Nodal Promotes Functional Luteolysis via Down-Regulation of Progesterone and Prostaglandins E2 and Promotion of PGF2α Synthetic Pathways in Mare Corpus Luteum

Abstract

In the present work, we investigated the role of Nodal, an embryonic morphogen from the TGFβ superfamily in corpus luteum (CL) secretory activity using cells isolated from equine CL as a model. Expression pattern of Nodal and its receptors activin receptor A type IIB (ACVR2B), activin receptor-like kinase (Alk)-7, and Alk4, as well as the Nodal physiological role, demonstrate the involvement of this pathway in functional luteolysis. Nodal and its receptors were immune localized in small and large luteal cells and endothelial cells, except ACVR2B, which was not detected in the endothelium. Nodal mRNA in situ hybridization confirmed its transcription in steroidogenic and endothelial cells. Expression analysis of the aforementioned factors evidenced that Nodal and Alk7 proteins peaked at the mid-CL (P < .01), the time of luteolysis initiation, whereas Alk4 and ACVR2B proteins increased from mid- to late CL (P < .05). The Nodal treatment of luteal cells decreased progesterone and prostaglandin (PG) E₂ concentrations in culture media (P < .05) as well as mRNA and protein of secretory enzymes steroidogenic acute regulatory protein, cholesterol side-chain cleavage enzyme, cytosolic PGE₂ synthase, and microsomal PGE₂ synthase-1 (P < .05). Conversely, PGF_2α secretion and gene expression of PG-endoperoxidase synthase 2 and PGF2α synthase were increased after Nodal treatment (P < .05). Mid-CL cells cultured with PGF_2α had increased Nodal protein expression (P < .05) and phosphorylated mothers against decapentaplegic-3 phosphorylation (P < .05). Finally, the supportive interaction between Nodal and PGF_2α on luteolysis was shown to its greatest extent because both factors together more significantly inhibited progesterone (P < .05) and promoted PGF_2α (P < .05) synthesis than Nodal or PGF_2α alone. Our results neatly pinpoint the sites of action of the Nodal signaling pathway toward functional luteolysis in the mare.

The transient nature of the corpus luteum (CL) and complexity of mechanisms governing its function have motivated its study for the last decades (1, 2). Overall, this endocrine organ presents a remarkably fast and complex sequence of events including nonvascular and vascular proliferation (2, 3) steroidogenic activity (4), and/or apoptosis (5). However, the exact nature of luteolysis (functional vs structural) is not fully understood. Recent reports associated the presence of Nodal, a TGFβ superfamily member with cancer progression (6). In fact, factors that are important during embryogenesis, regulating events like angiogenesis, cell migration, and invasion, might also play a relevant role in such a dynamic organ as the CL. In the reproductive tract, Nodal induces apoptosis (7), regulates folliculogenesis and follicular atresia (7), and modulates lactation (8) and placentation (9). Nevertheless, to the best of our knowledge, the functional role of Nodal in the CL has not been described so far.

The ligand Nodal may control different cellular processes including cell proliferation and differentiation, apoptosis, extracellular matrix remodeling, adhesion, and mobility (10, 11). Regarding its signaling pathway, Nodal binds to a type II receptor, activin receptor A type IIB (ACVR2B), which further activates a type I receptor, either activin-like kinase (Alk) 4 or Alk7. Activated type I receptor phosphorylates cytoplasmic phosphorylated mothers against decapentaplegic (Smad)-2/Smad3, promoting their interaction with Smad4 and consequent formation of transcriptional complexes in the nucleus (12).

In the present study, we hypothesized that the Nodal signaling pathway is expressed in equine CL, modulating its secretory activity throughout the luteal phase of the ovarian cycle. Moreover, we also addressed the putative regulation of the Nodal signaling components by the major mediators of luteal fate, the LH and the prostaglandin (PG) F_2α. Thus, we investigated the following: 1) the cellular localization and both mRNA and protein expression of Nodal and receptors ACVR2B, Alk4, and Alk7 in equine CL throughout the luteal phase; 2) the in vitro effect of Nodal on progesterone (P₄), PGE₂, and PGF_2α secretion and modulation of their secretory enzymes; 3) the effect of LH, PGF_2α, and PGE₂ on the Nodal signaling pathway components; and 4) the interaction between Nodal- and PGF_2α-promoting luteolysis.

Materials and Methods

Collection of equine CL

Luteal tissue and venous blood from the jugular vein were collected postmortem at a local abattoir from randomly selected cyclic Lusitano mares, aged from 3 to 8 years old, from May until the end of August. The mares were euthanized after stunning according to the European Legislation concerning welfare aspects of animal stunning and killing methods (European Food Safety Authority, AHAW/04–027) and the Portuguese legislation (DL 98/96, Article 1°) and as approved by the Ethics Committee of the Faculty of Veterinary Medicine, University of Lisbon (Lisbon, Portugal).

Soon after euthanasia of the mares, internal genitalia were collected. The luteal structures were classified as previously described (13, 14) in early CL, mid-Cl and late CL. Next, luteal samples were isolated and placed in the following: 1) RNAlater (AM7020, Ambion, Applied Biosystems) for gene and protein expression analysis; 2) transport media, M199 (M2154; Sigma) with 0.1% BSA (735078; Roche Diagnostics GmbH), 20 μg/mL gentamicin (G1397; Sigma), and 250 μg/mL amphotericin (A2942; Sigma) for in vitro studies; or 3) in buffered formaldehyde or frozen in −80°C for immunostaining studies and in situ hybridization.

Immunohistochemistry analysis

Immunostaining studies were performed on consecutive 5-μm histological sections (15). Primary mouse monoclonal antibody against Nodal (antihuman, 1:400, ab55675; Abcam) and primary rabbit polyclonal (RP) antibodies against ACVR2B (antihuman, 1:300, ab128544; Abcam), Alk4 (antihuman, 1:400, ab64813; Abcam), and Alk7 (antihuman, 1:100, ab71539; Abcam) were used. Negative control was performed by replacing the primary antibody by either RP IgG (ab27478; Abcam) for antibodies developed in the rabbit or by mouse IgG (550878; BD Bioscience) for antibodies developed in the mouse, at the same dilution and for the same time as for the primary antibody and PBS 0.1 M (pH 7.4). Immunostaining was assessed as a characteristic brown staining, with a light microscope (BX51; Olympus), on 10 random fields at different magnifications. Microscopic fields were photographed (DP11; Olympus).

In situ hybridization

Tissue sections (12 μm thickness) were processed for RNA in situ detection using cDNA antisense oligoprobe for Nodal (5′-CAGGCTCTGGATGTAGGCATGGTTGGTCGGATG GAACTCCTCCCCCAC-3′). The oligoprobe was 5′end, 3′end, and center labeled with digoxigenin (DIG) (Genomed). Hybridization followed the protocol previously described by Malek et al (16), with slight adaptations for luteal tissue. Hybridization with a DIG-labeled oligoprobe (0.1, 5, and 10 nM) was performed for 48 hours in a humid chamber at 37°C. After hybridization, sections were treated with deoxyribonuclease I (D5025; Sigma) for 30 minutes at 37°C (D5025; Sigma). Stringency washes were performed and slides were blocked (11096176001; Sigma) for 1 hour at room temperature. Incubation with the antibody anti-DIG alkaline phosphatase (AP)-labeled (11093274910; Roche) was performed overnight at room temperature. For each CL, two sections were hybridized with the antisense probe and one section was hybridized with the sense probe.

Western blotting

Protein expression was quantified by Western blotting. Radioimmunoprecipitation assay buffer (250 μL) (R0278; Sigma) with a protease inhibitor (complete) (pH 7.4) was added to the cells for general protein analysis. For Smad3 phosphorylation (Smad3P) analysis, phosphatase inhibitor (88662; Thermo Scientific) was also added. Protein concentration was determined using bicinchoninic acid assay (BCA-1; Sigma), and 10–40 μg of protein was loaded on an acrylamide gel (161–0155; Bio-Rad Laboratories) (17). Protein levels were evaluated with the antibodies used in immunohistochemistry (Nodal and Alk4, 1:400; ACVR2B and Alk7, 1:200). Activation of Nodal signaling was confirmed with the assessment of Smad3 (1:500, rabbit polyclonal, ab73942; Abcam) and Smad3P (1:1000, rabbit polyclonal, ab51451; Abcam). With respect to steroidogenic enzymes, RP antibody was used against steroidogenic acute regulatory protein (StAR; 1:1000, ab96637; Abcam); RP against 3β-hydroxysteroid dehydrogenase (3βHSD; 1:1000, ab80363; Abcam); and goat polyclonal antibody against cholesterol side-chain cleavage enzyme (CYP11A1; 1:300, sc-18043; Santa Cruz Biotechnology). For eicosanoids, RP against PGF2α synthase (PGFS; 1:400, ab84327; Abcam), RP against PG-endoperoxidase synthase 2 (PTGS2; 1:200, sc-7951; Santa Cruz Biotechnology), RP cytosolic PGE₂ synthase (cPGES; 1:200, 160150; Cayman), RP microsomal PGES-1 (mPGES1; 1:200, ab62050; Abcam), and RP microsomal PGES-2 (mPGES2; 1:500, 160145; Cayman) were used. All primary antibodies were incubated overnight at 4°C. Normalization was done with mouse monoclonal antibody against βactin (A5441; Sigma) diluted 1:10 000. Then proteins were detected by incubating the membrane with a secondary polyclonal antirabbit AP-conjugated (1:30 000, A3812; Sigma), polyclonal antimouse AP-conjugated (dilution 1:30 000, A3562; Sigma) or polyclonal antigoat AP-conjugated (dilution 1:6000, ab6722–1; Abcam) antibodies for 2 hours at room temperature. Immune complexes were visualized using the AP visualization procedure. Blots were scanned and specific bands quantified using Kodak 1D image analysis software (Eastman Kodak). At last, band density for each of the target protein was normalized against β-actin.

RNA isolation and cDNA synthesis

Total RNA was extracted from luteal tissue or cells after incubation with TRIZOL reagent (Invitrogen) following the manufacturer's instructions. As described before, 1 μg of total RNA was reverse transcribed using Reverse Transcriptase Superscript III enzyme (reference 18080093; Invitrogen, Gibco) and oligodeoxythymidine primer (27–7858-01; GE Healthcare).

Real-time polymerase chain reaction

Real-time PCR assays were performed in a 7900 real-time PCR system (Applied Biosystems) as shown before (17). In each real-time assay, both the target gene and a housekeeping gene, β2-microglobulin (B2MG; primers are listed in Table 1), were run simultaneously, and reactions were carried out in duplicate wells on a 384-well optical reaction plate (Applied Biosystems) in 12.5 μL reaction volume (17). Real-time results were analyzed with the real-time PCR miner algorithm (18).

Table 1

Specific Primer Sequences Used for Quantitative Real-Time PCR

Gene	Accession Number	Sequence (5′–3′)	Length, bp
Nodal	XM_001503737.2	Forward: TGTGAGGGCGAGTGTCCTAAC	89
		Reverse: CGTTTCAGCAGGCTCTGGAT
ACVR2B	XM_001488736.4	Forward: CCCGCCATTAAGGATCACTG	89
		Reverse: GCATCATGGTCCCAGCACTC
Alk4	XM_001494129.4	Forward: GGGTATCCAGGCTCTGCTGT	200
		Reverse: GCAGTGGGTGTTACGCAGGT
Alk7	XM_005601500.1	Forward: GGATTGGCTGTGAAGCATGA	186
		Reverse: CCTTCGGGCTATTTCCCAGT
StAR	XM_008507165.1	Forward: AGGCTCCACCTGTGTGTTGG	90
		Reverse: GTGGGACCATGCTCAGCTCT
3βHSD	NM_001081911.1	Forward: CAGACCAGAATTGCGGGAAG	143
		Reverse: CGATGATAGAGGCGGTGTGA
CYP11A1	XM_008543945.1	Forward: ACCTGTTTCGCTTCGCCTTC	81
		Reverse: GGTCCACTATCTCGTCCAGCA
PTGS2	NM_001081775.1	Forward: GCTCTTCCTCCTGTGGCTGA	218
		Reverse: CCTTTGGTGAAAGCTGGACCT
PGES	XM_005605567.1	Forward: CACGCTGCTGGTCATCAAGA	127
		Reverse: GGTCGTCCCGGTGAAACTG
PGFS	XM_001500921.1	Forward: TGGGTCACTTTCCTTCAACCA	200
		Reverse: CTTCTCCATTGCCTCCCATGT
Nur77	XM_001504364.4	Forward: AGCACGGCCAAATTGGACTA	106
		Reverse: AGGCGGAGAGCAGGTCGTAG
B2MG	X69083	Forward: CGGGCTACTCTCCCTGACTG	92
		Reverse: TTGGCTTTCCATTCTCTGCTG

Isolation and culture of luteal cells

Luteal cell isolation was described previously (13, 14). Briefly, after washing the CL in sterile PBS 0.1 M (pH 7.4), connective tissue and blood clots were removed and luteal tissue was minced into small pieces. Cells were collected by enzymatic digestion and were then resuspended in DMEM and Ham's F-12 (D/F) medium (1:1 [vol/vol], D-8900; Sigma) containing 10% fetal bovine serum (26140–079; Gibco) and gentamicin (20 μg/mL). Cell viability, which ranged between 82% and 87% live cells, was assessed by trypan blue exclusion dye (T8154; Sigma). Dispersed luteal cells were then cultured either in 1 mL of D/F medium in 24-well culture plates (142475; Nunc) (2.0 × 10⁵/mL) or in 5 mL D/F medium in T25 culture flasks (136196; Nunc) (5.0 × 10⁶/mL), with 10% fetal bovine serum, amphotericin (250 μg/mL), and gentamicin (20 μg/mL) at 37°C with 5% CO₂. A smear of each cell suspension was prepared for cytological examination and stained with Diff-Quick (Baxter Scientific) (19), showing the following: early CL, approximately 40% of large luteal cells (LLCs) and approximately 50% of small luteal cells (SLCs); mid-CL, approximately 60% LLCs and 20% SLCs; and late CL, 45% LLCs and 35% SLCs. Other cell types like endothelial cells and fibroblasts represented approximately 10% of the total cells. Once cells were adherent in a 85%–95% confluent monolayer, they were washed with M-199 with 0.1% BSA and phenol red-free D/F medium (1:1) (11039; Gibco) supplemented with 0.1% BSA, gentamicin (20 μg/mL), and transferrin (T1428; Sigma) (5 μg/mL) was added.

After the in vitro protocol, cells were resuspended in TRIZOL/reagent (15596–026; Invitrogen) for mRNA analysis or ice-cold radioimmunoprecipitation assay buffer for Western blotting analysis. Cells and media were then stored at −80°C.

Luteal cell viability and proliferation assessment

In all in vitro studies, luteal cells were also plated in 96-well cell culture plates (Corning) and treated with the same factors, as described before (15). Briefly, cells at 1.0 × 10⁴/mL were incubated for 24 hours at 37°C in a humidified, 5% CO₂ in air atmosphere. Cell viability was determined with Cell Titer 96 Aqueous One solution cell proliferation assay (G3581; Promega), according to the manufacturer's manual, at 490 nm.

Hormone determinations

Concentrations of P₄, PGE₂, and PGF_2α in luteal conditioned media and P₄ in plasma were determined by direct enzyme immunoassay. Assay conditions followed previously described methods (17). For P₄ evaluation, antiserum was used at a final dilution of 1:100 000, as described previously (20). Horseradish peroxidase-labeled P₄ was used at a final concentration of 1:75 000. The standard curve ranged from 0.39 to 100 ng/mL and the concentration of P₄ at 50% binding (ED₅₀) was 4.1 ng/mL. The intra- and interassay coefficients of variation (CVs) were 5.5% and 8.5%, respectively. For PGE₂ concentration assessment, the standard curve ranged from 0.39 ng/mL to 100 ng/mL, and the concentration at 50% binding (ED₅₀) was 6.25 ng/mL. The intra- and interassay CV were 1.9% and 10.8%, respectively. The PGF2_α standard curve ranged from 0.016 ng/mL to 4 ng/mL, and the ED₅₀ was 0.25 ng/mL. The intra- and interassay CVs were on average 6.5% and 11.3%, respectively.

Concentrations of hormones in culture media were normalized for the number of live cells, after viability assessment. In viability assessment, the assay baseline corresponded to the control level. In all colorimetric assessments, the effect of cell culture treatment was considered after normalizing hormone concentrations by the treatment control (culture media with treatment factors but no cells incubated in similar conditions as performed for cell cultures).

Statistical analysis

The data are shown as the mean ± SEM of values obtained in separate experiments, each performed in triplicate. The statistical analysis of data from Western blotting and real time was performed using a nonparametric one-way ANOVA Kruskal-Wallis test followed by Dunn's test (GraphPad Software, version 5). The statistical analysis of data from in vitro treatments was performed using a parametric one-way ANOVA followed by a Newman-Keuls comparison test. Analysis was done independently for each hormone and luteal stage. In all experiments different replicates contributed for the final n, which represents the number of biological replicates. Significance was defined as values of P < .05.

Results

Nodal signaling pathway characterization in the CL

In this experiment, samples from early, mid- and late CL were used. Transcription of Nodal and receptors ACVR2B, Alk4, and Alk7 mRNA (n = 6/stage) was analyzed by real-time PCR (primers shown in Table 1), whereas protein expression (n = 4/stage) was assessed by Western blotting.

Additionally, the aforementioned factors were localized in the CL by immunohistochemistry (n = 4 /stage) and Nodal mRNA by in situ hybridization (n = 4 /stage).

Nodal mRNA level consistently decreased from early to late CL (P < .05, Figure 1A). Whereas ACVR2B mRNA did not change, Alk4 presented an opposite profile to Nodal, increasing from early to late CL (P < .05, Figure 1C). The mRNA of Alk7 increased from early to mid-CL and decreased slightly in late CL (P < .05, Figure 1D).

Figure 1

Relative quantification of Nodal (A), ACVR2B (B), Alk4 (C), and Alk7 (D) mRNA level by real-time PCR (n = 6) in CL tissue throughout the luteal phase. Transcription normalized with the housekeeping gene B2MG. Bars represent mean ± SEM. Different letters indicate significant differences.

Protein analysis revealed similar pattern for Nodal and Alk7, with a sharp increase from early to mid-CL, followed by a slight decrease in late CL; noteworthy, the expression in late CL was higher than in early CL (Figure 2, A and D, P < .05). The type II receptor ACVR2B showed the lowest expression in mid-CL (Figure 2B, P < .05), and protein expression of Alk4 decreased from early to mid-CL and peaked in the late CL (Figure 2C, P < .05).

Figure 2

Protein expression of Nodal (A), ACVR2B (B), Alk4 (C), and Alk7 (D) throughout the luteal phase. Upper panels depict representative Western blotting (n = 4). Data normalized against β-actin density values. Bars represent mean ± SEM. Different letters indicate significant differences.

Immunohistochemistry and in situ hybridization demonstrated the presence of the ligand Nodal and receptors ACVR2B, Alk4, and Alk7 in luteal tissue. The in situ hybridization confirmed the localization of Nodal mRNA in steroidogenic (Figure 3A) and endothelial cells (Figure 3B). Regarding immunohistochemistry, Nodal presented a diffuse and homogeneous cytoplasmic staining in SLCs, LLCs, and luteal endothelial cells (LECs), with no nuclear staining (Figure 3C). The receptor ACVR2B showed a peripheral staining exclusively in the cytoplasm of LLCs (Figure 3D), with no staining either in SLCs or in LECs. Considering Alk4, a strong central staining was visible in both cytoplasm and nucleus of LLCs, SLCs, and LECs (Figure 3E), in association with a less intense marginal staining. Alk7 depicted a homogeneous and dense cytoplasm staining in LLCs, SLC, and LECs but no nuclear staining (Figure 3F). No staining was seen in negative controls for either RP (Figure 3G) or mouse polyclonal IgG (Figure 3H) in immunohistochemistry and the sense oligoprobe for in situ (Figure 3F).

Figure 3

Representative images of in situ hybridization of Nodal mRNA in mid-CL (A) and late CL (B) and immunostaining for the presence of Nodal in mid-CL (C), ACVR2B in late CL (D), Alk4 in mid-CL (E), Alk7 in early CL (F). Negative control with substitution of primary antibody by rabbit IgG in early CL (G) and mouse IgG in early CL (H) in immunohistochemistry or substitution of the antisense oligoprobe by the sense oligoprobe (I) for in situ hybridization. Black arrow indicates LLCs, blue arrow indicates SLCs, and yellow arrow indicates LECs. Because all factors stained equally throughout the oestrous cycle, pictures from each luteal phase were randomly assigned.

Nodal effect on luteal cells secretory activity and viability

The physiological role of Nodal in CL function was investigated. After a brief stabilization period, cells from early CL, mid-CL, and late CL (n = 6 CL/stage) were incubated in 24-well plates for 24 hours with the following: 1) no exogenous factor (negative control); 2) Nodal at different doses (0.1, 1, 10, and 50 ng/mL); or 3) equine LH (10 ng/mL, positive control). After incubation, cell-conditioned media were frozen at −80°C for further quantifications of P₄, PGE₂, and PGF_2α. The impact of Nodal on cell viability and proliferation was run in parallel cultures.

Treatment of isolated cells with all Nodal doses showed a consistent decrease in P₄ concentration in culture media from mid-CL (Figure 4B, P < .05) as well as for the doses of 0.1 and 1 ng/mL in media from late CL (Figure 4C, P < .05). A similar inhibitory effect was seen in PGE₂ output during the late CL, for which all doses reduced its secretion compared with the control (Figure 4C, P < .01). Lastly, Nodal at 0.1 and 1 ng/mL increased PGF_2α secretion in mid-CL (Figure 4B, P < .05), whereas in late CL, a similar result was found only for 0.1 ng/mL (Figure 4C, P < .05). The treatment of 1 ng/mL evidenced a trend to raise PGF_2α (Figure 4C, P = .06). The positive control (LH 10 ng/mL) confirmed the reliability of our culture system, with the expected increase in P₄ production in early (Figure 4A, P < .01) and mid-CL (Figure 4B, P < .01) and PGE₂ in mid-CL (Figure 4B, P < .05). Regarding cell viability, no effects were seen after Nodal treatment (Supplemental Figure 1).

Figure 4

Luteal cells in vitro production of P₄, PGE₂, and PGF_2α after 24 hours of treatment with no exogenous factor (control) or increasing doses of Nodal (N) (0.1–50 ng/mL) or LH (10 ng/mL) in early CL (A), mid-CL (B), and late CL (C) (n = 6/stage). Bars represent mean ± SEM. Asterisks indicate significant differences. +, P = .06; *, P < .05; **, P < .01; ***, P < .001.

Activation of the Nodal signaling pathway in luteal cells and modulation of P₄, PGE₂, and PGF_2α synthetic enzymes

To confirm the activation of Nodal signaling pathway in luteal cells, mid-CL cells (n = 3) were cultured in T25 flasks and treated with the following: 1) no exogenous factor (negative control); or 2) Nodal at different doses (0.1, 1, 10 ng/mL). After protein extraction, Smad3P and Smad3 expression was assessed by Western blotting at different time points (30 min, 60 min, 3 h, and 6 h). Subsequently, the protocol was repeated in T25 flasks (n = 5) and the treatments as follows: 1) no exogenous factor (negative control); 2) Nodal (0.1 ng/mL, based on significance level); or 3) LH (10 ng/mL, positive control). After 8 hours, cells were scraped and mRNA transcription of StAR, CYP11A1, 3βHSD, PTGS2, PGEFS, and mPGES1 was determined by real-time PCR (primers in Table 1). After 24 hours, protein concentrations of StAR, CYP11A1, 3βHSD, PTGS2, cPGES, mPGES1, mPGES2, and PGFS were quantified by Western blotting (Table 2).

Table 2

Antibody Table

Peptide/Protein Target	Antigen Sequence (if Known)	Name of Antibody	Manufacturer, Catalog Number, and/or Name of Individual Providing the Antibody	Species Raised (Monoclonal or Polyclonal)	Dilution Used
Nodal		Anti-Nodal	Abcam, ab55675	Mouse	400 times
ACVR2B		Anti-ACVR2B	Abcam, ab128544	Rabbit	200 times
Alk4		Anti-Alk4	Abcam, ab64813	Rabbit	400 times
Alk7		Anti-Alk7	Abcam, ab71539	Rabbit	200 times
Smad3		Anti-Smad3	Abcam, ab73942	Rabbit	500 times
Smad3P		Anti-Smad3P	Abcam, ab51451	Rabbit	1000 times
RP IgG		RP IgG	Abcam, ab27478	Rabbit	100 times
StAR		Anti-StAR	Abcam, ab96637	Rabbit	1000 times
3βHSD		Anti-3βHSD	Abcam, ab80363	Rabbit	1000 times
CYP11A1		Anti-CYP11A1	Santa Cruz Biotechnology, sc-18043	Goat	300 times
PTGS2		Anti-PTGS2	Santa Cruz Biotechnology, sc-7951	Rabbit	200 times
PGFS		Anti-PGFS	Abcam, ab84327	Rabbit	400 times
cPGE2 synthase		Anti-cPGES	Cayman Chemicals, 160150	Rabbit	200 times
mPGES-1		Anti-mPGES1	Abcam, ab62050	Rabbit	200 times
mPGES-2		Anti-mPGES2	Cayman Chemicals, 160145	Rabbit	500 times
β-Actin		Anti-β-actin	Sigma, A5441	Mouse	10 000 times
Antirabbit IgG (whole molecule)		Antirabbit AP-conjugated a ntibody	Sigma, A3812	Rabbit	30 000 times
Antimouse IgG (whole molecule)		Antimouse AP-conjugated antibody	Sigma, A3562	Mouse	30 000 times
Antigoat IgG (whole molecule)		Goat antirabbit IgG AP-conjugated antibody	Abcam, ab6722–1	Goat	6000 times

The concentration of Smad3P protein clearly increased between 30 and 60 minutes after cell treatment with Nodal (Figure 5A), confirming the activation of the Nodal signaling pathway in luteal cells. Smad3P protein was still elevated until 3 hours after treatment but decreased after 6 hours (Figure 5A). Levels of unphosphorylated Smad3 were constant throughout the assay (Figure 5A).

Figure 5

Nodal signaling pathway activation in mid-CL cells and modulation of P₄, PGE₂, and PGF_2α synthesis. Time-course characterization of Smad3P (A), after Nodal treatment (0.1–10 ng/mL) at 30 minutes, 60 minutes, 3 hours, and 6 hours (Cont, no treatment). Representative diagram and panels of Smad3P, Smad3, and β-actin Western blots. Effect of Nodal (0.1 ng/mL) on mRNA transcription by real-time PCR of the following: StAR (B), CYP11A1 (C), 3β-HSD (D), PTGS2 (E), mPGES1 (F), and PGFS (G). Transcription normalized with the housekeeping gene B2MG (n = 6). Effect of Nodal (0.1 ng/mL) on protein level of the following: StAR (H), CYP11A1 (I), 3β-HSD (J), PTGS2 (K), cPGES (L), mPGES1 (M), mPGES2 (N), and PGFS (O). Data were normalized against β-actin density values. In panel A, Smad3P and Smad3 were normalized against β-actin density values; treatments were compared with the control for each time point. Upper panels depict representative Western blotting (n = 4). Bars represent mean ± SEM. Asterisks indicate significant differences. *, P < .05; **, P < .01. Cont, negative control; LH, 10 ng/mL, positive control.

In the second part of the experiment, the inhibitory effect of Nodal on mRNA transcription of StAR (Figure 5B, P < .05), CYP11A1 (Figure 5C, P < .05), and mPGES1 (Figure 5F, P < .05) was confirmed. For 3βHSD mRNA, inhibition by Nodal presented a trend (Figure 5D, P = .06). The mRNA of PTGS2 and PGFS was amplified by Nodal (Figure 5, E and G, P < .05). Treatment with LH (positive control) consistently augmented the mRNA of all studied enzymes except PGFS (Figure 5, B–G, P < .05).

Results from protein quantification were in line with mRNA profile. Nodal treatment of mid-CL cells inhibited StAR (Figure 5H, P < .05) and CYP11A1 (Figure 5I, P < .05) proteins, but no changes were seen for 3βHSD (Figure 5J). Nodal decreased the expression of cPGES (Figure 5L, P < .05) and mPGES1 (Figure 5M, P < .01) but did not affect mPGES2 (Figure 5N). Synthesis of PGE₂ was modulated by cPGES and mPGES1. Conversely, the expression of PTGS2 and PGFS protein was promoted (Figure 5, K and O, P < .05). The positive control LH increased StAR (Figure 5H, P < .05), CYP11A1 (Figure 5I, P < .05), 3βHSD (Figure 5J, P < .01), and mPGES1 (Figure 5M, P < .05).

Modulation of the Nodal signaling by LH, PGF_2α, and PGE₂

In this experiment the effects of LH, PGF_2α and PGE₂ on Nodal, ACVR2B, Alk4, and Alk7 were investigated. Cells (n = 5 CL) were cultured in T25 flasks and treated with the following: 1) no exogenous factor (negative control); 2) PGF_2α (10⁻⁷ M); 3) LH (10 ng/mL); or 4) PGE₂ (10⁻⁷ M). After 8 hours, the mRNA of Nodal, ACVR2B, Alk4, Alk7, StAR, and 3βHSD was analyzed. The same protocol was repeated for 24 hours for protein quantification. Finally, phosphorylation of Smad 3 was assessed after 3 hours of treatment with PGF_2α (10⁻⁷ M).

No effect on Nodal mRNA was seen (Figure 6A), but ACVR2B decreased after PGF_2α treatment (Figure 6B, P < .05). The mRNA of Alk4 and Alk7 receptors was increased by LH (Figure 6, C and D, P < .05). The controls StAR and 3βHSD were reduced by PGF_2α (Figure 6, E and F, P < .05) and increased by LH (Figure 6, E and F, P < .01), showing a reliable response from our model. The eicosanoid PGE₂ also promoted StAR transcription (Figure 6E, P < .05). Regarding the protein, Nodal expression was augmented by PGF_2α treatment (Figure 6G, P < .05), whereas ACVR2B was diminished (Figure 6H, P < .05). Alk4 was significantly decreased by PGE₂ (Figure 6I, P < .05) and Alk7 was raised by LH (Figure 6J, P < .05). Finally, treatment of luteal cells with PGF_2α promoted the phosphorylation of Smad 3, compared with the control (Figure 6K, P < .05).

Figure 6

Role of LH (10 ng/mL), PGF_2α (10⁻⁷ M), and PGE₂ (10⁻⁷ M) in Nodal signaling pathway modulation. Assessment of mRNA level was determined by real-time PCR for the following: Nodal (A), ACVR2B (B), Alk4 (C), Alk7 (D), and the controls StAR (E) and 3β-HSD (F). Transcription was normalized with the housekeeping gene B2MG (n = 6). Effect of LH and PGs was determined on protein level of the following: Nodal (G), ACVR2B (H), Alk4 (I), and Alk7 (J). Data were normalized against β-actin density values. Upper panels depict representative Western blotting (n = 4). Confirmation of Nodal signaling activation by PGF_2α on the Smad3P level (K). Both Smad3P and Smad3 were normalized against β-actin density values. Panels depict representative Western blotting (n = 3). Asterisks indicate significant differences. *, P < .05; **, P < .01. Cont, control.

Supportive interaction between Nodal and PGF_2α during luteolysis

In the last experiment, the interaction between Nodal and the main luteolysin PGF_2α was tested. After plating mid-CL cells (n = 5 CL) in 24-well plates, treatments with no exogenous factor (negative control), Nodal (0.1 ng/mL), Nodal+LH (10 ng/mL), Nodal+PGF_2α (10⁻⁷ M), or PGF_2α were performed. After 8 hours of incubation, mRNA for StAR, CYP11A1, 3βHSD, nerve growth factor IB (Nur77), and PGFS were quantified, whereas the concentrations of P₄ and PGF_2α in culture media after 24 hours of incubation were assessed by an enzyme immunoassay (primers in Table 1).

Supportive interactions between Nodal and PGF_2α on functional luteolysis were demonstrated after analyzing P₄ and PGF_2α secretory activities. Considering factors affecting synthesis of P₄, treatments with Nodal, Nodal+PGF_2α, and PGF_2α decreased StAR mRNA (Figure 7A, P < .05), whereas LH in association with Nodal reversed the inhibitory effect described. Noteworthy, StAR mRNA was suppressed in a higher magnitude by the association Nodal+PGF_2α, compared with PGF_2α alone (Figure 7A, P < .05). The same holds true for CYP11A1 (Figure 7B, P < .05). A trend was seen between the treatments Nodal+PGF_2α and PGF_2α (Figure 7B, P = .06), with higher effectiveness on CYP11A1 inhibition for the association. Furthermore, Nodal+LH did not entirely reverse Nodal inhibitory effect alone on CYP11A1 mRNA (Figure 7B, P = .06). Regarding 3βHSD, its mRNA transcription was decreased by the association Nodal+PGF_2α and PGF_2α (Figure 7C, P < .05). Finally, P₄ in culture media reflected mRNA data because Nodal, Nodal+PGF_2α, and PGF_2α decreased its output (Figure 7D, P < .001). Although no statistical significances were found among treatments, the association of Nodal+PGF_2α produced a greater inhibition of P₄ than either Nodal or PGF_2α alone (Figure 7D, P < .05). Taking into consideration PGF_2α, the mRNA transcription of Nur77 and PGFS was amplified by Nodal, Nodal+PGF_2α, and PGF_2α (Figure 7, E and F, P < .05). The effect of Nodal+PGF2α on increasing PGFS and Nur77 mRNA levels was greater than with Nodal and PGF2α alone (Figure 7, E and F, P < .05). Concentration of PGF_2α in the medium was augmented, compared with the control (Figure 7G, P < .01), but no statistical significance was found between Nodal+PGF_2α and PGF_2α.

Figure 7

Interactions during functional luteolysis. Effects of Nodal (0.1 ng/mL), Nodal+LH (10 ng/mL), Nodal+PGF_2α, and PGF2_α (10⁻⁷ M) were determined on the following: StAR (A), CYP11A1 (B), 3β-HSD (C) mRNA, and P₄ concentration in the media (D); or Nur77 (E), PGFS (F) mRNA, and PGF_2α concentration in the media (G) (n = 4/stage). Bars represent mean ± SEM. Asterisks indicate significant differences. +, P = .06; *, P < .05; **, P < .01; ***, P < .001. Cont, control.

Discussion

The present report provides a comprehensive description of the Nodal signaling pathway component expression in equine CL throughout the luteal phase. Modulatory effects on P₄ and PGs indicate the participation of Nodal in functional luteolysis. Nodal was shown to suppress P₄ and PGE₂ and promote PGF_2α synthesis. Furthermore, PGF_2α activated Nodal expression, revealing an intraluteal mechanism for luteolytic signal amplification. Finally, supportive interactions between both factors on functional luteolysis clearly substantiated the role of Nodal in CL regression.

Former studies linked the pathway to ovarian cancer (21, 22), but no reference was made to CL function. Nodal is an embryonic morphogen from the TGFβ superfamily, and its role in embryo proximal-distal axis development (23, 24) and left–right asymmetry (12, 25, 26) has been well documented. In the last years. Nodal has been implicated in multiple biological processes in adults. However, controversial findings were produced, particularly in the reproductive tract. The Nodal/left right determination factor pathway was shown to be involved in endometrial regulation of extracellular matrix and connective tissue during the estrous cycle and pregnancy (27). As demonstrated by Cornet et al (28) (2002), it regulated the production of collagen, connective tissue growth factor, and several matrix metalloproteinases. Furthermore, Nodal was expressed in the endothelium of the bovine oviduct (29). The signaling pathway was expressed cyclically during mammary gland remodeling, with Nodal up-regulation being associated with proliferative alveolar expansion (30) and down-regulation with apoptosis (8, 31). The aforementioned reports represent a growing body of evidence concerning the role of Nodal in reproductive tract. Consistently, our work describes how Nodal mediates CL function.

We colocalized the ligand Nodal and both type I receptors (Alk4 and Alk7) in SLCs, LLCs, and LECs, and in situ hybridization results confirmed the transcription of Nodal in steroidogenic and endothelial cells. However, ACVR2B stained exclusively on SLCs and LLCs, with no detection in LECs. Jones et al (32) (2002) described the localization of all activin receptors, including ACVR2B, in endothelial cells from human endometrium. Nevertheless, a study from Quail et al(33) (2012) on breast cancer vascularization demonstrated that Nodal inhibited angiogenesis exclusively in the presence of other cells except the endothelium. The authors suggested the differential expression of Nodal receptor subtypes in breast cancer endothelium. Therefore, the actions of Nodal on luteal angiogenesis appear to involve luteal-endothelial or immune-endothelial interactions because this cytokine is unlikely to signal via ACVR2B in LECs. Alternatively, a noncanonical pathway activation of Nodal might be considered for LECs, although it has not yet been characterized in reproductive organs.

Nodal and the expression pattern of its receptors invite consideration of their involvement in luteal regression. Despite the opposite profiles between mRNA and protein for Nodal, ACVR2B, and Alk4, the availability of these proteins during mid- and late CL supports their involvement in luteolysis. Dissimilarities between the patterns of mRNA and protein during CL evolution were seen previously (15) and might be justified with the posttranscriptional modifications or specific mechanisms of self-regulated expression, in which Nodal modulates its own expression via feedback loops (23, 34). Ultimately, all components of the Nodal signaling pathway were increased during later stages of the luteal phase.

Initial results supported a postembryonic role for Nodal in the CL function, so we examined whether this cytokine was able to modulate luteal secretory activity. Using an in vitro cell culture system previously validated (13), the effect of Nodal on luteal cell secretion and viability was investigated. Interestingly, no reports have demonstrated the molecular effect of Nodal at a secretory level in the adult reproductive tract so far. The assessment of P₄ and PGs concentrations in culture media produced very consistent results from the physiological point of view, which motivated studies on their synthetic pathways. Surprisingly, Nodal did not affect luteal cell viability, and this led us to focus exclusively on functional luteolysis.

First, the assessment of the Smad 3P profile in luteal cell lysate confirmed the activation of the Nodal signaling pathway and its involvement in downstream-observed effects. Subsequently, we verified that Nodal inhibits P₄ by down-regulation of StAR and CYP11A1 genes transcription. A trend (P = .06) was observed for 3βHSD mRNA, but no changes were obtained at protein level. Correlating these findings with Nodal expression profile during the cycle, and its sharp protein increase in Mid-CL, one may infer the importance of Nodal translation for intraluteal P₄ suppression. Indeed, the P₄ luteoprotective role was previously shown to prevent apoptosis and structural luteolysis (35, 36), and maintenance of P₄ concentrations suppressed the initiation of luteolysis (37). Nodal decreased PGE₂ via down-regulation of cPGES and mPGES1. The luteotrophic role of PGE₂ in the mare is not as well established as in other species, such as human (38), cow (39), or dog (40). Nevertheless, our data are in line with previous reports because the in vitro treatment of equine CL cells with PGE₂ promoted mRNA transcription of StAR (Figure 6E).

Another interesting finding was the modulation of luteal PGE₂ by cPGES and mPGES1 isoforms. Waclawik et al (41) (2008) attributed the regulation of PGE₂ in porcine CL exclusively to mPGES1. However, cPGES was highly expressed in fully developed corpora lutea in mice (42). According to our interpretation, the equine CL expressed both cPGES and mPGES1 and their concentrations were decreased by Nodal during the initiation of luteolysis to abolish any putative PGE₂ luteotrophic effect. The confirmation of Nodal luteolytic role was given by PGFS and PTGS2 mRNA and protein, which were up-regulated after Nodal treatment. Interestingly, PTGS2 expression was closely associated with PGFS in the equine ovary, as shown for ruminant and porcine CL that had acquired the capacity for luteal regression (43). Thus, concerning its coordinated expression during the CL life cycle and further actions on physiologic responses, Nodal emerges as a supportive luteolysin, decreasing luteoprotective factors P₄ and PGE₂ and amplifying the luteolytic PGF2_2α.

Next, we were interested in the interactions between Nodal and the main regulators of luteal fate, LH and PGF2_2α. In another set of in vitro assays, we observed that LH in association with Nodal reversed the inhibitory effect of Nodal on P₄ synthetic pathway as well as the effectiveness of Nodal and PGF2_2α association on P₄ inhibition. The treatment of Nodal+PGF2_2α more significantly inhibited StAR transcription and P₄ secretion than Nodal and PGF_2α alone. Considering StAR and CYP11A1 transcription, a trend was seen when Nodal acted in concert with PGF2_2α, compared with PGF2_2α alone, what supports an integrative action of both factors toward the establishment of luteolysis. Similar interaction was obtained for the luteolysin PGF2_2α. The association Nodal+PGF2_2α promoted a higher transcription of Nur77 and PGFS than any of the factors alone. As previously reported, activation of the luteolytic cascade by PGF_2α involves the transcription of Nur77 (44–46). Indeed, we observed an increase in Nur77 mRNA in mid- and late CL, when compared with early CL, and the treatment of mid-CL cells with PGF_2α promoted Nur77 transcription (A.G., D.S., G.F.-D., unpublished data). Once again, the LH association with Nodal reversed its supportive effect on Nur77 and PGFS mRNA as well as the product PGF2_2α. In conclusion, we clearly made evident a stronger luteolytic effect when Nodal acted together with PGF_2α than PGF_2α alone.

The remaining experiment addressed the effect of LH, PGE₂, and PGF_2α on the expression of the Nodal signaling pathway components. Despite no significant increase in Nodal mRNA, PGF_2α promoted Nodal protein expression. We confirmed the interaction between Nodal and PGF_2α in the autoamplification of lutelytic signal, in that each protein promoted the expression of the other. Another important finding was the promotion of Smad3P activity by PGF_2α (Figure 6K), which definitely confirms the canonical activation of the pathway (47). However, less conclusive results were obtained for the receptor components of the pathway. Indeed, PGF_2α decreased ACVR2B and LH increased Alk7 proteins, which represents less consensual findings considering our initial hypothesis. Evidently the regulation of the expression of Nodal receptors is a complex mechanism, not exclusively dependent on LH, PGE₂, or PGF_2α. The present study did not consider additional coregulators of the pathway such as teratocarcinoma-derived growth factor 1 (Cripto-1) or other factors that might underlie the present results. Previous studies demonstrated that Cripto-1 protein binds not only to the Nodal signaling pathway elements (48) but also is able to influence their activity via activation of the ras/raf/MAPK and phosphatidylinositol 3-kinase/AKT pathways (31, 49). Activin receptors are subject to a very intricate mechanism of activity and regulation and are able to interact with multiple ligands of the TGFβ superfamily to modulate different functional roles (50). Further studies considering the integration of relevant factors such as Cripto-1 or left right determination factor 1, the last one an inhibitor of Nodal (47), are needed to better understand these interactions. Beyond the scope of this report was the assessment of the phosphorylation of Nodal receptors.

To conclude, the Nodal signaling pathway is expressed in equine CL, and its activation supports functional luteolysis. Nodal protein translation in the mid-CL down-regulates the activity of P₄ and PGE₂ secretory enzymes. In parallel, the up-regulation of PGF_2α jeopardizes functional activity and triggers luteolysis. An autoamplification system between Nodal and PGF_2α was also evident, in that the coexpression of both factors enhanced the luteolytic signal. Further studies are being pursued to clarify the involvement of Nodal in other aspects of luteolysis.

Acknowledgments

This work was supported by Maestro, National Polish Research Center (Grant 2011/02/A/NZ5/00338); and Program “Iuwentus Plus” (Grant IP2014011373) from the Polish Ministry of Science and Higher Education. A.G. was supported by the Portuguese Foundation for Science and Technology Fellowship SFRH/BPD/79001/2011. A.G., D.S., and G.F.-D. were supported by the bilateral scientific cooperation between Portuguese Foundation for Science and Technology and the Polish Ministry of Science (Poland), “Transforming Growth Factor-β Superfamily in the Physiopathology of Mare Endometrium and Ovary.”

Disclosure Summary: The authors have nothing to disclose.

Abbreviations

 
• ACVR2B

  activin receptor type IIB

 
• Alk

  activin receptor-like kinase

 
• AP

  alkaline phosphatase

 
• CL

  corpus luteum

 
• cPGES

  cytosolic PGE₂ synthase

 
• CV

  coefficient of variation

 
• CYP11A1

  cholesterol side-chain cleavage enzyme

 
• D/F

  DMEM and Ham's F-12

 
• DIG

  digoxigenin

 
• 3βHSD

  3β-hydroxysteroid dehydrogenase

 
• LEC

  luteal endothelial cell

 
• LLC

  large luteal cell

 
• mPGES1

  microsomal PGES-1

 
• mPGES2

  microsomal PGES-2

 
• P₄

  progesterone

 
• PG

  prostaglandin

 
• PGFS

  PGF2α synthase

 
• PTGS2

  PG-endoperoxidase synthase 2

 
• RP

  rabbit polyclonal

 
• SLC

  small luteal cell

 
• Smad

  phosphorylated mothers against decapentaplegic

 
• Smad3P

  Smad3 phosphorylation

 
• StAR

  steroidogenic acute regulatory protein.