Transcriptomic profile of genes encoding proteins responsible for regulation of cells differentiation and neurogenesis in vivo and in vitro – an oocyte model approach

The growth and development immature female egg cell -called the oocyte – is a complex process involving the varied expression of many proteins, metabolites and environment [1]. Our understanding of the mechanisms surrounding fertilization and in vitro fertilization (IVF) would benefit from more information on the oocyte such as potential bio-markers, stages of its growth, and what genes could potentially improve the competency of these cells.

The oocyte sits in the center of the ovarian follicle, surrounded by layers of granulosa cells and then thecal cells [2]. The granulosa cells are split in two populations: cumulus cells (CCs) and mural granulosa cells (GCs). To improve fertility and health of the follicle, interest is increasingly surrounding the bi-directional communication between the oocyte and CCs, a relationship identified as the cumulus-oocyte complex (COC).

The state of the cumulus-oocyte complex directly impacts the health and competence of the oocyte by cell nuclear regulation of meiosis [3]. The bi-directional communication between the gamete and somatic cell populations are conducted through gap junctions made up of connexins, and there is numerous evidence of molecules being exchanged to regulate the other’s activity such as proteins, oocyte-derived paracrine factors or glucose metabolic enzymes [4,5]. Poor fertility is often associated with a deficiency in communication from the CC side. Despite this co-operative relationship, not much is known about the gene expression of the COC.

Micro-array analysis and RT-qPCR can be used to study the transcriptome of the oocyte during its development, and identify molecular markers involved in neuron differentiation and negative regulation of cell differentiation [6]. In addition to building our understanding of the maturation of the ovarian follicle, these possible molecular markers could be used to try and manipulate CCs for their possible stemness qualities in a similar manner as Kossowska-Tomaszczuk’s lab were able to expose in vitro GCs differential capabilities [7].

In this study, the protein expression of genes belonging to ‘neuron differentiation’ and ‘negative regulation of cell differentiation’ will be chosen and analyzed in vitro and IVM of a porcine oocyte model. We aim to find if the genes are up- or down-regulated after IVM has taken place to better understand the microenvironment of the COC and establish the possible implications of the somatic and gamete cell communications.

Whole transcriptome profiling by Affymetrix microarray allowed us to analyze the gene expression changes in freshly isolated oocytes, before invitro procedure (“before IVM”), in relation to after in vitro maturation (“after IVM”). By Affymetrix® Porcine Gene 1.1 ST Array we have examined expression of 12258 porcine transcripts. Genes with fold change higher than |2| and with corrected p value lower than 0.05 were considered as differentially expressed. This set of genes consisted of 419 different transcripts. Subsequently, the genes were used for identification of significantly enriched GO BP terms.

DAVID (Database for Annotation, Visualization and Integrated Discovery) software was used for extraction of the genes belonging to regulation of “neuron differentiation” and “negative regulation of cell differentiation” gene ontology Biological Process term (GO BP). We found that 29 genes from these GO BP term were significantly represented in down-regulated gene sets. This set of genes was subjected to hierarchical clusterization procedure and presented as heat maps (Fig. 1).

Figure 1

Set of the differentially expressed genes belonging to “neuron differentiation” and “negative regulation of cell differentiation” GO BP terms with their official gene symbols, Entrez Gene IDs, ratio and corrected p values were shown in table 1.

The enrichment of each GO BP term as well KEGG pathway were calculated as z-score and shown on the circle diagram (Fig. 2)

Figure 2

Moreover, in Gene Ontology database genes that formed one particular GO group can also belong to other different GO term categories. By this reason we explore the gene intersections between selected GO BP terms. The relation between those GO BP terms was presented as circle plot (Fig. 3) as well as heatmap (Fig. 4).

Figure 3

Figure 4

STRING-generated interaction network was created with differentially expressed genes belonging to the “fatty acid metabolic process” ontology group. The intensity of the edges reflects the strength of interaction score (Fig. 5).

Figure 5

The cumulus-oocyte complex (COC) describes the barrier between cumulus cells (CCs) and the oocyte they surround [8]. Numerous hormonal secretions and changes in protein expression occur throughout growth and maturation of the oocyte, which makes up the inner layer of the ovarian follicle. CCs (with the higher metabolic activity) deliver energy and proteins necessary for the oocyte via gap junctions, and as the oocyte matures: gene expression, organelle distribution and the proteins present will be altered [8]. CCs are one of two populations of granulosa cells, the other being mural granulosa cells (GCs) which make up the outer layers of an ovarian follicle [9]. The essential processes for the growth and maturation of the ovarian follicle, called oogenesis and folliculogenesis, are controlled by granulosa cell and theca cell function [10]. GCs and CCs are waste material from IVF and are therefore accessible cells for studies [11].

To better understand how the growth and development of the ovarian follicle, studying the relationship between oocytes and CCs is important. One way to do this is by comparing the change of gene expression in an oocyte model before in vitro maturation (IVM) and after. Due to CCs involvement in oocytes in vitro growth and in vitro maturation (IVM), further study of the COC could shed light on the oocytes potential to differentiate into neuronal cells.

In this study, the expression of a group of marker genes in ontology groups ‘neuron differentiation’ and ‘negative regulation of cell differentiation’ were examined before and after IVM of the oocyte to evaluate the cell model’s stemness. After IVM, we found 29 down-regulated genes belonging to our chosen ontology groups, which can be used to suggest the change in oocyte function in this stage of development and how the oocyte interacts with CCs. The 10 most negatively regulated genes were selected and analyzed: VEGFA, BTG2, MCOLN3, EGR2, TGFBR3, GJA1, FST, CTNNA2, RTN4 and MDGA1.

VEGFA was the most downregulated selected and analyzed gene after IVM, and belongs to the ‘neuron differentiation’ ontology group. VEGFA plays an important role in inducing vascularization and angiogenesis, and its capability is emphasized by being able to induce neo-angiogenesis single-handedly [12]. The dramatic change in expression of the gene after IVM indicates an importance in regulating the development and maturation of the oocyte. Gao et al.’s study supports that VEGFA is down-regulated in oocytes after IVM [13]. VEGFA inhibits SMAD4, catalyzes KIT and ITGB1, and also binds to the latter gene.

The second most down-regulated gene was found to be BTG2 which regulates progenitor cell numbers and differentiation [14]. It will only act to induce differentiation of progenitor cells to orthosympathetic neurons in the presence of NGF. A study done by Li et al. suggest that the BTG family that BTG2 is a member of, has an essential role in regulating the cell cycle of granulosa and thecal cells [15]. If overexpression of the gene arrests the cell cycle, then this would explain why the maturation of the oocyte would downregulate this gene.

The next most negatively regulated gene that we selected and analyzed is MUCOLIPIN-3 (MCOLN3), which is a member of the mucolipin superfamily of transient receptor potential (TRP) channels [16]. The protein localizes to intracellular vesicles, in addition to the plasma membrane, and is involved in the endosomal pathway for calcium ion transport [17]. Found in both granulosa cells and ovaries, there is evidence that the gene helps drive differentiation in at least hair cells [18], in addition to being part of our ‘neuron differentiation’ ontology group.

EGR2 (early growth receptor 2) is the fourth most negatively expressed gene, and is necessary for peripheral myelin production [19]. This is shown by the process of myelination and nerve injury triggering the same expression profile of EGR2 protein [20]. Furthermore, EGR2 deficient mice did not express other genes that are associated with myelin [21]. Jin et al. have hypothesized that EGR2 is crucial for granulosa cells, as it interacts with gonadotropins to during folliculogenesis [22]. However, after IVM occurs, the expression of the gene has perhaps a less necessary function in the oocytes, which is supported by Brązert et al.’s study [23].

The fifth most negatively regulated gene is the first one from the ‘negative regulation of cell differentiation’ ontology group, and is TGFBR3. The activities of transforming growth factor-beta (TGFB) and inhibin are modulated by TGFBR3. In the presence of the receptor, TGFB will more readily bind to TGFBR2, and also inhibin more strongly repels the activity of activin and Bone Morphogenetic Proteins (BMPs). TGFBR3 null mice resulted in liver and heart defects, which suggestively led to neo-natal deaths [24]. Therefore, its expression would be essential for any endeavors to guide neurogenic differentiation in oocyte cells. TGFBR3’s change in expression in porcine oocytes before and after IVM has been examined before in Kranc et al.’s microanalysis study [25].

GJA1 (Gap Junction protein, alpha-1) is the sixth most negatively expressed gene and is part of the ‘neuron differentiation’ ontology group. Human connexin 43 (Cx43) is encoded by GJA1 [26]. Being responsible for cell-cell communication throughout the human organ systems, mutations in GJA1 are connected with ischemia and heart failure. Furthermore, negative neuronal differentiation and signaling is strongly linked to the protein’s activity [27]. In the ovaries, GJA1 form gap junctions for the COC so the microenvironment can communicate, and the ovaries of women who have PCOS show a lower expression of GJA1 [28].

FST (follistatin) is the seventh most downregulated gene, and the second in the chosen ‘negative regulation of cell differentiation’ ontology group. Follicle stimulating hormone is believed to be regulated by FST which inhibits the hormones activity [29]. Activin and inhibin, which negatively regulate each other, are also affected by FST, and downregulates activin’s activity in differentiating granulosa cells [30]. Multiple studies support that FST expression strongly affects the development and vitality of the oocyte and even the COC significantly [31, 32, 33]. In our study, we have evidence that FST binds with INHBA.

CTNNA2 (catenin cadherin-associated protein, alpha 2) is the eighth most downregulated gene, and is involved in ‘neuron differentiation’. The protein encoded is found to be mostly expressed in the central nervous system [34]. The absence of CTNNA2 results in problems for neuronal processes and cell migration [35]. Budna et al. also studied the expression of CTNNA2 in oocytes (originating from cows) before and after IVM, and related the gene to function as cell recognition and adherence [36]. They did not find any statistical significance in the change of expression between the two time periods, unlike in this study.

Nogo proteins are encoded by RTN4 (reticulon-4), which is the ninth most downregulated gene and also in our ‘neuron differentiation’ ontology group. These proteins have myelin-associated functions important for the endoplasmic reticulum proteins [37]. These functions include cell migration, metastasis, death and adhesion for myocytes, endothelial cells, and neurons. In studies of ischemic animal models, it has been suggested that the expression of RTN4 will increase to protect nerve function against the stress to a certain extent before the overexpression causes apoptosis of non-neurons [38]. In the study of Kranc et al., it was also noted RTN4 was less upregulated after IVM compared to before [39].

Our final gene is the ten most downregulated genes after IVM is MDGA1 (MAM domain containing glycosylphosphatidylinositol anchor 1 gene). Included in the ‘neuron differentiation’ group, MDGA1 is necessary for the migration of neural cells, such as to adhere to the top layer of cortical plates [40], and its expression can be used to detect neurons Reln-positive Cajal and Retzius [41]. There are little studies on MDGA1’s effect on oocyte or cumulus cells.

The least downregulated genes from our chosen GO BP terms were KIT, RYK, NOTCH2, RORA, SMAD4, ITGB1, SEMA5A, SMARCA1, WWTR1 and APP.

KIT, ITGB1, NOTCH2 and WWTR1 all belonged to the ‘negative regulation of cell differentiation’ ontology group. All of their functions contribute to the growth and development of the oocyte and would explain their higher expression to the other genes in our study. As mentioned before, our studies believe both KIT and ITGB1 catalyze VEGFA activity. In a follicle, KIT ligand prevents apoptosis of preantral cells so they can transition to the antral follicle stage [42], and, like VEGFA, its expression promotes proliferation of granulosa cells and the oocyte [43]. ITGB1 also furthers the development of the follicle in folliculogenesis, being involved in embryonic processes such as trophoblastic implantation [36]. Evidence from knockout mice studies and the association of the genes expression with PCOS, it is suggested that formation of the primordial follicle also relies on the expression of NOTCH2 (a member of the NOTCH family) in granulosa cells [44]. WWTR1 is a target gene in the Hippo pathway, which phosphorylates the protein to increase its promotional activity of proliferation and maturation of granulosa cells during ovulation [45].

Interestingly, most of the higher regulated genes are part of our ‘neuron differentiation’ GO BP terms group. These genes are RYK, RORA, SMAD4, SEMA5A, SMARCA1 and APP. One of the receptors for Wnt is receptor-like tyrosine kinase (RYK), and consequently acts to promote neurogenesis and axonal growth [46]. A study done by Budna et al. supports that expression of RYK is downregulated after IVM [47], and its role in follicular development should be further studied. Related morphan receptor A (RORA) is usually found in mesenchymal stem cells originating from bone marrow, and its function in the oocyte could be regulating its development [48]. Kawaguchi-Niida et al. have shown that, for the TGFβ/BMP signaling pathway, expression of SMAD4 is needed to ensure differentiated neural progenitor cells don’t reverse their transition [49]. The presence of SMAD4 was also studied by Xing et al., and in particular they found it mainly expressed in the oocyte, but also in its surrounding granulosa and theca cells, whilst its expressing being dependent on the stage of follicle development [50].

SEMA5A is a member of the semaphoring family and expressed by oligodendrocytes and their progenitor cells [51]; and its expression was also found in oocytes to be increased in Nawrocki et al.’s study comparing gene expression in vitro compared to in vitro [52]. SMARCA1 encodes a protein which regulates chromatin remodeling, and inhibiting its function increased brain size in mice [53]. Finally, our last selected and analyzed gene is amyloid precursor protein (APP). Neuronal and brain development is influenced by the protein, the former relying on the APPs cleavage in precursor and neuron cells [54].

The genes belonging to our ‘neuron differentiation’ ontology group term have relatively little knowledge on how they affect processes concerning the oocyte, and even less on its interaction with cumulus cells. Therefore, more research should be done on these noted genes as insight could be given from understanding the COC microenvironment. With these results, our analyzed genes could perhaps be used for the purpose oocyte maturation biomarkers.

To summarize, we have identified the downregulation of various genes after the transition from in vitro to in vitro maturation through an oocyte model. The identified genes were part of two chosen ontological groups: ‘negative regulation of differentiation’ or ‘neuron differentiation’, with roles in the microenvironment such as proliferation, migration and differentiation. Our findings would like to highlight that the expression of the genes occurred due to the cumulus cells and the cellular relationship between the oocytes. Rather than a merely an exchange of nutrients, the COC provides the proteins and cellular environment necessary for processes such as folliculogenesis and oogenesis. Therefore, in future work, further focus should be given to the COC in order to improve the competency and understanding of the development of the follicle. Stemness of the cumulus cells and oocyte should also be considered because of our ontology groups expression, as differentiation could be directed in future studies.