Characterization of NOBOX DNA Binding Specificity and Its Regulation of Gdf9 and Pou5f1 Promoters*

Nobox (newborn ovary homeobox gene) deficiency disrupts early folliculogenesis and the expression of oocyte-specific genes in mice. Here, we identified several cis-acting sites, TAATTG, TAGTTG, and TAATTA as NOBOX DNA binding elements (NBEs) using a library of randomly generated oligonucleotides by cyclic amplification of sequence target assay and mutation analyses. We show that NOBOX preferentially binds to the NOBOX binding elements with high affinity. In addition, we found that promoter regions of mouse Pou5f1 and Gdf9 contain one (–426) and three NOBOX binding elements (–786, –967, and –1259), respectively. NOBOX binds to these putative NOBOX binding elements with high affinity and augmented transcriptional activity of luciferase reporter driven by mouse Pou5f1 and Gdf9 promoters containing the NOBOX binding elements. In chromatin immunoprecipitation assays, DNA sequences from Pou5f1 and Gdf9 promoters co-precipitated with anti-NOBOX antibody. These results suggest that NOBOX directly regulates the transcription of Pou5f1 and Gdf9 in oocytes during early folliculogenesis.

Early ovarian folliculogenesis begins with the breakdown of germ cell clusters and the formation of primordial follicles, the reservoir of oocytes during reproductive lifespan of mammals. In mice, formation of primordial follicles begins soon after birth. Primordial follicles are the smallest ovarian follicle units and continuously recruited to grow into primary follicles. During early folliculogenesis, transcription of numerous oocytespecific genes such as zona pellucida (Zp) 2 1, 2, and 3 (Zp1, Zp2, and Zp3), growth differentiation factor 9 (Gdf9), and Pou5f1 (Oct4) commences (1)(2)(3)(4)(5)(6). However, the genetic mechanisms of early folliculogenesis are poorly understood (7). We recently reported a novel homeobox containing germ cell-specific transcription factor NOBOX (6,8). Nobox transcripts are detectable as early as E15.5 in the embryonic ovaries and beyond (6). Nobox mRNA and protein are preferentially expressed in oocytes of germ cell cysts and in primordial and growing oocytes throughout different stages of folliculogenesis (6,8). Nobox deletion accelerates postnatal oocyte loss and results in abnormal primordial and primary oocytes (6). Newborn ovary histology in Nobox knock-out and wild-type female mice is grossly similar, but differs significantly at the molecular level. Numerous genes preferentially expressed in the oocytes, including Gdf9 and Oct4 are down-regulated in oocytes that lack Nobox. However, it is unknown whether NOBOX directly or indirectly regulates transcription of these oocyte-specific genes during development of the ovarian follicle.
In the present study, we identified DNA sequences that NOBOX binds and show that NOBOX can transactivate luciferase reporters via the NOBOX DNA binding element (NBE). In addition, we describe identification of NBEs in promoters of mouse Oct4 and Gdf9. These observations suggest that NOBOX directly regulates the expression of oocyte-specific genes, Oct4 and Gdf9, during early folliculogenesis in the mouse ovary.

EXPERIMENTAL PROCEDURES
Expression and Purification of GST-NXHD-To create the pET41b-Nobox homeodomain protein (GST-NXHD) and bacterial expression construct, the insert was amplified by PCR using primers containing EcoRI and HindIII sites, and cloned into pET41b-EcoRI/HindIII (Novagen, Madison, WI). We verified the sequences of the resulting GST-NXHD fusion constructs to ensure that no mutations have been introduced during cloning. For creating GST-NXHD (N185Q) mutation proteins, the GST-NXHD construct vector was mutated using QuikChange multi site-directed mutagenesis kit (Stratagene, La Jolla, CA), and its mutated sequences were verified. We transformed BL21-pLysS Escherichia coli (Stratagene) with GST-NXHD and GST-NXHD(N185Q) constructs and expressed them by inoculating Luria-Bertani media containing 20 g ml Ϫ1 kanamycin with overnight culture (1:20 dilution) at 37°C to an A 600 of 0.5. Protein expression was induced at 30°C with 2 mM of isopropyl-1-thio-␤-D-galactopyranoside to A 600 of 1.0. We stored cell pellets at Ϫ80°C, thawed and lysed them with BugBuster lysis buffer (Novagen). We purified the GST-NXHD and GST-NXHD (N185Q) fusion proteins with GST bind resin (Novagen), dialyzed them three times in 1 liter of * This work was supported by Grant HD44858 from the National Institutes of Health and a March of Dimes Basil O'Connor Award (5-FY02-266) (to A. R.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1  phosphate-buffered saline and quantified them. Purified proteins were stored at Ϫ80°C until use. Tissue Extract-Ovaries from wild-type and Nobox-null newborn mice were collected and homogenized in 100 l of lysis buffer containing a final concentration of 20 mM HEPES (pH 7.9), 25% glycerol, 420 mM NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA. Homogenized ovary tissues were incubated on ice for 30 min and then centrifuged at 4°C for 30 min. After centrifuging, the supernatant was transferred into new tubes. Samples were stored at Ϫ80°C until use.
Plasmids-The pGL3-promoter vector with SV40 promoter (Promega) was used for constructing luciferase reporter vectors carrying 3xNBE and 3xmNBE. The pGL3-basic vector lacked SV40 promoter (Promega) was used to subclone amplified Oct4 and Gdf9 promoters. Oct4 or Gdf9 NBEs were mutated using QuickChange multi site-directed mutagenesis kit. Overexpression vector carrying the mouse Nobox were constructed by cloning the full-length of Nobox cDNA into pcDNA3 vector (Invitrogen) and sequenced. The pBIND vector (Promega) was used to construct plasmids carrying the full-length or truncated of Nobox cDNAs. Full-length and partial NOBOX cDNAs were fused after GAL4DBD in-frame and verified by sequencing. The expression of the constructs in mammalian cells was confirmed by Western blot analysis using anti-GAL4DBD monoclonal antibody (SantaCruz Biotechnology, Inc., Santa Cruz, CA).
Electrophoretic Mobility Shift Assay (EMSA)-EMSAs were performed in 20-l reaction mixtures at room temperature at a final concentration of 10 mM Tris, pH 7.5, 50 mM NaCl, 1.5 mM MgCl 2 , 2.5 mM dithiothreitol, 5% glycerol, 5 g/ml poly(dI)poly(dC), and 250 g/ml bovine serum albumin. EMSA probes were prepared by end-filling annealed primers with [␣-32 P]dCTP and Klenow polymerase (Invitrogen). Binding reactions were conducted by incubating 32 P-labeled probes (250,000 cpm/reaction) with 50 ng of purified proteins or 1 g of ovarian extract in the absence and presence of polyclonal anti-GST (Amersham Biosciences) or polyclonal anti-NOBOX antibody (6). For competition assay, purified proteins were incubated at room temperature with cold competitors for 15 min before addition of the 32 P-labeled probe.
Cyclic Amplification of Sequence Target (CAST) Analysis-The CAST assay was carried out essentially as previously described with some modifications (9). Purified GST-NXHD proteins and GST-bind resin (Novagen) were used for the CAST assay. Binding reactions for CAST were performed in DNA binding buffer (10 mM Tris. pH 7.5, 50 mM NaCl, 7.5 mM MgCl 2 , 1 mM EDTA, 5% glycerol, 5% sucrose, 0.1% Nonidet P-40, and 5 mg/ml bovine serum albumin). 0.5 pmol of short double-stranded DNA containing a 15-bp random sequence flanked by 20-bp fixed sequences, CAST oligonucleotides (Table 1), was incubated with 100 ng of purified GST-NXHD protein for 20 min at room temperature. Unbound DNAs were washed away with binding reaction buffer. Bound DNAs were amplified by PCR for the next round of CAST. A total of five cycles of CAST were performed. Final PCR products were cloned into the pGEM-T easy vector (Promega). The inserted DNAs were sequenced and scored to get the consensus sequences.

Name Sequences
CAST-oligo 5Ј-GAG TCC AGC GAA TTC TGT CG-(N) 15 Cell Culture and Reporter Assays-Human embryonic kidney cells (HEK293) were grown in Dulbecco's modified Eagle's medium with 10% fetal calf serum. For transient transfection, FuGENE6 (Roche Applied Science) was used according to the manufacturer's instructions. Following the transfection, the cells were incubated for 48 h before harvest. For each transfection, 200 ng of reporter construct, 200 ng of the indicated expression plasmid, and 20 ng of pRT-TK normalization plasmid were used per single well of a 12-well plate. Dual luciferse assays were carried out with total cell extracts as recommended by Promega. All transfection experiments were performed in triplicate, and results were normalized to the expression of the Renilla luciferase.
Chromatin Immunoprecipitation (ChIP)-Mouse newborn ovaries were collected and fixed directly by adding 27 l of 37% formaldehyde to 1 ml of Hank's balanced salt solution (Invitrogen). The fixed ovaries were washed with cold 1ϫ phosphatebuffered saline two times, homogenized into 4 ml of homogenization buffer (50 mM sodium phosphate buffer, pH 7.5, 10% glycerol, 10 mM ␤-mercaptoethanol, 50 mM phenylmethylsulfonyl fluoride, and protein inhibitors) and prepared for immunoprecipitation using the modified protocol of ChIP assay kit (Upstate Biotechnology Inc., Lake Placid, NY). The samples were pre-cleared with salmon sperm DNA/protein A-agarose (Upstate Biotechnology Inc.) and incubated with either 10 l of goat anti-Nobox antibody or IgG at 4°C overnight. Chromatin samples immunoprecipitated by salmon sperm DNA/protein beads were washed two times with 1 ml of low salt immune complex wash buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl, pH 8.1, 150 mM NaCl), once with 1 ml of high salt immune complex wash buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl, pH 8.1, 500 mM NaCl), once with 1 ml of LiCl immune complex wash buffer (0.25 M LiCl, IGE-PAL-CA630, 1% sodium deoxycholate, 1 mM EDTA, 10 mM Tris-HCl, pH 8.1), and two times with 1 ml of TE buffer (10 mM Tris-HCl, pH8.0, 1 mM EDTA). Immune complexes were eluted from the antibody by adding 500 l of elution buffer (0.1% SDS, 0.1 M NaHCO 3 ). The complex DNA cross-links were reversed by adding 20 l of 5 M NaCl and heating at 65°C for 4 h followed by adding 10 l of 0.5 M EDTA, 20 l of 1 M Tris-HCl, pH 6.5, and 2 l of proteinase K (10 mg/ml) for 1 h at 45°C. DNA was recovered by phenol/chloroform extraction and ethanol precipitation. Ethanol-precipitated DNA pellets were re-dissolved in 50 l of TE buffer. The supernatant of an immunoprecipitation reaction done in the absence of anti-NOBOX antibody was purified and used as a control. PCR analysis was carried out using primers corresponding to promoter regions of the Oct4 and Gdf9 genes. The sequences of the PCR primers are shown in Table 1. After 33 cycles of amplification (all primer sets), PCR products were run on 2% agarose gels and analyzed.

RESULTS
We determined a consensus DNA binding sequence for NOBOX using the CAST assay as previously described (10,11). We used a predicted homeodomain portion of NOBOX (aa 134 -214) fused to GST-NXHD. DNA sequences that bound GST-NXHD went through five cycles of CAST. The 38 selected sequences for GST-NXHD were aligned (Fig. 1A) and scored (Fig. 1B). The most frequently observed sequences are shown in Fig. 1B based on the percentage occurrence of each base at each position from 58 to 97%. CAST assay revealed a consensus sequence, TAATTG (Fig. 1B), as the NBE.
We next confirmed the interaction between NBE and GST-NXHD using EMSA. Using a competitive binding assay ( Fig. 2A), GST-NXHD protein bound to the labeled probe ( Fig. 2A, lane 2). The bound complex was competed out by 10-fold and 100-fold molar excess of unlabeled cold competitors containing TAATTG ( Fig. 2A, lanes 3  and 4, respectively). To further confirm the identity of these shifted bands, anti-GST antibody was added to the binding reaction in the absence or presence of GST-NXHD or GST protein. As shown in Fig. 2B, the anti-GST antibody further shifted the DNA-protein complex (Fig. 2B, lane 4). A number of homeoproteins including Msx-1, Nkx2.1, Nkx3.1, and Nkx3.2 contain a conserved asparagine residue (N) in the homeodomain (Fig. 2C). This conserved asparagine residue is critical for interaction of DNA base pairs and various homeodomains (12,13). A missense mutation of this asparagine (N) to glutamine (Q) has been shown to reduce DNA binding activity in many homeoproteins (9, 14 -17). NOBOX also contains this conserved N at residue 185 (Fig.  2C). A point mutation of NOBOX (N185Q) was generated and used to investigate whether GST-NXHD (N185Q) can bind to the NBE sequences in vitro. Purified GST-NXHD (N185Q) was incubated with 32 P-labeled NBE probe and binding was analyzed by EMSA (Fig.  2D). As shown in Fig. 2D, GST-NXHD (N185Q) failed to interact with the 32 P-labeled NBE probe, but GST-NXHD formed complexes with the NBE probe. These results suggest that the homeodomain is critical for DNA binding activity.
The NBE is identical to the core binding sequences of homeoprotein, MSX1 (18), but slightly different from the binding sites of other homeoproteins such as CAAGTG for TITF1 (19), TAAGTA for NKX3-1 (20) and TAAGTG for BAPX1 (9). We examined whether NOBOX can also bind to these consensus sequences. Increasing amounts of purified GST-NXHD protein incubated with 32 P-labeled probes containing these related sequences, TAATTG, CAAGTG,  Table 1.
TAAGTA, or TAAGTG were analyzed using EMSA. At high concentration (100 ng), GST-NXHD protein bound to these related sequences with low affinity compared with the NBE (Fig. 3A). In contrast, NOBOX bound to the NBE with high affinity relative to other related sequences at low concentration (Fig. 3A). These results suggest that NOBOX preferentially binds to the NBE sequence. To examine the effect of variation of each base pair of the NBE sequence, we incubated GST-NXHD with a labeled NBE probe (TAATTG) and added a 100fold molar excess of unlabeled oligonucleotides, which are different from NBE sequence by a single base pair substitution in each position. We found that the DNA-protein complex was competed out by two NBE-related sequences, TAGTTG and TAATTA, containing G in the 3rd position or A in the 6th position, respectively (Fig. 3B). In contrast, the binding specificity for the 1st, 2nd, 4th, and 5th position is invariant. The variation of 3rd position was not expected, however, the variation of 6th position was consistent with the result of CAST assay, because the occurrence of the A at the 6th position is 37% (Fig. 1B). These data suggest that TAGTTG and TAATTA are other consensus sequences for NOBOX DNA binding (Fig. 3C).
To confirm that endogeneous NOBOX from newborn ovaries binds to the NBE (TAATTG), the NBE was 32 P-labeled and incubated with ovary extracts from wild-type or Nobox-null mice. To identify these shifted bands, anti-NOBOX antibody or pre-immune serum was added to the binding reaction. As shown in Fig. 4, the NBE-protein complex formed using 32 P-labeled oligonucleotide containing TAATTG incubated with ovarian extract from wild-type newborn mice, but not from Nobox-null mice. Anti-NOBOX antibody shifted the DNA-protein complex (lane 3 in Fig. 4). We obtained similar results using NBErelated sequence, TAATTA.
To examine whether NOBOX can regulate the expression of NBEdriven reporter gene in vivo, we constructed two luciferase reporter vectors containing either three NBE (TAATTG), 3xNBE-pGL3, or mutant NBE (TAGGCG), 3xmNBE-pGL3 upstream of the luciferase reporter gene. These reporter vectors were co-transfected with either empty vector (pCDNA3.1) or NOBOX expression vector (pCDNA3.1-NX) into HEK293 cells. The relative luciferase activity of reporter constructs containing three copies of TAATTG was increased by 4-fold with NOBOX overexpression (Fig. 5A). However, the level of transcriptional activation of reporter constructs containing three copies of mutated NBE (TAGGCG) was not affected by NOBOX overexpression. These data suggest that NOBOX can transactivate reporter genes through the NBE.
To investigate the contribution of the various protein regions of NOBOX to transactivation, we generated a series of truncated NOBOX proteins fused to GAL4 DNA binding domain (GAL4DBD) (Fig. 5B). These constructs were co-transfected with 5xUAS-TATA-luciferase vector into HEK293 cells. The relative fold increase of luciferase activity is shown in Fig. 5D. GAL4-CT containing the C terminus of NOBOX significantly increased the luciferase activity. In contrast, the luciferase activity was not affected by GAL4-NT and GAL4-HD containing the N terminus and the homeodomain of NOBOX, respectively. Therefore, the C terminus possesses the activation domain of NOBOX.  Table 1. B, purified GST-NXHD proteins were incubated with 32 P-labeled probes containing TAATTG in the either the absence (lane 2) or the presence of a 100-fold excess of unlabeled oligonucleotides containing either the NBE or sequences different from the NBE sequence at each position. C, consensus DNA binding sequences for NOBOX are shown in a box.
We showed previously that Nobox deficiency disrupts the expression of several oocyte-specific genes, including Gdf9 and Oct4. Gdf9 is expressed early in the oocyte (3,6,21) and is essential for the growth and differentiation of the surrounding granulosa cells (22)(23)(24), antral follicle growth and ovulation (4,(25)(26)(27). Oct4 is expressed early in the development of the germ cell, but its transcription ceases after embryonic day 14.5 (E14.5) and reappears in the oocyte at birth (5,6). Thus, we examined the promoter region of mouse Gdf9 and Oct4 for the presence of the NBEs (TAATTG, TAGTTG or TAATTA). As summarized in Table 2, the 2.0-kb mouse Gdf9 promoter contains three putative NBEs (position Ϫ1259, Ϫ967, and Ϫ786) and the 2.0-kb mouse Oct4 promoter contains one putative NBE (position Ϫ426) relative to the translational start site (ϩ1). Interestingly, these putative NBEs contain the TAATTG sequence, but not the variant NBEs. Using a competitive binding assay (Fig. 6, A and C) and supershift assay (Fig. 6, B and D), we compared GST-NXHD to these putative NBEs containing TAATTG. We found that the DNA-protein complexes were competed out by 10-fold or 100-fold molar excess of unlabeled putative NBE sequences including Oct4 NBE (Ϫ462) and Gdf9 NBEs (Ϫ1259, Ϫ967 and Ϫ786) and were supershifted by antibody against GST.
To determine the effects of NOBOX on transactivation of mouse Oct4 and Gdf9, 293HEK cells were transiently transfected with a construct containing wild type or mutated NBE sequences. The 2.0-kb Gdf9 promoter and the 0.5-kb Oct4 promoter were cloned into the luciferase reporter vector. The NBE (Ϫ462) sequences in Oct4 promoter were mutated into GGATCC (BamHI) and three NBEs (Ϫ1259, Ϫ967, and Ϫ786) in Gdf9 promoter were mutated into GGATCC (BamHI), GCTAGC (NheI), and GGTACC (KpnI), respectively. These wild-type or mutated constructs were transiently transfected with Nobox expression vector or empty vector as a control. A Renilla luciferase expression vector was used as an internal control for normalization of transfection efficiency. The relative luciferase activities of the reporter plasmids containing wild-type NBE or mutated NBE are shown in Fig. 7, A and B. Nobox expression increased the activity of Oct4 and Gdf9 luciferase construct by 3.7-fold (Fig. 7A) and 6.2-fold (Fig. 7B), respectively. Mutation of the NBE sequences abolished Noboxdependent transactivation of Gdf9-luciferase construct (Fig.  7B), but NOBOX effect on mutated NBE containing Oct4-luciferase construct was partially decreased (Fig. 7A). These experiments suggest that Oct4 and Gdf9 are candidate genes for direct regulation by NOBOX.
To verify whether NOBOX directly binds to mouse Oct4 and Gdf9 promoters, we performed chromatin immunoprecipitation assays with protein extracts from newborn mouse ovaries or liver. Liver tissue does not express NOBOX and serves as a negative control. Extracts were cross-linked, sonicated, and immunoprecipitated with the anti-NOBOX antibody. PCR was used to amplify promoter regions containing NBE elements in the Oct4 and Gdf9 genes (P1) and regions that lacked NBE elements (C1) (Fig. 7C). NBE containing genomic sequences from the Oct4 and Gdf9 promoters co-precipitated with the anti-NOBOX antibody in the newborn ovary extracts (Fig. 7D). In contrast, co-immunoprecipitation was not detected in the liver extracts, which do not express Nobox or Oct4 (Fig. 7D).

DISCUSSION
Our previous study indicated that numerous oocyte-specific genes, including Gdf9 and Oct4, were down-regulated in ovaries that lack NOBOX (6). However, it is unknown whether NOBOX directly or indirectly regulates transcription of these oocyte-specific genes during development of the ovarian follicle. In the present study, we have identified a consensus sequence, TAATTG, as NOBOX DNA binding elements using purified proteins in a random DNA selection assay (Fig. 1). However, the sequences generated by a single base pair substitution in each position revealed that NOBOX bound two other NBE related sequences, TAGTTG and TAATTA (Fig. 3B). The variation of the 6th base was consistent with the result of CAST, but that of the 3rd base was unexpected. The TAATTG consensus site is shared with homeoprotein MSX1 (18), but TAATTA and TAGTTG appear unique to NOBOX. The NBE sequences are similar to other homeoprotein DNA binding elements such as CAAGTG for TITF1 (19), TAAGTA for NKX3-1 (20), and TAAGTG for BAPX1 (9).
NOBOX-interacting partners, tissue-specific expression, and other flanking nucleotide motifs are likely required additional elements for oocyte-specific effects of NOBOX. Msx1, Titf1, Nkx3-1, and Bapx1 are not expressed in the newborn ovaries (data not shown), which accounts for the dominant shift of TAATTG incubated with the newborn ovary extracts to be due to NOBOX (Fig. 4). Newborn Nobox-null ovary extracts lack the dominant shift band, in concordance with NOBOX being the  5); wild-type ovarian extract (lanes 2, 3, and 4); Nobox-null ovarian extract (lanes 6 -8); no antisera (lanes 2 and 6); anti-NOBOX antiserum (lanes 3 and 7); or preimmune goat serum (lanes 4 and 8). TAATTG containing NBE oligonucleotide sequence is shown in Table 1.
main binder of the TAATTG element in wild-type newborn ovaries.
Homeoproteins contain a highly conserved asparagines residue at position 51 in the homeodomain (28), and this conserved asparagine residue appears critical for the interaction of DNA base pairs and many homeodomains (12,13). NOBOX contains the conserved asparagine residue at position 51 in the homeodomain (position 185 in the protein) (Fig. 2C) and mutating this asparagines to glutamine abolished NOBOX homeodomain binding to NBEs. These results suggest that NOBOX binds DNA in a manner that is structurally similar to that of other homeoproteins.
Unlike many other homeproteins, NOBOX protein C terminus is proline-rich and contains a putative SH3 and WW domains (29). Esx1, a homeoprotein preferentially expressed in the placenta (30), also contains a proline-rich region that is important in localizing ESX1 to the nucleus and also contains a prolinerich SH3 domain that binds c-abl SH3 domain in vitro (31). Homeoproteins can execute their biological activity not only through protein-DNA interactions, but also proteinprotein interactions (18,(32)(33)(34)(35)(36)(37). This suggests that NOBOX prolinerich C terminus binds other, perhaps oocyte-specific proteins that are critical in NOBOX transcriptional activity.
We have shown that mouse Gdf9 and Oct4 promoters contain NOBOX DNA binding sites ( Table  2). OCT4 and GDF9 play a crucial role for the maintenance of germ cells and the formation of follicle, yet little is known about the direct regulation of expression of Oct4 and Gdf9 in the oocyte. GDF9 is an oocyte-derived growth factor required for somatic cell function. The lack of Gdf9 causes infertility secondary to an early block in follicular development at the primary one layer follicle stage (2). Oct4-null mutation results in peri-implantation lethality prior to the egg cylinder stage (38). However conditionally targeted Oct4 flox mice whereby Cre is under the control of TNAP (Tissue-nonspecific alkaline phosphatase) deplete oocytes by 6 weeks of age (39). These results suggest that GDF9 and OCT4 are required for follicle formation. . NOBOX transactivates through the NBE and the C terminus of NOBOX contains transactivational activity. A, the pCDNA3 or pCDNA3-NX was co-transfected with either 3xNBE-pGL3 or 3xmNBE-pGL3 luciferase reporter vector into HEK293 cells. Cell extracts were collected after 48 h of transfection and analyzed for luciferase activity. The black box indicates the wild-type 3xNBE (TAATTG) and the white box indicates the mutated 3xmNBE (TAGGCG). Oligonucleotide sequence is shown in Table 1. B, diagram of the full-length and truncated NOBOX proteins. Shown are the N terminus (NT), homeobox domain (HD), and the C terminus (CT). Numbers indicate the position of amino acid sequences. GAL4DBD represents GAL4 DNA binding domain. C, the full-length and truncated NOBOX proteins fused with GAL4DBD were transfected into HEK293 cells and cellular proteins were extracted in lysis buffer. Cell lysates were resolved by SDS-PAGE and visualized by Western blot analysis with anti-GAL4DBD monoclonal antibody. D, GAL4DBD vector containing the full-length and truncated NOBOX was co-transfected with 5xUAS-TATA-luciferase reporter vector into HEK293 cells. Cell extracts were collected after 48 h of transfection and analyzed for luciferase activity.

TABLE 2
Sequences of putative NBE in the promoter of mouse Oct4 and Gdf9 Putative NBE sequences are bold and underlined. Top strand shown for probes used in EMSA.

Oct4
Ϫ426 OCT4 plays a critical role in the basic biology of embryonic stem cells. Embryonic stem cells and germline share several important transcriptional regulators. The Oct4 transcript is differentially regulated by a TATA-less minimal promoter (40) and two cis-elements, distal enhancer (DE) and proximal enhancer (PE) (41). The DE is active in embryonic stem cells and the PE is active in the epiblast (41). Promoter elements in the Oct4 responsible for re-expression of Oct4 in the postnatal female germline are unknown.
In this study, we have shown that the promoters of mouse Oct4 (0.5 kb) and Gdf9 (1.5 kb) contain one (Ϫ426) and three NBEs (Ϫ786, Ϫ967, and Ϫ1259) from the ATG site (Table 2). Interestingly, these NBEs contain TAATTG sequences and they are not located in the PE or DE elements of the Oct4 promoter and the proximal region of the Gdf9 promoter. The NBEs are not conserved among other species, but several putative NBEs can be found in the proximal promoter regions of human, chimpanzee, pig, cow, and sheep Oct4 and Gdf9 genes (data not shown). Our results suggest that putative NBEs are candidate cis-acting elements for NOBOX activity during folliculogenesis. We have shown that NOBOX can bind putative mouse NBEs with high affinity (Fig. 5) and that NOBOX can trans-activate luciferase driven by Gdf9 (1.5 kb) and Oct4 (0.5 kb) promoters (Fig. 6). Because NOBOX is oocyte-specific, transactivation in oocytes may be even greater. Currently oocyte cell lines do not exist. Moreover, we have shown that antibodies against NOBOX can immunoprecipitate NBEs from the Oct4 and Gdf9 promoters (Fig. 6D). Nobox expression precedes closely the expression of Gdf9 and Oct4. All of these results implicate that NOBOX directly regulates the expression of Oct4 and Gdf9 in the oocyte during early folliculogenesis. FIGURE 7. NOBOX activates transcriptional activity of the promoter of mouse Oct4 and Gdf9. A, the pCDNA3-NX was HEK293 cells co-transfected with either pOct4NBE-pGL3 or pOct4mNBE-pGL3 luciferase reporter vector into HEK293 cells. Cell extracts were collected after 48 h of transfection and analyzed for luciferase activity. The black box indicates the wild-type NBE (TAATTG), and the white box indicates the mutated NBE (GGATCC). B, the pCDNA3-NX was HEK293 cells co-transfected with either pGdf9NBE-pGL3 or pGdf9mNBE-pGL3 luciferase reporter vector into HEK293 cells. Cell extracts were collected after 48 h of transfection and analyzed for luciferase activity. The black box indicates the wild-type NBE (TAATTG), and the white boxes indicate the mutated NBEs (GGATCC(Ϫ1259), GCTAGC(Ϫ967), GGTACC(Ϫ786)). C, PCR-amplified regions of the chromatin of mouse Oct4 and Gdf9 are described. P1 indicates the promoter region containing NBE. C1 indicates downstream region unrelated to NBE and used as a control. D, chromatin immunoprecipitation analysis of NOBOX binding for the mouse Oct4 and Gdf9 promoter in newborn ovary or liver extract. PCR amplifies DNA for P1 in mouse Oct4 and Gdf9 co-precipitated with Goat anti-NOBOX antibody. Input represents PCR product from chromatin pellets prior to immunoprecipitation. Samples incubated with 10 l of goat anti-NOBOX antibody (␣NOBOX) and the control sample without antibody (Bead) or IgG were used as template for PCR amplification. PCR primer are shown in Table 1.
Oocyte-specific pathways are critical for oogenesis. Transcription of numerous genes preferentially expressed in the oocyte, and critical for ovarian folliculogenesis and early embryogenesis, commences at birth when primordial ovarian follicles form. Transcriptional networks unique to oocytes are key to understanding and generating germ cells. NOBOX is one of the few germ cell-specific transcription factors that is essential in oogenesis and disrupts transcription of numerous oocyte-specific transcripts (6). Our findings indicate that transcriptional activity of NOBOX requires direct DNA binding through specific consensus sequences, the NBEs, and that NOBOX directly regulates the expression of oocyte-specific genes, Gdf9 and Oct4, through the NBE during folliculogenesis. These results also spotlight Oct4 as having a possible role in oogenesis, and future experiments are necessary to assess the function of Oct4 in early folliculogenesis and oogenesis. NOBOX transcriptional partners are unknown, and identification of NOBOX partners will be important to further investigate NOBOX specificity for genes preferentially expressed in the oocytes as well as in reprogramming somatic cells to germ cells.