help button home button Endocrine Society Endocrinology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology, doi:10.1210/en.2003-0476
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
144/10/4492    most recent
Author Manuscript (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Burns, K. H.
Right arrow Articles by Matzuk, M. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Burns, K. H.
Right arrow Articles by Matzuk, M. M.
Endocrinology Vol. 144, No. 10 4492-4507
Copyright © 2003 by The Endocrine Society

Expression Profiling Analyses of Gonadotropin Responses and Tumor Development in the Absence of Inhibins

Kathleen H. Burns, Gabe E. Owens, Samuel C. Ogbonna, John H. Nilson and Martin M. Matzuk

Departments of Pathology (K.H.B., S.O., M.M.M.), Molecular and Human Genetics (K.H.B., M.M.M.), and Molecular and Cellular Biology (M.M.M.), Baylor College of Medicine, Houston, Texas 77030; and Department of Pharmacology, Case Western Reserve University (G.E.O., J.H.N.), Cleveland, Ohio 44106

Address all correspondence and requests for reprints to: Dr. Martin M. Matzuk, Department of Pathology, One Baylor Plaza, Baylor College of Medicine, Houston, Texas 77030. E-mail: mmatzuk{at}bcm.tmc.edu.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results and Discussion
 References
 
Transgenic mice with engineered disruptions in bidirectional endocrine signaling between the pituitary and gonad have shed light on the specific effects of the loss of function of gonadotropins and inhibins. These models are valuable tools for studying ovarian biology because they phenocopy specific pathological states and have variations in ovarian tissue composition that allow us to identify genes expressed in specific cell types. We have used emerging mRNA expression profiling technologies to gain a more comprehensive view of genes that are expressed in the mammalian ovary and adrenal gland in the FSHß and inhibin {alpha} knockout mouse models. Oligonucleotide array hybridization experiments using Affymetrix GeneChip technology and NIA 15K murine cDNA microarray studies identified hundreds of transcripts differentially expressed compared with wild type, over 30 of which were selected for further characterization by Northern blot analyses. Additionally, we performed in situ hybridization studies to localize 10 mRNAs, melanocyte-specific gene 1, amino acid transporter SN2, overexpressed and amplified in teratocarcinoma (Bcat1), Forkhead box protein FOXO1, 24p3, vascular cell adhesion molecule, epiregulin, Bcl2-like10, PC3B, and retinoblastoma binding protein 7. These 10 genes have expression patterns and postulated functions suggesting that they mediate important processes in the physiology and pathology of ovarian and adrenal tissue.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results and Discussion
 References
 
THE MAMMALIAN OVARY is the site of complex biological processes regulated by cell autonomous, paracrine, and endocrine factors that together are essential for achieving female fertility and orchestrating appropriate cellular development and differentiation pathways (1, 2). Ovarian cell populations include granulosa cells, somatic cells that undergo dramatic proliferation followed by differentiation during follicle development; thecal and interstitial cells, which are crucial to steroidogenesis; as well as oocytes, female germ cells arrested in prophase I of meiosis. Pronounced structural rearrangements occur in the ovary during the process of folliculogenesis, and exquisite control of cellular phenotypes is maintained to regulate follicular recruitment, the events of oocyte maturation and meiosis, and the proliferation and organization of the surrounding somatic cells.

Completion of follicular development depends on the pituitary gonadotropins, FSH and LH (3). FSH and LH are glycoprotein {alpha}:ß heterodimers that share a common {alpha}-subunit in noncovalent association with a unique ß-subunit. Knockout mice homozygous for a null allele at the FSHß-encoding locus (Fshb-/-) exhibit female infertility due to a block in follicle development before the climax of granulosa cell proliferation and antrum formation that typify preovulatory follicles (4). Although these mice develop high levels of circulating LH, granulosa cells within these arrested follicles fail to become responsive to LH, do not express markers associated with luteinization, and are defective in the production of steroidogenic enzymes and peptide hormones (5). FSH receptor (Fshr) knockout females display a similar block in follicle development and also exhibit high levels of circulating LH (6, 7). It is believed that LH and LH-induced androgen synthesis in the Fshr model contribute to the development of ovarian sex cord stromal tumors, which are seen in 92% of these mice by 12 months of age. In these tumors there is a loss of granulosa cell proliferation control programs accompanied by an up-regulation of Sertoli cell markers, Mullerian inhibiting substance, and the GATA4 transcription factor (8).

Gonadal peptide hormones, inhibins and activins, normally elaborated by granulosa cells in females and Sertoli cells in males, comprise an endocrine feedback loop to the pituitary. Inhibins are {alpha}:ß heterodimers ({alpha}:ßA; {alpha}:ßB), and activins are ß:ß dimers (ßA:ßA; ßB:ßB; ßA:ßB), named for their effects, which inhibit and activate, respectively, pituitary production of FSH. Knockout mice with a deletion of the inhibin {alpha}-subunit locus, and thus deficient in inhibins, develop mixed granulosa/Sertoli cell tumors in the ovaries and testes (9). The progression of tumor development in these mice is associated with a cachexia-like wasting syndrome and high serum concentrations of the FSH and activins (9, 10). Together FSH and LH are required for the development of tumors in these mice (11), whereas FSH and sex steroids are important regulators of the process (Refs. 12 and 13 and our unpublished data). Castrated inhibin {alpha} knockout mice develop steroidogenic tumors of the adrenal cortex with similar histological features (10).

Granulosa cells normally up-regulate LH receptor expression in large preovulatory follicles and terminally differentiate or luteinize in response to the surge of circulating LH at ovulation (14, 15). Together with luteinizing follicular thecal cells, these cells form the corpus luteum and express factors critical to the maintenance of early pregnancy. Although this physiological LH exposure induces cell cycle withdrawal in the granulosa population and increased expression of cell cycle inhibitors, female LH-overexpressing transgenic mice develop tumors of the granulosa cell lineage in some genetic backgrounds (16, 17, 18). In these mice, a chimeric bovine LHß-human chorionic gonadotropin (hCG) transgene is driven in a tissue-specific manner to achieve a 10-fold increase in circulating LH immunoreactivity. The tumor phenotype in this model can be averted by administrating ovulatory doses of hCG, diverting granulosa cells to instead form luteomas that have a distinct histological appearance and transcriptome profile compared with the granulosa cell tumors (19).

All of the conditions modeled in the lines of transgenic mice considered above have analogies in human medicine (3). Women who are homozygous or compound heterozygous for inactivating FSHß ligand (frameshift/truncation and missense) or FSH receptor (missense) mutations exhibit normal preantral follicle development, but antral stage follicles capable of ovulation do not form (20). Granulosa cell tumors comprise nearly 10% of ovarian cancer cases in women, and they exhibit propensity for invasiveness, metastasis, and recurrence. Gain of function mutations of the LH receptor have been described in families with autosomal dominant precocious puberty (21), and a nonconservative amino acid change in inhibin {alpha} (Ala257Thr) has been associated with several cases of premature ovarian failure with elevated serum gonadotropins (22). However, to date neither activating mutations in gonadotropin receptors nor loss of inhibin function have been associated with human sex cord stromal tumor development.

To better appreciate the molecular mechanisms underlying ovarian and adrenal responses to gonadotropins and inhibins and the biological bases for the phenotypes of these transgenic models, we have undertaken a series of mRNA expression profiling experiments using Affymetrix 11K oligonucleotide array GeneChips and NIA 15K cDNA microarray slides. These approaches have revealed several novel candidate genes with potentially pivotal roles in ovarian and adrenal function and pathophysiology.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results and Discussion
 References
 
Animals and tissue collection for mRNA expression analyses
Mice were maintained in accordance with the NIH Guide for the Care and Use of Laboratory Animals, housed in a room with a 12-h light, 12-h dark cycle, and fed ad libitum. The generation of FSHß knockout (Fshb-/-), GDF9 knockout (Gdf9-/-), and inhibin {alpha} knockout (Inha-/-) mice and genotyping protocols have been described previously (4, 9, 23).

For RNA extraction, ovaries and adrenal glands were collected from 13–16 wild-type C57BL/6J/129S6/SvEv randomly cycling females, 2–3 months old, and pooled unless otherwise indicated. Ovarian tumors were collected from individual inhibin {alpha} knockout females and were not pooled. These mice were killed at approximately 8 wk of age when demonstrating signs of a cachexia-like wasting syndrome associated with tumor development (9). To study adrenal tumor gene expression, castrated inhibin {alpha} null females were killed upon weight loss at 22–35 wk of age (10). Tissues were immediately homogenized in RNA-STAT (Leedo Medical Laboratories, Houston, TX) and processed in accordance with the manufacturer’s protocol or were stored in RNAlater (Ambion, Austin, TX) before RNA extraction. For in situ hybridization experiments, ovaries, testes, or ovarian tumors were collected from mice, 2–3 months old, and immediately submerged in 4% paraformaldehyde fixative before processing and wax embedding.

Affymetrix gene chip mRNA expression analyses
Total RNA was used as a template for cDNA synthesis and biotinylated antisense cRNA probe preparation as described by the manufacturers of the SuperScript System kit (Invitrogen, Carlsbad, CA) and the ENZO BioArray HighYield RNA labeling kit (Enzo Diagnostics, Farmingdale, NY). Unincorporated nucleotides were removed from the riboprobe preparation using the RNeasy Mini kit (Qiagen, Valencia, CA). The integrity of the riboprobe was checked by gel electrophoresis.

The Affymetrix murine 11K oligonucleotide array set (Mu11KsubA, Mu11KsubB) was hybridized, washed, and scanned using Affymetrix equipment and protocols (Affymetrix, Santa Clara, CA). Data analysis was performed using Affymetrix software with the decision matrix default parameters recommended. For the purposes of this manuscript, only 3-fold or greater comparative changes in mRNA expression were considered significant and included in Tables 2–4GoGoGo below. More information detailing our protocols and data analyses and a listing of the genes surveyed in this analysis are available at http://www.affymetrix.com.


View this table:
[in this window]
[in a new window]
  		TABLE 2. Affymetrix GeneChip comparison of ovarian gene expression in Fshb-/- mice

 

View this table:
[in this window]
[in a new window]
  		TABLE 3. Affymetrix GeneChip comparison of ovarian gene expression in Fshb-/- mice treated with PMSG

 

View this table:
[in this window]
[in a new window]
  		TABLE 4. Affymetrix GeneChip comparison of ovarian gene expression in tumors of Inha-/- mice

 
cDNA microarray mRNA expression studies
Total RNA served as a template for the synthesis of aminoallyl-cDNA, which was then coupled to fluorescent dye esters (Cy3 and Cy5). Size exclusion column purification steps preceded and followed the dye-coupling reaction. Probe was prepared from RNA of pooled ovaries of 2- to 3-month-old, wild-type and Fshb-/- females as well as from a single ovarian tumor from an inhibin {alpha} knockout female; these RNA samples were identical to three of those used for the Affymetrix analysis.

Wild-type (Cy3) and mutant (Cy5) probes were hybridized to the NIA 15K cDNA clone set (http://lgsun.grc.nia.nih.gov/cDNA/15k.html) at the Population Center Gene Array Facility (http://depts.washington.edu/popctrma/index.shtml). After washing, slides were scanned using a GenePix scanner (Axon Instruments, Union City, CA), and returned values were analyzed with GeneSpring software (Silicon Genetics, Redwood City, CA). For the purposes of this manuscript, only changes in mRNA expression more than 2-fold (ratio of medians, >2 or <0.5) with less than 20% difference between the ratio of the medians, median of ratios, and the normalized ratio of medians were considered significant and included in Tables 5Go and 6Go below.


View this table:
[in this window]
[in a new window]
  		TABLE 5. cDNA microarray comparison of ovarian gene expression in Fshb-/- mice

 

View this table:
[in this window]
[in a new window]
  		TABLE 6. cDNA microarray comparison of ovarian gene expression in Inha-/- mice

 
Probe template preparation
RNA preparations from ovaries of wild-type mice and ovarian tumors from inhibin {alpha} knockout mice were reverse transcribed to cDNAs using the SuperScript RT kit (Invitrogen, Carlsbad, CA). The coding sequences of select genes were used to find primers 19–22 bp in length with predicted Tms of 60 C using the Cprimer program (http://iubio. bio.indiana.edu/soft/molbio/mac/). Sequences were then amplified by PCR using an annealing temperature equivalent to or 5 C lower than the lower predicted primer Tm, ligated into pGEM(R)-T or T-easy vector (Promega, Madison, WI), and checked by bidirectional sequencing before being used as templates to synthesize Northern blot and in situ hybridization probes. The sequences amplified are listed in Table 1Go.


View this table:
[in this window]
[in a new window]
  		TABLE 1. Northern blot and in situ hybridization probes

 
Northern blot analyses and in situ hybridizations
Northern blot analyses and in situ hybridizations were performed as previously described (5, 24). For Northern blot analyses, 15 µg total RNA were electrophoresed and transferred onto nylon membranes as detailed previously (25). [32P]Deoxy-ATP-labeled probes were synthesized from the templates listed (Table 1Go) using the Strip-EZ kit (Ambion), and blots were hybridized and washed as recommended. Probe hybridization was assessed by autoradiography and phosphorimaging with ImageQuant software (Molecular Dynamics, Sunnyvale, CA) (26). A background level for each blot was determined and subtracted. Blots were stripped and reprobed once, and then were probed for glyceraldehyde 3-phosphate dehydrogenase (Gapd) mRNA to allow for loading corrections when comparing the signal intensities between samples on the same membrane.

For in situ hybridization, [35S]UTP antisense and sense probes were transcribed from the template plasmids (Table 1Go) using T7 and SP6 polymerases (Promega, Madison, WI). Paraffin-embedded samples were cut into 5-µm-thick sections, which were dewaxed, pretreated, hybridized, and washed as previously detailed (27). Probe hybridization was detected by autoradiography using the NTB-2 emulsion (Eastman Kodak, Rochester, NY). Slides were developed as described by the manufacturer and were counterstained with hematoxylin.

Serum testosterone analysis
Ten randomly cycling, wild-type and Fshb-/- females, 3–4 months old, were anesthetized by isoflurane inhalation (Abbott Laboratories, North Chicago, IL), and blood was recovered by closed cardiac puncture. Serum was separated in Microtainer tubes (BD Biosciences, Franklin Lakes, NJ) and stored at -80 C before analysis. Testosterone measurements were made at the University of Virginia Ligand Core Facility (Specialized Cooperative Centers Program in Reproduction Research) and are reported as the average ± SE.


   Results and Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results and Discussion
References
 
Affymetrix profiling of mRNA expression in the FSHß knockout ovary
To survey changes in the expression of ovarian mRNAs in the absence of FSH, we prepared complex biotin-labeled riboprobes from the total ovarian RNA pooled from 2- to 3-month-old wild-type and Fshb-/- mice and hybridized them to the Affymetrix 11K murine GeneChips. Hybridization signals amplified with streptavidin-biotin conjugates were scanned, and comparison analysis was performed using Affymetrix software with the default decision matrix settings. Overall, the vast majority of gene products and all loading controls demonstrated essentially no change in the comparison, and no correction factors were employed. For the purposes of this report, comparisons yielding fold changes with absolute values less than 3 as well as data obtained for the expression of unmatched expressed sequence tags (ESTs) have been omitted. The results of this analysis are listed in Table 2Go. Although the majority of mRNAs identified as regulated in this screen are the products of genes never before characterized in the mammalian ovary, several findings from this analysis confirmed the results of our previous investigations. This list includes expression data for inhibin/activin ßA and ßB subunits, LH receptor, serum/glucocorticoid-regulated kinase, and P450 aromatase mRNAs, which are all dramatically down-regulated in the Fshb-/-ovary (5). These data are indicative of the failure of granulosa cells to function in paracrine/endocrine signaling pathways in the absence of FSH and underscore the potential for sequelae in other cell types (see below).

Profiling ovarian mRNA expression changes in FSHß knockout mice treated with pregnant mare serum gonadotropin (PMSG)
To identify potential gonadotropin-responsive gene products, we collected ovaries from 3-wk-old Fshb null females 8 h after administering 7.5 IU PMSG, an FSH receptor agonist that functions well regardless of the presence of endogenous FSH (4). Riboprobe prepared from total RNA was then hybridized to Affymetrix GeneChip oligonucleotide arrays, and the hybridization signals were compared with results obtained using untreated 3-wk-old Fshb-/- mice. The results of this comparison are shown in Table 3Go. Several of the identified changes in gene expression are consistent with our previous results (Table 2Go), including the values obtained for 24p3, inhibin/activin ßB, prostaglandin D synthetase, P450 aromatase, scavenger receptor class B type I, and steroidogenic acute regulatory protein. Thus, in select cases, aberrant gene expression in the Fshb knockout ovary can be readily reversed with PMSG administration.

Although PMSG is chiefly used to mimic FSH function in superovulatory regimes in rodents, it also has been demonstrated to activate the LH receptor (28, 29). Thus, several genes identified in this screen may be up-regulated as part of a periovulatory sequence in the ovary, as has been reported for epiregulin and steroidogenic acute regulatory protein, which are also induced after administration of the LH analog, hCG, to wild-type mice (30). Consistent with this, we found that up-regulation of epiregulin (Ereg), lymphocyte antigen 6 complex (Ly6d), and mouse 3T3 double minute 1 were recapitulated in wild-type mice only when PMSG was followed by treatment with hCG (see below).

Affymetrix profiling of mRNA expression in tumors of inhibin {alpha} knockout and LHß-overexpressing mice
Because it is feasible to recover large amounts of total RNA from the ovarian tumors that develop in the inhibin {alpha} knockout and LHß-overexpressing mouse models, and because it seemed conceivable that variation in gene expression patterns would probably exist between tumor samples and even within different regions of an individual sample, we performed Affymetrix analyses on individual tumors. A comparison analysis of one of the tumor samples recovered from an inhibin {alpha} knockout demonstrated hundreds of mRNA species with differential expression (>3-fold) compared with wild-type ovary (240 transcripts identified). The serial addition of a second inhibin {alpha} knockout tumor sample (262 transcripts identified), allowed us to compile a list of 118 transcripts with common directions of regulation in both samples. Table 4Go lists the common expression changes of 74 known genes shared between the two inhibin {alpha} knockout tumor samples (excludes EST and duplicate sequences). We annotated this table by adding expression changes observed in the tumor transcriptome of an LHß-overexpressing mouse.

After gonadectomy, inhibin {alpha} knockout mice develop sex steroidogenic tumors of the adrenal cortex. Although these adrenal tumors do not result from metastases, they bear histological resemblance to granulosa cell tumors and perhaps share a similar means of pathogenesis (10). We used the Affymetrix system to assess gene expression in a unilateral adrenal tumor recovered from a 22-wk-old Inha-/- female after ovariectomy at 5 wk of age. A column in Table 4Go indicates those gene products identified as present (P) vs. those identified as absent (A) in the absolute analysis of this sample. We speculate that gene products expressed in both gonadal and adrenal tumor samples have the potential to participate in transformation.

cDNA microarray analyses
To confirm the results of our Affymetrix studies and to identify additional changes in ovarian gene expression in FSHß and inhibin {alpha} knockouts, samples of ovarian RNA from 2- to 3-month-old, wild-type and Fshb-/- mice and ovarian tumor RNA from an inhibin {alpha} knockout female were studied by cDNA microarray analyses. Total RNA was used to transcribe fluorescently labeled cDNA, which was then hybridized to the NIA 15K murine microarray glass slides. Hybridization signals were scanned, and Cy5(mutant)/Cy3(wild-type) intensity ratios were compared. As expected, the majority of gene products and all loading controls demonstrated essentially equivalent hybridization signals and a Cy5/Cy3 ratio of approximately 1.0. For the purposes of this report, comparisons yielding fold changes with absolute values less than 2 as well as data obtained for the expression of ESTs have been omitted. The results of this analysis are listed in Tables 5Go and 6Go.

Comparing quantitative technologies: array analyses vs. Northern blot phosphorimaging
The array analyses described above provided us with quantitative data reflecting changes in the expression of many mRNA species in the ovaries of FSHß and inhibin {alpha} knockout mice vs. wild-type mice or in young FSHß knockout mice treated with PMSG. To better appreciate which of these findings were reproducible by standard molecular techniques and to what extent such analyses would provide comparable fold change values, we selected 32 genes for Northern blot expression analyses with phosphorimaging quantitation (26). The fold change values of the Affymetrix analyses with streptavidin conjugate amplification and/or of the cDNA microarray analysis as well as the Northern blot phosphorimaging fold changes are depicted in Table 7Go, and autoradiographs are shown in Figs. 1–4GoGoGoGo. Adult wild-type ovary, Fshb-/- ovary, and wild-type adrenal gland RNA samples were extracted from pooled tissues. Ovarian and adrenal tumors from inhibin {alpha} knockout mice were not pooled. To further investigate gene products identified as up-regulated by gene chip analysis after PMSG treatment, we performed Northern blot analyses on RNA of pooled ovaries of 3- to 4-wk-old mice both 46 h after PMSG treatment (7.5 IU) and after 5 h of subsequent hCG treatment (5 IU). Although the majority of high throughput gene expression profiling findings were verifiable, some results were not readily reproduced in related samples or were not reproduced in the tissue RNA sample first profiled. This underscores the need to interpret array data with caution even when similar profiling results are obtained by related experimental approaches or replicate sampling.


View this table:
[in this window]
[in a new window]
  		TABLE 7. Northern blot analyses

 


View larger version (17K):
[in this window]
[in a new window]
  		FIG. 1. Northern blot analysis of gene expression in FSHß knockout ovaries. Total pooled ovarian RNA from wild-type (WT), targeted Fshb heterozygote (+/-), and Fshb knockout (-/-) mice was electrophoresed, transferred to membranes, and probed with 32P-labeled cDNAs corresponding to the sequences shown. Gapd was used as a loading control. PhosphorImager quantitation of each signal with respect to this control is listed in Table 7Go.

 


View larger version (39K):
[in this window]
[in a new window]
  		FIG. 2. Northern blot analysis of ovarian gene expression after gonadotropin administration. Total pooled ovarian RNA from untreated wild-type (WT), PMSG-treated (PMSG) wild-type, and hCG-treated (hCG) wild-type mice was electrophoresed, transferred to membranes, and probed with 32P-labeled cDNAs corresponding to the sequences shown. Gapd was used as a loading control. PhosphorImager quantitation of each signal with respect to this control is listed in Table 7Go.

 


View larger version (20K):
[in this window]
[in a new window]
  		FIG. 3. Northern blot analysis of gene expression in ovarian tumors of inhibin {alpha} knockout females. The leftmost lane represents total RNA from pooled ovaries of wild-type (WT) mice; the three right lanes represent RNA extracted from individual ovarian tumors (OT) of inhibin {alpha} knockout mice. In most cases, Cdrap being a notable exception, there was little heterogeneity in gene expression detected between tumor samples. Northern blots were probed with 32P-labeled cDNAs corresponding to the sequences shown. Gapd was used as a loading control. Each signal is quantified in Table 7Go.

 


View larger version (21K):
[in this window]
[in a new window]
  		FIG. 4. Northern blot analysis of gene expression in adrenal tumors of inhibin {alpha} knockout females. The leftmost lane represents total RNA from pooled adrenal glands of wild-type (WT) female mice; the two right lanes represent RNA extracted from adrenal tumors (AT) of two inhibin {alpha} knockout females. Northern blots were probed with 32P-labeled cDNAs corresponding to the sequences shown. Gapd was used as a loading control. Each signal is quantified in Table 7Go.

 
Novel candidate determinants of the granulosa/Sertoli lineage
To localize the principle sources of expression of several mRNAs identified in this study and not previously investigated by such a technique in mammalian ovaries, we performed in situ hybridizations in fixed tissue samples (Fig. 5Go). We focused on mRNAs that encode cell cycle regulators or proteins implicated in tumor growth or metastasis. The results for 10 of these mRNAs are reported here; these were selected as they demonstrated high levels and cell type-specific patterns of expression. Four of these transcripts, melanocyte-specific gene 1 (Msg1), amino acid transporter SN2, overexpressed and amplified in teratocarcinoma (Bcat1), and the Forkhead box protein FOXO1-encoding gene, represent potential targets of FSH and inhibin signaling pathways as they are expressed in the granulosa cell lineage.



View larger version (156K):
[in this window]
[in a new window]
  		FIG. 5. In situ hybridization analysis of 10 ovarian-expressed genes. 35S-Labeled antisense riboprobes were hybridized to endogenous mRNA in ovaries of wild-type (A, C–H, and K–N), inhibin {alpha} knockout (B), Fshb knockout (I–J), and Gdf9 knockout (O–P) mice. Brightfield images (C, E, G, I, K, M, and O) serve to highlight the histology, and probe hybridization appears black; in darkfield images (A, B, D, F, H, J, L, N, and P), probe hybridization appears white. In several cases (C–H, K–L, and O–P), these are paired so that brightfield (left) and darkfield (right) images of the same section are shown. A and B, Msg1 mRNA is expressed in granulosa cells of ovarian follicles (A) and in regions of ovarian tumors (B; arrows). C and D, Amino acid transporter SN2 mRNA is localized to granulosa cells of growing follicles (arrow) and is not detected in corpora lutea (CL). E and F, Bcat1 mRNA is expressed principally in granulosa cells of developing follicles (arrow) and luteal cells (CL). G and H, In small follicles, Foxo1 mRNA is expressed exclusively in oocytes (upper two arrows), whereas in larger multilayered follicles it is expressed in both oocytes and granulosa cells (lower arrow). It is not detected in CL. I, 24p3 mRNA is detected only in Fshb knockout ovaries and localizes to the interstitial cells between follicles (black). An arrested secondary follicle is marked (arrow). J, VCAM mRNA also only detected in Fshb knockout ovarian interstitium. The hybridization signal is excluded from follicles (arrows). K and L, Ereg mRNA is expressed in interstitial cells and granulosa cells of antral follicles (arrow) after hCG administration, but is not detected in CL. M, Bcl2l10 is expressed exclusively in oocytes (black). N, PC3B mRNA is detected exclusively in oocytes (white). O and P, Rbbp7 mRNA is expressed predominantly in oocytes in all genotypes examined (data not shown). The signal is pronounced in the Gdf9 knockout ovaries depicted, which accumulate arrested primary stage follicles (23 ). Further descriptions of the hybridization signal are included within the text. In each case, controls probed with sense probes demonstrated no hybridization signal.

 
Msg1 (Cited1) in an X-linked gene that encodes a transcriptional coactivator implicated in both SMAD-mediated and estrogen receptor-mediated gene activation (31, 32). The protein was first described in melanocytes, and its expression is lost after ras oncogene induction in depigmented melanoma cells (33). Msg1 was discovered in our screen as up-regulated in tumor samples, but was found by Northern blot analysis to be down-regulated in tumor samples. In situ hybridization reveals Msg1 mRNA in only defined compartments of granulosa cell tumors (Fig. 5BGo), which probably explains the disparity between the results of our gene chip and Northern blot analyses. Msg1 is also expressed in slowly dividing granulosa cells of small follicles and atretic follicles (Fig. 5AGo) and corpora lutea in wild-type ovaries; Msg1 is also expressed in granulosa cells in the arrested follicles in the FSHß knockout mice (data not shown). The roles of Msg1 in regulating cell cycle progression in the granulosa/Sertoli cell lineage remain to be investigated.

The amino acid transporter SN2 was identified in these analyses as being up-regulated in ovaries and ovarian tumors of Fshb-/- and Inha-/- mice, respectively (Tables 5Go and 6Go). These findings were confirmed by Northern blot analyses, which also demonstrated up-regulation of SN2 mRNA expression in adrenal tumors of Inha-/- mice (Figs. 1Go, 3Go, and 4Go). In situ hybridization showed that ovarian SN2 mRNA is localized exclusively to granulosa cells of growing follicles and is not detectable in corpora lutea (Fig. 5Go, C and D). Thus, the increased level of ovarian SN2 mRNA in Fshb null mice reflects in part the relative enrichment of nonluteinized granulosa cells rather than direct gene regulation. SN2 is a neutral amino acid transporter that is known to be highly expressed in many tissues in rats and may be important in tumor development, because it makes glutamine available for catabolism or ammonia export (34, 35). Based on this, it seems that ovarian SN2 is transcribed in an FSH-independent manner and encodes a protein with potential roles in regulating or supporting the proliferation of granulosa cells.

Amino acid metabolism is regulated within cells by specific processing enzymes, including the branched chain aminotransferases (BCATs), which facilitate the first catabolic steps in the processing of leucine, isoleucine, and valine. One BCAT-encoding gene, Bcat1 or ECA39, is tightly linked to the Kras2 oncogene in both the mouse and human genomes. It was originally identified as being overexpressed and amplified in teratocarcinomas (36) and was further described as being a direct transcriptional target of the c-myc oncogene (37, 38). Bcat1 mRNA was identified in our microarray studies as being up-regulated in ovarian tumors, although by Northern blot analysis we found that this was true in adrenal tumors and was not readily reproducible in ovarian tumors (Table 5Go and Figs. 3Go and 4Go). Despite these findings, which suggest that BCAT1 promotes cell proliferation, our in situ hybridization shows that Bcat1 mRNA is highly expressed in terminally differentiated luteal cells (Fig. 5Go, E and F). There is other evidence that BCAT1 may function dichotomously in modulating cell cycle progression, as loss of the yeast ortholog leads to faster cell growth and a shortened G1 phase (39). Although their specific functions remain unclear, our identification of SN2 and BCAT1 calls attention to the pathways regulating granulosa cell amino acid import and metabolism and how they may participate as determinants of granulosa cell division and differentiation.

The transcription factor FOXO1 (FKHR) is classified as a Forkhead family member based on its three {alpha}-helix, two loop DNA-binding domain. There is accumulating evidence that these Forkhead family members are important regulators of cell death and cell cycle progression. Our observations of Foxo1 mRNA expression in oocytes of early follicle stages and in the granulosa cells of secondary and preovulatory follicles, but not corpora lutea (Fig. 5Go, G and H), corroborate findings recently reported by Richards et al. (40). Based on granulosa cell culture studies, this group reported that FOXO1 function seems to be antagonized at the transcriptional and posttranslational levels by FSH, which down-regulates the expression of Foxo1 mRNA and causes the phosphorylation and nuclear to cytoplasmic sequestration of the FOXO1 protein. Consistent with this premise, we observed high levels of Foxo1 mRNA expression in granulosa cells of FSHß knockout ovaries, although Foxo1 was not identified in our gene chip analyses for the subtle increase in expression seen in the absence of FSH (Fig. 1Go), which we ascribe to be largely due to enrichment of this cell population rather than direct gene derepression. Rather, Foxo1 mRNA was called to our attention because of its high level of expression in granulosa cell tumors in inhibin {alpha} knockout females despite high levels of circulating FSH, and this was confirmed by Northern blot analysis and in situ hybridization (Fig. 4Go and data not shown).

Interstitial cell effects of the gonadotropins
Just as these studies identified genes expressed in granulosa cells, as expected, a surprising number of interstitial cell gene products were brought to our attention. Rather than representing direct targets of FSH or PMSG, we hypothesize that transcriptional control of these genes reflects interstitial cell responses to elevated levels of LH in Fshb-/- mice or occurs in response to steroid production in the granulosa cell compartment. Two mRNAs encoding 24p3 and vascular cell adhesion molecule 1 (VCAM1), identified in the Affymetrix array analysis as dramatically up-regulated in the ovaries of FSHß-deficient mice, were studied by in situ hybridization analyses. Both transcripts were highly expressed in the ovarian interstitial cell population in Fshb-/- mice, and no signal was detected in granulosa cells or oocytes of ovarian follicles (Fig. 5Go, I and J). No expression was detectable in ovaries from age-matched, wild-type mice.

The 24p3 gene encodes a hydrophobic molecule transporter protein of the lipocalin family. Originally called an oncogene up-regulated in simian virus 40-transformed cells, 24p3 mRNA has been shown to be induced in response to glucocorticoid and retinoic acid (41) and is induced by steroid hormone signaling in the mouse uterus (42). 24p3 is also expressed in the epithelial cells of the mouse epididymis, and the protein associates with the acrosomes of caudal spermatozoa (43). We find by oligonucleotide array and Northern blot analyses that 24p3 is highly expressed in ovaries of FSHß knockout mice, although it is virtually undetectable in the wild-type ovary (Table 2Go and Fig. 1Go). In situ hybridization reveals that the mRNA species accumulates within the interstitial cell population between the arrested follicles (Fig. 5IGo); no signal was detected in wild-type ovary. Notably, 24p3 protein exhibits closest homology to prostaglandin D synthase, which was also identified in the Affymetrix analysis as being up-regulated in the absence of FSH (Table 2Go); both are down-regulated with PMSG administration (Table 3Go).

VCAM1 (CD106) is a transmembrane protein belonging to the immunoglobulin superfamily, which is known to mediate cell-cell interactions by binding to {alpha}4 integrins and has been implicated in inflammatory processes (44). Other binding partners for VCAM1 have not been described, and the significance of its expression in the Fshb-/- ovary is not clear, since integrin {alpha}4 is termed absent by Affymetrix analysis in both wild-type ovary (45) and Fshb-/- ovary. Targeted disruption of Vcam1 in vivo results in lethality between d 10–12 of mouse embryonic development due to defects in the placenta and heart, and this has precluded studies of its functions in other tissues (46, 47). Defects in VCAM1 expression are also seen in the embryonic lethal phenotypes of knockouts lacking SMAD1 and the Forkhead transcription factor FOXF1, although whether these changes reflect direct Vcam1 gene regulation is not clear (48, 49). Like 24p3, in situ hybridization reveals VCAM1 mRNA within the interstitial cell population of Fshb-/- ovaries (Fig. 5JGo).

Two additional interstitial cell expressed genes, Ereg and Ly6d, were identified in these studies as being up-regulated in Fshb-/- mice given PMSG. Epiregulin is a member of the epidermal growth factor family and may have effects that either inhibit or promote cell cycle progression depending on its context. Ereg has been identified previously as a granulosa cell expressed gene that is responsive to gonadotropins, and it has been speculated that epiregulin may be involved in enhancing cyclooxygenase 2 mRNA expression during the periovulatory period (30). Our in situ hybridization confirms that epiregulin transcript is expressed in granulosa cells of preovulatory follicles after hCG administration and also demonstrates high levels of expression in the thecal/interstitial cell population (Fig. 5Go, K and L). Based on the intensity of the in situ hybridization signal, Ereg does not seem to be differentially regulated between mural and cumulus granulosa cell populations. Ereg mRNA expression is not detectable in corpora lutea, indicating that the transcript is cleared upon ovulation and the differentiation of luteal cells.

LY6D (or ThB) is a cell surface glycoprotein that is involved in cell-cell interactions (50). Although its normal functions in vivo are not well understood, differential expression of Ly6d in tumor cell populations seems to be a determinant of how cancers interact with endothelial cells and blood vessel precursors to ultimately regulate the potential for tumor invasiveness, metastasis, and angiogenesis (51). By Northern blot and in situ hybridization analysis, Ly6d mRNA is weakly detected in the interstitial cells of wild-type ovaries after treatment with hCG (Fig. 2Go and data not shown). Ly6d is highly expressed in the outermost layer of the adrenal cortex of wild-type mice, and this expression is lost during adrenal tumorigenesis in the inhibin {alpha} knockout mouse model (Fig. 4Go and data not shown).

Androgen biosynthesis in the FSHß knockout mouse
Affymetrix oligonucleotide hybridization analyses of Fshb-/- ovaries indicated aberrant expression levels of a number of mRNAs encoding proteins associated with steroidogenic pathways, some of which we had previously studied (5). Additionally, this study identified two androgen-responsive genes encoding seminal vesicle F protein and seminal vesicle secretory protein as being up-regulated in the ovaries of Fshb-/- mice. Given the high levels of circulating LH in these mice (4) and their relative lack of aromatase mRNA for converting ovarian androgens to estrogens (5), we hypothesized that the lack of FSH would result in elevations in circulating androgens. To test this, we measured circulating testosterone in age-matched, adult wild-type and FSHß knockout mice and found that knockouts have a 4-fold increase in circulating testosterone over wild-type mice (0.94 ± 0.24 vs. 0.21 ± 0.07 ng/ml). Corroborating these findings, elevated testosterone levels have been recently reported for FSH receptor knockout mice (52). Thus, these studies indicate a potential for elevated androgens to contribute to the etiology of FSH deficiency syndromes in women.

Cell cycle control in oocytes: potential roles for Bcl210, PC3B, and Rbb7
In situ hybridization analyses demonstrated oocyte expression of mRNAs encoding the cell cycle regulators BCL2-like-10 (also known as BCLB, BOO, and DIVA), PC3B (also known as B cell translocation gene 4), and retinoblastoma binding protein 7 (RBBP7; also known as mRbAp46). The first two transcripts were identified as being down-regulated in tumor samples from inhibin {alpha} knockout mice; the last was identified as up-regulated in FSHß knockout ovaries. Their relative expression in each case reflects changes in tissue composition between these knockout models; there are no oocytes in advanced tumors, and there is a relative increase in oocytes in Fshb-/- ovaries (see below). In oocytes, these gene products may function to inhibit apoptotic programs and maintain meiotic arrest.

Bcl2l10 encodes an apoptotic protease-activating factor 1-binding protein that prevents apoptotic protease-activating factor 1 from activating downstream caspase enzymes (53, 54). Bcl2l10 mRNA is highly expressed in the mammalian ovary and epididymis based on our analyses and previous descriptions of the gene (53, 54). The observation of ovarian expression led to the speculation that BCL2-like-10 is a regulator of apoptotic programs that mediate follicle atresia. Our in situ hybridization analysis reveals that Bcl2l10 mRNA is expressed preferentially in oocytes, and that this expression persists in oocytes of primary to preovulatory stage follicles (Fig. 5MGo). This expression was also clearly visible in ovaries of Gdf9-/- and Fshb-/- mice (data not shown). Notably, there is no expression detectable in oocytes of primordial follicles; thus, BCL2-like-10 is more likely to influence oocyte apoptosis after follicle recruitment than the survival of primordial germ cells. Although a recent knockout study indicates that BCL2-like-10 is not alone necessary for ovarian development or function (55), it is possible that double mutant studies will indicate important, although redundant, in vivo roles for this protein. For example, BCL2-Like-10 may cooperate with BCLX or BCL2, both of which are important apoptotic inhibitors in mouse oocytes based on loss of function and gain of function mouse models (56, 57, 58).

Our in situ hybridization studies confirm the oocyte expression of PC3B, which was previously reported based on EST database and RT-PCR analyses (59). PC3B is a member of the PC3/B cell translocation gene family identified by EST database sequence analyses (59). Related gene products can inhibit cyclin D1 expression in a retinoblastoma-dependent manner and therefore limit the propensity of a cell to proceed from G1 to S phase (60). Similarly, PC3B is able to induce G1 arrest when expressed in NIH-3T3 cells (59). In addition to being expressed in oocytes, PC3B mRNA has been detected in testis and differentiating olfactory epithelium (59). We localized PC3B mRNA to growing oocytes from the primary stage throughout the remainder of follicle development (Fig. 5NGo and data not shown). There is no detectable expression in other ovarian cell types.

Finally, a third gene, Rbbp7, was identified in our studies as being preferentially expressed in mouse oocytes. Like Bcl2l10 and PC3B mRNAs, Rbbp7 transcript is detectable in the cytoplasm of oocytes at the primary follicle stage and in oocytes of more advanced ovarian follicles. It is also highly expressed in the arrested primary follicles of Gdf9 null ovaries (Fig. 5Go, O and P). RBBP7 was originally identified from HeLa cell lysates based on its binding affinity for RB, and ribonuclease protection assays indicated that it is expressed widely in mouse tissues, with highest expression in testis and ovary/uterus (61). RBBP7 (RbAp46) is closely related to RBBP4 (RbAp48), and both homologs are highly conserved between mouse and humans. Although this is the first report of Rbbp7 expression in mammalian oocytes, the Xenopus RbAp46/48 ortholog is known to complex with a 57-kDa histone deacetylase expressed only during oogenesis and early cleavage stage embryos (62). The functions of RBBP7 have not yet been identified in vivo, but may prove important in mediating chromatin configuration and accessibility to transcription factors in oocytes and early embryos.

Implications for future expression profiling endeavors in reproductive tissues
In a previous study, we used Affymetrix GeneChip analysis to define the direct downstream targets of GDF9 in granulosa cells (63). In these experiments we isolated granulosa cells from PMSG-treated females, placed them in culture, and maintained them for 5 h in the presence or absence of recombinant GDF9. These studies examined an isolated population of cells at one time point, the only difference between the two samples being GDF9 stimulation. We were able to confirm known targets of GDF9 (e.g. cyclooxygenase 2 and hyaluronan synthase 2) (64) and also identify novel downstream targets of GDF9 (e.g. pentraxin 3 and TNF-induced protein 6). In the present study, tissues composed of multiple cell lineages were used in their entirety, and several genes were identified as abnormally expressed in the absence of either FSH or inhibins. Although some of these genes may represent direct targets of FSH or participants in the tumor process, in situ hybridization analyses demonstrated that many transcripts were instead increased or decreased in representation because of structural alterations in the ovaries. Specifically, FSHß knockout ovaries have a block in folliculogenesis before antral follicle development, resulting in a relative increase in the number of secondary follicles and oocytes compared with wild-type adult ovaries, in which much of the ovary is filled with functional or regressing corpora lutea. The result is that genes expressed in oocytes, such as Zp1 and Rbbp7, are termed up-regulated in the absence of FSH. Similarly, markers of luteinization, including calcyclin and LH receptor, are termed down-regulated, as luteinization is precluded without FSH (5). Likewise, comparing ovarian tumors of inhibin {alpha} knockout mice and wild-type ovaries yields a host of down-regulated gene products specific to oocytes (e.g. Gdf9, Zp2, Zp3, Bcl2l10, and PC3B) and luteinized cells (e.g. calcyclin and LH receptor), which are not found in ovarian tumors. Thus, whether similar studies are being planned to investigate in vivo gene expression in the ovary or in other tissues, it is critical to perform in situ hybridizations to determine the principle cell population expressing a gene of interest. In addition, short-term in vitro experiments with single cell types and single ligands can be expected to simplify the results obtained and identify target genes more readily.

The data in this study have been compiled using two different expression profiling methodologies, Affymetrix oligonucleotide arrays and cDNA microarrays, and we find the results of these analyses to be complementary in that each has revealed different transcripts to be aberrantly expressed in analyses of the same RNA sample. Reasons for this include differences in the sequences represented on each array and may also extend to different sources of technical variability, as different strategies are employed by each to synthesize riboprobe, different conditions are used for the hybridization, and different quantifications and mathematical analyses to identify regulated transcripts are used. For each type of analysis, we did not perform replicate analyses on the same RNA sample, and data indicate that for genes that are highly or even moderately expressed provide good reproducibility in such analyses (65). We followed up specific changes by Northern blot analysis rather than combining multiple array studies for samples that were relatively homogeneous, such as ovarian samples pooled from a group of wild-type mice or FSHß knockout mice. We relied on the fact that these samples were pooled from a large group of genetically identical and age-matched mice to minimize changes in gene expression in response to stress effects (known to vary more widely in comparative studies of nonovarian tissues) (65) or other causes of variability from individual to individual. This concern presents another rationale for single cell analyses, as different samples of cultured cells demonstrate variability in gene chip studies comparable to those attributable to technical variability from chip to chip (65). We adopted a different strategy in analyzing tumor samples from inhibin {alpha} knockout mice, profiling two ovarian tumors, and subsequently annotating our results with gene changes shared by analysis of an adrenal tumor from an inhibin {alpha} knockout mouse and a study of an ovarian tumor from an LHß-overexpressing mouse. Using two tumor samples enabled us to narrow a list of approximately 250 gene products defined as regulated in either sample to a list half this size of gene expression changes shared by both samples. Consistent with our hypothesis that there are commonalities in the pathophysiology of granulosa cell tumors of different mouse models and between these ovarian tumors and the adrenal tumors that occur in the absence of inhibins, we find several examples of shared gene expression changes. Together, these studies suggest several avenues for collecting and verifying expression profiling data.

In conclusion, this work has identified a myriad of gene products, many of which have not been previously characterized in the mammalian ovary or adrenal gland or implicated in ovarian tumorigenesis. We focused more in-depth in situ hybridization analyses of gene expression on transcripts encoding putative cell cycle regulators and have uncovered mRNAs preferentially expressed in every ovarian cell type. Only a few similar studies have been undertaken and reported to date, and none with a more extensive series of follow-up expression studies (19, 66, 67). Together, these analyses will prompt future investigations into how specific gene products contribute to normal and aberrant ovarian functioning. Some of the differential gene expressions uncovered may prove important in the endocrinopathies modeled by our transgenic mice, whereas others, perhaps equally relevant to our understanding of ovarian biology, reflect the results of alterations in tissue composition between these models.


   Acknowledgments
 
We thank Drs. Ruth Keri, T. Rajendra Kumar, and Eli Adashi for their thought-provoking discussions; the Jameson Laboratory at Northwestern University Medical School for the gift of the Dax1 probe; and Julio Agno, Valerie Long, Hillary McGraw, and Kimberly Smith for their technical assistance and expertise.


   Footnotes
 
Sequencing was performed at the Mental Retardation Research Center core facility at Baylor College of Medicine, supported by Grant HD-24064. Microarray hybridization was performed at the Population Center Gene Array Facility, which receives support from the Specialized Cooperative Centers Program in Reproductive Research and the National Cooperative Program for Infertility Research. Serum sample analyses were performed by the UVA Ligand Core Facility, supported by the NICHD/NIH through Cooperative Agreement U54-HD-28934 as part of the Specialized Cooperative Centers Program in Reproduction Research. This work was supported by NIH Grants CA-60651 (to M.M.M.) and CA-086387 (to J.H.N.). K.H.B. and G.E.O. are students in the Medical Scientist Training Programs at Baylor College of Medicine and Case Western Reserve University School of Medicine supported by NIH Grants T32-GM-07330 and T32-GM-007250, respectively.

Abbreviations: Bcat, Branched chain aminotransferase 1; Ereg, epiregulin; EST, expressed sequence tag; Gapd, glyceraldehyde 3-phosphate dehydrogenase; hCG, human chorionic gonadotropin; Ly6d, lymphocyte antigen 6 complex; Msg1, melanocyte-specific gene 1; PMSG, pregnant mare serum gonadotropin; RBBP7, retinoblastoma binding protein 7; VCAM1, vascular cell adhesion molecule 1.

Received April 16, 2003.

Accepted for publication June 17, 2003.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results and Discussion
 References
 

  1. Matzuk MM, Burns K, Viveiros MM, Eppig J 2002 Intercellular communication in the mammalian ovary: oocytes carry the conversation. Science 296:2178–2180[Abstract/Free Full Text]
  2. Matzuk MM, Lamb DJ 2002 Genetic dissection of mammalian fertility pathways. Nat Med 8:S41–S49
  3. Burns KH, Matzuk MM 2002 Genetic models for the study of gonadotropin actions. Endocrinology 143:2823–2835[Abstract/Free Full Text]
  4. Kumar TR, Wang Y, Lu N, Matzuk MM 1997 Follicle stimulating hormone is required for ovarian follicle maturation but not male fertility. Nat Genet 15:201–204[CrossRef][Medline]
  5. Burns KH, Yan C, Kumar TR, Matzuk MM 2001 Analysis of ovarian gene expression in follicle-stimulating hormone ß knockout mice. Endocrinology 142:2742–51[Abstract/Free Full Text]
  6. Dierich A, Sairam MR, Monaco L, Fimia GM, Gansmuller A, LeMeur M, Sassone-Corsi P 1998 Impairing follicle-stimulating hormone (FSH) signaling in vivo: targeted disruption of the FSH receptor leads to aberrant gametogenesis and hormonal imbalance. Proc Natl Acad Sci USA 95:13612–13617[Abstract/Free Full Text]
  7. Abel MH, Wootton AN, Wilkins V, Huhtaniemi I, Knight PG, Charlton HM 2000 The effect of a null mutation in the follicle-stimulating hormone receptor gene on mouse reproduction. Endocrinology 141:1795–1803[Abstract/Free Full Text]
  8. Danilovich N, Roy I, Sairam MR 2001 Ovarian pathology and high incidence of sex cord tumors in follitropin receptor knockout (FORKO) mice. Endocrinology 142:3673–3684[Abstract/Free Full Text]
  9. Matzuk MM, Finegold MJ, Su J-GJ, Hsueh AJW, Bradley A 1992 {alpha}-Inhibin is a tumor-suppressor gene with gonadal specificity in mice. Nature 360:313–319[CrossRef][Medline]
  10. Matzuk MM, Finegold MJ, Mather JP, Krummen L, Lu H, Bradley A 1994 Development of cancer cachexia-like syndrome and adrenal tumors in inhibin-deficient mice. Proc Natl Acad Sci USA 91:8817–8821[Abstract/Free Full Text]
  11. Kumar TR, Wang Y, Matzuk MM 1996 Gonadotropins are essential modifier factors for gonadal tumor development in inhibin-deficient mice. Endocrinology 137:4210–4216[Abstract]
  12. Kumar TR, Palapattu G, Wang P, Woodruff TK, Boime I, Byrne MC, Matzuk MM 1999 Transgenic models to study gonadotropin function: the role of follicle-stimulating hormone in gonadal growth and tumorigenesis. Mol Endocrinol 13:851–865[Abstract/Free Full Text]
  13. Shou W, Woodruff TK, Matzuk MM 1997 Role of androgens in testicular tumor development in inhibin-deficient mice. Endocrinology 138:5000–5003[Abstract/Free Full Text]
  14. Richards JS, Russell DL, Ochsner S, Hsieh M, Doyle KH, Falender AE, Lo YK, Sharma SC 2002 Novel signaling pathways that control ovarian follicular development, ovulation, and luteinization. Recent Prog Horm Res 57:195–220[Abstract/Free Full Text]
  15. Robker RL, Richards JS 1998 Hormonal control of the cell cycle in ovarian cells: proliferation versus differentiation. Biol Reprod 59:476–482[Free Full Text]
  16. Risma KA, Clay CM, Nett TM, Wagner T, Yun J, Nilson JH 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[Abstract/Free Full Text]
  17. Keri RA, Lozada KL, Abdul-Karim FW, Nadeau JN, Nilson JH 2000 Luteinizing hormone induction of ovarian tumors: oligogenic differences between mouse strains dictates tumor disposition. Proc Natl Acad Sci USA 97:383–387[Abstract/Free Full Text]
  18. Nilson JH, Abbud RA, Keri RA, Quirk CC 2000 Chronic hypersecretion of luteinizing hormone in transgenic mice disrupts both ovarian and pituitary function, with some effects modified by the genetic background. Recent Prog Horm Res 55:69–91
  19. Owens GE, Keri RA, Nilson JH 2002 Ovulatory surges of human CG prevent hormone-induced granulosa cell tumor formation leading to the identification of tumor-associated changes in the transcriptome. Mol Endocrinol 16:1230–1242[Abstract/Free Full Text]
  20. Themmen APN, Huhtaniemi IT 2000 Mutations of gonadotropins and gonadotropin receptors: elucidating the physiology and pathophysiology of pituitary-gonadal function. Endocr Rev 21:551–583[Abstract/Free Full Text]
  21. Shenker A, Laue L, Kosugi S, Merendino Jr JJ, Minegishi T, Cutler Jr GB 1993 A constitutively activating mutation of the luteinizing hormone receptor in familial male precocious puberty. Nature 365:652–654[CrossRef][Medline]
  22. Shelling AN, Burton KA, Chand AL, van Ee CC, France JT, Farquhar CM, Milsom SR, Love DR, Gersak K, Aittomaki K, Winship IM 2000 Inhibin: a candidate gene for premature ovarian failure. Hum Reprod 15:2644–2649[Abstract/Free Full Text]
  23. Dong J, Albertini DF, Nishimori K, Kumar TR, Lu N, Matzuk MM 1996 Growth differentiation factor-9 is required during early ovarian folliculogenesis. Nature 383:531–535[CrossRef][Medline]
  24. Elvin JA, Yan C, Wang P, Nishimori K, Matzuk MM 1999 Molecular characterization of the follicle defects in the growth differentiation factor-9-deficient ovary. Mol Endocrinol 13:1018–1034[Abstract/Free Full Text]
  25. Mahmoudi M, Lin VK 1989 Comparison of two different hybridization systems in Northern transfer analysis. BioTechniques 7:331–332[Medline]
  26. Johnston RF, Pickett SC, Barker DL 1990 Autoradiography using storage phosphor technology. Electrophoresis 11:355–360[CrossRef][Medline]
  27. Albrecht U, Eichele G, Helms JA, Lu HC 1997 Visualization of gene expression patterns by in situ hybridization. In: Daston GP, ed. Molecular and cellular methods in developmental toxicology. Boca Raton, FL: CRC Press; 23–48
  28. McIntosh JE, Kaethner M, Stewart F, Allen WR, Moor RM 1976 Proceedings: FSH AND LH activites of PMSG fractions isolated from serum and from fetal trophoblast cells maintained in culture. J Reprod Fertil 46:521–522[CrossRef]
  29. Stewart F, Allen WR, Moor RM 1976 Pregnant mare serum gonadotrophin: ratio of follicle-stimulating hormone and luteinizing hormone activities measured by radioreceptor assay. J Endocrinol 71:471–482[Abstract]
  30. Espey LL, Richards JS 2002 Temporal and spatial patterns of ovarian gene transcription following an ovulatory dose of gonadotropin in the rat. Biol Reprod 67:1662–1670[Abstract/Free Full Text]
  31. Yahata T, de Caestecker MP, Lechleider RJ, Andriole S, Roberts AB, Isselbacher KJ, Shioda T 2000 The MSG1 non-DNA-binding transactivator binds to the p300/CBP coactivators, enhancing their functional link to the Smad transcription factors. J Biol Chem 275:8825–8834[Abstract/Free Full Text]
  32. Yahata T, Shao W, Endoh H, Hur J, Coser KR, Sun H, Ueda Y, Kato S, Isselbacher KJ, Brown M, Shioda T 2001 Selective coactivation of estrogen-dependent transcription by CITED1 CBP/p300-binding protein. Genes Dev 15:2598–2612[Abstract/Free Full Text]
  33. Shioda T, Fenner MH, Isselbacher KJ 1996 msg1, a novel melanocyte-specific gene, encodes a nuclear protein and is associated with pigmentation. Proc Natl Acad Sci USA 93:12298–12303[Abstract/Free Full Text]
  34. Nakanishi T, Kekuda R, Fei YJ, Hatanaka T, Sugawara M, Martindale RG, Leibach FH, Prasad PD, Ganapathy V 2001 Cloning and functional characterization of a new subtype of the amino acid transport system N. Am J Physiol 281:C1757–C1768
  35. Bode BP, Fuchs BC, Hurley BP, Conroy JL, Suetterlin JE, Tanabe KK, Rhoads DB, Abcouwer SF, Souba WW 2002 Molecular and functional analysis of glutamine uptake in human hepatoma and liver-derived cells. Am J Physiol 283:G1062–G1073
  36. Niwa O, Kumazaki T, Tsukiyama T, Soma G, Miyajima N, Yokoro K 1990 A cDNA clone overexpressed and amplified in a mouse teratocarcinoma line. Nucleic Acids Res 18:6709[Free Full Text]
  37. Benvenisty N, Leder A, Kuo A, Leder P 1992 An embryonically expressed gene is a target for c-Myc regulation via the c-Myc-binding sequence. Genes Dev 6:2513–2523[Abstract/Free Full Text]
  38. Ben-Yosef T, Eden A, Benvenisty N 1998 Characterization of murine BCAT genes: Bcat1, a c-Myc target, and its homolog, Bcat2. Mamm Genome 9:595–597[CrossRef][Medline]
  39. Schuldiner O, Eden A, Ben-Yosef T, Yanuka O, Simchen G, Benvenisty N 1996 ECA39, a conserved gene regulated by c-Myc in mice, is involved in G1/S cell cycle regulation in yeast. Proc Natl Acad Sci USA 93:7143–7148[Abstract/Free Full Text]
  40. Richards JS, Sharma SC, Falender AE, Lo YH 2002 Expression of FKHR, FKHRL1, and AFX genes in the rodent ovary: evidence for regulation by IGF-I, estrogen, and the gonadotropins. Mol Endocrinol 16:580–599[Abstract/Free Full Text]
  41. Garay-Rojas E, Harper M, Hraba-Renevey S, Kress M 1996 An apparent autocrine mechanism amplifies the dexamethasone- and retinoic acid-induced expression of mouse lipocalin-encoding gene 24p3. Gene 170:173–180[CrossRef][Medline]
  42. Huang HL, Chu ST, Chen YH 1999 Ovarian steroids regulate 24p3 expression in mouse uterus during the natural estrous cycle and the preimplantation period. J Endocrinol 162:11–19[Abstract]
  43. Chu ST, Lee YC, Nein KM, Chen YH 2000 Expression, immunolocalization and sperm-association of a protein derived from 24p3 gene in mouse epididymis. Mol Reprod Dev 57:26–36[CrossRef][Medline]
  44. Osborn L, Hession C, Tizard R, Vassallo C, Luhowskyj S, Chi-Rosso G, Lobb R 1989 Direct expression cloning of vascular cell adhesion molecule 1, a cytokine-induced endothelial protein that binds to lymphocytes. Cell 59:1203–1211[CrossRef][Medline]
  45. Burns KH, Owens GE, Fernandez JM, Nilson JH, Matzuk MM 2002 Characterization of integrin expression in the mouse ovary. Biol Reprod 67:743–751[Abstract/Free Full Text]
  46. Kwee L, Baldwin HS, Shen HM, Stewart CL, Buck C, Buck CA, Labow MA 1995 Defective development of the embryonic and extraembryonic circulatory systems in vascular cell adhesion molecule (VCAM-1) deficient mice. Development 121:489–503[Abstract]
  47. Gurtner GC, Davis V, Li H, McCoy MJ, Sharpe A, Cybulsky MI 1995 Targeted disruption of the murine VCAM1 gene: essential role of VCAM-1 in chorioallantoic fusion and placentation. Genes Dev 9:1–14[Abstract/Free Full Text]
  48. Lechleider RJ, Ryan JL, Garrett L, Eng C, Deng C, Wynshaw-Boris A, Roberts AB 2001 Targeted mutagenesis of Smad1 reveals an essential role in chorioallantoic fusion. Dev Biol 240:157–167[CrossRef][Medline]
  49. Mahlapuu M, Ormestad M, Enerback S, Carlsson P 2001 The Forkhead transcription factor Foxf1 is required for differentiation of extra-embryonic and lateral plate mesoderm. Development 128:155–166[Abstract]
  50. Gumley TP, McKenzie IF, Sandrin MS 1995 Tissue expression, structure and function of the murine Ly-6 family of molecules. Immunol Cell Biol 73:277–296[Medline]
  51. Eshel R, Zanin A, Kapon D, Sagi-Assif O, Brakenhoff R, van Dongen G, Witz IP 2002 Human Ly-6 antigen E48 (Ly-6D) regulates important interaction parameters between endothelial cells and head-and-neck squamous carcinoma cells. Int J Cancer 98:803–810[CrossRef][Medline]
  52. Tam J, Danilovich N, Nilsson K, Sairam MR, Maysinger D 2002 Chronic estrogen deficiency leads to molecular aberrations related to neurodegenerative changes in follitropin receptor knockout female mice. Neuroscience 114:493–506[CrossRef][Medline]
  53. Song Q, Kuang Y, Dixit VM, Vincenz C 1999 Boo, a novel negative regulator of cell death, interacts with Apaf-1. EMBO J 18:167–178[CrossRef][Medline]
  54. Inohara N, Gourley TS, Carrio R, Muniz M, Merino J, Garcia I, Koseki T, Hu Y, Chen S, Nunez G 1998 Diva, a Bcl-2 homologue that binds directly to Apaf-1 and induces BH3-independent cell death. J Biol Chem 273:32479–3286[Abstract/Free Full Text]
  55. Russell HR, Lee Y, Miller HL, Zhao J, McKinnon PJ 2002 Murine ovarian development is not affected by inactivation of the bcl-2 family member diva. Mol Cell Biol 22:6866–6870[Abstract/Free Full Text]
  56. Rucker III EB, Dierisseau P, Wagner KU, Garrett L,