists of a 2.2 kb genomic Prop1 fragment under the control of a 5 kb mouse GSU promoter (Fig. 2A). A quantity of 4.6 kb of the GSU promoter is sufficient for high levels of developmentally regulated and cell-specific expression (20,21). This promoter drives high levels of transgene expression from E9.5 onwards. In early development, transgenes are expressed throughout the anterior and intermediate pituitary primordia, whereas expression in the adult gland is restricted to the thyrotropes and gonadotropes. Six of the 10 lines generated had germline transmission of the GSU-Prop1 transgene (for a description of breeding, see Materials and Methods). To identify lines that express the GSU-Prop1 transgene at a high level in the adult pituitary gland, in situ hybridization was performed using a Prop1 riboprobe. As expected, no Prop1 expression is detected in an adult non-transgenic control (Fig. 2B). The lines with the highest levels of GSU-Prop1 transgene expression in the anterior lobe of the adult pituitary gland (D4 and D6) were selected for all subsequent experiments (Fig. 2B). Despite constitutive Prop1 expression, GSU-Prop1 transgenic mice appear to develop normally. There are no obvious differences in the physical aspects or growth of the transgenic mice compared with non-transgenic littermates. No obvious defects in fertility exist in the male transgenics used to maintain the lines.
|
To demonstrate that the
GSU-Prop1 transgene is expressed at a level high enough to exert a biologic effect on the animal, the transgene was crossed onto the Ames dwarf background (Prop1df/+). Transgenic Prop1df/+ mice were backcrossed to Prop1df/+ animals and the F2 progeny were analyzed for growth and thyrotrope differentiation. Due to deficiencies in TSH and GH, Ames dwarf (Prop1df/df) mice are approximately one-third of the size of their wild-type littermates (22). Treatment of Prop1df/df mice with supraphysiologic doses of TSH increases body weight by 
45%, resulting in mice that are 50–70% the size of wild-type controls (22). The residual growth insufficiency is due to the lack of GH. If the GSU-Prop1 transgene could promote thyrotrope expansion and function, transgenic Ames dwarf animals would exhibit an increase in body size similar to that observed in TSH-treated Prop1df/df mice, as well as an increase in thyrotropes, compared with non-transgenic dwarf controls.
At weaning, transgenic Prop1df/df mice are approximately one-third the size of normal mice (Prop1+/+ and Prop1df/+) and the same size as their non-transgenic dwarf counterparts (Prop1df/df). However, as the mice age, the growth of the transgenic dwarf mice improves relative to non-transgenic dwarf littermates. By 5 months of age, the body size of the transgenic dwarf mice (n = 6) is 
30% greater than non-transgenic dwarf controls (n = 6) (Fig. 3A). This increase in size in transgenic dwarf animals is an increase in linear growth and not an increase in body fat, suggesting that the transgene corrected the thyrotrope defect.
|
Pituitary thyrotropes normally constitute
5% of the total anterior pituitary cells (23). In contrast, the pituitaries of Prop1df/df mice have occasional clusters of Tshß-positive cells representing <1% of the normal number of thyrotropes (Fig. 3B) (8). To compare the thyrotrope populations in transgenic and non-transgenic Prop1df/df mice, in situ hybridization using a Tshb riboprobe was performed. The thyrotrope population is substantially expanded in the pituitaries of GSU-Prop1 transgenic dwarf animals relative to non-transgenic controls (Fig. 3C). Thus, the GSU-Prop1 transgene corrected the block in thyrotrope expansion, enhancing TSH production and resulting in a growth increase in transgenic dwarf mice. As the transgene is not expressed in the somatotropes, the remaining growth insufficiency can be attributed to a deficiency in GH. These data demonstrate that the transgene in line D4 is being expressed at a level that has biological consequences.
Persistent Prop1 expression results in delayed gonadotrope differentiation
To assess the need for tightly regulated Prop1 expression during pituitary gland development, the anterior pituitary cell populations of transgenic and non-transgenic mice were compared at birth. In situ hybridization using probes for each pituitary hormone was performed on the pituitary glands of GSU-Prop1 transgenic mice at postnatal day 1 (P1). High levels of Prop1 expression were confirmed in GSU-Prop1 transgenics at P1 whereas no expression was detected in non-transgenic controls (Fig. 4A and B). As the transgene is expressed only in the thyrotropes and gonadotropes of the mature anterior pituitary, it might not alter the cell specification and/or proliferation of the somatotropes, corticotropes or lactotropes. Consistent with this expectation, the expression of Gh and Pomc is the same in GSU-Prop1 transgenic and non-transgenic neonates (data not shown). In addition, Tshb expression in the neonatal transgenics is similar to that of the non-transgenic controls (Fig. 4C and D). Although no obvious effects on the thyrotrope population were noted at P1, adult GSU-Prop1 transgenic mice exhibit mild hypothyroidism. GSU-Prop1 transgenics, aged from 4 weeks to 1 year, have decreased T4 levels compared with age-matched, non-transgenic controls [2.4 ± 0.205 µg/dl (n = 10) versus 4.2 ± 0.594 µg/dl (n = 9), respectively; P = 0.0074, unpaired t-test].
|
In contrast to the other hormone markers, expression of both Lhb and Fshb is nearly absent in the anterior pituitaries of
GSU-Prop1 transgenic neonates (Fig. 4E–H). LHß, FSHß and the gonadotropin-releasing hormone receptor (GnRH-R) are the three definitive markers of the gonadotrope lineage, with Lhb and Fshb expression marking terminal differentiation (24). An absence of all three markers indicates a lack of gonadotrope cells. A lack of Lhb and Fshb expression, with normal Gnrhr expression, suggests that gonadotrope precursors are present but have not fully differentiated. Gnrhr in situ hybridization revealed the presence of Gnrhr transcripts in GSU-Prop1 transgenic neonates, although at a reduced level compared with the non-transgenics (Fig. 4I and J). This indicates that some precursors to the gonadotrope lineage are present in the pituitaries of transgenic mice at P1 but that gonadotrope differentiation is developmentally delayed. Thus, persistent Prop1 expression interferes with the activation of Lhb and Fshb gene transcription during pituitary gland development.
Despite the delay in gonadotrope differentiation, GSU-Prop1 transgenic males are fertile. This suggests that gonadotrope differentiation must be completed by adulthood. To confirm this, LHß immunostaining was performed on pituitary glands from 4-week-old GSU-Prop1 transgenic mice. At this age, the LHß-expressing gonadotrope cell population in the transgenic pituitaries does not differ from that of non-transgenic controls (Fig. 4K and L). This indicates that gonadotrope differentiation and gonadotropin production are established in GSU-Prop1 transgenic mice within the first 4 weeks of life.
Development of the gonads is delayed in GSU-Prop1 transgenic animals
Adequate levels of gonadotropin secretion are required for the timely onset of puberty (25). In females, the development of a vaginal opening is indicative of the onset of puberty (26). This generally occurs at the age of 4 weeks, although premature expression of gonadotropins accelerates the onset of puberty by 2 days (27,28). In addition, reduced gonadotropins cause a variable delay in the opening of the vagina (29,30). All of the non-transgenic females analyzed (n = 4) had easily detectable vaginal openings at 4 weeks of age. Only two of five GSU-Prop1 transgenic females had vaginal openings at the same age. At 5 weeks of age, one of the GSU-Prop1 transgenic females still had no vaginal opening. These data indicate that, due to the delayed gonadotropin production, the onset of puberty in GSU-Prop1 transgenic mice is variably delayed (0 to >7 days).
Expression of Lhb and Fshb, and secretion of LH and FSH, are essential for the postnatal development and function of the ovaries and testes. Delayed gonadotropin production has the expected physiological effect on the gonads of pubertal and adult transgenic animals. Four-week-old GSU-Prop1 transgenic females (n = 4) have smaller ovaries and thinner uteri than their non-transgenic counterparts, consistent with hypogonadotropic hypogonadism (Fig. 5A). Histological analysis of the ovaries from these mice demonstrates that, despite the decrease in ovary size, the development of follicles in the GSU-Prop1 transgenics is similar to that of the non-transgenic controls (Fig. 5C and D). In both sets of animals, primordial, primary and secondary follicles are present throughout the ovary. Mature, Graafian follicles are absent in normal prepubertal females and in the transgenic mice. Four-week-old transgenic males (n = 2) have smaller testes compared with non-transgenic controls (Fig. 5B). Histological analysis of these testes revealed that the transgenic males have delayed sperm development and reduction in the size of the seminiferous tubules (Fig. 5E and F). Spermatids and mature sperm cells are present in the seminiferous tubules of normal mice but absent in most tubules of the transgenics.
|
The
GSU-Prop1 transgenic animals were also examined at the onset of sexual maturity. At 6 weeks of age, both the ovaries and testes of the transgenic mice appear histologically the same as those of their non-transgenic littermates. In the females, follicular development in the transgenic mice parallels that of the non-transgenic mice, with mature, Graafian follicles present in both (Fig. 5G and H). Spermatogenesis in GSU-Prop1 transgenic males has caught up with that of their non-transgenic counterparts by 6 weeks. Mature sperm cells, along with developing spermatocytes and spermatids, are present in both sets of mice (Fig. 5I and J). This rapid recovery of gonad development evident in 4- and 6-week-old transgenic animals is consistent with a recovery of gonadotropin gene expression in the anterior pituitaries, despite the persistent expression of the transgene.
GSU-Prop1 transgene expression increases pituitary adenoma susceptibility
To assess the long-term consequences of persistent Prop1 expression, GSU-Prop1 transgenic mice were aged 1 year. Whole pituitary glands and pituitary sections from the transgenic animals were examined for morphological abnormalities. Eight transgenic mice from the D4 line and six transgenic mice from the D6 line were analyzed and compared with seven non-transgenic controls. Both lines of GSU-Prop1 transgenic mice had an increased propensity for histological changes relating to cell proliferation, including adenomatous hyperplasia, tumors, massive Rathke’s cleft cysts and pyknotic nuclei. In addition, some of the pituitaries had hypertrophied cells, some of which demonstrated ‘signet ring’ changes, with the nuclei pushed to the edge of the cell by the expanded cytoplasm (31). Thirteen of 14 transgenic animals (7/8 D4 mice and 6/6 D6 mice) examined demonstrated histopathologic changes including tumors, massive cysts and extensive cellular hypertrophy and hyperplasia. In contrast, none of the non-transgenics exhibited these changes. Three of the 7 non-transgenics displayed minor histopathologic changes, including microcysts and minimal cellular hypertrophy.
Three out of eight D4 transgenic pituitaries exhibited gross anatomical changes which distorted the lobular structure of the pituitary (Fig. 6A and data not shown). Analysis of sections from these pituitaries confirmed the presence of a pituitary tumor in the anterior lobe of each gland. To further characterize these tumors, sections were immunostained with antibodies against each of the pituitary hormone markers. All three tumors were negative for TSHß, LHß, PRL, GH, ACTH and GSU expression (Fig. 6B and data not shown). This indicates that the tumors which result from persistent Prop1 expression produce little or no pituitary hormones. In humans, tumors which have no hormone immunoreactivity are referred to as null cell adenomas (32,33).
|
Expression of the
GSU-Prop1 transgene also leads to an increased susceptibility for hyperplasia, hypertrophy and Rathke’s cleft cysts. Rathke’s cleft cysts are generally identified by their location within the pituitary, in the remnants of Rathke’s cleft between the intermediate and anterior lobes. Both symptomatic and asymptomatic Rathke’s cleft cysts have been reported in humans (34,35). Of seven D4 transgenic pituitaries, three developed large Rathke’s cleft cysts (Fig. 7B and data not shown), one had regions with numerous pyknotic nuclei (Fig. 7C), and six demonstrated widespread hyperplasia and hypertrophy (Fig. 7D), including the presence of ‘signet ring’ cells. Of six D6 transgenic pituitaries, one exhibited focal thyrotrope hyperplasia (Fig. 6C), one had a small cyst and six demonstrated extensive cellular hypertrophy including ‘signet ring’ cells. In contrast, of the seven non-transgenic controls, two had microcysts, two had areas of minor hypertrophy and none developed hyperplasia or tumors.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
In the murine pituitary, Prop1 expression is restricted to the period when undifferentiated precursors are proliferating (5,7). No expression is detectable in neonates or adult animals. PROP1 deficiency in both Ames dwarf mice and humans with CPHD causes multiple hormone deficiencies as well as pituitary hypoplasia (7,12,36). In humans, pituitary hypoplasia is diagnosed by magnetic resonance imaging, whereas hormone deficiency is assessed by measuring circulating levels in the serum. The accessibility of tissue in Ames dwarf mice has permitted the investigation of PROP1 deficiency at a molecular level, during both gestation and adulthood. Taken together, studies involving Ames dwarf mice and CPHD patients demonstrate that PROP1 function is required for the expansion of undifferentiated precursor cells in the pituitary primordium and of thyrotropes, somatotropes and lactotropes in the adult anterior lobe (7,8). In addition, PROP1 clearly plays a role in the gonadotrope lineage in humans and is important for normal gonadotropin production in mice (9,10). The role of Prop1 during pituitary gland ontogeny includes appropriate repression of Rpx and activation of Pit1 (17,18). The decrease in Prop1 expression when Pit1 is turning on suggested Pit1 as a likely candidate for the repressor of Prop1 expression (3). However, we have shown here that Prop1 expression does not persist in Pit1-deficient Snell dwarves, indicating that PIT1 is not required to extinguish Prop1 (Fig. 8). Thus, the mechanism for silencing Prop1 expression remains to be determined. Prop1 target genes, which probably include transcription factors that are involved in the specialization of cells within both the Pit1 and gonadotrope lineages, are candidates for Prop1 repressors.
|
Although the effects of PROP1 deficiency on anterior pituitary cell differentiation are clear, it was unknown whether persistent expression of Prop1 had deleterious effects on pituitary cell specification or proliferation. We demonstrate here that extinguishing Prop1 expression is as critical for normal pituitary function as the activation of Prop1 transcription. Constitutive expression of Prop1 in the gonadotropes and thyrotropes of the pituitary interferes with the differentiation and function of these cells.
Two lines of GSU-Prop1 transgenic mice express the transgene at levels that are sufficient to exert a biologic effect on the animal. Transgene expression corrects the portion of the growth defect in Ames dwarf (Prop1df/df) mice that is due to TSH deficiency by increasing the population of TSH-producing thyrotropes. The subsequent increase in TSH production results in an increased body size compared with non-transgenic dwarf controls. On the normal mixed genetic background (Prop1df/+ and Prop1+/+), the GSU-Prop1 transgene does not result in obvious defects in growth. However, thyrotrope function must be compromised in GSU-Prop1 transgenic animals, since adult transgenics exhibit hypothyroidism. The thyrotropes in GSU-Prop1 transgenic mice may be resistant to thyrotrophin-releasing hormone stimulation or are less responsive to feedback regulation by thyroid hormone. This suggests that elevated levels of PROP1 could be a contributing factor to adult onset hypothyroidism in humans, a common disorder affecting 
10% of mature women (37).
Constitutive Prop1 expression leads to a delay in the terminal differentiation of anterior pituitary gonadotropes. LHß, FSHß and GnRH-R are the main markers of the gonadotrope lineage in the anterior lobe of pituitary, with gonadotropin gene expression identifying terminally differentiated gonadotropes (24). While transcripts can be detected as early as E16 and E17 in wild-type mice (6), expression of the gonadotropin genes Lhb and Fshb is not detectable in the pituitaries of GSU-Prop1 transgenic mice at P1. In contrast, transcripts of another gonadotrope marker, Gnrhr, are evident in the transgenic neonates at reduced levels. The lack of Lhb and Fshb expression, and the presence of Gnrhr expression, suggests that GnRH-responsive gonadotrope precursors are present in the pituitary glands of transgenic neonates but that these precursors have yet to differentiate into mature LH- and FSH-producing gonadotropes.
Several genetically engineered mouse models support the existence of gonadotrope precursor cells. Mice that carry an SV40 T antigen transgene under the control of a 1.8 kb human GSU promoter develop tumors that consist solely of gonadotrope precursors (24,38). These tumors, and cell lines derived from them, express Gsu and Gnrhr. They do not, however, synthesize either LHß or FSHß. Mice that lack functional EGR1, a zinc finger transcription factor, express Gsu, Gnrhr and Fshb, but not Lhb (39,40). These models suggest that gonadotropin gene transcription represents the final step of differentiation in the gonadotrope lineage.
The delay in gonadotrope differentiation which occurs in GSU-Prop1 transgenic mice causes a delay in gonad development and puberty. At 4 weeks of age, the ovaries and testes of transgenic animals are much smaller than those of their non-transgenic littermates. Spermatogenesis is delayed in transgenic males, while follicle maturation in the ovaries of transgenic females is similar to that of non-transgenic controls. By 6 weeks of age, gonad development in both transgenic males and females has recovered and sexual maturity is achieved on the same schedule as that of normal mice.
The rapid recovery of GSU-Prop1 transgenic mice from neonatal gonadotropin deficiency may involve feedback regulation of target organs on the pituitary gland. The gonads, thyroid gland and adrenal glands all exhibit feedback control on the production of anterior pituitary hormones (41). These target organs send hormone signals that affect hypothalamic inputs to the pituitary and exert direct effects on the pituitary gland by regulating pituitary hormone gene transcription, hormone secretion and cell proliferation. These regulatory mechanisms allow the pituitary to respond to the continually changing needs of the organism. Specifically, LH and FSH production is enhanced by GnRH stimulation and suppressed by gonadal steroid hormone feedback. Chronic reduction in gonadal steroids causes gonadotrope hypertrophy and very limited hyperplasia (42,43). Prolonged GnRH supplementation can overcome a gonadotropin deficiency caused by disruption of the orphan nuclear hormone receptor SF1 and induce terminal gonadotrope differentiation (44). The compensation for persistent Prop1 expression which allows the hypogonadal mice to become sexually mature and fertile probably involves gonadal feedback regulation at the level of the hypothalamus and the pituitary gland.
The homeodomain transcription factor gene Otx1 may also be involved in the neonatal compensation for persistent Prop1 expression. Otx1-deficient mice also exhibit a transient delay in gonadotropin expression and gonad development (45). OTX1 is made in the pituitary gland from birth through adulthood, and it transactivates the Gh, Gsu, Lhb and Fshb genes. Young Otx1-deficient mice (3–4 weeks) lack LHß, FSHß and GH and exhibit hypogonadism and dwarfism. The receptors for both GnRH and GHRH are unaffected in these mice, indicating that the gonadotropic and somatotropic lineages are present. By 4 months of age, the Otx1–/– mice have completely recovered from their hormone deficits. Thus, activation of Otx1 at birth may override the effects of persistent Prop1 expression in the Prop1 transgenic mice. We propose that PROP1 is required for initiation of gonadotrope differentiation, by stimulating the proliferation of gonadotrope precursors, and that PROP1 must be extinguished for gonadotropes to complete the differentiation process between E14.5 and P1. OTX1 is critical for the maintenance of gonadotrope function from birth to 4 months, and some other unknown factor must predominate from 4 months through the rest of the life of the animal. The dependence of gonadotropes on different transcription factors at different times in development was not expected, but it parallels the cascade of transcription factors required for muscle differentiation at each stage of development (46).
Constitutive Prop1 expression has deleterious effects in aging mice as well. GSU-Prop1 transgenic mice exhibit increased susceptibility to cysts, cellular hypertrophy, hyperplasia and pituitary tumors. This variety of histopathologic changes is consistent with the multiple ‘hits’ hypothesis of tumor formation (47,48). Constitutive Prop1 expression may cause an increase in cell proliferation which enhances the probability of a tumor-initiating mutation. The spectrum of abnormalities observed could represent the progression of transgene-expressing cells from hypertrophy to hyperplasia to adenomatous hyperplasia and subsequent tumor development as additional hits are accumulated. This type of progression has been observed in genetically engineered TSH-deficient mice (42,43) (M.L. Brinkmeier and S.A. Camper, unpublished data).
Pituitary tumors are the most common type of intracranial tumor found in humans (49). Despite the high incidence of pituitary tumors, the genetics of pituitary adenoma initiation and progression are poorly understood (50). Null cell adenomas and prolactinomas are the two most common types of human pituitary tumors. Null cell adenomas present with mass effects, including headaches and visual disturbances, as opposed to endocrinopathies, because they do not secrete hormones. In this respect the pituitary tumors in GSU-Prop1 transgenic animals are analogous to human null cell adenomas. The human tumors are thought to result either from uncontrolled proliferation of early, undifferentiated pituitary precursor cells or incompletely differentiated gonadotropes (49,51). Both of these ideas are consistent with the critical role of PROP1 in stimulating the expansion of undifferentiated precursor cells in Rathke’s pouch and in influencing gonadotrope differentiation and function. Several recent studies have detected PROP1 expression in both human pituitary adenomas and normal adult pituitary tissue, although quantitative studies have not yet been done (52–55). The transgenic mice we report here suggest that chronic expression of the PROP1 homeobox gene could be an initiating event in the etiology of null cell adenomas.
Aged GSU-Prop1 transgenics exhibit an increased propensity for the formation of Rathke’s cleft cysts. Two transgenic mouse models of Rathke’s cysts have been generated by expressing leukemia inhibitory factor (LIF) under the control of two different pituitary-specific promoters (Gsu and Gh) (56,57). LIF is a cytokine that functions within the pituitary to regulate hormone gene transcription and cell proliferation. In vitro, LIF expression inhibits embryonic stem cell differentiation (58). These two lines of transgenic mice developed Rathke’s cysts as a result of constitutive LIF expression. It is possible that the expression of Prop1 in the GSU-Prop1 transgenics activates LIF, or other differentiation-inhibiting factors, resulting in the formation of Rathke’s cleft cysts.
The overall phenotype of the GSU-Prop1 transgenic mice demonstrates the importance of regulated homeobox gene expression during organogenesis. Although there are many examples of inappropriate homeobox gene expression that enhance tumor susceptibility in other organ systems, none have been reported for the pituitary gland (59). Loss and gain of function models for PAX2, another ‘paired’-like homeodomain transcription factor, suggests its role in kidney development is similar to the role of PROP1 in pituitary development (60–62). Hypomorphic or null mutations in Pax2 lead to proliferative defects at sites of early Pax2 expression, including an absence of metanephric kidneys, hypoplastic kidneys and malformations of the metanephric mesenchyme (63–65). Human mutations in PAX2 result in renal-coloboma syndrome with various degrees of kidney defects (66). Persistent expression of Pax2 also leads to kidney defects in mice (67) and patients with juvenile cystic and dysplastic kidneys (68). These two precedents suggest that inappropriate activation of homeobox gene expression is as deleterious as a deficiency in gene function.
In summary, unregulated expression of the Prop1 homeobox gene causes delayed differentiation and reduced pituitary cell function, as well as defects in the control of cell proliferation. The phenotype of the GSU-Prop1 transgenic mice suggests the possibility that the misregulation of PROP1 expression in humans may account for some endocrine deficiencies, including hypothyroidism and hypogonadism, and may increase pituitary adenoma susceptibility.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Transgene construction
To generate the
GSU-Prop1 transgene, a 2.4 kb fragment containing the Prop1 coding sequences was PCR-amplified from 129sV/J genomic DNA using BamHI-linkered primers (5'-AAG GGA TCC AGA AGC AGA GAA AGA CTT CGG GG-3' and 5'-GCG GGA TCC ACC CCC GTT TCT AGG TTC ACC CT-3'). The reaction was performed for 10 cycles of denaturation at 92°C for 15 s, annealing at 55°C for 30 s and extension at 72°C for 90 s, followed by 15 cycles of denaturation at 92°C for 15 s, annealing at 55°C for 30 s and extension at 72°C for 120 s, with a final extension at 72°C for 10 min. Approximately 200 ng of genomic DNA, 0.5 pmol/µl primers, 2.5 mM MgSO4 and 0.1 U/µl Pwo Taq DNA polymerase (Roche Molecular Biochemicals) were used in this reaction. The Prop1-containing fragment was isolated by agarose gel electrophoresis and purified using the QIAEX II Gel Extraction Kit (Qiagen). It was subcloned into pGEM-T (Promega) and then re-isolated by BamHI digestion. This fragment was then inserted into a BamHI site downstream of the GSU (or Cga) promoter (–5000 to +43) in pGEM7Zf+ (20). Restriction mapping and partial DNA sequencing were used to confirm the identity of this plasmid.
Generation, genotyping and breeding of transgenic mice
An 8.0 kb GSU-Prop1 fragment was generated by partial digestion of the GSU-Prop1 plasmid with EcoRI. The insert was isolated by agarose gel electrophoresis and purified with the SpinBind DNA Recovery System (FMC Bioproducts). Purified DNA (2–3 ng/µl) was microinjected into F2 zygotes from C57BL/6J x SJL/J parents (69). Embryos at the two-cell stage were transferred to pseudopregnant CD-1 females 0.5 days post coitum. Genomic DNA was prepared from tail biopsies of all progeny born and then screened for the GSU-Prop1 transgene. Transgenic mice were identified by PCR using oligonucleotides that amplify a 438 bp product which spans the junction of the GSU promoter and the Prop1 coding sequence (5'-ATG GCT CCT TCT TTG AGC TTC-3' and 5'-TCA ACT TTC AGG ATG TTT TGT ATA A-3'). Reactions were performed for 35 cycles of denaturation at 92°C for 30 s, annealing at 56°C for 30 s and extension at 72°C for 30 s, with a final extension at 72°C for 10 min. Reactions were carried out under standard conditions using 100–200 ng of genomic DNA, 0.5 pmol/µl primers, 2.5 mM MgCl2 and 0.02 U/µl Taq DNA polymerase per reaction.
Transgenic founders and their male progeny were bred to the Ames dwarf background (DF/B-Prop1df/+) to establish and maintain lines. Six lines were generated with the GSU-Prop1 transgene (D1–D6). The D4 and D6 lines were officially named TgN(Cga-Prop1)D4Sac and TgN(Cga-Prop1)D6Sac. To detect the Prop1df mutation, a 137 bp product was amplified from genomic DNA using Prop1-specific primers (5'-GAG CTG GGG AGA CCT AAG CTT TGC C-3' and 5'-GCC CAG ATC TCA GGA TAC TG-3'). PCR products were digested with 2 U HinfI for 2 h at 37°C. The PCR product generated from wild-type mice was cleaved to produce 97 and 40 bp fragments. The PCR product generated from mice with the Prop1df mutation remains uncut following incubation with HinfI. For a subset of experiments, transgenic mice from lines D4 and D6 were bred to C57BL/6J mice.
C57BL/6J x SJL/J F1 mice, C57BL/6J (The Jackson Laboratory), and CD-1 mice (Charles River) were purchased and bred at the University of Michigan. DF/B-Prop1df/+ mice were obtained from Dr Andrzej Bartke (Southern Illinois University, Carbondale, IL) and bred at the University of Michigan. All procedures involving mice were approved by the University of Michigan Committee on Use and Care of Animals. All experiments were conducted in accord with the principles and procedures outlined in the NIH guidelines for the Care and Use of Experimental Animals.
Histology and in situ hybridization
Pituitaries from aged mice and ovaries were fixed for 2–6 h in 4% paraformaldehyde in phosphate-buffered saline (PBS, pH 7.2). Testes were fixed overnight in Bouin’s fixative. All samples were washed in PBS, dehydrated and embedded in paraffin. Five- to six-micron sections were prepared and stained with hematoxylin and eosin. A periodic acid-Schiff stain was used on testis sections.
The pituitary cell populations in 4-week-old and aged mice were analyzed by immunohistochemistry with antibodies against each of the pituitary hormone markers. Immunostaining was carried out with polyclonal antisera against rat PRL (1:2000, GenBank accession no. AFP1050B), rat GH (1:1000, accession no. AFP411S), rat LHß (1:1000, accession no. AFP22238790GPOLHB), rat TSHß (1:1000, accession no. AFP1274789) (National Hormone and Pituitary Program, NIDDK, Bethesda, MD), and human ACTH (1:1000, Dako). Biotinylated secondary antibodies were used in conjunction with avidin and biotinylated peroxidase (Vectastain guinea pig, rabbit and human kits; Vector Laboratories). Diaminobenzidine, which produces a brown precipitate, was used as the chromogen.
Heads were collected from mice at P1 and frozen in 2-methylbutane at –30°C. To obtain embryos, timed pregnancies were generated by mating D6 GSU-Prop1 transgenic males with C57BL/6J females. Noon on the day the copulatory plug was detected was considered E0.5. Embryos were harvested at E14.5 and frozen in 2-methylbutane. Twenty-micron sections of P1 heads and E14.5 embryos were prepared and used for in situ hybridization. Sections were fixed in 4% paraformaldehyde in PBS (pH 7.2) for 15 min. They were then subjected to proteinase K digestion (0.1 µg/ml in 100 mM Tris–HCl, 50 mM EDTA pH 8.0) for 5 min at 37°C. To acetylate sections, slides were placed in a 0.1 M triethanolamine, 0.3% acetic anhydride solution for 10 min. Sections were prehybridized in hybridization buffer (50% formamide, 5x SSC, 2% blocking powder (Roche Molecular Biochemicals), 0.1% Triton X-100 (Sigma), 0.5% CHAPS (Sigma), 1 mg/ml yeast tRNA, 5mM EDTA pH 8.0, and 50 µg/ml heparin). Sections were then hybridized with the various probes diluted in hybridization buffer. Prop1, Lhb, Tshb and Fshb probes were hybridized at 57°C and the Gnrhr probe was hybridized at 55°C.
The Prop1 riboprobe was generated by PCR amplification (5'3') of a 222 bp product contained in the coding region 3' of the homeodomain. This fragment was subcloned into pGEMT (Promega). The clone was linearized by digestion with SacI to generate the antisense probe and with SpeI to generate the sense probe. The Tshb riboprobe was obtained from Dr David Gordon (University of Colorado Health Sciences Center, Denver, CO). We obtained the Lhb riboprobe from Dr John Nilson (Case Western Reserve University) and the Fshb riboprobe from Dr Richard Maurer (Oregon Health Sciences University). The GnRH-R riboprobe was originally obtained from Dr Keith Parker (University of Texas Southwestern Medical Center). All riboprobes were generated and labeled with digoxigenin (Roche Molecular Biochemicals) following standard procedures (70).
Thyroid hormone analysis
Serum was collected by cardiac puncture, allowed to clot overnight at room temperature and centrifuged for 10 min. The serum supernatant was collected and stored at –20°C. Free T4 levels were assessed using a 125I radioimmunoassay (RIA) to determine the amount of bioactive thyroid hormone present in the serum. The set of animals tested was comprised of five transgenics and four non-transgenics aged 4 weeks, three transgenics and three non-transgenics aged 6 months and two transgenics and two non-transgenics aged 1 year. Results are reported ± standard error. The limit of detection for this experiment, which was carried out in the laboratory of Dr Rich Miller (University of Michigan), was 1.5 µg/dl. This set was performed using an ICN thyroxine, Total (T4) 125I RIA kit (ICN Pharmaceuticals). Alterations to the kit protocol included the addition of a total count tube and an outside standard, and all reagents were used at one-quarter the volume.
|
ACKNOWLEDGEMENTS |
|---|
We thank Dr David Gordon, Dr Rich Miller and Steve Pinkosky for performing the thyroid hormone assays, and Mark Berard, Dr Maggie VanKeuren, Dr Thom Saunders and the University of Michigan Transgenic Animal Model Core for generating the transgenic mice. We are grateful to Dr David Gordon, Dr Keith Parker, Dr John Nilson and Dr Richard Maurer for riboprobes, the National Hormone and Pituitary Program (NIDDK) for antibodies, and Dr Kathleen Mahon for helpful suggestions. We also thank Pallavi Eswara for contributions to the early stages of this work and Hoonkyo Suh for help with in situ hybridization experiments. This work was funded by NICHD grant 30428 (to S.A.C.), a Loeb Cancer Biology Fellowship (to L.J.C.), a Rackham Pre-Doctoral Fellowship (to D.E.W.-C.) and the Endocrinology and Metabolism Training Grant (to L.T.R.).
|
FOOTNOTES |
|---|
+ To whom correspondence should be addressed at: 4301 MSRB III, 1500 West Medical Center Drive, University of Michigan Medical School, Ann Arbor, MI 48109-0638, USA. Tel: +1 734 763 0682; Fax: +1 734 763 7672; Email: scamper@umich.edu
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
1 Procter, A.M., Phillips, J.A.,III and Cooper, D.N. (1998) The molecular genetics of growth hormone deficiency. Hum. Genet., 103, 255–272.[ISI][Medline]
2 Asa, S. and Ezzat, S. (1998) The cytogenesis and pathogenesis of pituitary adenomas. Endocr. Rev., 19, 798–827.
3 Watkins-Chow, D.E. and Camper, S.A. (1998) How many homeobox genes does it take to make a pituitary gland? Trends Genet., 14, 284–290.[ISI][Medline]
4 Dasen, J.S. and Rosenfeld, M.G. (1999) Combinatorial codes in signaling and synergy: lessons from pituitary development. Curr. Opin. Genet. Dev., 9, 566–574.[ISI][Medline]
5 Carbajo-Perez, E., Motegi, M. and Watanabe, Y.G. (1989) Cell proliferation in the anterior pituitary of mice during growth. Biomed. Res., 10, 275–281.
6 Japon, M.A., Rubinstein, M. and Low, M.J. (1994) In situ hybridization analysis of anterior pituitary hormone gene expression during fetal mouse development. J. Histochem. Cytochem., 42, 1117–1125.[Abstract]
7 Sornson, M.W., Wu, W., Dasen, J.S., Flynn, S.E., Norman, D.J., O’Connell, S.M., Gukovsky, I., Carrière, C., Ryan, A.K., Miller, A.P. et al. (1996) Pituitary lineage determination by the Prophet of Pit-1 homeodomain factor defective in Ames dwarfism. Nature, 384, 327–333.[Medline]
8 Gage, P.J., Roller, M.L., Saunders, T.L., Scarlett, L.M. and Camper, S.A. (1996) Anterior pituitary cells defective in the cell autonomous factor, df, undergo cell lineage specification but not expansion. Development, 122, 151–160.[Abstract]
9 Tang, K., Bartke, A., Gardiner, C.S., Wagner, T.E. and Yun, J.S. (1993) Gonadotropin secretion, synthesis, and gene expression in human growth hormone transgenic mice and in Ames dwarf mice. Endocrinology, 132, 2518–2524.[Abstract]
10 Wu, W., Cogan, J.D., Pfaffle, R.W., Dasen, J.S., Frisch, H., O’Connell, S.M., Flynn, S.E., Brown, M.R., Mullis, P.E., Parks, J.S. et al. (1998) Mutations in PROP1 cause familial combined pituitary hormone deficiency. Nature Genet., 18, 147–149.[ISI][Medline]
11 Parker, G.E., Sloop, K.W. and Rhodes, S.J. (2001) Transciptional control of the development and function of the hypothalamic-pituitary axis. In Eugster, E.A. and Pescovitz, O.H. (eds). Developmental Endocrinology: for Research to Clinical Practice. Humana Press, Totowa, NJ, in press.
12 Dattani, M.T. and Robinson, I.C. (2000) The molecular basis for developmental disorders of the pituitary gland in man. Clin. Genet., 57, 337–346.[ISI][Medline]
13 Cogan, J., Wu, W., Phillips, J.I., Arnhold, I., Agapito, A., Fofanova, O., Osorio, M., Bircan, I., Moreno, A. and Mendonca, B. (1998) The PROP1 2-base pair deletion is a common cause of combined pituitary hormone deficiency. J. Clin. Endocrinol. Metab., 83, 3346–3349.
14 Pernasetti, F., Toledo, S.P., Vasilyev, V.V., Hayashida, C.Y., Cogan, J.D., Ferrari, C., Lourenco, D.M. and Mellon, P.L. (2000) Impaired adrenocorticotropin-adrenal axis in combined pituitary hormone deficiency caused by a two-base pair deletion (301–302delAG) in the prophet of Pit-1 gene. J. Clin. Endocrinol. Metab., 85, 390–397.
15 Flück, C., Deladoey, J., Rutishauser, K., Eblé, A., Marti, U., Wu, W. and Mullis, P.E. (1998) Phenotypic variability in familial combined pituitary hormone deficiency caused by a PROP1 gene mutation resulting in the substitution of ArgCys at codon 120 (R120C)*. J. Clin. Endocrinol. Metab., 83, 3727–3734.
16 Hermesz, E., Mackem, S. and Mahon, K.A. (1996) Rpx: A novel anterior-restricted homeobox gene progressively activated in the prechordal plate, anterior neural plate, and Rathke’s pouch of the mouse embryo. Development, 122, 41–52.[Abstract]
17 Anderson, B., Pearse, R.V.,II, Jenne, K., Sornson, M., Lin, S.-C., Bartke, A. and Rosenfeld, M.G. (1995) The Ames dwarf gene is required for Pit-1 gene activation. Dev. Biol., 172, 495–503.[ISI][Medline]
18 Gage, P.J., Brinkmeier, M.L., Scarlett, L.M., Knapp, L.T., Camper, S.A. and Mahon, K.A. (1996) The Ames dwarf gene, df, is required early in pituitary ontogeny for the extinction of Rpx transcription and initiation of lineage specific cell proliferation. Mol. Endocrinol., 10, 1570–1581.[Abstract]
19 Li, S., Crenshaw, E.B.,III, Rawson, E.J., Simmons, D.M., Swanson, L.W. and Rosenfeld, M.G. (1990) Dwarf locus mutants lacking three pituitary cell types result from mutations in the POU-domain gene Pit-1. Nature, 347, 528–533.[Medline]
20 Kendall, S.K., Gordon, D.F., Birkmeier, T.S., Petrey, D., Sarapura, V.D., O’Shea, K.S., Wood, W.M., Lloyd, R.V., Ridgway, E.C. and Camper, S.A. (1994) Enhancer-mediated high level expression of mouse pituitary glycoprotein hormone -subunit transgene in thyrotropes, gonadotropes, and developing pituitary gland. Mol. Endocrinol., 8, 1420–1433.[Abstract]
21 Burrows, H.L., Birkmeier, T.S., Seasholtz, A.F. and Camper, S.A. (1996) Targeted ablation of cells in the pituitary primordia of transgenic mice. Mol. Endocrinol., 10, 1467–1477.[Abstract]
22 Bartke, A. (1965) The response of two types of dwarf mice to growth hormone, thyrotropin, and thyroxine. Gen. Comp. Endocrinol., 5, 418–426.[ISI][Medline]
23 Kulig, E., Camper, S.A., Kuecker, S., Jin, L. and Lloyd, R.V. (1998) Remodeling of hyperplastic pituitaries in hypothyroid -subunit knockout mice after thyroxine and 17ß-estradiol treatment: role of apoptosis. Endocr. Pathol., 9, 261–274.[ISI][Medline]
24 Alarid, E.T., Windle, J.J., Whyte, D.B. and Mellon, P.L. (1996) Immortalization of pituitary cells at discrete stages of development by directed oncogenesis in transgenic mice. Development, 122, 3319–3329.[Abstract]
25 Jameson, J.L. (1996) Inherited disorders of the gonadotropin hormones. Mol. Cell. Endocrinol., 125, 143–149.[ISI][Medline]
26 Risma, K.A., Clay, C.M., Nett, T.M., Yun, J. and Nilson, J.H. (1995) Targeted overexpression of luteinizing hormone in transgenic mice leads to infertility, polycystic ovaries, and ovarian tumors. Proc. Natl Acad. Sci. USA, 92, 1322–1326.
27 Shenker, A., Laue, L., Kosugi, S., Merendino, J.J.,Jr, Minegishi, T. and Cutler, G.B.,Jr (1993) A constitutively activating mutation of the luteinizing hormone receptor in familial male precocious puberty. Nature, 365, 652–654.[Medline]
28 Risma, K.A., Hirshfield, A.N. and Nilson, J.H. (1997) Elevated luteinizing hormone in prepubertal transgenic mice causes hyperandrogenemia, precocious puberty, and substantial ovarian pathology. Endocrinology, 138, 3540–3547.
29 Lyon, M.F., Cattanach, B.M. and Charlton, H.M. (1981) In Austin, C.R. and Edwards, R.G. (eds), Mechanisms of Sex Differentiation in Animals and Man, Academic Press, New York, NY, pp. 329–275.
30 Charlton, H.M., Halpin, D.M., Iddon, C., Rosie, R., Levy, G., McDowell, I.F., Megson, A., Morris, J.F., Bramwell, A., Speight, A. et al. (1983) The effects of daily administration of single and multiple injections of gonadotropin-releasing hormone on pituitary and gonadal function in the hypogonadal (hpg) mouse. Endocrinology, 113, 535–544.[ISI][Medline]
31 Nishizono, H., Soji, T. and Herbert, D.C. (1993) Intercellular communication within the rat anterior pituitary gland: V. Changes in cell-to-cell communications as a function of the timing of castration in male rats. Anat. Rec., 235, 577–582.[Medline]
32 Lloyd, R.V. (1993) Ultrastructural diagnosis of pituitary adenomas and hyperplasias. In Livolsi, V.A. (ed.), Surgical Pathology of the Pituitary Gland, Vol. 27. W.B. Saunders Co., Philadelphia, PA, 52–84.
33 Kovacs, K. and Asa, S.L. (1991) Functional Endocrine Pathology, Vol. 1. Blackwell Scientific Publications, Boston, MA.
34 Yoshida, J., Kobayashi, T., Kageyama, N. and Kanzaki, M. (1977) Symptomatic Rathke’s cleft cyst. Morphological study with light and electron microscopy and tissue culture. J. Neurosurg., 47, 451–458.[ISI][Medline]
35 Steinberg, G.K., Koenig, G.H. and Golden, J.B. (1982) Symptomatic Rathke’s cleft cysts. Report of two cases. J. Neurosurg., 56, 290–295.[ISI][Medline]
36 Fofanova, O., Takamura, N., Kinoshita, E., Vorontsov, A., Vladimirova, V., Dedov, I., Peterkova, V. and Yamashita, S. (2000) MR imaging of the pituitary gland in children and young adults with congenital combined pituitary hormone deficiency associated with PROP1 mutations. Am. J. Roentgenol., 174, 555–559.
37 Helfand, M. and Redfern, C.C. (1998) Clinical guideline, part 2. Screening for thyroid disease: an update. American College of Physicians [Published erratum appears in Ann. Intern. Med. (1999), 130, 246.]. Ann. Intern. Med., 129, 144–158.
38 Windle, J.J., Weiner, R.I. and Mellon, P.L. (1990) Cell lines of the pituitary gonadotrope lineage derived by targeted oncogenesis in transgenic mice. Mol. Endocrinol., 4, 597–603.[Abstract]
39 Lee, S.L., Sadovsky, Y., Swirnoff, A.H., Polish, J.A., Goda, P., Gavrilina, G. and Milbrandt, J. (1996) Luteinizing hormone deficiency and female infertility in mice lacking the transcription factor NGF1-A (Egr-1). Science, 273, 1219–1221.[Abstract]
40 Topilko, P., Schneider-Maunoury, S., Levi, G., Trembleau, A., Gourdji, D., Driancourt, M.-A., Rao, C.V. and Charnay, P. (1998) Multiple pituitary and ovarian defects in Krox-24 (NGF1-A, Egr-1)-targeted mice. Mol. Endocrinol., 12, 107–122.
41 Tortora, G.J. and Grabowski, S.R. (1996) In Roesch, B. (ed.), Principles of Anatomy and Physiology, HarperCollins Publishers, New York, NY, pp. 501–550.
42 Kendall, S.K., Samuelson, L.C., Saunders, T.L., Wood, R.I. and Camper, S.A. (1995) Targeted disruption of the pituitary glycoprotein hormone -subunit produces hypogonadal and hypothyroid mice. Genes Dev., 9, 2007–2019.
43 Stahl, J., Kendall, S., Brinkmeier, M., Greco, T., Watkins-Chow, D., Canpos-Barros, A., Lloyd, R. and Camper, S. (1999) Thyroid hormone is essential for pituitary somatotropes and lactotropes. Endocrinology, 140, 1884–1892.
44 Luo, X., Ikeda, Y. and Parker, K.L. (1994) A cell-specific nuclear receptor is essential for adrenal and gonadal development