Changes in Mouse Granulosa Cell Gene Expression during Early Luteinization
http://www.100md.com
内分泌学杂志 2005年第1期
Department of Veterinary Preclinical Studies, University of Glasgow Veterinary School, University of Glasgow, Glasgow G61 1QH, United Kingdom
Address all correspondence and requests for reprints to: P. J. O’Shaughnessy, Institute of Comparative Medicine, University of Glasgow Veterinary School, Bearsden Road, Glasgow G61 1QH, United Kingdom. E-mail: p.j.o’shaughnessy@vet.gla.ac.uk.
Abstract
Changes in gene expression during granulosa cell luteinization have been measured using serial analysis of gene expression (SAGE). Immature normal mice were treated with pregnant mare serum gonadotropin (PMSG) or PMSG followed, 48 h later, by human chorionic gonadotropin (hCG). Granulosa cells were collected from preovulatory follicles after PMSG injection or PMSG/hCG injection and SAGE libraries generated from the isolated mRNA. The combined libraries contained 105,224 tags representing 40,248 unique transcripts. Overall, 715 transcripts showed a significant difference in abundance between the two libraries of which 216 were significantly down-regulated by hCG and 499 were significantly up-regulated. Among transcripts differentially regulated, there were clear and expected changes in genes involved in steroidogenesis as well as clusters of genes involved in modeling of the extracellular matrix, regulation of the cytoskeleton and intra and intercellular signaling. The SAGE libraries described here provide a base for functional investigation of the regulation of granulosa cell luteinization.
Introduction
THE TERMINAL DIFFERENTIATION event of ovarian follicles that avoid atresia is luteinization to form the corpus luteum. Studies using hypophysectomized, mutant, and knockout animals have shown that FSH regulates granulosa cell proliferation and development from the secondary stage to the preovulatory stage (1, 2, 3). Luteinization of these cells occurs after the ovulatory LH surge and is dependent upon expression of LH receptors on the developing granulosa cells. The regulatory mechanisms inducing luteinization remain uncertain, but it is clear that granulosa cells begin to luteinize before ovulation and the process may be linked to disintegration of cell-to-cell communication between granulosa cells and the oocyte (4, 5).
After the LH surge and ovulation, the corpus luteum rapidly develops to become a highly active steroidogenic tissue and it plays an essential role in the establishment and maintenance of pregnancy. Several functional differences between presurge granulosa cells and luteal cells have been identified but the underlying, early molecular events that occur during terminal granulosa cell differentiation remain unclear. Earlier studies using microarray, differential display and subtractive hybridization techniques have identified a number of genes that show changes in expression during early luteinization (6, 7). This work has provided valuable insight into the luteinization process, but it is likely that genes identified in these studies represent a small fraction of those undergoing change. As a step toward understanding the process of luteinization, and to identify genes undergoing regulation during the early phase of terminal differentiation, we have used the technique of serial analysis of gene expression (SAGE) (8) to provide a comprehensive profile of gene expression in granulosa cells before and after a luteinizing dose of human chorionic gonadotropin (hCG).
Materials and Methods
nimals and treatments
The mice used in this study were bred at the University of Glasgow Veterinary School and were maintained as required under United Kingdom Home Office regulations. Normal mice, bred on a C3H/Heh-101/H genetic background, were derived from stock animals originally obtained from the MRC Radiobiology Unit (now the MRC, Mammalian Genetics Unit, Harwell, UK).
To generate a preovulatory granulosa cell SAGE library, immature mice (aged 18–22 d) were injected ip with a single dose of 5 IU pregnant mare serum gonadotropin (PMSG) and granulosa cells were collected 48 h later. To generate a SAGE library of granulosa cells undergoing luteinization, immature mice were treated with PMSG followed 48 h later by hCG, and granulosa cells were isolated after a further 12 h. To isolate the cells, ovaries were placed in DMEM (Invitrogen Ltd., Paisley, Scotland, UK) and granulosa cells were released from large antral follicles using needles. No attempt was made to remove oocytes from the isolated granulosa cells. Cells were stored in liquid N2 until RNA extraction.
Construction and analysis of SAGE libraries
The RNA from PMSG and PMSG/hCG-treated granulosa cells was extracted using TRIzol reagent (Invitrogen Ltd.). The SAGE libraries were constructed from RNA pooled from granulosa cells of 30 different mice. Poly-A+ RNA was isolated using oligo deoxythymidine cellulose (Invitrogen Ltd.) and SAGE libraries were constructed and analyzed as previously described (9). Fourteen nucleotide SAGE tags, representing individual transcripts, were extracted from the sequence data and initially analyzed using SAGE 2000 software (http://www.sagenet.org/sage_protocol.htm). Differences in tag frequency between SAGE libraries were statistically analyzed using the 2 test (10). Tags were identified through mouse SAGEmap database build 136. The original SAGE data from each library described here have been deposited in the National Center for Biotechnology Information (NCBI) public gene expression database (http://www.ncbi.nlm.nih.gov/geo/) accession nos. GSM30721 and GSM30722.
Rapid amplification of cDNA ends (RACE)
For those tags which could not be identified by comparison to the NCBI database, it was necessary to generate additional 3' sequence to identify the source mRNA species. This was done using a 3' RACE technique similar to that described previously (11) but using biotinylated primer TTCTAGAATTCAGCGGCCGC(T)30(AGC)(AGCT) to prime reverse transcription.
Real-time PCR
Levels of specific mRNA species were measured by real-time PCR using the Taqman method as described previously (12). The tissue used to generate RNA for real-time PCR was prepared using the same protocols as above for generation of SAGE libraries (see Construction and analysis of SAGE libraries). The sequences of primers and probes used for real-time PCR were as shown in Table 1 and in previous studies (12, 13).
TABLE 1. Sequences of real-time PCR primers and probes used in this study
Results
Ovarian response to hormone treatment
After a single injection of PMSG, the ovaries of immature mice contained numerous large antral follicles which provided granulosa cells for the PMSG SAGE library. These follicles ovulated around 13–14 h after a subsequent injection of hCG and luteinizing granulosa cells were, therefore, collected from intact follicles 12 h after hCG injection to generate the PMSG/hCG library.
SAGE libraries
The total number of tags sequenced in the PMSG library (treatment with PMSG alone) was 51,528, and the total number sequenced in the PMSG/hCG library (treatment with PMSG followed by hCG) was 53,696. The combined total of 105,224 tags corresponded to 40,248 unique transcripts, of which 9,877 were represented by two or more tags. Of the transcripts represented by more than one tag, 5,689 were shared between both libraries, 1,806 were unique to the PMSG library, and 2,382 were unique to the PMSG/hCG library. Using 2 analysis to test for significant differences in tag abundance between those tags with greater than five transcripts present in the combined libraries, 499 tags were significantly up-regulated by hCG treatment (P < 0.05), and 216 tags were significantly down-regulated (P < 0.05).
Abundant, differentially expressed SAGE tags
The most abundant tags that were shown to be differentially expressed between SAGE libraries by 2 test are shown in Table 2. Most of these tags match unambiguously to known genes and a number of them are known to be expressed in granulosa cells during FSH-dependent development [e.g. inhibin ?B and follistatin (14, 15)] or to show a change in expression after luteinization [e.g. P450 11a1, 17?-hydroxysteroid dehydrogenase type 1 and scavenger receptor class B type 1 (SRB1) (16, 17, 18)]. This list also contains 36 tags, including six of the eight most abundant tags, which have no match, are linked to sequence of unknown function or have multiple assignments. A number of these tags, without confirmed or reliable matches in the SAGEmap database, were analyzed further using 3'RACE (indicated by an asterisk in Table 2). The transcript associated with the tag CAGTCAATAC is of interest because it shows abundant, differentially regulated expression. The 3'RACE data from this tag matched it to Unigene cluster Mm.290944, which shows 95% homology with a noncoding human mRNA sequence. This tag, in addition, is not highly expressed in other mouse SAGE libraries and appears to show a degree of selectivity for the granulosa cell. The expression pattern for this sequence has been confirmed using real-time PCR (Fig. 1, noncoding RNA). Other tags extended by 3'RACE could be linked to Unigene clusters or to individual expressed sequence tags (ESTs). These ESTs may be linked to Unigene clusters (e.g. TTGTTGCTAC matches EST BY419395, which is located on the genome close to Mm.190648) or may represent unique transcribed regions.
TABLE 2. Most abundant differentially expressed tags
TABLE 2A. Continued
FIG. 1. Comparison of SAGE and real-time PCR data. Results show changes in expression of specific genes in granulosa cells of PMSG-treated ovaries from mice before or after injection with hCG. Gene expression was measured by real-time PCR or by SAGE and results are expressed as fold increase or fold decrease in expression after hCG. The real-time PCR data show the mean of six pools of RNA in each group, and each RNA pool was derived from granulosa cells isolated from two or more animals. The SAGE data were derived from that shown in Tables 2 and 3. arom, Cytochrome P450 aromatase; ?-act, ?-actin; FSHR, FSH receptor; kit, kit ligand; LHR, LH receptor; ncRNA, noncoding RNA; scc, cytochrome P450 side chain cleavage; wbscr1, Williams-Beuren syndrome chromosome region 1.
Functional groups
Table 3 shows the frequency of expression of a number of tags in the PMSG and PMSG/hCG libraries, which represent genes linked with established functional groups. The list is not exhaustive and does not, in general, include genes with low levels of expression. This table is designed to show changes (or lack of them) in the expression of selected genes during granulosa cell luteinization. In each of the categories, there are genes listed that have been shown previously to be regulated during luteinization or which are know to be expressed in the corpus luteum [e.g. gap junction protein Cx43 (1), steroidogenic acute regulatory protein (StAR), adrenodoxin, tissue inhibitor of metalloproteinase 1 (TIMP 1), and cathepsin L (6, 19, 20, 21, 22, 23)]. In addition, there are a number of genes listed which have not previously been shown to be regulated during luteinization such as cofilin, seizure-related 6 homolog (mouse)-like 2 (Sez6l2), and the listed transcription factors.
TABLE 3. Tags associated with specific functional groups
TABLE 3A. Continued
To determine the contribution of oocyte mRNA to the SAGE libraries, the frequency of tags representing oocyte-specific genes (24, 25, 26) was determined in the two libraries. There were no tags representing the genes Nobox, factor in the germline- (Fig), and ataxia-telangiectasia mutated (ATM) present in either library, whereas zona pellucida (ZP)-2 and ZP-3 were represented by a single tag each in one library. The signaling molecules bone morphogenic protein 15 (BMP15) and growth differentiation factor 9 (GDF9) were represented by a maximum of four and 11 tags, respectively.
Real-time PCR
Real-time PCR was used to determine whether differences in gene expression between granulosa cell SAGE libraries could be confirmed for a number of transcripts. To normalize the data from real-time PCR, the housekeeping gene GAPDH (glyceraldehyde-3-phosphate dehydrogenase) was used because it is expressed equally in both SAGE libraries (Table 3). Results in Fig. 1 show that, for these selected transcripts, there is good correlation between data obtained by real-time PCR and data derived from SAGE libraries.
Discussion
The SAGE method allows for a quantitative, qualitative, and comprehensive measurement of gene expression patterns in cell populations and tissues. We have applied SAGE, in this study, to an examination of the changes in transcript expression patterns in granulosa cells from mature antral follicles before and after the start of gonadotropin-induced luteinization. One potential problem with the SAGE technique is that it can be time consuming, and this can impose a practical limitation on the number of libraries generated in any one particular study. Thus, in this report we have limited identification of genes involved in luteinization to a single time after induction by hCG. Nevertheless, the data produced from these studies provide both a base for studies into granulosa cell development and function and a comprehensive list of genes undergoing regulation during early granulosa cell luteinization.
A number of previous studies have examined changes in gene expression in granulosa cells during luteinization (6, 18, 19, 21, 22, 27, 28). In general, these studies examined changes in a single or limited number of genes over the time course of luteinization and have served to provide a valuable insight into the development of the corpus luteum. Genes that have been shown to alter expression during luteinization include, for example, the steroidogenic enzymes and associated proteins [e.g. P45011a1, adrenodoxin, P45019a1, StAR, and SRB1 (6, 18, 21, 22)], extracellular matrix modifiers [e.g. A disintegrin-like and metalloprotease (reprolysin type) with thrombospondin type 1 motif 1 (ADAMTS-1), cathepsin L (6, 19)], signaling molecules [e.g. epiregulin, secreted frizzled-related protein-4 and wnt 4 (27, 28)] and others [e.g. regulator of G protein signaling protein-2 and cell surface antigen CD63 (6)]. In general, results from these previous studies correlate well with the SAGE data reported here. When taken alongside results using real-time PCR (Fig. 1), this is a further indication that the SAGE results are a reliable measure of changes in gene expression at the selected stage of luteinization. Previous SAGE studies have also shown that the technique is highly reproducible and accurate (29).
The SAGE libraries reported in this study were generated predominantly from granulosa cell mRNA, but oocytes were present in the isolated cell mix and would have contributed to the total mRNA pool used in library construction. The expression of tags representing the oocyte-specific genes (24, 25, 26) Nobox, Fig, ATM, ZP-3, and ZP-2 was very low or undetectable, and only BMP15 and GDF9 were represented by more than a single tag. Both ZP-3 and ATM are known to be highly expressed within the oocyte (24), and it is unlikely, therefore, that oocyte-derived genes will contribute significantly to the overall tag number. Care in interpretation may be required for low abundance tags, however, because these could have an oocyte origin.
Genes known to be associated with luteinization
The most fundamental change in granulosa/luteal cell function after induction of luteinization is a marked increase in steroidogenic activity by the developing corpus luteum. It has been shown previously that this is associated with increased expression of P45011a1, SRB1, StAR, ferredoxin, and low-density lipoprotein receptor and decreased expression of P45019a1 (aromatase) and 17?-hydroxysteroid dehydrogenase type 1 (18, 21, 22). The overall effect is a reduction in estrogen production and a marked increase in progesterone production through increased availability of substrate and converting enzymes. As described above in second paragraph of Discussion, each of these previously reported changes in gene expression correlates well with changes in the tag numbers associated with the SAGE libraries generated for this study. One possible anomaly, however, between the SAGE data and previous studies is the expression of 3?-hydroxysteroid dehydrogenase (3?HSD) during luteinization. It has previously been shown that 3?HSD activity and expression increase in the developing corpus luteum in line with the general increase in steroidogenic activity (30, 31). Data from the SAGE libraries, in contrast, indicate that 3?HSD type 1 expression is reduced in granulosa cells 12 h after hCG treatment. This change was confirmed by real-time PCR and it appears likely, therefore, that luteinization leads to a temporary decrease in 3?HSD expression, perhaps through down-regulation induced by exposure to high gonadotropin levels, which reverses as the corpus luteum develops.
Data from the SAGE libraries show that the expression levels of several cytoskeletal-associated transcripts including actin, vinculin, cofilin, tubulin, and tropomyosin, are differentially regulated during luteinization. There is also significant differential expression of actin and tubulin isoforms. It is known that gonadotropins will regulate cytoskeletal gene expression in granulosa cells (32, 33) and cytoskeletal remodeling during luteinization is likely to be an essential part of the movement and morphological development of the cells.
Several membrane-binding and communication-related components such as clusterin, annexin A2, and the gap junction membrane channel proteins (connexins) are known to show fluctuations in expression levels during granulosa cell development and luteinization. The overwhelming level of Cx 43 expression in the PMSG SAGE library supports current thinking that this is the primary means of intercellular communication between granulosa cells and reinforces the hypothesis of a functional granulosa cell syncytium throughout folliculogenesis (34). Expression of Cx43 decreases markedly after induction of luteinization, as shown previously (23), although it may continue to be expressed in the developing corpus luteum (35).
Secreted acidic cysteine-rich glycoprotein (SPARC, osteonectin, basement membrane protein 40) is a highly expressed transcript which undergoes a 7-fold up-regulation during luteinization. SPARC is expressed in vivo where cells are undergoing proliferative or reorganizational activity (36), and it has previously been identified in follicular granulosa cells after the LH surge (37). It is possible that SPARC may play an essential role in the development of the corpus luteum because specific peptide fragments of the protein are strongly angiogenic (38). In the follicle, SPARC is found in both granulosa cells and oocytes, although expression in the oocyte may derive from adjacent granulosa cells (37). Calmodulin, a protein with strong structural and functional similarities to SPARC, has been implicated in the resumption of meiosis in the starfish oocyte (39), and it is possible that up-regulation of SPARC after hCG may play a role in allowing resumption of meiosis in the oocyte.
Genes with a poorly defined role in luteinization
In addition to confirming changes in gene expression previously reported or predicted, the SAGE data set reported here also identifies a number of genes of interest that have not previously been linked to the process of early luteinization. This list is made up of genes with unknown function and genes with known function but no previous association with granulosa luteinization. Genes with known function and differentially regulated during luteinization include syndecan-1, secreted phosphoprotein 1 (Spp1), prosaposin, and vanin 1.
Syndecan-1 is a heparin sulfate-rich integral membrane proteoglycan, and it is expressed in a developmental and cell type-specific pattern (40), but it has not previously been identified as having a role in folliculogenesis or luteinization. Syndecans have major roles as matrix and cell surface receptors, coreceptors for growth factor signaling, internalization receptors, and soluble paracrine effectors (40). In addition, syndecan-1 appears to be capable of independent signaling and may play a role in regulation of Wnt signaling (40). Syndecan-1 is not expressed in the PMSG-stimulated library but shows a high level of transcript expression after hCG administration. Syndecans have a well-established involvement in the regulation of cytoskeletal organization (40) and a likely function of syndecan-1 in the luteinizing follicle is in the regulation of cytoskeletal assembly.
Spp1 (also known as osteopontin or Eta-1) is among the most highly up-regulated tags after hCG administration. It is a multifunctional protein expressed in various cell types and involved in a number of physiological and pathological process including biomineralization, inflammation, leukocyte recruitment, cell survival, tissue repair, cell proliferation, and proliferation of vascular smooth muscle cells (41). It is possible, therefore, that Spp1 has multiple functions during luteinization. For example, it may act as a survival factor preventing onset of apoptosis during the critical phase of ovulation and luteinization. Equally, the effects on vascular smooth muscle suggest a possible role in the angiogenic process that accompanies formation of the corpus luteum.
Prosaposin is expressed at a medium level in the antral follicle but shows a 6-fold up-regulation after hCG administration. The protein is either secreted or acts as a precursor of smaller saposins, and it has been shown to have diverse functions including involvement in the MAPK and Akt signaling pathways and maintenance of cell growth, differentiation, and survival (42). The likely role of prosaposin in development of the corpus luteum is uncertain but may, again, relate to overall function as a survival factor.
Vanin 1 is a glycosylphosphatidylinositol-anchored cell surface molecule involved in thymic and gonadal development (43, 44). As with Spp1, it is also highly up-regulated in granulosa cells after hCG administration. Vanin 1 is expressed specifically in the Sertoli cells of the developing fetal gonad, and it has been suggested that it may be involved in the migration of mesenchynmal cells from the mesonephros into the developing gonad (43). The likely function of vanin 1 in the developing corpus luteum is unclear but, by analogy with developing thymic and gonadal tissue, it is possible that it may be involved in the cell migration that occurs early in corpus luteum formation to integrate both thecal and endothelial cells into the developing tissue.
For those genes already discussed in this section, it is possible to hypothesize a possible function in ovulation and corpus luteum development based on known properties and functions of the genes in other tissues. The SAGE libraries described here also contain many other highly expressed or differentially expressed transcripts for which this not currently possible. Included among this latter group are (Sez6l2), proline-rich protein MP-5, and epithelial V-like antigen. In addition, the list of tags differentially regulated after hCG contains a considerable number that are unmatched in the SAGEmap database or are matched only to EST clusters or uncharacterized transcripts. The challenge now will be to identify those genes from this list that are fundamentally involved in the process of luteinization and those genes that have a more downstream role in the development of the corpus luteum.
References
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
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
Halpin DM, Jones A, Fink G, Charlton HM 1986 Postnatal ovarian follicle development in hypogonadal (hpg) and normal mice and associated changes in the hypothalamic-pituitary ovarian axis. J Reprod Fertil 77:287–296
Smith MF, McIntush EW, Smith GW 1994 Mechanisms associated with corpus-luteum development. J Anim Sci 72:1857–1872
Simon AM, Goodenough DA, Li E, Paul DL 1997 Female infertility in mice lacking connexin 37. Nature 385:525–529
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
Leo CP, Pisarska MD, Hsueh AJ 2001 DNA array analysis of changes in preovulatory gene expression in the rat ovary. Biol Reprod 65:269–276
Velculescu VE, Zhang L, Vogelstein B, Kinzler KW 1995 Serial analysis of gene expression. Science 270:484–487
O’Shaughnessy PJ, Fleming L, Baker PJ, Jackson G, Johnston H 2003 Identification of developmentally-regulated genes in the somatic cells of the mouse testis using serial analysis of gene expression. Biol Reprod 69:797–808
Man MZ, Wang X, Wang Y 2000 POWER_SAGE: comparing statistical tests for SAGE experiments. Bioinformatics 16:953–959
Chen JJ, Rowley JD, Wang SM 2000 Generation of longer cDNA fragments from serial analysis of gene expression tags for gene identification. Proc Natl Acad Sci USA 97:349–353
O’Shaughnessy PJ, Willerton L, Baker PJ 2002 Changes in Leydig cell gene expression during development in the mouse. Biol Reprod 66:966–975
Johnston H, Baker PJ, Abel M, Charlton HM, Jackson G, Fleming L, Kumar TR, O’Shaughnessy PJ 2004 Regulation of Sertoli cell number and activity by follicle-stimulating hormone and androgen during postnatal development in the mouse. Endocrinology 145:318–329
O’Shaughnessy PJ, Gray SA 1995 Gonadotropin-dependent and gonadotropin-independent development of inhibin subunit messenger ribonucleic acid levels in the mouse ovary. Endocrinology 136:2060–2065
Jorgez CJ, Klysik M, Jamin SP, Behringer RR, Matzuk MM 2004 Granulosa cell-specific inactivation of follistatin causes female fertility defects. Mol Endocrinol 18:953–967
Oonk RB, Krasnow JS, Beattie WG, Richards JS 1989 Cyclic AMP-dependent and AMP-independent regulation of cholesterol side-chain cleavage cytochrome P450 (P450scc) in rat ovarian granulosa cells and corpora lutea-cDNA and deduced amino-acid sequence of rat P450scc. J Biol Chem 264:21934–21942
Ghersevich S, Nokelainen P, Poutanen M, Orava M, Autio-Harmainen H, Rajaniemi H, Vihko R 1994 Rat 17?-hydroxysteroid dehydrogenase type1: primary structure and regulation of enzyme expression in rat ovary by diethylstilbestrol and gonadotropins in vitro. Endocrinology 135:1477–1487
Rajapaksha WRAKJS, McBride M, Robertson L, O’Shaughnessy PJ 1997 Sequence of the bovine HDL-receptor (SR-BI) cDNA and changes in receptor mRNA expression during granulosa cell luteinization in vivo and in vitro. Mol Cell Endocrinol 134:59–67
Oksjoki S, Soderstrom M, Vuorio E, Anttila L 2001 Differential expression patterns of cathepsins B, H, K, L and S in the mouse ovary. Mol Hum Reprod 7:27–34
Smith MF, Mcintush EW, Ricke WA, Kojima FN, Smith GW 1999 Regulation of ovarian extracellular matrix remodelling by metalloproteinases and their tissue inhibitors: effects on follicular development, ovulation and luteal function. J Reprod Fertil Suppl 54:367–381
Pescador N, Soumano K, Stocco DM, Price CA, Murphy BD 1996 Steroidogenic acute regulatory protein in bovine corpora lutea. Biol Reprod 55:485–491
Rodgers RJ, Waterman MR, Simpson ER 1986 Cytochrome P-450scc, P45017, adrenodoxin, and reduced nicotinamide adenine dinucleotide phosphate- cytochrome P-450 reductase in bovine follicles and corpora lutea. Changes in specific contents during the ovarian cycle. Endocrinology 118:1366–1374
Itahana K, Morikazu Y, Takeya T 1996 Differential expression of four connexin genes, Cx-26, Cx-30.3, Cx-32, and Cx-43, in the porcine ovarian follicle. Endocrinology 137:5036–5044
Neilson L, Andalibi A, Kang D, Coutifaris C, Strauss JF, Stanton JL, Green DPL 2000 Molecular phenotype of the human oocyte by PCR-SAGE. Genomics 63:13–24
Rajkovic A, Pangas SA, Ballow D, Suzumori N, Matzuk MM 2004 NOBOX deficiency disrupts early folliculogenesis and oocyte-specific gene expression. Science 305:1157–1159
Soyal SM, Amleh A, Dean J 2000 FIG, a germ cell-specific transcription factor required for ovarian follicle formation. Development 127:4645–4654
Hsieh M, Mulders SM, Friis RR, Dharmarajan A, Richards JS 2003 Expression and localization of secreted frizzled-related protein-4 in the rodent ovary: evidence for selective up-regulation in luteinized granulosa cells. Endocrinology 144:4597–4606
Park JY, Su YQ, Ariga M, Law E, Jin SL, Conti M 2004 EGF-like growth factors as mediators of LH action in the ovulatory follicle. Science 303:682–684
Blackshaw S, Kuo WP, Park PJ, Tsujikawa M, Gunnersen JM, Scott HS, Boon WM, Tan SS, Cepko CL 2003 MicroSAGE is highly representative and reproducible but reveals major differences in gene expression among samples obtained from similar tissues. Genome Biol 4:R17
Juengel JL, Guy MK, Tandeski TR, McGuire WJ, Niswender GD 1994 Steady-state concentrations of messenger ribonucleic acid encoding cytochrome P450 side-chain cleavage and 3?-hydroxysteroid dehydrogenase/ 5,4 isomerase in ovine corpora lutea during the estrous cycle. Biol Reprod 51:380–384
Conley AJ, Kaminski MA, Dubowsky SA, Jablonka-Shariff A, Redmer DA, Reynolds LP 1995 Immunohistochemical localization of 3?-hydroxysteroid dehydrogenase and P450 17-hydroxylase during follicular and luteal development in pigs, sheep, and cows. Biol Reprod 52:1081–1094
Sasson R, Rimon E, Dantes A, Cohen T, Shinder V, Land-Bracha A, Amsterdam A 2004 Gonadotrophin-induced gene regulation in human granulosa cells obtained from IVF patients. Modulation of steroidogenic genes, cytoskeletal genes and genes coding for apoptotic signalling and protein kinases. Mol Hum Reprod 10:299–311
Grieshaber NA, Ko C, Grieshaber SS, Ji I, Ji TH 2003 Follicle-stimulating hormone-responsive cytoskeletal genes in rat granulosa cells: class I ?-tubulin, tropomyosin-4, and kinesin heavy chain. Endocrinology 144:29–39
Ackert CL, Gittens JE, O’Brien MJ, Eppig JJ, Kidder GM 2001 Intercellular communication via connexin43 gap junctions is required for ovarian folliculogenesis in the mouse. Dev Biol 233:258–270
Khan-Dawood FS, Yang J, Dawood MY 1998 Hormonal regulation of connexin-43 in baboon corpora lutea. J Endocrinol 157:405–414
Lane TF, Sage EH 1994 The biology of SPARC, a protein that modulates cell-matrix interactions. FASEB J 8:163–173
Bagavandoss P, Sage EH, Vernon RB 1998 Secreted protein, acidic and rich in cysteine (SPARC) and thrombospondin in the developing follicle and corpus luteum of the rat. J Histochem Cytochem 46:1043–1049
Reed MJ, Puolakkainen P, Lane TF, Dickerson D, Bornstein P, Sage EH 1993 Differential expression of SPARC and thrombospondin 1 in wound repair: immunolocalization and in situ hybridization. J Histochem Cytochem 41:1467–1477
Santella L, Kyozuka K 1997 Association of calmodulin with nuclear structures in starfish oocytes and its role in the resumption of meiosis. Eur J Biochem 246:602–610
Bellin R, Capila I, Lincecum J, Park PW, Reizes O, Bernfield MR 2002 Unlocking the secrets of syndecans: transgenic organisms as a potential key. Glycoconj J 19:295–304
Mazzali M, Kipari T, Ophascharoensuk V, Wesson JA, Johnson R, Hughes J 2002 Osteopontin—a molecule for all seasons. QJM 95:3–13
Morales CR, Zhao Q, Lefrancois S, Ham D 2000 Role of prosaposin in the male reproductive system: effect of prosaposin inactivation on the testis, epididymis, prostate, and seminal vesicles. Arch Androl 44:173–186
Bowles J, Bullejos M, Koopman P 2000 A subtractive gene expression screen suggests a role for vanin-1 in testis development in mice. Genesis 27:124–135
Aurrand-Lions M, Galland F, Bazin H, Zakharyev VM, Imhof BA, Naquet P 1996 Vanin-1, a novel GPI-linked perivascular molecule involved in thymus homing. Immunity 5:391–405(R. S. McRae, H. M. Johnst)
Address all correspondence and requests for reprints to: P. J. O’Shaughnessy, Institute of Comparative Medicine, University of Glasgow Veterinary School, Bearsden Road, Glasgow G61 1QH, United Kingdom. E-mail: p.j.o’shaughnessy@vet.gla.ac.uk.
Abstract
Changes in gene expression during granulosa cell luteinization have been measured using serial analysis of gene expression (SAGE). Immature normal mice were treated with pregnant mare serum gonadotropin (PMSG) or PMSG followed, 48 h later, by human chorionic gonadotropin (hCG). Granulosa cells were collected from preovulatory follicles after PMSG injection or PMSG/hCG injection and SAGE libraries generated from the isolated mRNA. The combined libraries contained 105,224 tags representing 40,248 unique transcripts. Overall, 715 transcripts showed a significant difference in abundance between the two libraries of which 216 were significantly down-regulated by hCG and 499 were significantly up-regulated. Among transcripts differentially regulated, there were clear and expected changes in genes involved in steroidogenesis as well as clusters of genes involved in modeling of the extracellular matrix, regulation of the cytoskeleton and intra and intercellular signaling. The SAGE libraries described here provide a base for functional investigation of the regulation of granulosa cell luteinization.
Introduction
THE TERMINAL DIFFERENTIATION event of ovarian follicles that avoid atresia is luteinization to form the corpus luteum. Studies using hypophysectomized, mutant, and knockout animals have shown that FSH regulates granulosa cell proliferation and development from the secondary stage to the preovulatory stage (1, 2, 3). Luteinization of these cells occurs after the ovulatory LH surge and is dependent upon expression of LH receptors on the developing granulosa cells. The regulatory mechanisms inducing luteinization remain uncertain, but it is clear that granulosa cells begin to luteinize before ovulation and the process may be linked to disintegration of cell-to-cell communication between granulosa cells and the oocyte (4, 5).
After the LH surge and ovulation, the corpus luteum rapidly develops to become a highly active steroidogenic tissue and it plays an essential role in the establishment and maintenance of pregnancy. Several functional differences between presurge granulosa cells and luteal cells have been identified but the underlying, early molecular events that occur during terminal granulosa cell differentiation remain unclear. Earlier studies using microarray, differential display and subtractive hybridization techniques have identified a number of genes that show changes in expression during early luteinization (6, 7). This work has provided valuable insight into the luteinization process, but it is likely that genes identified in these studies represent a small fraction of those undergoing change. As a step toward understanding the process of luteinization, and to identify genes undergoing regulation during the early phase of terminal differentiation, we have used the technique of serial analysis of gene expression (SAGE) (8) to provide a comprehensive profile of gene expression in granulosa cells before and after a luteinizing dose of human chorionic gonadotropin (hCG).
Materials and Methods
nimals and treatments
The mice used in this study were bred at the University of Glasgow Veterinary School and were maintained as required under United Kingdom Home Office regulations. Normal mice, bred on a C3H/Heh-101/H genetic background, were derived from stock animals originally obtained from the MRC Radiobiology Unit (now the MRC, Mammalian Genetics Unit, Harwell, UK).
To generate a preovulatory granulosa cell SAGE library, immature mice (aged 18–22 d) were injected ip with a single dose of 5 IU pregnant mare serum gonadotropin (PMSG) and granulosa cells were collected 48 h later. To generate a SAGE library of granulosa cells undergoing luteinization, immature mice were treated with PMSG followed 48 h later by hCG, and granulosa cells were isolated after a further 12 h. To isolate the cells, ovaries were placed in DMEM (Invitrogen Ltd., Paisley, Scotland, UK) and granulosa cells were released from large antral follicles using needles. No attempt was made to remove oocytes from the isolated granulosa cells. Cells were stored in liquid N2 until RNA extraction.
Construction and analysis of SAGE libraries
The RNA from PMSG and PMSG/hCG-treated granulosa cells was extracted using TRIzol reagent (Invitrogen Ltd.). The SAGE libraries were constructed from RNA pooled from granulosa cells of 30 different mice. Poly-A+ RNA was isolated using oligo deoxythymidine cellulose (Invitrogen Ltd.) and SAGE libraries were constructed and analyzed as previously described (9). Fourteen nucleotide SAGE tags, representing individual transcripts, were extracted from the sequence data and initially analyzed using SAGE 2000 software (http://www.sagenet.org/sage_protocol.htm). Differences in tag frequency between SAGE libraries were statistically analyzed using the 2 test (10). Tags were identified through mouse SAGEmap database build 136. The original SAGE data from each library described here have been deposited in the National Center for Biotechnology Information (NCBI) public gene expression database (http://www.ncbi.nlm.nih.gov/geo/) accession nos. GSM30721 and GSM30722.
Rapid amplification of cDNA ends (RACE)
For those tags which could not be identified by comparison to the NCBI database, it was necessary to generate additional 3' sequence to identify the source mRNA species. This was done using a 3' RACE technique similar to that described previously (11) but using biotinylated primer TTCTAGAATTCAGCGGCCGC(T)30(AGC)(AGCT) to prime reverse transcription.
Real-time PCR
Levels of specific mRNA species were measured by real-time PCR using the Taqman method as described previously (12). The tissue used to generate RNA for real-time PCR was prepared using the same protocols as above for generation of SAGE libraries (see Construction and analysis of SAGE libraries). The sequences of primers and probes used for real-time PCR were as shown in Table 1 and in previous studies (12, 13).
TABLE 1. Sequences of real-time PCR primers and probes used in this study
Results
Ovarian response to hormone treatment
After a single injection of PMSG, the ovaries of immature mice contained numerous large antral follicles which provided granulosa cells for the PMSG SAGE library. These follicles ovulated around 13–14 h after a subsequent injection of hCG and luteinizing granulosa cells were, therefore, collected from intact follicles 12 h after hCG injection to generate the PMSG/hCG library.
SAGE libraries
The total number of tags sequenced in the PMSG library (treatment with PMSG alone) was 51,528, and the total number sequenced in the PMSG/hCG library (treatment with PMSG followed by hCG) was 53,696. The combined total of 105,224 tags corresponded to 40,248 unique transcripts, of which 9,877 were represented by two or more tags. Of the transcripts represented by more than one tag, 5,689 were shared between both libraries, 1,806 were unique to the PMSG library, and 2,382 were unique to the PMSG/hCG library. Using 2 analysis to test for significant differences in tag abundance between those tags with greater than five transcripts present in the combined libraries, 499 tags were significantly up-regulated by hCG treatment (P < 0.05), and 216 tags were significantly down-regulated (P < 0.05).
Abundant, differentially expressed SAGE tags
The most abundant tags that were shown to be differentially expressed between SAGE libraries by 2 test are shown in Table 2. Most of these tags match unambiguously to known genes and a number of them are known to be expressed in granulosa cells during FSH-dependent development [e.g. inhibin ?B and follistatin (14, 15)] or to show a change in expression after luteinization [e.g. P450 11a1, 17?-hydroxysteroid dehydrogenase type 1 and scavenger receptor class B type 1 (SRB1) (16, 17, 18)]. This list also contains 36 tags, including six of the eight most abundant tags, which have no match, are linked to sequence of unknown function or have multiple assignments. A number of these tags, without confirmed or reliable matches in the SAGEmap database, were analyzed further using 3'RACE (indicated by an asterisk in Table 2). The transcript associated with the tag CAGTCAATAC is of interest because it shows abundant, differentially regulated expression. The 3'RACE data from this tag matched it to Unigene cluster Mm.290944, which shows 95% homology with a noncoding human mRNA sequence. This tag, in addition, is not highly expressed in other mouse SAGE libraries and appears to show a degree of selectivity for the granulosa cell. The expression pattern for this sequence has been confirmed using real-time PCR (Fig. 1, noncoding RNA). Other tags extended by 3'RACE could be linked to Unigene clusters or to individual expressed sequence tags (ESTs). These ESTs may be linked to Unigene clusters (e.g. TTGTTGCTAC matches EST BY419395, which is located on the genome close to Mm.190648) or may represent unique transcribed regions.
TABLE 2. Most abundant differentially expressed tags
TABLE 2A. Continued
FIG. 1. Comparison of SAGE and real-time PCR data. Results show changes in expression of specific genes in granulosa cells of PMSG-treated ovaries from mice before or after injection with hCG. Gene expression was measured by real-time PCR or by SAGE and results are expressed as fold increase or fold decrease in expression after hCG. The real-time PCR data show the mean of six pools of RNA in each group, and each RNA pool was derived from granulosa cells isolated from two or more animals. The SAGE data were derived from that shown in Tables 2 and 3. arom, Cytochrome P450 aromatase; ?-act, ?-actin; FSHR, FSH receptor; kit, kit ligand; LHR, LH receptor; ncRNA, noncoding RNA; scc, cytochrome P450 side chain cleavage; wbscr1, Williams-Beuren syndrome chromosome region 1.
Functional groups
Table 3 shows the frequency of expression of a number of tags in the PMSG and PMSG/hCG libraries, which represent genes linked with established functional groups. The list is not exhaustive and does not, in general, include genes with low levels of expression. This table is designed to show changes (or lack of them) in the expression of selected genes during granulosa cell luteinization. In each of the categories, there are genes listed that have been shown previously to be regulated during luteinization or which are know to be expressed in the corpus luteum [e.g. gap junction protein Cx43 (1), steroidogenic acute regulatory protein (StAR), adrenodoxin, tissue inhibitor of metalloproteinase 1 (TIMP 1), and cathepsin L (6, 19, 20, 21, 22, 23)]. In addition, there are a number of genes listed which have not previously been shown to be regulated during luteinization such as cofilin, seizure-related 6 homolog (mouse)-like 2 (Sez6l2), and the listed transcription factors.
TABLE 3. Tags associated with specific functional groups
TABLE 3A. Continued
To determine the contribution of oocyte mRNA to the SAGE libraries, the frequency of tags representing oocyte-specific genes (24, 25, 26) was determined in the two libraries. There were no tags representing the genes Nobox, factor in the germline- (Fig), and ataxia-telangiectasia mutated (ATM) present in either library, whereas zona pellucida (ZP)-2 and ZP-3 were represented by a single tag each in one library. The signaling molecules bone morphogenic protein 15 (BMP15) and growth differentiation factor 9 (GDF9) were represented by a maximum of four and 11 tags, respectively.
Real-time PCR
Real-time PCR was used to determine whether differences in gene expression between granulosa cell SAGE libraries could be confirmed for a number of transcripts. To normalize the data from real-time PCR, the housekeeping gene GAPDH (glyceraldehyde-3-phosphate dehydrogenase) was used because it is expressed equally in both SAGE libraries (Table 3). Results in Fig. 1 show that, for these selected transcripts, there is good correlation between data obtained by real-time PCR and data derived from SAGE libraries.
Discussion
The SAGE method allows for a quantitative, qualitative, and comprehensive measurement of gene expression patterns in cell populations and tissues. We have applied SAGE, in this study, to an examination of the changes in transcript expression patterns in granulosa cells from mature antral follicles before and after the start of gonadotropin-induced luteinization. One potential problem with the SAGE technique is that it can be time consuming, and this can impose a practical limitation on the number of libraries generated in any one particular study. Thus, in this report we have limited identification of genes involved in luteinization to a single time after induction by hCG. Nevertheless, the data produced from these studies provide both a base for studies into granulosa cell development and function and a comprehensive list of genes undergoing regulation during early granulosa cell luteinization.
A number of previous studies have examined changes in gene expression in granulosa cells during luteinization (6, 18, 19, 21, 22, 27, 28). In general, these studies examined changes in a single or limited number of genes over the time course of luteinization and have served to provide a valuable insight into the development of the corpus luteum. Genes that have been shown to alter expression during luteinization include, for example, the steroidogenic enzymes and associated proteins [e.g. P45011a1, adrenodoxin, P45019a1, StAR, and SRB1 (6, 18, 21, 22)], extracellular matrix modifiers [e.g. A disintegrin-like and metalloprotease (reprolysin type) with thrombospondin type 1 motif 1 (ADAMTS-1), cathepsin L (6, 19)], signaling molecules [e.g. epiregulin, secreted frizzled-related protein-4 and wnt 4 (27, 28)] and others [e.g. regulator of G protein signaling protein-2 and cell surface antigen CD63 (6)]. In general, results from these previous studies correlate well with the SAGE data reported here. When taken alongside results using real-time PCR (Fig. 1), this is a further indication that the SAGE results are a reliable measure of changes in gene expression at the selected stage of luteinization. Previous SAGE studies have also shown that the technique is highly reproducible and accurate (29).
The SAGE libraries reported in this study were generated predominantly from granulosa cell mRNA, but oocytes were present in the isolated cell mix and would have contributed to the total mRNA pool used in library construction. The expression of tags representing the oocyte-specific genes (24, 25, 26) Nobox, Fig, ATM, ZP-3, and ZP-2 was very low or undetectable, and only BMP15 and GDF9 were represented by more than a single tag. Both ZP-3 and ATM are known to be highly expressed within the oocyte (24), and it is unlikely, therefore, that oocyte-derived genes will contribute significantly to the overall tag number. Care in interpretation may be required for low abundance tags, however, because these could have an oocyte origin.
Genes known to be associated with luteinization
The most fundamental change in granulosa/luteal cell function after induction of luteinization is a marked increase in steroidogenic activity by the developing corpus luteum. It has been shown previously that this is associated with increased expression of P45011a1, SRB1, StAR, ferredoxin, and low-density lipoprotein receptor and decreased expression of P45019a1 (aromatase) and 17?-hydroxysteroid dehydrogenase type 1 (18, 21, 22). The overall effect is a reduction in estrogen production and a marked increase in progesterone production through increased availability of substrate and converting enzymes. As described above in second paragraph of Discussion, each of these previously reported changes in gene expression correlates well with changes in the tag numbers associated with the SAGE libraries generated for this study. One possible anomaly, however, between the SAGE data and previous studies is the expression of 3?-hydroxysteroid dehydrogenase (3?HSD) during luteinization. It has previously been shown that 3?HSD activity and expression increase in the developing corpus luteum in line with the general increase in steroidogenic activity (30, 31). Data from the SAGE libraries, in contrast, indicate that 3?HSD type 1 expression is reduced in granulosa cells 12 h after hCG treatment. This change was confirmed by real-time PCR and it appears likely, therefore, that luteinization leads to a temporary decrease in 3?HSD expression, perhaps through down-regulation induced by exposure to high gonadotropin levels, which reverses as the corpus luteum develops.
Data from the SAGE libraries show that the expression levels of several cytoskeletal-associated transcripts including actin, vinculin, cofilin, tubulin, and tropomyosin, are differentially regulated during luteinization. There is also significant differential expression of actin and tubulin isoforms. It is known that gonadotropins will regulate cytoskeletal gene expression in granulosa cells (32, 33) and cytoskeletal remodeling during luteinization is likely to be an essential part of the movement and morphological development of the cells.
Several membrane-binding and communication-related components such as clusterin, annexin A2, and the gap junction membrane channel proteins (connexins) are known to show fluctuations in expression levels during granulosa cell development and luteinization. The overwhelming level of Cx 43 expression in the PMSG SAGE library supports current thinking that this is the primary means of intercellular communication between granulosa cells and reinforces the hypothesis of a functional granulosa cell syncytium throughout folliculogenesis (34). Expression of Cx43 decreases markedly after induction of luteinization, as shown previously (23), although it may continue to be expressed in the developing corpus luteum (35).
Secreted acidic cysteine-rich glycoprotein (SPARC, osteonectin, basement membrane protein 40) is a highly expressed transcript which undergoes a 7-fold up-regulation during luteinization. SPARC is expressed in vivo where cells are undergoing proliferative or reorganizational activity (36), and it has previously been identified in follicular granulosa cells after the LH surge (37). It is possible that SPARC may play an essential role in the development of the corpus luteum because specific peptide fragments of the protein are strongly angiogenic (38). In the follicle, SPARC is found in both granulosa cells and oocytes, although expression in the oocyte may derive from adjacent granulosa cells (37). Calmodulin, a protein with strong structural and functional similarities to SPARC, has been implicated in the resumption of meiosis in the starfish oocyte (39), and it is possible that up-regulation of SPARC after hCG may play a role in allowing resumption of meiosis in the oocyte.
Genes with a poorly defined role in luteinization
In addition to confirming changes in gene expression previously reported or predicted, the SAGE data set reported here also identifies a number of genes of interest that have not previously been linked to the process of early luteinization. This list is made up of genes with unknown function and genes with known function but no previous association with granulosa luteinization. Genes with known function and differentially regulated during luteinization include syndecan-1, secreted phosphoprotein 1 (Spp1), prosaposin, and vanin 1.
Syndecan-1 is a heparin sulfate-rich integral membrane proteoglycan, and it is expressed in a developmental and cell type-specific pattern (40), but it has not previously been identified as having a role in folliculogenesis or luteinization. Syndecans have major roles as matrix and cell surface receptors, coreceptors for growth factor signaling, internalization receptors, and soluble paracrine effectors (40). In addition, syndecan-1 appears to be capable of independent signaling and may play a role in regulation of Wnt signaling (40). Syndecan-1 is not expressed in the PMSG-stimulated library but shows a high level of transcript expression after hCG administration. Syndecans have a well-established involvement in the regulation of cytoskeletal organization (40) and a likely function of syndecan-1 in the luteinizing follicle is in the regulation of cytoskeletal assembly.
Spp1 (also known as osteopontin or Eta-1) is among the most highly up-regulated tags after hCG administration. It is a multifunctional protein expressed in various cell types and involved in a number of physiological and pathological process including biomineralization, inflammation, leukocyte recruitment, cell survival, tissue repair, cell proliferation, and proliferation of vascular smooth muscle cells (41). It is possible, therefore, that Spp1 has multiple functions during luteinization. For example, it may act as a survival factor preventing onset of apoptosis during the critical phase of ovulation and luteinization. Equally, the effects on vascular smooth muscle suggest a possible role in the angiogenic process that accompanies formation of the corpus luteum.
Prosaposin is expressed at a medium level in the antral follicle but shows a 6-fold up-regulation after hCG administration. The protein is either secreted or acts as a precursor of smaller saposins, and it has been shown to have diverse functions including involvement in the MAPK and Akt signaling pathways and maintenance of cell growth, differentiation, and survival (42). The likely role of prosaposin in development of the corpus luteum is uncertain but may, again, relate to overall function as a survival factor.
Vanin 1 is a glycosylphosphatidylinositol-anchored cell surface molecule involved in thymic and gonadal development (43, 44). As with Spp1, it is also highly up-regulated in granulosa cells after hCG administration. Vanin 1 is expressed specifically in the Sertoli cells of the developing fetal gonad, and it has been suggested that it may be involved in the migration of mesenchynmal cells from the mesonephros into the developing gonad (43). The likely function of vanin 1 in the developing corpus luteum is unclear but, by analogy with developing thymic and gonadal tissue, it is possible that it may be involved in the cell migration that occurs early in corpus luteum formation to integrate both thecal and endothelial cells into the developing tissue.
For those genes already discussed in this section, it is possible to hypothesize a possible function in ovulation and corpus luteum development based on known properties and functions of the genes in other tissues. The SAGE libraries described here also contain many other highly expressed or differentially expressed transcripts for which this not currently possible. Included among this latter group are (Sez6l2), proline-rich protein MP-5, and epithelial V-like antigen. In addition, the list of tags differentially regulated after hCG contains a considerable number that are unmatched in the SAGEmap database or are matched only to EST clusters or uncharacterized transcripts. The challenge now will be to identify those genes from this list that are fundamentally involved in the process of luteinization and those genes that have a more downstream role in the development of the corpus luteum.
References
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
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
Halpin DM, Jones A, Fink G, Charlton HM 1986 Postnatal ovarian follicle development in hypogonadal (hpg) and normal mice and associated changes in the hypothalamic-pituitary ovarian axis. J Reprod Fertil 77:287–296
Smith MF, McIntush EW, Smith GW 1994 Mechanisms associated with corpus-luteum development. J Anim Sci 72:1857–1872
Simon AM, Goodenough DA, Li E, Paul DL 1997 Female infertility in mice lacking connexin 37. Nature 385:525–529
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
Leo CP, Pisarska MD, Hsueh AJ 2001 DNA array analysis of changes in preovulatory gene expression in the rat ovary. Biol Reprod 65:269–276
Velculescu VE, Zhang L, Vogelstein B, Kinzler KW 1995 Serial analysis of gene expression. Science 270:484–487
O’Shaughnessy PJ, Fleming L, Baker PJ, Jackson G, Johnston H 2003 Identification of developmentally-regulated genes in the somatic cells of the mouse testis using serial analysis of gene expression. Biol Reprod 69:797–808
Man MZ, Wang X, Wang Y 2000 POWER_SAGE: comparing statistical tests for SAGE experiments. Bioinformatics 16:953–959
Chen JJ, Rowley JD, Wang SM 2000 Generation of longer cDNA fragments from serial analysis of gene expression tags for gene identification. Proc Natl Acad Sci USA 97:349–353
O’Shaughnessy PJ, Willerton L, Baker PJ 2002 Changes in Leydig cell gene expression during development in the mouse. Biol Reprod 66:966–975
Johnston H, Baker PJ, Abel M, Charlton HM, Jackson G, Fleming L, Kumar TR, O’Shaughnessy PJ 2004 Regulation of Sertoli cell number and activity by follicle-stimulating hormone and androgen during postnatal development in the mouse. Endocrinology 145:318–329
O’Shaughnessy PJ, Gray SA 1995 Gonadotropin-dependent and gonadotropin-independent development of inhibin subunit messenger ribonucleic acid levels in the mouse ovary. Endocrinology 136:2060–2065
Jorgez CJ, Klysik M, Jamin SP, Behringer RR, Matzuk MM 2004 Granulosa cell-specific inactivation of follistatin causes female fertility defects. Mol Endocrinol 18:953–967
Oonk RB, Krasnow JS, Beattie WG, Richards JS 1989 Cyclic AMP-dependent and AMP-independent regulation of cholesterol side-chain cleavage cytochrome P450 (P450scc) in rat ovarian granulosa cells and corpora lutea-cDNA and deduced amino-acid sequence of rat P450scc. J Biol Chem 264:21934–21942
Ghersevich S, Nokelainen P, Poutanen M, Orava M, Autio-Harmainen H, Rajaniemi H, Vihko R 1994 Rat 17?-hydroxysteroid dehydrogenase type1: primary structure and regulation of enzyme expression in rat ovary by diethylstilbestrol and gonadotropins in vitro. Endocrinology 135:1477–1487
Rajapaksha WRAKJS, McBride M, Robertson L, O’Shaughnessy PJ 1997 Sequence of the bovine HDL-receptor (SR-BI) cDNA and changes in receptor mRNA expression during granulosa cell luteinization in vivo and in vitro. Mol Cell Endocrinol 134:59–67
Oksjoki S, Soderstrom M, Vuorio E, Anttila L 2001 Differential expression patterns of cathepsins B, H, K, L and S in the mouse ovary. Mol Hum Reprod 7:27–34
Smith MF, Mcintush EW, Ricke WA, Kojima FN, Smith GW 1999 Regulation of ovarian extracellular matrix remodelling by metalloproteinases and their tissue inhibitors: effects on follicular development, ovulation and luteal function. J Reprod Fertil Suppl 54:367–381
Pescador N, Soumano K, Stocco DM, Price CA, Murphy BD 1996 Steroidogenic acute regulatory protein in bovine corpora lutea. Biol Reprod 55:485–491
Rodgers RJ, Waterman MR, Simpson ER 1986 Cytochrome P-450scc, P45017, adrenodoxin, and reduced nicotinamide adenine dinucleotide phosphate- cytochrome P-450 reductase in bovine follicles and corpora lutea. Changes in specific contents during the ovarian cycle. Endocrinology 118:1366–1374
Itahana K, Morikazu Y, Takeya T 1996 Differential expression of four connexin genes, Cx-26, Cx-30.3, Cx-32, and Cx-43, in the porcine ovarian follicle. Endocrinology 137:5036–5044
Neilson L, Andalibi A, Kang D, Coutifaris C, Strauss JF, Stanton JL, Green DPL 2000 Molecular phenotype of the human oocyte by PCR-SAGE. Genomics 63:13–24
Rajkovic A, Pangas SA, Ballow D, Suzumori N, Matzuk MM 2004 NOBOX deficiency disrupts early folliculogenesis and oocyte-specific gene expression. Science 305:1157–1159
Soyal SM, Amleh A, Dean J 2000 FIG, a germ cell-specific transcription factor required for ovarian follicle formation. Development 127:4645–4654
Hsieh M, Mulders SM, Friis RR, Dharmarajan A, Richards JS 2003 Expression and localization of secreted frizzled-related protein-4 in the rodent ovary: evidence for selective up-regulation in luteinized granulosa cells. Endocrinology 144:4597–4606
Park JY, Su YQ, Ariga M, Law E, Jin SL, Conti M 2004 EGF-like growth factors as mediators of LH action in the ovulatory follicle. Science 303:682–684
Blackshaw S, Kuo WP, Park PJ, Tsujikawa M, Gunnersen JM, Scott HS, Boon WM, Tan SS, Cepko CL 2003 MicroSAGE is highly representative and reproducible but reveals major differences in gene expression among samples obtained from similar tissues. Genome Biol 4:R17
Juengel JL, Guy MK, Tandeski TR, McGuire WJ, Niswender GD 1994 Steady-state concentrations of messenger ribonucleic acid encoding cytochrome P450 side-chain cleavage and 3?-hydroxysteroid dehydrogenase/ 5,4 isomerase in ovine corpora lutea during the estrous cycle. Biol Reprod 51:380–384
Conley AJ, Kaminski MA, Dubowsky SA, Jablonka-Shariff A, Redmer DA, Reynolds LP 1995 Immunohistochemical localization of 3?-hydroxysteroid dehydrogenase and P450 17-hydroxylase during follicular and luteal development in pigs, sheep, and cows. Biol Reprod 52:1081–1094
Sasson R, Rimon E, Dantes A, Cohen T, Shinder V, Land-Bracha A, Amsterdam A 2004 Gonadotrophin-induced gene regulation in human granulosa cells obtained from IVF patients. Modulation of steroidogenic genes, cytoskeletal genes and genes coding for apoptotic signalling and protein kinases. Mol Hum Reprod 10:299–311
Grieshaber NA, Ko C, Grieshaber SS, Ji I, Ji TH 2003 Follicle-stimulating hormone-responsive cytoskeletal genes in rat granulosa cells: class I ?-tubulin, tropomyosin-4, and kinesin heavy chain. Endocrinology 144:29–39
Ackert CL, Gittens JE, O’Brien MJ, Eppig JJ, Kidder GM 2001 Intercellular communication via connexin43 gap junctions is required for ovarian folliculogenesis in the mouse. Dev Biol 233:258–270
Khan-Dawood FS, Yang J, Dawood MY 1998 Hormonal regulation of connexin-43 in baboon corpora lutea. J Endocrinol 157:405–414
Lane TF, Sage EH 1994 The biology of SPARC, a protein that modulates cell-matrix interactions. FASEB J 8:163–173
Bagavandoss P, Sage EH, Vernon RB 1998 Secreted protein, acidic and rich in cysteine (SPARC) and thrombospondin in the developing follicle and corpus luteum of the rat. J Histochem Cytochem 46:1043–1049
Reed MJ, Puolakkainen P, Lane TF, Dickerson D, Bornstein P, Sage EH 1993 Differential expression of SPARC and thrombospondin 1 in wound repair: immunolocalization and in situ hybridization. J Histochem Cytochem 41:1467–1477
Santella L, Kyozuka K 1997 Association of calmodulin with nuclear structures in starfish oocytes and its role in the resumption of meiosis. Eur J Biochem 246:602–610
Bellin R, Capila I, Lincecum J, Park PW, Reizes O, Bernfield MR 2002 Unlocking the secrets of syndecans: transgenic organisms as a potential key. Glycoconj J 19:295–304
Mazzali M, Kipari T, Ophascharoensuk V, Wesson JA, Johnson R, Hughes J 2002 Osteopontin—a molecule for all seasons. QJM 95:3–13
Morales CR, Zhao Q, Lefrancois S, Ham D 2000 Role of prosaposin in the male reproductive system: effect of prosaposin inactivation on the testis, epididymis, prostate, and seminal vesicles. Arch Androl 44:173–186
Bowles J, Bullejos M, Koopman P 2000 A subtractive gene expression screen suggests a role for vanin-1 in testis development in mice. Genesis 27:124–135
Aurrand-Lions M, Galland F, Bazin H, Zakharyev VM, Imhof BA, Naquet P 1996 Vanin-1, a novel GPI-linked perivascular molecule involved in thymus homing. Immunity 5:391–405(R. S. McRae, H. M. Johnst)