| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Obstetrics and Gynecology,3 Baylor College of Medicine, Houston, Texas 77030
Center for Reproductive Medicine,4 Shandong Provincial Hospital of Shandong University, Jinan, 250012, China
Division of Genetics,5 Harvard University Graduate Biophysics Program,6 Department of Pathology,7 and Harvard/MIT Division of Health Sciences and Technology,8 Cambridge, Massachusetts 02138
ABSTRACT
Nobox is a homeobox gene expressed in oocytes and critical in oogenesis. Nobox deficiency leads to rapid loss of postnatal oocytes. Early oocyte differentiation is poorly understood. We hypothesized that lack of Nobox perturbs global expression of genes preferentially expressed in oocytes as well as microRNAs. We compared Nobox knockout and wild-type ovaries using Affymetrix 430 2.0 microarray platform. We discovered that 28 (74%) of 38 of the genes downregulated more than 5-fold in the absence of Nobox were preferentially expressed in oocytes, whereas only 5 (15%) of 33 genes upregulated more than 5-fold in the absence of Nobox were preferentially expressed in oocytes. Protein-binding microarray helped identify nucleotide motifs that NOBOX binds and that several downregulated genes contain within putative promoter regions. MicroRNA population in newborn ovaries deficient of Nobox was largely unaffected. Genes whose proteins are predicted to be secreted but were previously unknown to be significantly expressed in early oogenesis were downregulated in Nobox knockouts and included astacin-like metalloendopeptidase (Astl), Jagged 1 (Jag1), oocyte-secreted protein 1 (Oosp1), fetuin beta (Fetub), and R-spondin 2 (Rspo2). In addition, pluripotency-associated genes Pou5f1 and Sall4 are drastically downregulated in Nobox-deficient ovaries, whereas testes-determining gene Dmrt1 is overexpressed. Our findings indicate that Nobox is likely an activator of oocyte-specific gene expression and suggest that the oocyte plays an important role in suppressing expression of male-determining genes, such as Dmrt1.
follicular development,, folliculogenesis,, gamete biology,, microarray,, microRNA,, Nobox,, oocyte,, oocyte development,, oogenesis,, ovarian failure,, ovary,, transcription factor
Oocyte differentiation begins in the mouse embryonic gonad after oogonia enter meiosis around 13.5 days postcoitum (dpc). In the mouse embryonic gonad, oocytes exist as clusters also known as cysts [1, 2]. Within a few days of birth, oocyte clusters break down, and individual oocytes become enveloped by flat somatic pregranulosa cells. The early germ cell-somatic units are called primordial follicles, and oocytes in these units are usually less than 20 µm in diameter. Several signaling molecules, such as Kitl, Fgf2, Fgf7, neurotrophins, Lif, and steroids, have been implicated in the primordial to primary and growing follicle transition [3–9]. Oocyte-somatic cell communications are paramount for the successful differentiation of the follicular unit, but little is known about the molecules that initiate germ cell cluster breakdown [10]. Factor in the germline alpha (Figla) was the first oocyte-specific basic helix-loop-helix transcription factor that was shown to affect primordial follicle formation in mammals [11]. However, downstream targets of Figla that are critical for the formation of primordial follicles are largely unknown.
We identified Nobox, a homeobox gene preferentially expressed in the germline, as critical in the early oocyte differentiation [12, 13]. Nobox knockout females are infertile and lose oocytes quickly, so that by 2 weeks of postnatal life the ovary is depleted of oocytes. At the time of birth, Nobox knockout females contain oocytes that grossly do not differ from the wild-type ovaries. We therefore used whole-mouse genome expression arrays and microRNA arrays to identify genes and microRNAs perturbed significantly by the lack of Nobox. In addition, we used protein binding microarray (PBM) technology [14] to identify NOBOX binding preferences and identified a subset of genes that contain such elements in their promoter regions.
Newborn ovaries were pooled separately from wild-type and Nobox–/– animals, and total RNA was isolated using RNeasy mini kit (Qiagen). Newborn ovaries were collected within 12 h of delivery. Animal experimentation was approved by the Institutional Animal Care and Use Committee of Baylor College of Medicine. Three independently pooled RNA samples from wild-type and Nobox–/– newborn ovaries were used to generate biotinylated cRNA. Biotinylated cRNA was hybridized to GeneChip Mouse Expression Set 430 2.0 (Affymetrix Inc.). Since three independent experiments were performed from three independent pools of wild-type and Nobox–/– RNA, signal intensities for particular genes were averaged between the three chips, and ratio of the wild-type over the knockout signal was calculated. Signal less than 100 was considered background. The raw data in the Affymetrix CEL files were normalized by the RMA method (robust multiarray analysis) [15, 16]. Then, the null hypothesis that there is no significant change in gene expression between the treatment pairs was tested. This was done by LIMMA [17] and the pooled local error (LPE) method [18]. The raw P values were adjusted by the Benjamini-Hochberg method for the false-discovery rate of 5% [19]. We used the DAVID database [20] to aid in functional annotation of genes affected by Nobox deficiency.
Total RNA was isolated for microRNA microarray, size fractionated (<200 nt) using mirVana RNA isolation kit (Ambion), and labeled with Cy3 (wild-type) and Cy5 (Nobox knockout). Microarray microRNA assays were performed on a µParaFlo microfluidics chip (LC Sciences LLC) containing 373 distinct microRNAs based on Sanger miRBase Release 9.0. The melting temperature of the detection probes was balanced by the incorporation of a varying number of modified nucleotides with increased binding affinities. The maximal signal level of background probes was less than 250.
RNA Isolation, RT-PCR, Quantitative PCR, and In Situ Hybridization
Multitissue RT-PCR was performed as previously described [21]. Oligonucleotides corresponding to Nlrp14 (also known as Nalp14), Nlrp4a (also known as Nalp4a), Nlrp4f (also known as Nalp4f), Nlrp4c (also known as Nalp4c), Oas1c, Oas1d, Oas1e, Oas1a, Bax, Bcl2, Bcl2l2, and Casp2 were selected using Primer 3 software (http://www.genome.wi.mit.edu/cgi-bin/primer/primer3_www.cgi) to generate an approximately 500-bp nucleotide fragment that is interrupted by an intron within the mouse genome. The sequences of these primers are available on request from A.R. Mouse actin-specific primers were used to verify cDNA synthesis from RNA isolated from each tissue. Polymerase chain reaction was carried out for 28 cycles. For each cycle, the template was denatured at 94°C for 30 seconds, annealed at 55°C for 30 seconds, and extended at 72°C for 30 seconds, and PCR products were resolved on 2% agarose gels. RT-PCR results were verified on three independently collected pools of newborn ovaries.
Quantitative real-time PCR was performed on the ABI Prism 7500 Sequence Detection System (ABI, Foster City, CA) using Assays-On-Demand (ABI) PCR primer and probe sets for each gene and mouse Gapdh (VIC-labeled MGD probe; primer limited; ABI) as the endogenous control. RT-PCR was performed using the TaqMan Universal PCR Master Mix (ABI) in 20 µl. Each sample was analyzed in duplicate from at least three independent newborn wild-type and Nobox–/– cDNA samples. Two nontemplate controls (RNase-free water) samples were included on each plate for each primer-probe set. The relative amount of transcripts was calculated by the 
CT method as described by Applied Biosystems using the ABI 7500 System Software (V.1.2.3) and was normalized to the endogenous reference (Gapdh). One wild-type sample was randomly chosen to serve as the reference sample to which all other samples were normalized. The average and standard error were calculated for the triplicate measurements, and the relative amount of target gene expression for each sample was plotted. Significance was performed using Student t-test using Excel (Microsoft).
Mouse cDNA fragments corresponding to Oas1d, Nlrp14, Nlrp4f, and AU016322 were subcloned into the pGEM-T Easy vectors (Promega) and used to generate antisense and sense strands by labeling with [
-35S]UTP using the Riboprobe T7/SP6 Combination System (Promega) [22]. In situ hybridization was carried out as previously described [21]. In situ hybridizations were performed on ovarian sections derived from at least two different animals.
We fused NOBOX amino acids 124–205 (containing the homeodomain region) with the glutathionine S transferase protein in the pet-41B vector (Novagen) to create a GST-NxHD protein. GST-NxHD protein fusion was purified on the GST bind resin column (Novage) and dialyzed against 1x PBS. Protein-binding microarrays were performed and enrichment scores calculated essentially as described previously [14, 23–25].
Nobox Deficiency Disproportionately Affects Expression of Genes Preferentially Expressed in the Oocytes
We showed previously that several genes preferentially expressed in oocytes, such as Gdf9, Pou5f1 (also known as Oct4), and Mos, are drastically downregulated in ovaries deficient in NOBOX [13]. In addition, our data suggest that NOBOX directly binds Gdf9 and Pou5f1 promoters, and may therefore directly regulate transcription of Gdf9 and Pou5f1 [26]. In order to better characterize effects of Nobox deficiency on the global newborn ovary transcriptome, we analyzed Nobox knockout and wild-type newborn ovaries with the whole-genome mouse expression arrays. We chose to compare newborn ovary transcriptomes due to the grossly similar histology and oocyte numbers in both wild-type and Nobox-deficient mouse ovaries. We used Affymertrix 430 2.0 mouse microarray because it is the most comprehensive array system that allows expression inquiry for more than 34 000 well-substantiated mouse genes and expressed sequence tags (ESTs). Analysis of three independent experiments comparing Nobox-deficient and wild-type newborn ovaries revealed that a total of 1432 genes showed significantly different expression levels (Supplementary Tables 1–4, available online at www.biolreprod.org). Among these 1432 genes were genes that we previously identified as downregulated [13], except for Dnmt1o and Fgf8, which were not detected in our microarray analysis. The oocyte-specific isoform of Dnmt1 (Dnmt1o) is not represented on the Affymetrix 430 chip, and as such will not be detected. Fgf8 is not abundantly expressed and therefore is detectable only by RT-PCR but not by microarrays in the newborn ovaries.
Genes that showed greater than 2-fold expression differences in Nobox-deficient ovaries are involved in many different biologic processes (Supplementary Table 2). We chose to further analyze a 5-fold difference arbitrarily, thinking that such genes are more likely to be directly regulated by Nobox. A total of 38 genes were downregulated more than 5-fold in the Nobox-deficient newborn ovaries (Table 1 and Supplementary Tables 3 and 4). We used Unigene EST database and GNF Sym Atlas resources to determine which genes are preferentially expressed in germ cells [27]. A total of 28 (74%) of 38 genes that were downregulated more than 5-fold showed preferential expression in oocytes (Table 1), whereas only 5 (15%) of 33 genes that were upregulated more than 5-fold showed preferential expression in oocytes (Table 2).
|
|
Crabp2 (cellular retinoic acid binding protein 2) and Stra8 (stimulated by retinoic acid gene 8) are both involved in the retinoic acid metabolism and are among the few upregulated genes expressed in oocytes (Fig. 1G and Table 2). Crabp2 is not essential for oogenesis, because Crabp2-deficient mice are fertile and grossly normal [28]. Stra8 is essential for fertility and is expressed in premeiotic germ cells [29]. Stra8 may be essential for the onset of meiosis, but it disappears as oocytes become arrested in meiosis in wild-type ovaries, and it is not detectable at the newborn ovary stage. Stra8 persistent expression in the Nobox knockout newborn ovary (Fig. 1G) signifies that Nobox-deficient oocytes are immature at the molecular level, despite forming primordial-like follicles.
|
Genes Essential for Pluripotency Are Downregulated in Nobox Knockout Ovaries
Pou5f1 and Sall4 transcription factors are downregulated in Nobox knockout ovaries more than 30- and 20-fold, respectively (Fig. 1, A and D) [13]. Pou5f1 is a POU domain-encoding homeobox gene that is essential for pluripotency and critical in primordial germ cell biology [30]. Pou5f1 expression diminishes around Embryonic Day 15.5 (E15.5), when oocytes enter meiosis, and reappears after birth [13, 31]. The role of Pou5f1 in folliculogenesis is unknown, but its reexpression in the newborn ovary may be essential in establishing reprogramming potential to the egg. Indeed, another transcription factor, Sall4, is a zinc finger transcriptional regulator known to be essential for pluripotency [32], expressed in oocytes [27], and to regulate Pou5f1 expression [32]. Sall4 is drastically downregulated in Nobox knockout ovaries (Fig. 1D) and may also play an important role in conferring oocytes the ability to reprogram. Other transcriptional regulators implicated in pluripotency, such as Sox2, Nanog, and Nr6a1, were not detectable in the wild-type mouse newborn ovary microarrays (data not shown). Lin28 is a developmental regulator preferentially expressed in oocytes, early embryos, and embryonic stem cells [27, 33]. Lin28 is expressed in wild-type newborn ovaries, but its expression is not affected by the lack of Nobox (Fig. 1F). The role of Lin28 in the pluripotency of the stem cell biology is not clear and awaits further investigations. The expression of Pou5f1, Sall4, and Lin28 in the newborn oocytes as well as the downregulation of Pou5f1 and Sall4 in the Nobox-deficient ovaries signifies that these essential pluripotency networks lie downstream of Nobox.
Kit, Kitl, and multiple other growth factors have been implicated in the promotion of primordial to primary follicles, but little is understood about oocyte-specific secretory factors that are critical in the formation of germ cell cysts and primordial follicles [9]. We therefore examined whether proteins predicted to be involved in signaling were affected by Nobox deficiency. Transcripts corresponding to astacin-like metalloendopeptidase (Astl), jagged 1 (Jag1), growth differentiation factor 9 (Gdf9), fetuin beta (Fetub), oocyte-secreted protein 1 (Oosp1), and R-spondin 2 homolog (Rspo2) were downregulated at least 3-fold (Fig. 1, B and E). Gdf9 plays a central role in oocyte signaling to the somatic cells and is downregulated in Nobox-deficient ovaries [13]. Gdf9 knockout experiments show primary follicle arrest [34], and Gdf9 does not appear to be essential in the formation of primordial follicles.
Little is known about the roles of Astl, Oosp1, Rspo2, Jag1, and Fetub in oogenesis. Astl is a member of the astacin family of metalloproteinases and is preferentially expressed in human and mouse oocytes [35]. Although its function has not been determined by knockout, its location and structure suggest that Astl is similar to hatching proteases in arthropods and fish. It is therefore likely that Astl is a maternal effect gene and does not play significant function in early folliculogenesis. Oosp1 is exclusively expressed in oocytes, but its function is unknown [36]. Functions of Fetub, Jag1, and Rspo2 in oogenesis are currently unknown, and future studies are necessary to determine the role of these genes in early folliculogenesis. Therefore, it is clear that transcripts encoding several secreted proteins are downregulated in Nobox-deficient ovaries, and some of these proteins may be critical in oocyte-somatic communication required for the breakdown of germ cell cysts and the formation of primordial follicles. Future studies on these proteins may improve the efficiency of in vitro maturation of oocytes.
Nlrp and Oas Family of Gene Expression in Nobox Knockout Ovaries
Nlrp genes belong to a large family of genes that encode proteins involved in apoptotic and inflammatory signaling pathways. Several Nlrp genes play critical functions in humans and mice. Nlrp5 (also known as Nalp5 and Mater) is a maternal effect gene essential for embryonic development past the two-cell stage [37]. NLRP7 (also known as NALP7) is a human maternal effect gene important in recurrent spontaneous abortions and molar pregnancies [38], whereas Nlrp3 (also known as Nalp3) dysfunction causes multisystemic inflammation. Several Nlrp genes are expressed preferentially in germ cells and include Nlrp14, Nlrp4f, and Nlrp4c [39]. Nlrp14 is preferentially expressed in oocytes and spermatogonia, and mutations in Nlrp14 can be found in men with spermatogenic failure [40, 41]. However, little is known about the in vivo roles for Nlrp14, Nlrp4f, and Nlrp4c in germ cell biology, as knockout mouse models have not been reported. Nlrp14 is drastically downregulated in Nobox newborn ovaries (Fig. 2, A and B) as well as Nlrp4f and Nlrp4c (Figs. 2A and 1C). Whether Nlrp14, Nlrp4f, or Nlrp4c is critical in early folliculogenesis or, like Nlrp5, acts as a maternal effect gene, awaits generation of animals that lack these genes.
|
Oas1d, Oas1c, Oas1e, and Oas1h belong to a 2',5'-oligoadenylate synthetase (OASs) family of genes. OAS family expression is highly induced by alpha/beta interferons as a response to viral infection. OAS proteins bind double-stranded RNA produced by virally infected cells, which ultimately leads to global inhibition of protein synthesis that blocks viral proliferation [42]. The unique feature of Oas1d, Oas1c, Oas1e, and Oas1h genes is their restricted expression to oocytes [27, 43]. In addition, Oas1d, Oas1c, Oas1e, and Oas1h comprise a subfamily of OAS proteins that can bind double-stranded RNA but lack OAS activity [43]. It therefore appears that these oocyte-specific OASs act by blocking OASs involved in interferon response and protect oocyte depletion from interferon-induced cell death, a property that oocytes share with embryonic stem cells [44]. Oas1d, Oas1c, Oas1e, and Oas1h proteins may therefore act to blunt oocytes' responses to an inflammatory milieu that characterizes folliculogenesis and ovulation, and therefore to stem oocyte loss to interferon-like responses. All four oocyte-specific OASs are downregulated in Nobox knockout newborn ovaries (Fig. 1, A and B), and it is tempting to speculate that such downregulation predisposes Nobox-deficient oocytes to premature death. Our findings that oocyte-specific OASs and NLRP proteins are downregulated in Nobox-deficient mice may signify that OAS and NLRP proteins play critical functions in oocyte survival. Bax, Bcl2l2, and Casp2 expressions do not significantly differ between Nobox knockout and wild-type ovaries (Fig. 2A), and oocyte-specific OAS proteins and NLRP proteins may be more important than other members of the classic apoptotic pathways in regulating oocyte death and survival.
Nobox Binding Elements Discovered with Protein-Binding Microarray
We used protein-binding microarrays to identify nucleotide motifs that NOBOX homeodomain preferentially binds [14, 23, 24]. NOBOX homeodomain was expressed as a GST fusion protein, and affinity to all possible 10-mers was assessed by protein binding microarrays. A position weight matrix shows a consensus for NOBOX-binding sequence (Fig. 3A). We analyzed 2 kb of nucleotide sequences upstream of genes downregulated more than 5-fold in the Nobox knockout ovaries for the presence of NOBOX DNA-binding sequences discovered by the protein-binding microarray. We have previously shown that NOBOX can bind these elements in the promoters of Gdf9 and Pou5f1 promoters [26]. In addition to Gdf9 and Pou5f1, a number of other downregulated genes, including Nlrp14, Nlrp4c, and Sall4, as well as secretory molecules Oosp1 and Astl, contain NOBOX DNA-binding sequences (Fig. 3C).
|
Among downregulated genes that contain NOBOX-binding elements is Olfr976. Olfr976 is an olfactory-like receptor preferentially expressed in oocytes [27] and downregulated more than 5-fold in the Nobox knockout ovaries. Olfactory receptors have been described in spermatozoa [45], although their precise role in sperm chemotaxis is unknown. Unlike testicular olfactory receptors, little is known about olfactory receptors in oogenesis, and expression of Olfr976 in the cell type specific fashion argues that it may have an important function in the biology of the oocyte. Whether Olfr976 acts during folliculogenesis or as a maternal effect gene is not known.
Genes Upregulated in Nobox-Deficient Ovaries
We analyzed genes that are significantly upregulated in the Nobox knockout ovaries (Table 2). The number of genes preferentially expressed in germ cells that were upregulated more than 5-fold was 15%, compared with 74% that were downregulated. A number of genes that are normally not expressed in wild-type newborn ovaries are significantly expressed in Nobox knockout newborn ovaries. Dmrt1, doublesex, and mab-3-related transcription factor 1 are not expressed in wild-type newborn ovaries. Dmrt1 is exclusively expressed in the somatic component of the genital ridge prior to sexual differentiation and plays a critical role in male gonadal development [46, 47]. Dmrt1 is overexpressed in Nobox knockout ovaries, which argues that oocyte-specific transcription factor NOBOX can repress male-determining gene expression (Fig. 1H). Other genes preferentially expressed in testes, such as Tekt2, 1110020C03Rik, and 1700019D03Rik [27], are upregulated more than 5-fold in the Nobox knockout compared with the wild-type ovary (Table 2). Tekt2 is essential for male fertility [48], whereas the roles of 1110020C03Rik and 1700019D03Rik are unknown. Other male-determining genes such as Wt1, Dhh, Sox9, Gata4, and Fgf9, do not show significant differences between the wild-type and Nobox knockout ovaries. Dmrt1 is expressed in germ cells, unlike other sex-determining genes such as Sox9 and Gata4, whose expression is confined to the somatic cells [49]. Dmrt1, Tekt2, and 1700019D03Rik are preferentially expressed in adult testis germ cells [27] and contain NOBOX-binding elements upstream of the transcription initiation sites.
Reexpression of male-determining genes has been observed in Foxl2 knockout ovaries [50]. Foxl2 is preferentially expressed in the pregranulosa and granulosa soma of the ovary and causes BPES I and II syndromes [51]. Our results indicate that transcription factors preferentially expressed in oocytes may repress male-determining gene expression.
The LH-responsive hydroxysteroid (17-beta) dehydrogenase 2 (Hsd17b2) is not significantly expressed in newborn mouse ovaries, but in the Nobox knockout ovaries Hsd17b2 expession rises more than 5-fold. Hsd17b2 plays an important role in the interconversion of estradiol and estrone. It therefore appears that the wild-type postnatal oocyte represses expression of Hsd17b2 either directly or indirectly. The overexpression of mainly somatic genes in the Nobox knockout ovary likely represents a homeostatic response to the globally dysfunctional oocytes and disrupted signaling pathways.
MicroRNA Expression in Nobox-Deficient Ovaries
MicroRNAs (miRNAs) are known to play important functions in vertebrate differentiation and development. MicroRNAs are usually 21–25 nucleotides long and appear to regulate mRNA functions on various levels, including translation [52]. Little is known about miRNA expression in newborn mouse ovaries or transcriptional regulators that control miRNA expression. We therefore studied whether miRNA expression in Nobox knockout ovaries differs significantly from that in wild-type newborn ovaries. We performed microarray analysis on 373 known mouse miRNAs, which represent known miRNA transcripts from the Sanger miRBase Release 9.0 [53]. A total of 177 of 373 miRNAs were significantly expressed in wild-type mouse newborn ovaries (signal greater than 250). The most abundant miRNAs include mir-709, which was recently detected as a novel miRNA during microRNA profiling of mouse embryos [54]. Let-7, a founding member of the miRNA family, is critical in cell fate determination in Caenorhabditis elegans [55]. Let-7 miRNA family transcripts are highly expressed in the newborn ovary (Table 3). An approximately 2-fold decrease in the Nobox knockout compared with wild-type ovaries was shown in mir-801, let-7d, mir-346, and mir-699 (Table 4). Changes in miRNA are relatively small compared with Nobox effects on the mRNA expression. Comprehensive sequencing of miRNA transcripts in various tissues is ongoing, and mouse oocyte miRNA transcriptome is currently unknown. Tissue-specific miRNAs exist [56], and oocyte-specific miRNAs are therefore likely.
|
|
The microarray analysis of the wild-type and Nobox knockout newborn ovary reveals disruption in the expression of many germ cell-specific genes, with wide-ranging functions, many novel and yet uncharacterized. The function of a large subset of these germ cell-specific genes, such as Nlrp14, Nlrp4f and Nlrp4c, Olfr976, and Astl, remains to be determined. Some of these genes may function as maternal effect genes and, as such, play little of a role in early folliculogenesis. Other genes, such as Gdf9 and Bmp15, which are downregulated in Nobox knockouts, play significant roles during the growth phase of ovarian follicles [34, 57].
Pluripotency genes Sall4 and Pou5f1 are expressed early during oogenesis. Whether these transcription factors are necessary for folliculogenesis or postfertilization is unknown. Mammalian oocytes can differentiate into many tissue types via parthenogenesis [58], and therefore molecules such as Sall4 and Pou5f1 may be important mediators of oocyte's totipotency. Whether other unknown molecules are distinctly necessary for oocyte totipotency versus embryonic stem cell pluripotency remains to be determined.
Using protein-binding microarrays we have discovered 8-mer sequences that NOBOX homeodomain preferentially binds. The top ten 8-mer sequences can be found in 18 (49%) of the 37 most downregulated genes and include genes such as Sall4, oocyte-specific Nlrp genes, Pou5f1, and Olfr976. Whether these elements are functional is difficult to assess due to the lack of oocyte cell lines. We have previously reported that Pou5f1 and Gdf9 promoters can bind NOBOX by immunoprecipitating wild-type newborn ovary chromatin using anti-NOBOX antibodies and amplifying genomic regions corresponding to putative NOBOX binding sites in the Pou5f1 and Gdf9. It is therefore likely that a number of potential NOBOX binding sites are functional. Preferential downregulation of numerous oocyte-specific genes in Nobox-deficient ovaries indicates that NOBOX is a transcriptional activator of oocyte-specific gene expression.
Genes essential to female determination are not well understood. Both somatic and germline components are likely to play key roles in ovarian determination. Foxl2 is a transcription factor preferentially expressed in pregranulosa and granulosa cells in the mammalian ovary, and its deficiency causes ovarian failure as well as reexpression of genes such as Sox9, Fgf9, and Dmrt1 [50]. It is unclear whether Foxl2 is truly required for commitment to ovarian differentiation, or whether observed reexpression of male-determining genes in Foxl2-deficient ovaries results from general response to disrupted oocyte development. Unlike Foxl2, Nobox is exclusively expressed in oocytes [12]. Reexpression of Dmrt1 (>10-fold increase over wild-type newborn ovary) in the Nobox newborn knockout ovaries, which still contain many oocytes, indicates that oocyte-specific transcription pathways also play a key role in repressing male sexual differentiation pathways. The role of germ cell-specific transcription factors in sexual differentiation is poorly understood [59]. Identification and characterization of transcription factors that act upstream of Nobox will help delineate early germ cell-specific pathways that contribute to the molecular basis of ovarian determination.
ACKNOWLEDGMENTS
We thank Stephanie Pangas for critical reading of the manuscript.
FOOTNOTES
1Supported by grant HD044858, and a March of Dimes grant (6-FY05-70) to A.R. 
Correspondence: 2Aleksandar Rajkovic, Department of Obstetrics and Gynecology, Baylor College of Medicine, 1709 Dryden Rd., Suite 1100, Houston, TX 77030. FAX: 713 798 2744; e-mail: rajkovic{at}bcm.edu
Received: 1 February 2007.
First decision: 5 March 2007.
Accepted: 30 April 2007.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  Y. Choi, D. Yuan, and A. Rajkovic Germ Cell-Specific Transcriptional Regulator Sohlh2 Is Essential for Early Mouse Folliculogenesis and Oocyte-Specific Gene Expression Biol Reprod, December 1, 2008; 79(6): 1176 - 1182. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. E. Newburger and M. L. Bulyk UniPROBE: an online database of protein binding microarray data on protein-DNA interactions Nucleic Acids Res., October 8, 2008; (2008) gkn660v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Choi, D. J. Ballow, Y. Xin, and A. Rajkovic Lim Homeobox Gene, Lhx8, Is Essential for Mouse Oocyte Differentiation and Survival Biol Reprod, September 1, 2008; 79(3): 442 - 449. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
