Our study was first focused on a specific group of gene transcripts and proteins accumulated during the oocyte growth to become necessary, after fertilisation, for successful embryogenesis. These gene products and their effects are well documented in organisms such as Drosophila melanogaster or Xenopus laevis 1920, but in recent years, increasing evidence has been produced that maternal-effect genes are crucial in pre- 212223242526272829 and post-implantation 283031 development.

We analysed the profile of expression of the following five genes, known for playing a role during the oocyte-to-embryo transition: Zar1, Npm2, Stella (Dppa3), Smarca4 (Brg1) and Prei3 22262729.

The relative number of transcripts of the genes analysed resulted equally represented in the two types of MII oocytes, with the exception of Stella that was 1.4-fold down-regulated (p < 0.05) in MII^NSN oocytes (Figure 1). Consistently with the down-regulation of gene expression, immunostaining with an antibody against Stella did not detect the presence of the protein either in NSN antral or MII^NSN oocytes (Figure 2D, J), whilst in SN antral oocytes Stella co-localised around the nucleolus (Figure 2A) and in MII^SN oocytes was dispersed within the whole ooplasm (Figure 2G), as already shown in ovulated MII oocytes 22.

Stella down-regulation is a good candidate to explain the developmental block at the 2-cell stage that MII^NSN oocytes encounter following fertilisation 22. For this reason,, we wondered what induces the down-regulation of Stella in MII^NSN oocytes and whether other genes could be involved in its regulation.

Stella is positively regulated by Oct-4 and Nanog based on mouse Oct-4 ChiP-pet data 32, and it is also down-regulated upon RNAi-mediated Oct-4 knockdown in human embryonic stem (ES) and embryo carcinoma (EC) cells 3334. A recent study has demonstrated a strong link between Stella and Oct-4, showing that the expression of Stella in mouse ES cells is regulated by Oct-4 which is necessary for maintaining a specific chromatin structure within the locus containing Stella 35. The authors observed collapse of higher order chromatin structure throughout the locus and down-regulation of Stella expression following depletion of the Oct-4 gene. Additional studies using microarray analysis of gene transcription in ES cells, have described Oct-4 involvement in the regulation of chromatin modelling and chromatin-mediated transcription regulation 36.

Next in our investigation, we addressed the question whether the pattern of Oct-4 gene and protein expression in developmentally incompetent MII^NSN and competent MII^SN oocytes, in concert with the knowledge already acquired on this gene, could establish Oct-4 as a key regulator of Stella expression during the late phases of oocyte maturation.

Oct-4 (Pou5f1), a nuclear transcription factor that belongs to the POU family, is involved in the regulation of the expression of a number of developmental genes and is required for the maintenance of cell pluripotency [for a review see 37]. Oct-4 is expressed throughout folliculogenesis 38; following fertilisation, it remains of maternal origin until the 2-cell embryo, then at the 4- to 8-cell stage is first expressed by the embryonic genome 404142.

In an earlier study, we identified a 2-fold down-regulated expression of Oct-4 in NSN antral oocytes compared to SN antral oocytes 39. In the present work, we found that down-regulation of Oct-4 expression in MII^NSN oocytes correlated with a down-regulated expression of its encoded protein as revealed by immunocytochemistry. Figure 2M–2R illustrates the immunostaining pattern of the Oct-4 protein in fully grown antral SN and NSN oocytes. While in SN oocytes the protein was intensely present and mostly localised around the nucleolus (Figure 2M), in NSN antral oocytes it was undetectable (Figure 2P). Following in vitro maturation to the MII phase, Oct-4 continued to be absent in MII^NSN oocytes (Figure 2V–2X), whereas MII^SN oocytes maintained the presence of the transcription factor localised around the MII plate and partially within the cytoplasm (Figure 2S–2U).

In summary, since a sound experimental evidence has demonstrated that Oct-4 regulates the expression of Stella, Oct-4 and Stella down-regulation in MII^NSN oocytes may contribute to the inability of these gametes to develop beyond the 2-cell stage. Our results suggest a novel role of maternal Oct-4 in the regulation of the developmental competence of the female gamete.

Oct-4-dependent transcriptional networks have been described regulating self-renewal and pluripotency in human 33344344 and mouse 3236 ES and EC cells and in human mesenchymal cells 44. These studies demonstrate that Oct-4 may interact with other genes to regulate specific biological process.

We next asked the question, is Oct-4 a key regulator of genes associated with or implicated in establishing developmental competence on oocytes? To address this, we first analysed by microarrays, global differences in gene expression between the two types of oocytes, and then searched for genes known to be regulated by Oct-4 (based on ChiP-on-chip and ChiP-Pet) in the array-derived list of regulated genes.

The microarray platform that we used allowed the analysis of a total of 39,450 genes and gene sequences (from now on they will be all referred to as genes). Of these, 1,421 genes were expressed only in MII^NSN oocytes; 1,967 genes were expressed specifically in MII^SN oocytes and 5,646 genes were expressed in both types of oocytes. A fold-change cut-off of at least 1.5 and a p value < 0.05 was set to filter the genes whose expression was differentially regulated. In the comparison between MII^NSN and MII^SN oocytes, we identified a total of 380 regulated genes, of these 303 were up-regulated in MII^NSN oocytes, and 77 down-regulated (Additional file 1, see the 'MII-NSN vs. MII-SN' data sheet). Using the Gene Ontology (GO) annotations implemented in the Illumina platform (Illumina^®, 2007) and also by DAVID (Database for Annotation, Visualization and Integrated Discovery) 45, we identified 12 major biological themes that characterised the group of regulated and annotated genes (Figure 3A): transcription regulation (18%, mainly negative regulation), protein biosynthesis (18%), cellular transport (16%), metabolism (11%), development (8%), signal transduction (8%), cell cycle (5%, mainly negative regulation), carbohydrate metabolism (4%), DNA metabolism (3%), cytoskeleton organisation (3%), apoptosis (3%), chromatin/nuclear organisation (2%) and various GO (1%); also, some genes fell into a group with undetermined GO (162 unannotated genes). An inherent weakness of the bioinformatic programs available to analyse the patterns of gene expression obtained following microarrays analysis is that they are limited to those genes with assigned annotations or published relationships, and depend on the accuracy of these annotations. The number of down-regulated and up-regulated genes in each of the 12 categories and the list of these genes are shown in Figure 3B and Additional file 2, respectively.

To determine whether and how individual genes are inter-related or interact with each other and to search for biological pathways and the inter-relationships between network genes, we used the Ingenuity Pathways Analysis (IPA) version 5.5 analysis tool (Ingenuity^® Systems, http://www.ingenuity.com). The genes of each gene group (called focus genes) were uploaded as tab-delimited text files and IPA queried the Ingenuity Pathways Knowledge Base for interactions between focus genes and all other objects stored in the knowledge base and generated a set of networks. Those networks with a score ≥ 2 were considered for further analyses (see Additional File 3 for details).

This analysis identified a total of 25 target genes. When comparing MII^NSN with MII^SN oocytes, all these genes, with the exception of Odz2 and Zfp36l1 (down-regulated 1.6 and 2.3-fold, respectively), were up-regulated in MII^NSN oocytes with fold changes ranging from 1.7 (Iqgap1) to 62.4 (Pgm2) (Figure 4A). Eighteen of these genes were part of gene expression networks 1 to 8 generated from the IPA of up-regulated genes in MII^NSN oocytes (Table 1, genes highlighted in green), and thus involved in the activation of the same biochemical pathways described above. When IPA was specifically interrogated with the list of these 25 genes, including Oct-4, it generated 6 networks (Table 3) with a score ≥ 2, of these, the first two contained 12 and 7 (including Oct-4 itself) focus genes respectively, and had top functions centred around biological themes such as gene expression, cell death, cancer and reproductive system disease. Oct-4 was part of network 2 (Figure 4B), which included the TNF (tumor necrosis factor) pathway, a multifunctional proinflammatory cytokine that acts on several different signalling pathways to regulate apoptosis (directly activating the caspase cascade or through a mitochondria-mediated apoptosis [for a review see 47]), NF-kB activation of inflammation, and activation of stress-activated protein kinases (SAPKs).

Interestingly, Zfp36l1 (Zinc finger protein 36), one of the two Oct-4-regulated genes that was down-regulated in MII^NSN oocytes, is a member of the Tis11 family of early response genes, of which Zfp36l2, another maternal-effect gene crucial for early embryonic development, is also a member 48. Perhaps, Zfp36l1 is a novel maternal-effect gene whose expression is regulated by Oct-4 and whose down-regulation may affect the developmental competence of MII^NSN oocytes.

The microarray analysis revealed a 1.8-fold down-regulation of Oct-4 expression in MII^NSN oocytes. However, this data was not included in the list of regulated genes (Additional file 1) because its p value was equal to 0.051, thus just slightly above the set cut-off limit of p < 0.05. To understand how the expression of the Oct-4 gene was related with the adverse gene expression networks that we found activated in MII^NSN oocytes, we next included the expression data for Oct-4 obtained by microarray analysis (despite its p value was 0.051) into the list of up-regulated or down-regulated genes and performed an IPA analysis. With the group of down-regulated genes, Oct-4 was included in network 1 (which has top functions as gene expression, cancer and cell cycle; see Additional file 5); with the up-regulated genes, Oct-4 was integrated into network 2, which has top functions as post-translational modification, cancer, genetic disorder and activates adverse pathways such as oxidative phosphorylation and mitochondrial dysfunction, see Table 1 and Table 2). In both cases the gene fell into the two top networks highlighting the central role it plays.

In summary, down-regulation of Oct-4 expression in MII^NSN oocytes induces the up-regulation of a group of 18 Oct-4-regulated genes that are part of the top gene expression networks involved in the activation of adverse pathways, i.e., oxidative phosphorylation, mitochondrial dysfunction and apoptosis.

Four to six week-old B6C3F1 female mice (Charles River, Como, Italy) were injected with 3.5 I.U. PMSG (Pregnant Mare Serum Gonadotropin, Folligon, Intervet Srl, Italy) and sacrificed 48 hr later. Ovaries were isolated and antral oocytes with a diameter > 70 nm were collected in M2 medium by puncturing the ovarian surface with a sterile needle, they were transferred into subsequent washes of M2 medium, until they appeared, under an inverted microscope, completely free of cumulus cells. Then, they were classified into NSN and SN oocytes and matured to the MII stage (for details, see Additional file 3 and 61). The majority of SN (85%) and NSN (65%) antral oocytes matured to the MII stage. Approximately 50 MII^NSN or MII^SN oocytes were transferred in single 0.2 ml eppendorf tubes containing 50 μl Trizol Reagent (Invitrogen), for microarray analysis. For RT-PCR (Retro Transcription-Polymerase Chain Reaction) analysis, individual MII oocytes were transferred in 0.5 μl medium to 0.2 ml eppendorf tubes containing 1.5 μl lysis buffer 17.

Relative amounts of transcripts of the five maternal-effect genes (Zar1, Stella, Smarca4, Npm2, and Prei3) and of the MII^NSN-specific (Jam2 and Ogfr) and MII^SN-specific (Crebbp and Fnip1) genes were determined by using a semi-quantitative RT-PCR assay (for a detailed description of the assay, see Additional file 3 and ref. 39). For each experiment, at least 5 single MII^NSN or MII^SN oocytes were analysed. Statistical analysis was done using the SigmaStat 3.0 software and comparing samples with the t-test. Differences were considered significant when p < 0.05.

Oocytes were fixed with freshly prepared 4% paraformaldehyde for 30 min and then permeabilised with 0.5% Triton X-100 for 15 min at 4°C. To suppress non-specific binding of antibodies, the gametes were incubated with 0.5% blocking reagent (Roche, Boston, MA) in TNT buffer (0.1 M Tris-HCl, pH 7.5, 0.15 M NaCl, 0.05% Tween-20) for 20 min at 4°C. Immunostaining was performed with rabbit anti-Oct-4 polyclonal antibody (Stanta Cruz Biotechnology; sc 9081, 1:500 dilution) or mouse anti-Stella monoclonal antibody (Millipore; MAB 4388, 1:1000 dilution). Oocytes were incubated with primary antibodies for 1 h at 37°C. The primary antibodies were detected using appropriated secondary antibodies diluted in TNT for 1 h at 37°C: Alexa Fluor488-goat anti-rabbit IgG (Molecular Probes; 1:6000 dilution) for Oct-4 detection and Alexa Fluor488-goat anti-mouse IgG (1:10000 dilution) for Stella detection. After immunostaining, oocytes were washed in three changes of TNT for 15 min at 4°C, and then counterstained with DAPI (0.2 μg/ml in PBS for 5 min) and mounted in Vectashield (Vector).

Two independent pools of MII^NSN or MII^NSN oocytes consisting of approximately 50 oocytes per pool were used as the basis of our study. Messenger RNA was isolated using Oligo-dT-linked magnetic beads as described in Adjaye et al. 6263. In order to generate enough RNA for the subsequent microarray analysis, a two-round linear amplification protocol was adopted to generate biotin-labelled cRNA employing a linear amplification kit (Ambion, Austin, TX, United States). 1.5 μg of cRNA was used for the hybridisation reaction. Washing, Cy3-streptavidin staining, and scanning were performed on the Illumina BeadStation 500 (Illumina, San Diego, CA, United States) platform using reagents and following protocols supplied by the manufacturer. cRNA samples were hybridised onto Illumina mouse-6 BeadChips.

All basic expression data analysis was carried out using the manufacturer's software BeadStudio 1.0. Raw data were background-subtracted and normalised using the "rank invariant" algorithm. Normalised data were then filtered for significant expression on the basis of negative control beads. Selection for differentially expressed genes was performed on the basis of arbitrary thresholds for fold changes plus statistical significance according to the Illumina t-test error model.

Gene annotation was first performed with the BeadStudio software, combined with DAVID annotation tool 45. File management, automated annotation and other statistical analyses not reported in the paper were performed with Matlab^® Bioinformatics Toolbox. The networks and pathways were generated with Ingenuity Pathways Analysis (Ingenuity^® Systems, http://www.ingenuity.com; for details see Additional file 3).