| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Ovary; |
UIM en Biología del Desarrollo,3 Centro Médico Nacional Siglo XXI, Instituto Mexicano del Seguro Social, Mexico City, Mexico
Division of Neuroscience,4 Oregon National Primate Research Center, Beaverton, Oregon 97006
Departamento de Biología Celular,5 Instituto de Investigaciones Biomédicas, Universidad Nacional Autonoma de Mexico, Mexico City, Mexico
ABSTRACT
In recent years, several factors required for follicular assembly and/or early growth of newly formed primordial follicles have been characterized, but additional factors likely remain to be identified. We have used cDNA arrays to compare gene expression in the neonatal mouse ovary at 48 h (when primordial follicles are being assembled) and at 96 h (when early follicular growth is taking place) after birth to that of ovaries collected <24 h after birth (when follicles have not yet been formed). Segregating genes according to their pattern of expression revealed the presence of one cluster of 24 genes for which expression consistently increased at 48 and 96 h. The top increaser in this cluster encodes a
1.5-kb mRNA containing an open reading frame of 1401 bp that encodes a protein of 466 amino acids. The predicted 52.3-kDa protein is a member of the F-box-only (FBXO) protein family, termed FBXW15 or FBXO12J. It has a cytoplasmic localization that includes the endoplasmic reticulum. Expression of Fbxw15/Fbxp12J mRNA is oocyte-specific; the mRNA is first detected on Gestational Day 18, decreasing thereafter to minimal levels on the day of birth. The prevalence of Fbxw15/Fbxp12J mRNA increases again at 48 and 96 h after birth, coinciding with the time of follicular assembly and the initiation of early follicular growth, respectively. The specific expression of Fbxw15/Fbxp12J in oocytes and its developmental pattern of expression suggest a role for this gene in the regulation of oocyte physiology.
F-box proteins, follicular development, oocyte development, oocytes, ovarian development
Assembly of oocytes, epithelial cells, and mesenchymal cells into follicular structures and the gonadotropin-independent growth of newly formed primordial follicles are complex processes that require a precise coordination between germ cells and somatic cells. Initiation and progression of these processes are governed by a variety of genetically unrelated, but functionally connected, genes (for review, see [1–4]). A fundamental concept that has emerged from these studies is that oocytes play an essential role in directing both follicle formation and subsequent follicular development, because in the absence of oocytes, follicular assembly does not occur [2, 5]. Oocytes may fail to develop or die because of the isolated deficiency of one of several proteins. For instance, stem cell factor (SCF) supports germ cells while they migrate toward the genital ridge [6], and WNT4 maintains oocyte viability once germ cells have reached their final destination in the gonad [7]. Two oocyte-specific transcription factors also play a critical role in oocyte survival: factor in the germline 
(FIGLA) [8], and newborn ovary homeobox-encoding gene (NOBOX) [9]. Absence of these genes results in oocyte death and prevents the subsequent formation of primordial follicles because of a failure of the ovigerous cords to become organized into follicular structures [9, 10].
Many locally produced proteins contribute to facilitating differentiation and growth of newly assembled follicles. These factors are produced either by granulosa cells or by the oocyte itself, and they include KL (the KIT ligand [11, 12]), basic fibroblast growth factor (FGF2) [13], leukemia inhibitory factor [14], and the neurotropins nerve growth factor, neurotropin 4/5 (NTF5), and brain-derived neurotropic factor (BDNF) [15–17]. Counteracting the effect of these proteins is anti-mullerian hormone, which inhibits follicular growth [18]. One protein expressed only in oocytes (GDF9) and one protein predominantly expressed in these cells (TRKB, the high-affinity receptor for BDNF and NTF5) are also implicated in the control of early follicular growth. In the absence of GDF9, which is first produced by oocytes of the primary follicle, the oocytes continue to grow, but follicular development is arrested at the primary stage [19]. In the absence of full-length TRKB receptors, the oocytes die before follicular assembly, and this is followed by loss of follicular organization [17]. In the absence of both full-length and truncated TRKB receptors, however, oocytes only die after the follicles have already formed [16]. In addition to intraovarian proteins, maternal steroids also affect primordial follicular assembly and the differentiation of primordial into primary follicles [20].
Despite the progress made toward the identification of key molecules controlling the assembly and growth of primordial follicles, additional genes likely remain to be identified [21]. To search for genes with an ovarian expression that is developmentally regulated, we interrogated a fraction of the mouse genome via hybridization of neonatal ovarian RNA to DNA microarrays and identified an oocyte-specific gene for which expression increases biphasically during fetoneonatal development. This gene, which has been termed Fbxw15/Fbxp12J [22, 23], is a novel member of the rapidly expanding family of F-box proteins [24, 25]. Members of this family are called F-box proteins because they contain a conserved, 50-amino-acid motif first identified in cyclin F [26], a cell-cycle protein. Three subfamilies of F-box proteins are known: FBW, the members of which contain tryptophan-aspartate (WD-40) repeats near their carboxy terminus; FBL, the members of which contain leucine-rich repeats; and FBXO, which consists of proteins that contain potential protein-protein interaction domains not yet identified as being present in canonical F-box proteins [25]. Despite its early classification as a FBXW-encoding protein, Fbxw15/Fbxo12J has been reported to encode an F-box-only (FBXO) protein [23], and the present results indicate this as well. A unique feature that distinguishes Fbxw15/Fbxp12J from many other F-box protein-encoding genes is that it is specifically expressed in oocytes. A partial report of these findings has been published [27].
Experiments were conducted using C57BL/6J mice obtained from The Jackson Laboratory and CD1 mice maintained as an inbred colony at the Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México. The mice were housed under a controlled photoperiod (12L:12D, lights-on at 0700 h) and temperature (23–25°C) and were given free access to rodent chow and tap water. They were used in accordance with the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals. The experimental protocols were approved by the Animal Care and Use Committee of the Oregon National Primate Research Center and the Bioethics Committee of the Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México. The ovaries from C57BL/6J mice were collected <24, 48, and 96 h after birth for histology, cDNA array analysis, semiquantitative PCR, and Northern blot analysis as well as at Postnatal (PN) Days 6 and 12 for in situ hybridization. In one experiment (conducted in Mexico), ovaries were collected from CD1 mouse fetuses at 13, 15, and 18 days postcoitum (dpc) and from neonatal CD1 mice at PN Days 1 and 2 (for semiquantitative PCR) in addition to PN Day 4 (for whole-mount in situ hybridization). After mating, the presence of a vaginal plug was monitored to estimate the day of conception, and the pregnant animals were killed by cervical dislocation at different gestational times (13, 15, 18, and 19 dpc). On collection, the tissues were either frozen on dry ice for RNA extraction or fixed for in situ hybridization (see below).
Total RNA was isolated using the RNeasy Mini Kit (Qiagen) following the manufacturer's instructions. Briefly, the tissue was homogenized in TRI Reagent (Molecular Research Center), and the aqueous and organic phases were separated by addition of one volume of bromo-3-chloropropane (Sigma Chemicals), followed by centrifugation at 13 000 rpm for 15 min at 4°C. Thereafter, 350 µl of 70% ethanol (Sigma) were added to all samples, and each sample was applied to an RNeasy minicolumn, followed by washing by centrifugation at 8000 x g for 2 min with buffers containing guanidine and ethanol. To elute the RNA, 30 µl of RNase-free water were added directly onto the silica-gel membrane of the columns, which were then centrifuged for 1 min at 13 000 rpm. The RNA was quantitated by measuring absorbance at 260 nm and was stored at –85°C until used. The quality of each RNA sample was assessed on 2% formaldehyde denaturing agarose gels.
To analyze differentially expressed mRNA profiles at the time of follicle formation and early follicular growth, we employed a two-dye spotted cDNA microarray containing 8400 gene probes generated from a mouse NIA 15K gene set, printed in duplicate on glass slides by the Gene Microarray Shared Resource Facility of the Oregon Health & Science University.
Synthesis of cDNA from RNA of ovaries collected at <24, 48, and 96 h after birth was carried out using the Micromax TSA labeling and detection kit (NEN Life Science Products) as follows: 5 µg of total RNA were denatured at 65°C for 10 min, then annealed to oligo(dT)12–18 primer (0.1 mM) at 25°C for 5 min, followed by incubation for 1 h at 42°C with either fluorescein-conjugated dUTP (10 mM; in the case of samples from ovaries collected <24 h after birth) or with biotin-conjugated dUTP (samples from 48- and 96 h-old ovaries) plus dithiothreitol (DTT) (0.1 M), 1 µl of RNaseOUT (40 U/µl), and 20 U of Superscript II reverse transcriptase (200 U/µl; Invitrogen Life Technologies). Reactions were then purified on Microcon filters (Microcon YM-100; Millipore). Afterward, each of the two biotin-conjugated, reverse-transcribed samples from 48- and 96-h-old ovaries was mixed seperately with a fraction of the fluorescein-conjugated sample from ovaries collected <24 h postpartum. Before hybridization, microarray glass slides were prehybridized for 1 h at room temperature with anti-fluorescein/alkaline phosphatase or anti-streptavidin/alkaline phosphatase conjugates and blocked for 30 min with 600 µl of buffer containing Tris-HCl (0.1 M, pH 7.5), NaCl (0.15 M), and goat serum (10%; Invitrogen Life Technologies). Hybridization was performed at 65°C for 16 h. The slides were then washed with 20x sodium citrate/sodium chloride buffer (1:1). Array hybridization was visualized using the Tyramine Signal Amplification (TSA) protocol (PerkinElmer). In brief, the signals were generated by a series of room-temperature reactions that included 1) binding of an anti-fluorescein antibody conjugated to horseradish peroxidase (HRP) to the fluorescein-containing DNA, 2) addition of cyanine 3-tyramine (Cy3) in solution for the HRP to catalyze the deposition of Cy3 onto the array, 3) inactivation of HRP, 4) addition of streptavidin conjugated to HRP for binding to the biotin-containing DNA, and 5) addition of cyanine 5-tyramine (Cy5) so that the HRP catalyzed the deposition of Cy5 onto the array.
In this design, the 48- and 96-h samples were labeled with Cy5, and the <24-h samples were labeled with Cy3. No dye swap was necessary, because the TSA protocol incorporated fluorescein and biotin-labeled dUTP into the nascent cDNA from the target RNA instead of Cy5- and Cy3-modified nucleotides. Fluorescein and biotin-labeled nucleotides do not exhibit the incorporation bias of dye-modified nucleotides, thereby obviating control dye-swap experiments. Changes in transcript abundance derived from arrays using this labeling protocol have been shown previously to reproduce faithfully those obtained by alternate means, including Northern blot analysis and in situ hybridization [28].
Signal detection resulted from the precipitation of these dyes. Finally, microarrays were scanned (ScanArray 4000XLs; PerkinElmer), and the acquired images were saved as TIFF files for further analysis. The signal intensities were analyzed using print tip group lowess, as recommended by Yang et al. [29] and implemented by Sandrine Dutoit in BioConductor (http://www.bioconductor.org/). To identify genes that are differentially expressed at 48 and 96 h after birth in comparison to <24 h postpartum, genes showing a change of less than 1.8-fold in either direction between these three developmental phases were removed from further analysis [30]. Thereafter, the data were merged using OmniViz software, and genes showing a similar temporal pattern of expression were clustered and visualized via a K-means clustering algorithm using J-Express 2.0 software (http://www.ii.uib.no/bjarted/jexpress/) [31]. The array results have been deposited in the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO; http://www.ncbi.nlm.nih.gov/geo/) and are accessible through GEO Series accession number GSE8528.
To detect changes in Fbxw15/Fbxo12J mRNA levels, total RNA was first reverse transcribed employing the Superscript First-Strand kit (Invitrogen Life Technologies) following the manufacturers' instructions. All reactions were carried out in a total volume of 20 µl. Initially, 200 ng of total RNA were annealed at 65°C for 5 min to 0.5 µg of oligo(dT)12–18 primer (0.5 µg/µl) and 1 µl of a dNTP cocktail (10 mM). The annealed RNA-primer samples were incubated for 1 h at 42°C with RT buffer (10x), MgCl2 (25 mM), RNaseOUT (40 U/µl), and 50 U of Superscript II reverse transcriptase (50 U/µl). Reactions were terminated by incubation at 70°C for 15 min, followed by incubation at 37°C for 20 min with 2 U of Escherichia coli RNase H (2 U/µl).
Semiquantitative PCR reactions were performed using the QuantumRNA 18S Internal Standard Kit (Ambion). The PCR amplification of the reverse-transcribed products was carried out in a total volume of 25 µl, using either 2.5 µl of 10x PCR buffer, dNTPs (0.1 mM), and 0.15 µl of Taq polymerase (5 U/µl; HotStar Taq; Qiagen) or 22 µl of PCR Supermix (Invitrogen Life Technologies), plus 2.5 µl of a mixture of 18S primers/competimers at a ratio of 3:7 and 1 µl of cDNA template annealed to 10 pmol of one of three sets of primers (Table 1). These primers were directed against the sequence of AK087669, a predicted gene found to be the top increaser in the DNA array analysis and that we subsequently named Fbxw15/Fbxo12J (see Results). The PCR conditions used were 5 min at 94°C to activate HotStar Taq enzyme, followed by 30 cycles of 1 min of denaturing at 94°C, 1 min of annealing at 58°C, and 1 min of extension at 72°C, followed by a 10-min final extension at 72°C. Equal volumes of the PCR reactions were electrophoresed on 2% agarose gels stained with ethidium bromide. Afterward, the gels were scanned electronically, and the images were quantitated by densitometry using image-analysis software (Quantity One; Bio-Rad Laboratories).
|
Total RNA was obtained from different tissues derived from C57BL/6J 2 mice at PN Day 2. PolyA+ RNA was isolated using the Micropoly A pure kit (Ambion) following the instructions provided by the manufacturer. The resulting polyA+ RNA (750 ng) from each tissue was electrophoresed on a 1% formaldehyde/formamide agarose gel in 1x MOPS buffer (N-morpholinopropanesulfonic acid; Sigma) and transferred to a Hybond XL membrane (Amersham Biosciences) by capillary blotting using 10x SSC (1x SSC: 0.15 M sodium chloride and 0.015 M sodium citrate buffer) to drive the transfer. To irreversibly cross-link the nucleic acids to the membrane, the membranes were exposed for 35 sec to ultraviolet rays at 120 000 µJ using a cross-linker apparatus (Stratagene). The cRNA probe used for Northern blot hybridization was generated by in vitro transcription of a linearized, 346-bp AK087669 cDNA template (500 ng) (Table 1) cloned into the pGEM-T Easy vector, 2 µl of Transcription buffer (5x), 1 µl of DTT (1 M), 2 µl of rNTPs, 1 µl of UTP (100 mM), 0.5 µl of RNase inhibitor (RNaseOUT, 40 U/µl), 1 µl of T7 RNA polymerase (20 U/µl; Invitrogen Life Technologies), and 100 µCi of [32P]UTP. The reaction was incubated for 45 min at 37°C. The resulting cRNA was purified using Sephadex NICK Columns (Pharmacia Biotech). Afterward, the membrane was hybridized overnight at 65°C to 1.5 x 106 cpm/ml of probe and washed with 0.1x SSC and 0.1% SDS at 65°C with mild agitation, four times, for 15 min each time. The hybridization signal was developed by exposing the blot to x-ray film (Kodak) for 2–7 days at –85°C. The membrane was then stripped by incubation in 0.1% SDS for 5 min at 90°C and rehybridized to a mouse cyclophilin cRNA probe (1.5 x 106 cpm/ml) to correct for procedural variabilities.
The putative full-length coding region of Fbxw15/Fbxo12J mRNA was cloned by amplification of reverse-transcribed samples from PN Day 2 ovaries. The reaction was carried out in a final volume of 25 µl that included 1 µl of cDNA template, 22 µl of PCR SuperMix (Invitrogen Life Technologies), and 10 pmol of the forward and reverse primers from set B (Table 1). The PCR conditions used were as described above. The PCR products were electrophoresed on a 2% agarose gel, and the corresponding bands were purified with the QIAEX II Gel Extraction Kit (Qiagen). The PCR products expected to contain the Fbxw15/Fbxo12J coding region were then sequenced on an Applied Biosystems DNA Sequencer (model 377) using the Big Dye Terminator Sequencing Reaction kit (version 3.1; Applied Biosystems). Sequencing results were compared against the GenBank sequence database by means of the BLAST algorithm of the NCBI. The predicted amino acid sequence was analyzed using the Pfam program (http://pfam.janelia.org/) [32]. Multiple sequence alignment was done using the Clustal W program [33].
A tagged Fbxw15/Fbxo12J construct was generated by PCR-amplifying the Fbxw15/Fbxo12J coding region from ovarian RNA with a sense primer (5'-acgctgctagcactgaagatggcgatccatttac-3') and an antisense primer (5'-cggtctagatcacttgtcgtcatcgtctttgtagtcagagcagatgttca-3') containing a Flag epitope-coding sequence (in italics); NheI and XbaI sequences added to the sense and antisense primers, respectively, are underlined. The resulting construct was ligated into pcDNA-Zeo (Invitrogen).
To determine whether the open reading frame of Fbxw15/Fbxo12J mRNA translates a protein, COS-7 cells were transiently transfected with the Fbxw15/Fbxo12J-Flag expression vector described above. The cells were plated into 10-cm dishes (2.5 x 106 cells/dish) and grown to 80% confluence in Dulbecco modified Eagle-F12 medium containing 10% fetal bovine serum (HyClone) and no antibiotics. Thereafter, the cells were transfected with either 500 ng/well of Fbxw15/Fbxo12J-Flag-pcDNA or with the empty vector (pcDNA 3.1 Zeo) using Lipofectamine 2000 (Invitrogen Life Technologies). Forty-eight hours later, the cells were harvested by scrapping and collected by centrifugation at 600 x g for 5 min. The cell pellets were lysed and solubilized using a low-ionic-force buffer containing Hepes (10 mM), MgCl2, (1.5 mM), and KCl (10 mM), supplemented with aprotinin (10 µg/ml), PMSF (1 mM), DTT (0.5 mM), and the protease-inhibitor 1,10-phenantroline (10 mM; Sigma). Protein concentrations were determined using the Bradford assay (Bio-Rad Laboratories), and 30 µg/well of protein were electrophoresed on a 4–12%, precast, Tris-Glycine SDS-PAGE gel (Invitrogen Life Technologies), transferred onto Immobilon-P membranes (Millipore), and blocked with freshly prepared TBS-T buffer (10 mM Tris [pH 8.0], 150 mM NaCl, and 0.05% Tween 20) containing 5% nonfat dry milk at room temperature for 1 h. The membrane was immunoblotted overnight at 4°C with mild agitation using a mouse monoclonal antibody against Flag (M2; Sigma) diluted 1:1000, followed by incubation with an anti-mouse HRP-conjugated immunoglobulin (Ig) G (1:25 000; Invitrogen) for 1 h at room temperature. After stripping the membrane for 2 h at 55°C in a 62.5 mM Tris-HCl buffer (pH 6.8) containing 100 mM 2-mercaptoethanol and 2% SDS, the membrane was reblocked and reprobed using an anti-glyceraldehyde phosphate dehydrogenase mouse monoclonal antibody (AbCAM) at 1:40 000 dilution and detected with a anti-mouse HRP-conjugated IgG (Invitrogen) at a dilution of 1:25 000 at room temperature for 2 h. The reaction was visualized with enhanced chemiluminescence reagents (PerkinElmer Life Science).
Histology was performed to select the ages of the ovaries to be compared via cDNA arrays, based on the degree of follicular development achieved in C57Bl/6J mice during the first 4 days after birth. The ovaries were collected on the day of delivery and at 48 and 96 h after birth (n = 3 mice/age group). The ovaries were then fixed in Kahle fixative, embedded in paraffin, serially sectioned (thickness, 6 µm), and stained with Weigert iron hematoxylin, as described previously [15].
The in situ hybridization procedure employed was based on the method described by Simmons et al. [34] with some modifications [35]. Mouse Fbxw15/Fbxo12J riboprobes labeled with either [35S]UTP (for tissue-section hybridization) or digoxigenin (DIG; for whole-mount hybridization) were generated as described above for Northern blot analysis, transcribing 500 ng of cDNA template with 250 µCi of [35S]UTP (PerkinElmer) or 750 ng of template with DIG-UTP, using a DIG RNA Labeling Mix Kit (Roche Diagnostics). The hybridization reactions using [35S]UTP were performed on 14-µm cryostat sections derived from ovaries collected at PN Days 0, 2, 4, 6, and 12 and immediately fixed by immersion in 4% paraformaldehyde-0.1 M sodium borate buffer (pH 9.5, overnight at 4°C). Thereafter, the ovaries were placed in 10% sucrose-PBS for 24 h at 4°C, embedded in optimal cutting temperature compound (O.C.T.; Miles), and frozen on dry ice before sectioning. After an overnight hybridization at 55°C, the sections were washed [35] and exposed to NTB-2 emulsion (Roche Diagnostics). The hybridization signal was developed 3 wk later, and the sections were counterstained with hematoxylin (Sigma), dehydrated in ascending alcohols, and coverslipped for microscopic examination. Control sections were incubated with a 35S-labeled sense Fbxw15/Fbxo12J RNA probe transcribed from the same cDNA template used to prepare the antisense probe, but in the opposite direction.
Whole-Mount In Situ Hybridization
Hybridization was performed as recommended by Wilkinson [36] with some modifications. Unless stated otherwise, all reagents used for this experiment were obtained from Sigma. Ovaries from CD1 mice at PN Day 4 were fixed in 4% paraformaldehyde-PBS (pH 7.4) overnight at 4°C. Thereafter, the gonads were rinsed in PBS, dehydrated in increasing concentrations (25–100%) of methanol-PBS-Tween 20, and stored at –20°C overnight. The next day, the tissue was rehydrated and pretreated with proteinase K (40 µg/ml) for 20 min at 25°C, followed by incubation in hybridization buffer (50% formamide, 5x SSC, 1% SDS, and 50 µg/ml of heparin) for 3 h at 65°C. The tissue was then hybridized to 1 µg/ml of DIG-UTP-labeled Fbxw15/Fbxo12J cRNA in 5 ml of hybridization buffer at 63°C overnight. Before incubation with an anti-DIG alkaline phosphatase antibody diluted 1:2000 in TBS-T containing 5 mM levamisole, the tissue samples were treated with 10% goat serum (Gibco-Invitrogen) in TBS-T with 2 mM levamisole for 3 h at room temperature to block nonspecific binding. The incubation was performed overnight at 4°C with slight agitation. Before adding the alkaline phosphatase substrate (BM-Purple; Roche Diagnostics), the tissues were washed 17 times (10 min each time) using TBS-T with 2 mM levamisole and for 30 min using NTMT (50 mM Tris-HCl [pH 9.5], 50 mM MgCl2, 20 mM NaCl, and 1% Tween 20) with 1 mM levamisole. The reaction was terminated when a dark-purple precipitate was observed in the samples hybridized to the antisense probe; the substrate was then replaced with NTMT twice for 15 min, followed by three washes with PBS. The ovaries were stored in glycerol/PBS (1:1, v/v), and the hybridization reaction was visualized using a light microscope (Carl Zeiss). Detection of the precipitate at the cellular level was performed on 5-µm, paraffin-embedded sections.
To determine the intracellular localization of FBXW15/FBXO12J, COS-7 cells (1.5 x 105 cells/well in six-well plates) were transfected with either 450 ng/well of Fbxw15/Fbxo12J-Flag-pcDNA or with the empty vector (pcDNA 3.1 Zeo) using Lipofectamine 2000. Twenty-four hours later, the cells were fixed in acetone (1 min at –20°C). The FBXW15/FBXO12J-Flag was detected with an anti-Flag monoclonal antibody (M2) at a dilution of 1:1000, and the reaction was developed using an Alexa 488-conjugated donkey anti-mouse gamma globulin antibody (1:500; Molecular Probes). To visualize the endoplasmic reticulum (ER), we used rabbit polyclonal antibodies (sc-20132, 1:250; Santa Cruz Biotechnology) against oxidoreductase-protein disulfide isomerase (PDI), an ER-specific protein. The PDI reaction was developed with Alexa 594 donkey anti-rabbit gamma globulin (1:500). After completion of the immunohistochemical procedure, cell nuclei were stained with Hoescht 33258 (1:10 000) for 1 min. Fluorescent images were acquired with a Marianas imaging workstation (Intelligent Imaging Innovations).
Expression of Fbxw15/Fbxo12J mRNA was quantitated by densitometry, with the results expressed as Fbxw15/Fbxo12J mRNA:18S ribosomal RNA ratios. The results were analyzed by one-way ANOVA followed by the Student-Newman-Keuls multiple comparison test for unequal replications. A P value of less than 0.05 were considered to be statistically significant.
Follicular Development in C57BL/6J Mice
Because the initiation of follicular assembly varies in different strains of mice [37], we performed a histological evaluation of C57BL/6J mouse ovaries at <24, 48, and 96 h after birth. As shown in Figure 1, primordial follicles were absent in ovaries collected <24 h after birth, with clusters of oocytes distributed throughout the inner region of the ovary (Fig. 1A, arrowheads; higher magnification in Fig. 1D). By 48 h, primordial follicles were clearly discernible (Fig. 1B, long arrows; higher magnification in Fig. 1E), and at 96 h, primary follicles were also present, as evidenced by the presence of cuboidal granulosa cells surrounding single oocytes (Fig. 1C, short arrows; higher magnification in Fig. 1F).
|
With this information in hand, we isolated RNA from C57Bl/6J mice ovaries collected <24 h after birth (when follicles have not yet formed), at 48 h after birth (when assembly of primordial follicles is initiated), and at 96 h after birth (when primary and some secondary follicles are developing). The RNA from three independent pools of 15–20 ovaries per age was used to interrogate 8400 cDNA probes spotted on duplicate arrays, employing a total of 12 arrays. Using a K-means clustering algorithm provided by the J-Express 2.0 software, genes expressed at these three postnatal ages were clustered according to their pattern of expression. One cluster (cluster 2-3) (Fig. 2) contained 24 sequences for which expression consistently increased at both 48 and 96 h after birth in comparison to that at <24 h postpartum. Of these 24 genes, five were repeated two or three times within the cluster, a transcript redundancy that provides internal consistency to the array results.
|
Most of the sequences composing the cluster encode genes of known function. A prominent member of this group was Gdf9, which is required for the transition from primary to secondary follicles [19]. The cluster also contained sequences encoding three oligoadenylate synthetase proteins (OAS1C, OAS1D, and OAS1E), which are members of the family of interferon-induced antiviral proteins [38], an oocyte-specific H1 histone [39], and an egg-restricted phospholipase A2, possibly involved in membrane remodeling [40]. Of seven sequences encoding genes of unknown function, one (BG068427) was identical to a germ cell-specific, leucine-rich repeat protein of the NACHT/NLR family [41], and another (BG067959) was 100% homologous to Ndg1, a gene encoding a protein downstream of Nur77, involved in caspase activation and induction of apoptosis [42]. A third sequence (BG068435) was 93% identical to Oog1, a gene expressed in oocytes and early cleavage-stage embryos [43]. No similarities to known genes were found for three other sequences encoding predicted genes (BG080498, BG068530/BG069167/BG068104, and BG067232/BG068334/BG067812). The sequence of the top increaser, expressed sequence tag (EST) BG083112, was found to be identical to that of AK087669, a cDNA first found in a 2-day-pregnant, adult female mouse ovary library. Analysis of the AK087669 sequence indicated that it encodes an F-box protein. Although AK087669 was earlier annotated as Fbxw15 [22], more recently it was named Fbxo12J [23], because it was identified as belonging to a cluster of genes encoding proteins with an N-terminus F-box domain and no other recognizable domain [23]. These genes map to mouse chromosome 9F2 and are exclusively expressed in mouse egg libraries. In addition, AK087669 has high similarity (64% at the amino acid level) to another F-box protein-encoding gene termed Fbx12/Fbxo12 [23], also known as Fbxw14 (NM_0015793) [22].
The K-means algorithm identified two additional clusters of genes showing increased expression at 48- and 96-h postpartum (58 genes in each cluster). Because the main difference between these two groups was merely the magnitude of change observed at 48 and 96 h in comparison with that at <24 h, we analyzed them as a single group. Most of the transcripts listed in these clusters are ESTs and RIKEN cDNAs. Of interest, among the genes with functional annotations are those encoding ubiquitins B and C, two proteins involved in proteosomal degradation [44], and several genes with either enriched or specific expression in oocytes, such as the B-cell translocation gene 4 [45]; four members (Nlrp4a, Nlrp4f, Nlrp9a, and Nlrp9b) of the NACHT/NLR leucine repeat pyrin domain-containing family, which has similarity to the oocyte-specific gene Mater [46]; Bmp15, an oocyte-specific gene, and a gene (epad) encoding ePAD (a peptidylarginine deiminase-like protein) that has been shown to be abundantly expressed in oocytes [47]. Whereas Bmp15 encodes a bone morphogenetic protein involved in the control of follicular growth [48], ePAD has been implicated in the regulation of cytoskeletal reorganization in eggs and early embryos [47]. Additional genes with functional annotations include those encoding exonuclease 1, a cAMP-activated protein kinase-1; BCL2-like 10, which encodes an antiapoptotic protein [49]; and 2'-5' oligoadenylate synthetase 1C, which was also present in cluster 2-3.
A cluster containing 62 genes showing a 
1.8-fold decrease at 48 and 96 h after birth with respect to that at <24 h postpartum was also identified. Nineteen of these genes have functional annotations, with most of them encoding proteins involved in protein-protein interactions, cell metabolism, and ion transport. They include syntrophin 1, centrin 4, serine protease inhibitor 1-5, zinc finger protein 42, F-box and leucine-rich repeat protein 10, and T-cell-specific GTPase. Two particularly interesting downregulated genes were CD44 and insulin-like growth factor 2 (Igf2). CD44 is a hyaluronic acid receptor expressed in cumulus cells, where it plays a role in oocyte maturation but not in cumulus expansion [50]. Earlier, Igf2 was found to be prominently downregulated at the time of follicular assembly in an Affymetrix-based analysis of the rat ovarian transcriptome [21].
Expression Profile of BG083112 mRNA
To validate the array results, the changes in BG083112 mRNA expression that occur during the first 4 days of life were examined by semiquantitative PCR using a set of primers that amplify the BG083112 sequence (Table 1, primer set A). As shown in Figure 3A, BG083112 mRNA abundance increased significantly (P < 0.05) between <24 and 48 h after birth, and it increased even further (P < 0.01) at 96 h postpartum. These developmental changes raised the question as to the expression of BG083112 mRNA during fetal ovarian development. To address this issue, we used primers that amplifed the entire predicted coding region of AK087669 (Table 1, primer set B), and we found that AK087669 mRNA was absent in the ovary between 13 and 15 dpc (Fig. 3B). At this time of gestation, some oogonia are still dividing, but others have already initiated meiosis [51]. The abundance of AK087669 mRNA increased strikingly at 18 dpc (P < 0.01) (Fig. 3B), decreasing to very low levels 1 day later (19 dpc) (Fig. 3B). Minimal expression was seen on the day of birth, but as seen in C57 BL/6J mice, AK087669 mRNA prevalence began to increase again at 48 h postpartum (not shown). We next determine if AK087669 mRNA is expressed in tissues other than the ovary, and we used RT-PCR with a set of primers that targeted a segment of the AK087669 coding region (Table 1, primer set C) to examine nine different tissues. The results showed that expression of AK087669 mRNA was restricted to the ovary (Fig. 3C). Although a Unigene in silico analysis indicated that AK087669 was expressed in mouse skin, we were unable to amplify its coding region in RT-PCR experiments using skin RNA from 2-day-old mice (Fig. 3D). In contrast, the same primers readily amplified the AK087669 sequence from ovarian RNA obtained from the same animals. These findings suggest that at least in neonatal mice, expression of AK087669 (henceforth referred to as Fbxw15/Fbxp12J) is restricted to the female gonad.
|
FBXW15/FBXO12J Is Encoded by a Single mRNA Transcript Expressed Exclusively in the Ovary
To determine the approximate size of Fbxw15/Fbxo12J mRNAs and to verify that Fbxw15/Fbxo12J transcripts are expressed in the ovary, we isolated mRNA from different tissues collected from PN Day 2 C57Bl/6J mice and subjected them to Northern blot analysis. In agreement with the mRNA size predicted in silico, a single Fbxw15/Fbxo12J mRNA transcript of 1.5–1.6 kb was observed (Fig. 4). Consistent with an earlier analysis based on EST expression profiles [23] and with the PCR results of the present study, this single transcript was only detected in the ovary.
|
Fbxw15/Fbxo12J Is an Oocyte-Specific Gene
Hybridization histochemistry using a [35S]UTP-labeled Fbxw15/Fbxo12J cRNA demonstrated that Fbxw15/Fbxo12J mRNA was specifically expressed in oocytes of growing follicles. Sections from ovaries collected <24 h after birth showed little, if any, detectable hybridization signal (Fig. 5A), but by 48 h after birth, the signal became apparent in some primordial follicles (Fig. 5A). By 96 h, Fbxw15/Fbxo12J mRNA abundance was distinct in growing follicles, with the signal remaining strong at both 144 h after birth (Fig. 5A) and PM Day 12 (Fig. 5B).
|
A more detailed examination of the cellular sites of Fbxw15/Fbxo12J mRNA showed that Fbxw15/Fbxo12J transcripts were present exclusively in oocytes (Fig. 5C). Whole-mount in situ hybridization using a nonradioactive probe on 96-h-old ovaries showed that Fbxw15/Fbxo12J mRNA was highly abundant in follicles of the medullary region of the ovary (Fig. 5D). At this age, the ovarian cortex contains primordial and primary follicles, but the first wave of secondary follicles can already be detected in the medullary region. Neither the isotopic or nonisotopic in situ hybridization procedures revealed any hybridization signal in somatic cells (Fig. 5). Likewise, no hybridization was detected using sense probes (not shown).
Fbxw15/Fbxo12J/Fbxo12J belongs to a cluster of genes encoding F-box proteins that is localized on chromosome 9F2 [23]. These proteins belong to a family of F-box-containing proteins that display, in addition to the F-box domain, either other types of protein interaction domains or no recognizable domains [26]. The Fbxw15/Fbxo12J homology to these genes is high (86–99% at the nucleotide level and 64–77% at the protein level) (Table 2).
|
The RT-PCR of ovarian RNA from ovaries collected 48 h after birth and sequencing of the resulting PCR product demonstrated that the Fbxw15/Fbxo12J coding region is 1401 bp in length (Fig. 6). Comparison of this sequence with the genomic sequence of contig NT-095756.1 revealed that the sequence encoding Fbxw15/Fbxo12J mRNA spans 16 452 kb of genomic DNA and consists of 10 exons ranging in size from 37 to 210 bp (Fig. 6).
|
A search (http://www.pfam.sanger.ac.uk/) [52] for conserved motifs in the encoded Fbxw15/Fbxo12J protein sequence revealed that the first three exons (Fig. 6) encode an amino-terminus F-box motif extending from amino acids 3 to 46 (Fig. 7A). The F-box motif in FBXW15/FBXO12J is similar to that of F-box proteins containing WD-40 repeats, such as FBXW12 (Fig. 7A), but it is most similar to that of the FBXO protein FBXO12/FBXW14 (89% homology). Like all other members of the cluster, it also has similarity to the classical F-box domain of cyclin F (Fig. 7A). Despite of the similarity between FBXW15/FBXO12J and FBXW14, we did not identify a number of WD-40 repeats in the carboxy terminus of FBXW15/FBXO12J that would meet the threshold for classifying FBXW15/FBXO12J as an FBXW protein (Fig. 7B). Therefore, we submitted Fbxw15/Fbxo12J to the NCBI as a sequence (DQ067445) encoding a member of the FBXO protein class [24, 26]. Our classification is in agreement with that (FBXO12J) proposed by others [23], but not with the notion that it is an FBXW protein (FBXW15) [22].
|
FBXW15/FBXO12J Is a Cytosolic Protein Localized to the ER
To determine the size of FBXW15/FBXO12J, we transfected COS-7 cells with a plasmid encoding the coding region of Fbxw15/Fbxo12J mRNA tagged with a Flag epitope, and we extracted the cellular proteins 48 h later for immunoblot analysis using antibodies against Flag. As shown in Figure 8A, the cells transfected with Fbxw15/Fbxo12J-Flag expressed a protein of 52–54 kDa, which closely corresponds to the size of 52.3 kDa predicted by translation of the Fbxw15/Fbxo12J mRNA coding region. Further analysis of the protein (http://www.cbs.dtu.dk/services/SignalP) [53, 54] predicted the absence of a canonical signal peptide, indicating that FBXW15/FBXO12J is not a secreted protein. To identify the cellular sites where FBXW15/FBXO12J is localized, we transfected COS-7 cells with Fbxw15/Fbxo12J-Flag and examined the cells 24 h later by immunohistofluorescence-confocal microscopy using antibodies against the Flag epitope and against PDI, an ER-specific marker. As shown in Figure 8, B–D, FBXW15/FBXO12J immunoreactivity was cytoplasmic and also colocalized with PDI, indicating that some FBXW15/FBXO12J was associated to the ER.
|
The present study demonstrates that expression of a gene termed Fbxw15/Fbxo12J increases biphasically in the mouse ovary, first on 18 dpc and then at the time after birth (2–4 days) when assembly, differentiation, and growth of primordial follicles are initiated. Our results also show that expression of Fbxw15/Fbxo12J is oocyte-specific, and they support earlier findings identifying its protein product as a member of the F-box family of proteins [22, 23]. Because a search for conserved motifs indicated that the only recognizable domain in FBXW15/FBXO12J is the F-box motif, our results agree with the conclusion that FBXW15/FBXO12J is an FBXO protein [23]—that is, a protein that lacks additional, canonical protein-protein interaction domains [24].
A direct comparison between our cDNA array results and those previously reported using RNA from neonatal rat ovaries [21] is difficult, because the two studies used different array platforms. Whereas Kezele et al. [21] hybridized rat ovarian RNA to Affymetrix arrays containing sets of oligonucleotides representing 14 000 transcripts, our array was based on the use of cDNAs representing only 8400 transcripts. Kezele et al. also used a different paradigm, comparing PN Day 0 rat ovaries with ovaries collected at either Day 4 or after Day 7 in culture. Despite these differences, a striking similarity exists between the two studies, because in both cases, the number of upregulated and downregulated genes was broadly similar (148 vs. 140 upregulated genes, and 50 vs. 62 downregulated genes, in Kezele et al. [21] vs. the present study, respectively). Notwithstanding these important yet broad similarities, the identity of individual genes showing a change in expression is dissimilar, with one notable expression. In both studies, expression of Igf2 was downregulated at the time of follicular assembly. Further studies are required to determine the biological significance of this change.
The postnatal pattern of Fbxw15/Fbxo12J mRNA expression detected by cDNA arrays was similar to that of a cluster of 23 additional genes. Of note, all genes with known function within this cluster are specifically expressed in oocytes, highlighting the importance of the oocyte in the control of early follicular development. An earlier cDNA analysis of fertilized oocytes during the preimplantation stages demonstrated the existence of cohorts of genes with similar patterns of expression, leading to the suggestion that these genes not only may act in a stage-specific manner but also may be coordinately regulated [55]. The synchronous upregulation of several germ cell-specific genes seen in newborn mouse ovaries [56] reinforces this concept, and it raises the intriguing possibility that members of the cluster we have now identified by microarray analysis play complementary roles in either primordial follicle formation or early follicular differentiation and growth. Directly relevant to this concept is the finding that Fbxw15/Fbxo12J belongs to a cluster of oocyte-specific Fbxo genes localized on chromosome 9F2 [23]. Like Fbxw15/Fbxo12J, the only two members of this cluster thus far analyzed by in situ hybridization (Fbxo12B and Fbxo12D) were selectively expressed in oocytes of newly formed follicles.
Gene clusters are thought to arise from duplication of a common ancestral gene [57]. Because of their closeness, the individual members of these clusters likely are coregulated by common cis-regulatory elements. The chromosome 9F2 F-box gene cluster has a subtelomeric localization, which has been recognized as a genomic region containing rapidly evolving DNA sequences [58]. As such, the individual genes of a cluster become specialized to perform different yet complementary functions, in addition to providing a degree of redundancy required for the integrity of the cellular processes they subserve. That the chromosome 9F2 F-box genes are effectively silenced everywhere in the organism except the oocytes strongly suggests these genes are involved in oocyte-specific functions.
The nature of these functions has not been elucidated. Originally, F-box proteins were described as components of E3 ubiquitin protein ligase complexes termed SCFs, because their basic components include SCP1, cullin, and an F-box protein [59]. The SCFs mediate the phosphorylation-dependent ubiquitination of proteins, and they appear to be essential components of proteolytic events regulating not only cycle progression but also signal transduction [26, 59]. Within the SCF complex, F-box proteins provide substrate specificity to the complex by recognizing and recruiting proteins to the complex. The recruited substrates are then targeted by ubiquitin and degraded by the 26S proteasome [24, 59]. Despite the identification of more than 70 genes encoding F-box proteins in the mammalian genome, only six substrates have been identified, the more recent being the clock genes Cry1 and Cry 2, which are targeted by an FBXL protein [60, 61].
Because of the absence of a canonical signal peptide in the FBXW15/FBXO12J protein sequence, the colocalization of expressed FBXW15/FBXO12J-Flag protein with an ER marker is surprising. This ER localization, however, is consistent with earlier findings that some F-box proteins are associated with the ER membrane, where they function in the ubiquitination of ER-associated degradation substrates immediately after these substrates are retro-translocated to the cytosol for proteolysis [62]. Neither the substrates nor the function of the novel family of oocyte-specific FBXO proteins have been identified. That their encoding genes are clustered in the genome suggests elucidation of their function by gene targeting or siRNA-mediated mRNA knock-down of individual genes might be difficult, because rapid compensation for loss of function would be expected to occur. It would appear, therefore, that searching for regulatory mechanisms affecting the entire cluster, such as those provided by trans-regulatory control of gene transcription by specific transcription factors, or control of mRNA prevalence by common miRNAs, may be more fruitful avenues of inquiry.
The biphasic mode of Fbxw15/Fbxo12J expression observed in the mouse ovary suggests the participation of FBXW15/FBXO12J in different, developmentally regulated processes. The remarkable increase in Fbxw15/Fbxo12J mRNA abundance at 18 dpc may be related to attainment of the zygotene and pachytene stages of the first meiotic prophase by germ cells at this time of development [63]. Mitotic proliferation of germ cells ends by Embryonic Day 13.5 and is followed by meiosis [64, 65]. During this process, germ cells progress through the first three stages of prophase I of the first meiotic division (leptotene, zygotene, and pachytene), with most of them reaching the zygotene/pachytene stages by 18 dpc. They do not advance to the final (diplotene/dictate) stages until approximately the time of birth. Therefore, FBXW15/FBXO12J might function at this time to prevent oocytes from reaching the diplotene stage and, thus, impede premature termination of the long prophase that characterizes the first meiotic division of germ cells. On reaching the dictate stage of diplotene shortly after birth, the oocytes remain arrested in this stage until the time of ovulation [66].
In contrast to the transient increase in Fbxw15/Fbxo12J mRNA prevalence seen in fetal ovaries, the postnatal increase in expression is sustained and not easily related to oocyte apoptosis, which is rarely seen in mouse ovaries after birth. This sustained increase in Fbxw15/Fbxo12J mRNA abundance possibly contributes to preventing oocytes from exiting meiotic prophase I, a function ascribed to other ubiquitin ligases, such as RFPL4 [67]. Alternatively, FBXW15/FBXO12J may function as a regulator of signaling events required for oocyte-granulosa cell communication. Though it needs to be interpreted with caution, the cytoplasmic localization of FBXW15/FBXO12J-Flag suggests that FBXW15/FBXO12J may not be directly involved in the regulation of cell nucleus biology. Further studies, involving the development of specific antibodies, are necessary to elucidate the functions of FBXW15/FBXO12J during fetal and neonatal development of the ovary.
ACKNOWLEDGMENTS
We thank Ms. Maria Costa for her expert technical assistance in performing the in situ hybridization and immunohistofluorescence studies. We also thank Dr. Ricardo López and Dr. Miguel Angel Palomino for their kind assistance with the whole-mount in situ hybridization experiments and Dr. Norma Moreno for providing us with CD1 mouse fetuses.
FOOTNOTES
1Supported by National Institutes of Health grants HD-24870 and RR-00163 for the operation of the Oregon National Primate Research Center, the National Institute of Child Health and Human Development (NICHD) through cooperative agreement U54 HD18185 as part of the Specialized Cooperative Centers Program in Reproduction and Infertility Research (S.R.O.), and the Fondo para el Fomento a la Investigación from the Instituto Mexicano del Seguro Social grant 2005/1/I/166 (E.D.L.C.). B.K. was a postdoctoral research fellow (from the P. Universidad Catolica de Chile, Santiago, Chile) supported in part by a fellowship from NICHD TW/HD00668 Fogarty International Training and Research in Population and Health grant. E.D.L.C. was a graduate student of the Program en Ciencias Biológicas, Universidad Nacional Autónoma de Mexico. The array results have been deposited in the National Center for Biotechnology Information Gene Expression Omnibus (GEO; http://www.ncbi.nlm.nih.gov/geo/) and are accessible through GEO Series accession number GSE8528. 
Correspondence: 2Sergio R. Ojeda, Division of Neuroscience, Oregon National Primate Research Center/Oregon Health & Science University, 505 NW 185th Ave., Beaverton, OR 97006. FAX: 503 690 5384; e-mail: ojedas{at}ohsu.edu
Received: 29 June 2007.
First decision: 25 July 2007.
Accepted: 27 November 2007.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  M. Zuccotti, V. Merico, L. Sacchi, M. Bellone, T. C. Brink, M. Stefanelli, C. A. Redi, R. Bellazzi, J. Adjaye, and S. Garagna Oct-4 regulates the expression of Stella and Foxj2 at the Nanog locus: implications for the developmental competence of mouse oocytes Hum. Reprod., May 28, 2009; (2009) dep191v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. A. Fowler, S. Flannigan, A. Mathers, K. Gillanders, R. G. Lea, M. J. Wood, A. Maheshwari, S. Bhattacharya, E. S. R. Collie-Duguid, P. J. Baker, et al. Gene Expression Analysis of Human Fetal Ovarian Primordial Follicle Formation J. Clin. Endocrinol. Metab., April 1, 2009; 94(4): 1427 - 1435. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
