Transcriptomic analysis of cyclic AMP response in bovine cumulus cells

Abstract

Acquisition of oocyte developmental competence needs to be understood to improve clinical outcomes of assisted reproduction. The stimulation of cumulus cell concentration of cyclic adenosine 3′5′-monophosphate (cAMP) by pharmacological agents during in vitro maturation (IVM) participates in improvement of oocyte quality. However, precise coordination and downstream targets of cAMP signaling in cumulus cells are largely unknown. We have previously demonstrated better embryo development after cAMP stimulation for first 6 h during IVM. Using this model, we investigated cAMP signaling in cumulus cells through in vitro culture of cumulus-oocyte complexes (COCs) in the presence of cAMP raising agents: forskolin, IBMX, and dipyridamole (here called FID treatment). Transcriptomic analysis of cumulus cells indicated that FID-induced differentially expressed transcripts were implicated in cumulus expansion, steroidogenesis, cell metabolism, and oocyte competence. Functional genomic analysis revealed that protein kinase-A (PKA), extracellular signal regulated kinases (ERK1/2), and calcium (Ca²⁺) pathways as key regulators of FID signaling. Inhibition of PKA (H89) in FID-supplemented COCs or substitution of FID with calcium ionophore (A23187) demonstrated that FID activated primarily the PKA pathway which inhibited ERK1/2 phosphorylation and was upstream of calcium signaling. Furthermore, inhibition of ERK1/2 phosphorylation by FID supported a regulation by dual specific phosphatase (DUSP1) via PKA. Our findings imply that cAMP (FID) regulates cell metabolism, steroidogenesis, intracellular signaling and cumulus expansion through PKA which modulates these functions through optimization of ERK1/2 phosphorylation and coordination of calcium signaling. These findings have implications for development of new strategies for improving oocyte in vitro maturation leading to better developmental competence.

folliculogenesis is characterized by cumulus-oocyte communication, which heralds better oocyte competence. This communication is mediated by gap junctions and affects accumulation of maternal factors within the oocyte that are essential to support early embryonic development (49). Also, through gap junctions, cumulus cells transfer cyclic adenosine 3′5′-monophosphate (cAMP) to the oocyte where one role of cAMP is to maintain arrest at meiotic prophase-I (12, 50). Oocyte resumes meiotic division (oocyte maturation) either in response to preovulatory luteinizing hormone (LH) surge or on its removal from its follicular environment (13, 18). Additionally, a decline in intracellular cAMP can induce oocyte meiotic resumption, which presages a progressive closure of cumulus-oocyte cross talk and transfer of maternal factors to the oocyte (5, 7). This suggests a vital role of cAMP in oocyte competence acquisition (1).

Primarily, in the oocyte cAMP is involved in inhibition of oocyte maturation (12). The downstream pathways through which cAMP modulates meiotic maturation are not fully characterized in bovine (5). Several pathways have been implicated in cAMP signaling that communicate through intricate cross talk. Principally, cAMP prompts protein kinase A (PKA) (6), which stimulates the activity of CDK1 involved in oocyte maturation inhibition (27). Also, activation of PKA results in activation of key transcription factors like SP1/SP3 and CREB, which eventually regulate transcription of downstream genes involved in cellular growth, steroidogenesis, and cell morphology (45). In different cellular systems PKA has been implicated in calcium signaling through scaffold protein mAKAP (muscle A-kinase anchoring protein) (for review Ref. 17). Additionally, cAMP interacts with the cGMP pathway. The cGMP from granulosa cells diffuses through the gap junctions and maintains high cAMP levels by inhibiting PDE3A in the oocyte (37). These data suggest that cAMP activates multiple pathways to inhibit oocyte maturation, which involves cumulus-oocyte communication.

In parallel, cAMP rise in the nonoocyte compartment has been implicated in oocyte maturation promotion. During in vivo maturation, a preovulatory LH surge causes an acute rise in cAMP concentration within the mural granulosa cells (15). Subsequently, this cAMP induces transient expression of the epidermal growth factor (EGF)-like peptides amphiregulin (AREG), epiregulin, and betacellulin, which are key paracrine mediators of cumulus expansion, oocyte maturation, and ovulation (38). EGF-like peptides' secretion results in activation of EGF receptors and subsequently provokes phosphorylation of ERK1/2 and MAPK14 (p38) (14, 19, 39, 52). In follicle, activation of ERK1/2 is central for induction of oocyte maturation by the EGF pathway since inhibition of the EGF receptor abrogates cumulus expansion, oocyte maturation, and ovulation (46, 47). Recently, it has been shown that cumulus cell derived EGF-like peptide (AREG) regulates translation of a subset of transcripts within the mouse oocyte during maturation, and perturbation of this factor affects oocyte quality (10). These reports indicate an important role of cAMP in oocyte maturation and acquisition of developmental competence through the EGF pathway.

Oocyte in vitro maturation (IVM) is an important assisted reproductive technology with a wide range of research and clinical applications. This technology is often used to generate mature oocytes for human infertility treatment but is increasingly used to generate high-quality livestock embryos as well as in cloning and transgenic technologies. However, it is generally accepted that in vitro-produced embryos are afflicted with lower developmental potential compared with their in vivo counterparts, primarily due to inappropriate follicular priming (8, 35) and IVM culture conditions (20). Therefore, it seems that the rapid reactivation of the meiotic cycle does not provide the opportunity to various components of the culture media to influence the oocyte developmental potential. Considering the role of cAMP in follicular physiology, several studies have shown that artificial modulation of cAMP during IVM significantly improves oocyte developmental competence (1, 2, 30). Generally, high levels of intracellular cAMP are maintained either by phosphodiesterase (PDE) inhibitors to prevent cAMP degradation or through sustained activation of adenylate cyclase (AC) with agents such as forskolin (23). It is supposed that during in vitro culture, inhibition of an early onset of oocyte maturation through cAMP modulation may delay closure of cumulus-oocyte communication and partially compensates the compromised oocyte developmental competence. However, the use of AC stimulator alone will increase the intracellular concentration of cAMP but, without controlling PDE activity, will result in a higher variability to cAMP response. Therefore, the use of AC stimulators along with PDE inhibitors will generate better stability to high levels of intracellular cAMP. Besides, it is unclear whether or not cAMP is differently regulated during in vitro and in vivo maturation; the precise mechanism by which cAMP improves oocyte competence during IVM remains to be elucidated.

Acquisition of oocyte competence is a complex process that involves interplay between oocyte-derived paracrine and extrafollicular (endocrine) factors. In fact, cumulus cell properties are modulated by both the oocyte and the external factors. It is of worth that the cumulus gene expression pattern has been associated with the developmental competence of the oocyte contained therein (3, 26). Therefore, it is postulated that acquisition of the proper gene expression pattern endows cumulus cells with special properties that eventually presage oocyte developmental competence. Assuming the differences between methods of oocyte maturation in vivo or in vitro, the precise gene networks through which cAMP modulates cumulus gene expression pattern and heralds oocyte competence are currently unknown. Therefore, considering the roles of cAMP in inhibition of oocyte nuclear maturation, cumulus expansion, and improvement of blastocyst yield along with transduction of intracellular FSH signaling, we have studied cAMP signaling networks at the end of 6 h of stimulation with cAMP-raising agents.

MATERIALS AND METHODS

In this study all biological material were handled according to the guidelines of the Canadian Council on Animal Care. Cumulus-oocyte complexes (COCs) were obtained by aspirating follicles from the slaughterhouse ovaries, and no animal handling was done on the university campus. All chemicals were purchased from Sigma Aldrich (St. Louis, MO), unless otherwise stated.

Study Design

In this study, we explored cAMP-induced gene networks by analyzing gene expression patterns of cumulus cells from COCs cultured with forskolin (AC stimulator), IBMX (general PDE inhibitor), and dipyridamole [inhibitor of PDE8 that is the most active PDE family in bovine cumulus cells (43)] (treatment here called FID) for 6 h (CC-FID-6H) with similar controls without FID (CC-6H). The concentrations for FID components were chosen from previous experiments that studied the effect of cAMP on either oocyte competence or meiotic inhibition in bovine (2, 7). Also, we analyzed temporal dynamics of cAMP-responsive genes by doing an in situ analysis of our earlier published data that charted standard temporal dynamics of bovine cumulus cells transcriptome by comparing gene expression of cumulus cells derived from COCs cultured in vitro for 6 h (CC-6H) with expression patterns of cumulus cells of COCs immediately after collection (CC-0H) (GSE67969). Later, implications of various cellular pathways in cAMP signaling cells were ascertained through a specific inhibitor or stimulator using gene expression and ERK1/2 activation as functional end points. Last, we investigated regulation of dual specific phosphatase 1 (DUSP1) by different pathways and its correlation with ERK1/2 phosphorylation in various conditions.

IVM and Cumulus Cell Collection

Cumulus cells were collected from COCs with or without IVM for 6 h as described earlier (9). In brief, COCs were obtained by aspirating 2–6 mm diameter-size follicles from the slaughterhouse ovaries. Following washes in HEPES-buffered Tyrode lactate solution (TLH), groups of up to 30 selected COCs with at least five layers of cumulus were either subjected to mechanical cumulus isolation (Ct-0H) or were placed in 500 μl of medium and maturated for 6 h at 39°C with 5% CO₂ and 100% humidity, and then cumulus cells were separated mechanically. The maturation medium used was modulated according to the experimental design as 1) Control conditions: TCM-199 supplemented with 10% fetal calf serum (Medicorps), 0.33 mM of pyruvic acid, and 50 μg/ml of gentamycin; 2) FID supplementation: Maturation medium of control conditions was supplemented with forskolin (100 μM, BioMol), IBMX (500 μM), and dipyridamol (50 μM); 3) Effect of inhibitors: The inhibitor of PKA or stimulator of intracellular calcium were added as a) 20 μM of PKA inhibitor (H89) in FID-supplemented culture medium or b) 10 μM calcium ionophore (A23187) in control medium without FID.

Determination of Differential Transcript Levels in Cumulus Cells

Microarray analysis.

Total RNA of cumulus cells from three independent biological repetitions was extracted and purified with the Arcturus PicoPure RNA Isolation kit (Life Technologies, Burlington, ON, Canada) according to the manufacturer's instructions. Purified RNA was amplified using T7 in vitro transcription, and subsequently the resultant aRNA was labeled with Cy3 and Cy5, as described previously (9). The labeled aRNA (825 ng/replicate) was hybridized on EmbryoGENE bovine microarray slides (42) in a two-color dye swap design. Microarray slides were scanned with PowerScanner (Tecan, Mannedorf, Switzerland), and features were extracted with ArrayPro 6.3 (MediaCybernetics, Rockville, MD). Intensity files were analyzed by FlexArray 1.6.1 (http://genomequebec.mcgill.ca/FlexArray), which consisted of raw data correction by background subtraction and normalization within (Loess) and between (quantile) arrays. The statistical comparison of various conditions was carried out with the Limma package, and differentially expressed targets were considered to have twofold change and P < 0.05 after false discovery rate correction. Data were deposited into the National Center for Biotechnology Information Gene Expression Omnibus with accession no. GSE67968.

Reverse transcriptase quantitative PCR.

Next, total RNA from three biological replicates of cumulus cells, other than the ones used for microarray, was reverse-transcribed with qScript Flex cDNA Synthesis Kit (Quanta Biosciences, Gaithersburg, MD) using oligo-dT primers according to the manufacturer's instructions. Specific primers for each selected gene were designed with Primer3PLus available at http://www.bioinformatics.nl/cgi-bin/primer3plus/primer3plus.cgi. Reverse transcriptase quantitative PCRs (RT-qPCRs) were performed using LC Fast Start DNA Master SYBR Green I in LightCycler 2.0 (Roche, Laval, QC, Canada). Each primer pair was first validated for a single product through melting-curve analysis and running the PCR product on a gel, and PCR template was used to prepare a standard curve that was included in each reaction. The data were normalized through GeNORM (51) by using the values of two reference genes (ACTB and GAPDH). Primer sequences and annealing temperatures are shown in Table 1.

Table 1.

List of primers used for qRT-PCR

Sr No.	Name	Primer Sequences	Annealing Temperature, °C
1.	ADCYAP1	F_TACGACGAGGACGGAAACC	58
		R_CCAGCACTTTGCGGTAGG
2.	DUSP1	F_GAACTGTCCCAACCATTTCG	57
		R_CAAACTCAAAGGCCTCATCC
3.	HK2	F_AATGTTGAGCTGGTGGAAGG	57
		R_GCTTGGTGAAATCGATGAGG
4.	MAP2K3	F_TCCCAAGTGCCAAAGAACC	57
		R_ACCAATGCCAGGAGAGAGG
5.	SLC2A1	F_TGATGCGGGAGAAGAAGG	57
		R_TAGCGACACGACAGTGAAGG
6.	SLC2A3	F_CCTCCAACCTCAAACTCACC	57
		R_TGAAGCACCTCCCTCTAACC
7.	NR4A1	F_CCAGCCCACCTACTCACC	60
		R_CCAGACGGAGGATAAAGAGC
8.	AREG	F_CCACAGTGCTGACAGATTTGAGGT	60
		R_GTGAAGGTCATGGCAGAGACGAAA
9.	TP53	F_TCGGGAGAGGTCAGAATGTGTTCC	60
		R_CTTTGGCACTGAGGTTCACCAAGG
10.	EGR1	F_AGAGGCATACCAAGATCCACTT	55
		R_GGATGGGTAGGTCGAGGAG
11.	FSHR	F_AGAGCAAGGTGACAGAGATTCC	55
		R_ATGTAGTTTGGGCAGGTTGG
12.	CCK	F_ATAAGTGACCGGGACTACATGG	57
		R_ATTTCTGCCAAGGTGTGAGG
13.	AKT2	F_CAGACGAGAGGGAAGAGTGG	55
		R_GCACCAGGATGACCTTGC
14.	PGR	F_GATTCAGAAGCCAGCCAGAG	55
		R_TACCTTCCATTGCCCTCTTAAA
15.	PDGFB	F_GGATCACAGTTCACACACACCA	55
		R_GTAATGGAGGCCACATGGACAA
16.	HAS2	F_ATAAATGTGGCAGGCGGAAGAAGG	60
		R_GTCTTTGTTCAAGTCCCAGCAGCA
17.	STAR	F_CCCAGCAGAAGGGTGTCATC	58
		R_TGCGAGAGGACCTGGTTGAT
18.	GFPT2	F_TTTCCCATCCTCTCTTGGCTGCT	57
		R_TAGGGCAAGGTAGCCCACTTGAAA

RT-qPCR, reverse transcriptase quantitative PCR; F, forward; R, reverse.

Analysis of Upstream Regulators:

To ascertain which biological functions were affected by FID in cumulus cells and by what mechanisms, FID-sensitive genes were subjected to functional annotation with Ingenuity Pathway Analysis. Owing to a large data set (∼1,900 FID-sensitive transcripts), we developed a robust analysis strategy to identify the core regulatory networks which emanate from FID stimulation. We identified major upstream regulators that affect the transcription of differentially expressed genes in our data. We performed Ingenuity upstream regulator analysis, which was based upon previously known effects of different molecules on their target genes. Ingenuity attributes each upstream molecule an overlap P value and an activation Z-score. The activation score is a global score based upon the known effects (upregulation or downregulation) of a molecule on each of its target genes. Consequently, an upstream regulator is attributed an “activated” (Z-score >2), “inhibited” (Z-score <2), or uncertain state, based upon the observed changes in gene expression. In concordance between the activated status of most significant upstream regulators and differentially expressed genes in our data, we built Circos plots (29). This enabled us to simultaneously visualize activated upstream regulators and their downstream targets that are differentially expressed genes in our data.

Dynamic Features of FID-induced Gene Expression Pathway

In this regard, we constructed gene interaction networks in Ingenuity by taking into account the molecules affected by major upstream regulators. Furthermore, considering a putative biased signaling by FID, the genes in our differential expression data that belonged to intracellular networks already implicated in FID signaling, like PKA, PKB (AKT), PKC, EGF receptor, calcium, ERK1/2, as well as their upstream and downstream targets, were included in this quest. Additionally, to complete the missing linkages, we also included a few molecules that initially were not differentially expressed in our analysis, like proteins of the MAPK pathway. It is worth mentioning that our proposed pathway of FID signaling was based upon transcriptomic and RT-qPCR results. Moreover, the activated or inhibited status of different molecules or networks was determined through Ingenuity database, according to the status of upstream and downstream molecules.

Western Blot Analysis

Groups of 10 bovine COC were run on 4–15% SDS-polyacrylamide gels. The proteins were transferred to Hybond-P membranes (GE Healthcare) using the Mini-Protean 3 cell apparatus (Bio-Rad). Membranes were blocked for 1 h with Tris-buffered saline (TBS) containing 0.05% (vol/vol) Tween 20 and 5% skimmed milk. Incubation was performed overnight at 4°C in TBS-Tween containing the primary antibody, that is, either rabbit anti-ERK1/2 (Cell Signaling Technology, Danvers, MA) diluted 1:2,000, rabbit anti-p-ERK1/2 (Cell Signaling Technology) diluted 1:1,000, rabbit anti-MKP-1 (DUSP1) (Santa Cruz Biotechnology, Dallas, TX) diluted 1:1,000, mouse anti-α-tubulin (Sigma, St. Louis, MO) diluted 1:50,000; or rabbit anti-α-Tubulin (Cell Signaling Technology). The membranes were then washed three times in TBS-Tween and hybridized for 1 h at room temperature with the secondary antibody, namely peroxidase-conjugated goat anti-mouse IgG (Upstate) diluted 1:20,000 in the blocking buffer or peroxidase-conjugated goat anti-rabbit IgG (Jackson Laboratories, West Grove, PA) diluted 1:30,000. Detection was performed using the ECL kit, and the membranes were exposed to a Fusion FX7 digital camera (Vilber Lourmat, France). Images were analyzed with Fusion FX software.

Statistical Analysis

The analysis of microarray data was performed as described above in this section. The results of qRT-PCR and Western blot were analyzed by ANOVA followed by a Bonferroni test or t-test using the software GraphPad Prism 5 (GraphPad Software, San Diego, CA).

RESULTS

FID (high cAMP) Modulated Features of Gene Expression Pattern during Early In Vitro Maturation

The objective of this study was to elucidate cAMP/(FID)-induced gene networks in cumulus cells along with their temporal dynamics during early maturation. In this analysis, 1,897 transcripts were differentially expressed between FID-6H and Ct-6H with 828 upregulated and 1,069 downregulated.

Principle component analysis (PCA) categorized the three kinds of cumulus samples (FID-6H, Ct-6H, and Ct-0H) separately, indicating significant changes of cumulus transcriptome during early maturation as well as extensive gene expression modifications induced by FID (Fig. 1A). Likewise, the hierarchical clustering showed that each of the three categories of cumulus cells (FID-6H, Ct-6H, and Ct-0H) had distinct transcriptomic profiles (Fig. 1B) with the deepest segregation between cumulus cells before (Ct-0H) and after maturation (FID-6H and Ct-6H), affirming a larger influence of maturation on cumulus transcriptome. Although Ct-6H and FID-6H clustered closer, they remained categorically apart, indicating significant transcriptomic shift due to FID addition.

Fig. 1.

In vitro maturation instigated dynamic transcriptome changes in cumulus cells, whereas FID (forskolin-IBMX-dipryidamole treatment) modulated features of gene expression. Transcriptome analyses of cumulus cells derived from cumulus-oocyte-complexes (COCs) before in vitro maturation (Ct-0H) and after 6 h of in vitro maturation in control conditions (Ct-6H) and with FID supplementation (FID-6H) displayed as principle component analysis (A), hierarchical clustering (indicating correlation between different conditions) (B), and dynamic expression changes (C) of 1,897 FID-sensitive transcripts differentially expressed between Ct-6H vs. FID-6H. The increase or decrease in expression for each transcript was observed relative to expression level before in vitro culture (Ct-0H).

To illustrate FID-induced temporal kinetics of gene expression, we subjected the normalized intensity values of FID-sensitive (1,897) transcripts to gene clustering in an in silico analysis (Fig. 1C). This analysis grouped FID-sensitive transcripts into eight clusters with cluster 1 (367 transcripts) as the largest. The transcripts included in cluster 1 increased their expression during maturation (Ct-6H vs. Ct-0H), whereas FID further increased their expression. Cluster 2 (309) had a reciprocal pattern to cluster 1. The cluster 3 (164) and cluster 4 (179) also increased and decreased, respectively, during maturation; however, FID reduced the effect of maturation. Cluster 5 (257) and cluster 6 (275) contained transcripts that did not change expression during control maturation, while FID increased expression of cluster 5 transcripts and declined cluster 6 transcripts. Furthermore, the transcripts in cluster 7 (206) and cluster 8 (140) showed dynamic expression changes during control in vitro maturation since their expression levels increased and decreased, respectively; however, addition of FID abrogated the change in their gene expression.

Biofunction Analysis and Upstream Regulators of FID-sensitive Genes

To identify FID-sensitive genes, we focused our analysis on the differentially regulated transcripts between FID-supplemented vs. control group. FID-sensitive genes were subjected to functional annotation with Ingenuity Pathway Analysis. Here, forskolin, CREB1 (cAMP-dependent transcription factor), EGF (ERK1/2 activator), H89 (PKA inhibitor), and calcium were identified as the major upstream regulators of FID-induced differentially expressed genes. Considering ovarian physiology, these upstream regulators signified the implication of PKA, MAPK (ERK1/2), and calcium pathways as main signaling networks in cumulus cells. The transcription of 67 genes was regulated by these upstream regulators. Most of these upstream regulators showed an overlap, and many genes were cogoverned by more than one factor (Fig. 2).

Fig. 2.

Analysis of upstream regulators of FID-sensitive genes in cumulus cells. Ingenuity Pathway Analysis of FID-sensitive transcripts was used to ascertain major upstream regulators with their corresponding differentially expressed genes. Data are presented in the form of Circos plots for better visualization and overlap between different upstream regulators and biological functions.

Additionally, we conducted gene search to identify putative oocyte competence markers. We compared the expression patterns of our differentially expressed genes with previously published data (36). In fact, Nivet et al. (36) showed the gene expression patterns of granulosa cells that contained oocytes with better developmental potentials. Our data analysis indicates that FID induced expression of 81 genes, which is similar to Nivet et al. in cumulus cells (Supplementary Table S1).¹ Interestingly, many of these genes were not previously known to be involved in oocyte competence acquisition. These genes are interesting candidates to be analyzed as potential oocyte quality markers.

Validation of FID-sensitive Gene Expression Patterns

To validate the microarray expression data and temporal gene dynamics through RT-qPCR, we compared the expression levels of 18 candidate genes in cumulus cells across all three conditions (Ct-0H, Ct-6H, and FID-6H). In the choice of candidate gene, particular attention was given to include genes from all categories of different analyses performed earlier like 1) upstream regulators and 2) dynamic expression changes during maturation. Notably, expression patterns of all genes tested validated the expression directionalities observed in the clustering analysis and were statistically consistent with the array data except one (TP53) (Fig. 3). In particular, genes that increased expression level during standard maturation and were upregulated to greater extent by FID include the intracellular signaling protein kinase MAP2K3, which is an upstream activator of MAPK14 (p38), glucose metabolism (GFPT2), and steroidogenic factor (STAR). Furthermore, a gene that decreased expression during maturation and was further reduced by FID was transcription factor TP53, which is involved in cellular proliferation and differentiation. No difference was observed between Ct-0H and Ct-6H; however, FID significantly reduced TP53 expression compared with Ct-0H. The cluster 3 genes increased during maturation; however, in FID they did not increased as much as under control condition. These include cumulus expansion gene (HAS2) and ligand of EGF receptor (AREG). The transcripts from cluster 4 include transcription factor (EGR1) and DUSP1. The cluster 5 transcripts that did not change during maturation and augmented expression with FID addition include modulators of intracellular calcium level (CCK), glucose metabolism (HK2), and glucose transport (SLC2A1). The cluster 6 transcripts that did not show dynamic changes in control maturation but were downregulated by FID comprise a member of the AKT pathway (AKT2). The transcripts that changed expression level during control maturation include cAMP signaling pathway molecule ADCYAP1 and growth factor PDGFB. Cluster 8 genes were downregulated during maturation but increased after FID addition; they include glucose transporter SLC2A3 and NR4A1 (also called NUR77), involved in cell transport and steroidogenesis/ovulation, respectively.

Fig. 3.

Validation of FID-sensitive gene expression patterns in cumulus cells. Expression profiles of selected genes of cumulus cells derived from COCs before in vitro maturation (Ct-0H) and after 6 h of in vitro maturation in control conditions (Ct-6H) along with FID supplementation (FID-6H) were charted with RT-qPCR. The results on y-axes are relative transcripts normalized by a normalization factor calculated through GeNORM by using the values of 2 reference genes (ACTB and GAPDH). Values are means ± SE, and statistically significant differences are presented as different letters.

Features of FID-induced Gene Expression Pathway

Next, to gain insight into interaction networks involved in FID signaling, we submitted gene lists to Ingenuity Pathway Analysis. The gene interaction network revealed that FID excited multiple pathways. Moreover, our data suggested a cross talk between these pathways (Fig. 4). Remarkably, there was downregulation of AC, which indicates a negative feedback effect due to a rise of cAMP under FID influence. Conversely we observed activation of the PKA pathway that upregulated various important transcription factors (STAT3, FOS, EGR1, and JUND), growth factors receptors (FGFR1, FGFR2), and molecules involved in steroidogenesis (NR4A1, NR4A2, STAR, NR5A2). Also, FID upregulated intracellular calcium through CCK and activated IP3 and PKC pathways. These pathways (PKA, PKC, and IP3) then interacted with the MAPK pathway to result in ERK1/2 activation. Interestingly, FID upregulated the enzymes involved in cleavage of EGF receptor-extracellular domain (ADAMTS1, ADAM17). However, downstream targets of EGF pathway, like genes of extracellular matrix formation (HAS2, PTX3), were downregulated. This signifies that cumulus expansion was temporarily inhibited by FID. Among other intracellular signaling pathways the PKB (AKT) network was shown to be inhibited.

Fig. 4.

Dynamic features of FID-induced gene expression pathway: differentially expressed transcripts between cumulus cells derived from COCs after 6 h of in vitro maturation (Ct-6H) and with FID-supplementation (FID-6H) were subjected to Ingenuity Pathway Analysis. Considering the genes implicated in cumulus physiology, gene expression pathway revealed activation of PKA, ERK1/2, and calcium pathways. The genes involved in cumulus expansion, ovulation, steroidogenesis, glucose metabolism, intracellular signaling, and oocyte competence were upregulated. Molecular relationships are based upon the Ingenuity knowledge base.

FID-induced Gene Expression Pattern Required PKA Pathway

The Ingenuity Pathway Analysis proposed modulation of several intracellular networks by FID, particularly, PKA, ERK1/2, and calcium in cumulus cells. To explore these networks in FID-supplemented COCs we did a second experiment and investigated the expression of nine genes by RT-qPCR. To better define the effect of calcium signaling, we supplemented the control culture medium with a calcium ionophore (A23187), whereas PKA implication was studied by culture of COCs in FID-supplemented culture medium with or without a PKA inhibitor (H89). Primarily, PKA mediated most of FID-specific gene expression patterns, whereas calcium seemed to be downstream of PKA since it had a partial effect (Fig. 5). To be precise, inhibition of PKA abrogated FID-specific expression patterns (whether upregulation or downregulation) of CCK, GFPT2, DUSP1, MAP2K3, HK2, ADCYAP1, and STAR. Interestingly, HAS2 and AREG were downregulated by FID, but their expressions were further reduced by PKA inhibition, suggesting a contribution of cAMP through the PKA pathway. In parallel, the calcium ionophore mimicked FID stimulation for HK2 and STAR, showing a partial role of calcium in FID-induced gene expression patterns of genes involved in glucose and lipid metabolism.

Fig. 5.

The FID-induced gene expression pattern is abrogated by inhibiting PKA in cumulus cells. Expression profiles of selected genes derived from COCs after 6 h of in vitro maturation in control conditions (Ct-6H) and FID-supplementation (FID-6H) with or without inhibitor of PKA (H89) or calcium ionophore without FID (A23187) were charted with qRT-PCR. The results on y-axes are relative transcripts normalized by a normalization factor calculated through GeNORM by using the values of 2 reference genes (ACTB and GAPDH). Values are means ± SE and statistically significant differences are presented as different letter.

Interactive Pathway of FID-induced cAMP Signaling

It has been demonstrated that PKA prompted specific gene expression patterns through regulation of ERK1/2 activation via DUSP1. Next, we explored FID-induced ERK1/2 phosphorylation and DUSP1 expression involving PKA and calcium pathways in cumulus cells. We analyzed cumulus cells obtained from COCs under the following conditions: 1) control conditions, 2) FID supplementation, 3) FID supplementation with PKA inhibitor, and 4) calcium ionophore supplementation without FID (Fig. 5). This analysis revealed that FID inhibited ERK1/2 phosphorylation in cumulus cells, at least in part, through PKA pathway (Fig. 6A). Indeed, PKA inhibition partially rescued the decrease in ERK1/2 phosphorylation by FID. Calcium ionophore did not significantly affect ERK1/2 phosphorylation. Furthermore, we studied PKA's effect on DUSP1. Intriguingly, DUSP1 transcripts were regulated by FID through PKA (Fig. 5); however, the effects of different conditions on DUSP1 protein were not significantly different (Fig. 6B). This implies that FID-induced PKA could play a significant role in decreasing ERK1/2 phosphorylation possibly through DUSP1 phosphatase.

Fig. 6.

Analysis of ERK1/2 phosphorylation and DUSP1 protein level in cumulus cells showed that FID treatment inhibited ERK1/2 phosphorylation through PKA-dependent way which remained unchanged by calcium ionophore and no treatment affected DUSP1. Cumulus cells derived from COCs after 6 h of in vitro maturation in control conditions (Ct-6H) and FID supplementation (FID-6H) with or without inhibitors of either PKA (H89) and calcium ionophore (A23187 without FID) were analyzed and shown as representative Western blots of ERK1/2, phosphorylated (p)-ERK1/2 and densitometric ratio of p-ERK1/2/total ERK1/2 (A), representative Western blots and densitometric values of DUSP1 protein (B). Values are means ± SE, and statistically significant differences are presented as different letter.

DISCUSSION

cAMP has been implicated in the acquisition of oocyte developmental competence in multiple species; however, molecular coordination of this process has not been elucidated (1, 2, 30). In this study we analyzed signaling networks prompted by cAMP in cumulus cells through addition of forskolin (AC stimulator), IBMX (nonspecific PDE inhibitor), and dipyridamole [inhibitor of PDE8, which is the most active PDE family of bovine cumulus cells (43)] (treatment here called FID) for the first 6 h of IVM. The current study demonstrates that cAMP modulated gene expression dynamics of cumulus cells during IVM and activated PKA, which regulated genes involved in cell metabolism, steroidogenesis, intracellular signaling, and cumulus expansion through the regulation of ERK1/2 phosphorylation and the coordination of calcium signaling.

Oocytes' developmental competence may be affected during oocyte maturation via cumulus-oocyte communication (49). This affirms the significant roles of cumulus cells, which they accomplish through de novo mRNA synthesis during oocyte maturation (31, 48). Here, we were particularly interested in cAMP-evoked signaling pathways. Results show that cAMP-modifying factors (FID) significantly and specifically affected gene expression of cumulus cells during early maturation. In particular, differentially expressed genes include those involved in cell metabolism (SLC2A1, SLC2A3, HK2, GFPT2), cell communication (GJA1), signal transduction (MAP2K3, AKT2), steroidogenesis (STAR, NR4A1), cell survival (TP53), and extracellular matrix formation (HAS2, PTX3). This is in concomitance with previous reports that suggest similar gene dynamics during maturation and after cAMP stimulation through different excitatory mechanisms (3, 22, 36). Notably, we observed upregulation of cell metabolism genes, which has been suggested as major mechanisms involved in oocyte competence acquisition after cAMP stimulation (53).

To validate microarray data, we considered different gene clusters and important functions in cumulus physiology and chose 18 candidate genes. Our findings confirmed previous reports with different cAMP agonists (2, 3). In particular, a gene involved in modulation of the cAMP-PKA pathway like AC-activating protein 1 (ADCYAP1), also called PACAP, was of interest (4). Downregulation of this gene after FID supplementation suggested the existence of a negative feedback loop between level of intracellular cAMP and adenylate activation. Furthermore, cAMP-dependent downstream targets of PKA pathway genes include those involved in transcriptional activity (EGR1), steroidogenesis (STAR), carbohydrate uptake (SLC2A1, SLC2A3) and metabolism (HK2, GFPT2), intracellular signaling (MAP2K3, AKT2), and oocyte maturation (CDKN1, CDKN2B) (2). In parallel, cAMP increased expression of cholecystokinin (CCK), involved in modulation of intracellular calcium levels. It is well characterized that cAMP interacts with the calcium pathway and activation of CCK after FID treatment proposed a role of intracellular calcium in this process (33). Additionally, the orphan nuclear receptor involved in steroidogenesis (NR4A1) was upregulated after cAMP stimulation. In fact, NR4A1 is an LH-response gene (32) that has been shown to be regulated by FSH during IVM. This reinforced the idea that cAMP mimicked, at least in part, the gonadotropin effects in cumulus cells (3). Interestingly, gonadotropin effects during oocyte maturation involved activation of EGF pathway, which is essential for cumulus expansion, oocyte maturation, and ovulation (14, 38). Here, we observed upregulation of enzymes involved in cleavage of the EGF receptor-extracellular domain (ADAMTS1, ADAM17). It has been shown that the promoter activity of ADAMTS1 increased after forskolin stimulation in a PKA-dependent manner (44). Concomitantly, in our analysis we observed downregulation of two genes (HAS2, AREG) from the EGF pathway after FID supplementation. We suppose that this could be due to either 1) an overactivation of the PKA pathway creating negative feedback at 6 h or 2) the FID treatment period of 6 h being too early to read any effect of ADAMTS1 stimulation, which had not yet trickled down to result in AREG activation. AREG is considered to be essential for transcription of EGF ligands (11). Interestingly, our findings corroborate an earlier observation that a forskolin exposure of >2 h resulted in downregulation of EGF ligands in cumulus cells (41). These data support the significance of cAMP in cumulus expansion through a cross talk with the EGF pathway at multiple points within the cell.

To identify FID-signaling pathways in cumulus cells, microarray data were analyzed through Ingenuity Pathway Analysis. Interestingly, this analysis revealed upstream regulators like forskolin, CREB1, EGF, calcium pathway, and H89. These results affirmed activation of three major pathways including PKA, ERK1/2, and calcium, confirming previous findings (3, 17). To elucidate the cAMP-induced signaling cascade, we evaluated the effects of inhibition of PKA (in FID-supplemented IVM) and substitution of FID with calcium ionophore on FID-induced gene expression patterns. The inhibition of PKA abolished FID-induced gene expression changes of nine genes evaluated in cumulus cells. This was in concordance with previous reports (5, 6). Calcium ionophore, on the other hand, could mimic the FID stimulation for only HK2 and STAR. In fact, calcium stimulation could enhance HK2 (25) and STAR (34) expression in different cell types as shown in earlier studies. The other investigated genes were not affected by calcium. Although these findings affirmed that cAMP primarily activated the PKA pathway, which appeared to manifest most of the genes induced by FID during IVM, it is possible that calcium signaling worked complementarily with PKA to acquire oocyte competence after cAMP stimulation.

Next, we analyzed the intracellular kinases that may be implicated in signal transduction of cAMP action. In fact, activation of mitogen-activated protein kinase kinase 1 (MEK1)-ERK1/2 by EGF receptor is one of the best documented among several pathways manifested during oocyte maturation (24). Moreover, it is known that some cumulus properties depend upon activation of ERK1/2 by either FSH or EGF for cumulus expansion (16). Interestingly, our results showed a downregulation of various EGF pathway genes (AREG, HAS2), which were implicated in cumulus expansion after cAMP stimulation. Although an increase in ERK1/2 phosphorylation has been suggested with forskolin (leading to indirect activation of PKA) (41), others have proposed a PKA-dependent decrease in ERK1/2 phosphorylation (21, 40). Therefore, we analyzed ERK1/2 phosphorylation in response to FID stimulation. Our results indicate that FID stimulation decreased ERK1/2 phosphorylation, which could partly be increased with PKA inhibition, suggesting a possible negative regulatory role of PKA in this process (Fig. 6). The calcium ionophore did not have a significant effect on ERK1/2 activity. Interestingly, various studies have proposed activation of DUSP1, which dephosphorylates ERK1/2 through PKA pathway. Therefore, we determined the levels of DUSP1 transcripts and protein. Our results affirm that the transcripts of DUSP1 increased after cAMP stimulation; however, protein of DUSP1 did not change significantly. Actually, the increase in DUSP1 protein required the EGF pathway activation in addition to PKA in fibroblast cells (NIH3T3) (40), and forskolin-mediated EGF pathway activation required endogenous activation of AREG in granulosa cells (46). It can be hypothesized that since there has been downregulation of AREG after cAMP stimulation, probably this factor contributed to prevention of significant DUSP1 downregulation, and hence, no difference in DUSP1 levels was observed. It is, therefore, possible to propose that cAMP-stimulated PKA, which regulated ERK1/2 activity, may involve DUSP1.

Conclusions

Our analysis proposes a new and more comprehensive pathway of cAMP signaling networks in cumulus cells (Fig. 7). Indeed, the maturation of oocyte involves complex cumulus-oocyte cross talk that culminates in special cumulus functions like cumulus expansion and steroidogenesis. These cumulus properties are acquired by characteristic gene expression patterns that can be modified by the culture conditions. In particular, addition of cAMP-modifying factors has been associated with increased COC oxygen consumption and oocyte oxidative metabolism (53), an initial increase and subsequently decrease in EGF pathway genes (41), steroidogenesis (28), and intracellular signaling (21), which are all implicated in oocyte quality. In this perspective, we propose that cAMP (FID) coordinated important cumulus functions through PKA, which possibly regulated 1) calcium signaling to activate genes of steroidogenesis (STAR) and glucose metabolism (HK2), 2) MAP2K3, which through MAPK14 (p38) signaling upregulated cumulus expansion genes, and 3) ERK1/2 phosphorylation. This network established a working model of the impact of cAMP not only on oocyte competence when COCs were cultured in vitro, but also on inhibition of oocyte nuclear maturation, cumulus cell expansion, and cAMP transduction of FSH intracellular signaling.

Fig. 7.

Proposed pathway of cAMP signaling after FID addition during in vitro maturation. Primarily, FID incited PKA, which collaborated closely with calcium and ERK1/2 pathways. PKA upregulated cholecystokinin (CCK) and putatively increased the intracellular calcium pool. Calcium was involved in the regulation of genes implicated in glucose metabolism and steroidogenesis. In parallel, PKA inhibited cumulus expansion genes through the ERK1/2 pathway with concomitant stimulation of their expression via MAPK14 (p38). (?) PKA regulated DUSP1 transcripts, suggesting its possible implication in ERK1/2 phosphorylation.

GRANTS

This study was supported by grants from NSERC-CRD and the EmbryoGENE Network of Canada (to M.A. Sirar and F. J. Ricard) and FSAA, Université Laval (to D. R. Khan).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: D.R.K. and C.G. performed experiments; D.R.K. and C.G. analyzed data; D.R.K., M.-A.S., and F.J.R. interpreted results of experiments; D.R.K. prepared figures; D.R.K. drafted manuscript; D.R.K., M.-A.S., and F.J.R. edited and revised manuscript; D.R.K., C.G., M.-A.S., and F.J.R. approved final version of manuscript; M.-A.S. and F.J.R. conception and design of research.