Role of bone morphogenetic protein signaling in bovine early embryonic development and stage specific embryotropic actions of follistatin

Abstract

Characterization of the molecular factors regulating early embryonic development and their functional mechanisms is critical for understanding the causes of early pregnancy loss in monotocous species (cattle, human). We previously characterized a stage specific functional role of follistatin, a TGF-beta superfamily binding protein, in promoting early embryonic development in cattle. The mechanism by which follistatin mediates this embryotropic effect is not precisely known as follistatin actions in cattle embryos are independent of its classically known activin inhibition activity. Apart from activin, follistatin is known to bind and modulate the activity of the bone morphogenetic proteins (BMPs), which signal through SMAD1/5 pathway and regulate several aspects of early embryogenesis in other mammalian species. Present study was designed to characterize the activity and functional requirement of BMP signaling during bovine early embryonic development and to investigate if follistatin involves BMP signaling for its stage specific embryotropic actions. Immunostaining and western blot analysis demonstrated that SMAD1/5 signaling is activated after embryonic genome activation in bovine embryos. However, days 1–3 follistatin treatment reduced the abundance of phosphorylated SMAD1/5 in cultured embryos. Inhibition of active SMAD1/5 signaling (8–16 cell to blastocyst) using pharmacological inhibitors and/or lentiviral-mediated inhibitory SMAD6 overexpression showed that SMAD1/5 signaling is required for blastocyst production, first cell lineage determination as well as mRNA and protein regulation of TE (CDX2) cell markers. SMAD1/5 signaling was also found to be essential for embryotropic actions of follistatin during days 4–7 but not days 1–3 of embryo development suggesting a role for follistatin in regulation of SMAD1/5 signaling in bovine embryos.

BMP signaling during peri-/post-compaction embryo development is required for blastocyst development and follistatin embryotropic actions in bovine.

Introduction

Poor oocyte quality severely restricts the viability of full-term pregnancy in cattle and humans largely due to failure in fertilization and early embryonic loss [1–3]. Over the last two decades, a number of studies have been carried out to understand the molecular mechanisms of poor oocyte quality and found that transcripts/proteins accumulated during oogenesis determine an oocytes competence to become fertilized, cleave, undergo blastocyst formation, and establish a viable full-term pregnancy in mammals [4–6]. Characterization of these oocyte factors and their functional mechanisms have been the subject of intense study to understand the etiology of unexplained early pregnancy loss and design better therapeutic treatment for infertility in livestock animals and human [7–11]. Previously, we characterized follistatin (a secreted glycoprotein) as a molecular determinant of oocyte competence, which was highly abundant in good quality oocytes (adult vs pre-pubertal) and embryos (early vs late cleaving) in cattle, and siRNA knockdown at zygotic stage resulted in reduced early embryonic development [12, 13]. Moreover, exogenous supplementation of follistatin during in vitro embryo culture exhibited stage specific embryotropic effects on bovine early embryonic development [14]. Treatment with exogenous follistatin during the pre-compaction period (days 1–3, 1 cell to 8–16 cell) of embryo culture increased the number of embryos cleaving early, developing to 8–16 cell and blastocyst stages, and blastocyst total and TE cell number. We also observed positive, but less potent than days 1–3, embryotropic effect on blastocyst production and cell number (total, TE) when follistatin was supplemented during peri-/post-compaction period (days 4–7, 8–16 cell to day 7 blastocyst) [14]. Furthermore, embryotropic actions of follistatin treatment were not specific to the bovine IVF embryos, but were also observed in bovine SCNT embryos [15] and in non-human primate IVF embryos [16].

Follistatin is a TGF-beta superfamily binding protein, which was first discovered as an activin binding protein, but also has been shown to function via binding other TGF-beta superfamily proteins such as inhibin and select bone morphogenetic proteins (BMPs 4, 7, and 15), and modulate their types 1 and II serine threonine kinase receptors activity [17]. Follistatin binding inhibits the interaction of these TGF-beta ligands to their receptors, which signal through SMAD1/5 (BMPs) and SMAD2/3 (TGF-beta, activin and nodal) pathways [5, 18]. In our previous study, we demonstrated that the actions of follistatin on bovine early embryonic development are not mediated by inhibition of activin activity [13]. We also found that follistatin days 1–3 embryotropic actions were completely ablated in the absence of SMAD4, a common SMAD linked to TGF-beta superfamily signaling through SMAD2/3 (activin, TGF-beta and nodal) and SMAD1/5 (BMP) pathways, or when SMAD2/3 signaling is inhibited [19, 20]. These results suggest that SMAD2/3 signaling is required for days 1–3 follistatin effect with implications for BMP (SMAD1/5) signaling. However, whether activity of SMAD1/5 signaling is required for stage specific (days 1–3 or 4–7), follistatin actions of promoting bovine embryo development is not known.

BMP signaling is known to play an important regulatory role in early embryonic development and TE cell lineage determination in mouse and humans. Activation of BMP signaling requires BMP ligands to bind to their kinase receptors that then phosphorylate SMAD1/5 proteins, which dimerize with co-SMAD4 and move to nucleus to regulate the transcription of downstream genes. SMAD6 is a natural inhibitor in this process, which competes with SMAD4 for binding to receptor activated SMAD1/5 proteins and inhibits BMP signaling (Supplementary Figure S1). Gene expression studies in bovine embryos revealed that BMP ligands (BMPs 2, 3, 4 7, 10, and 15), receptors (ALKs, ACVR2), and their signaling proteins (SMAD1, 4, 5, 6) are dynamically regulated during early embryogenesis [21, 22]. Moreover, BMP4 supplementation with exogenous BMP4 during embryo culture decreased in vitro development of fertilized cattle embryos [23]. These findings raise the possibility that BMP signaling is active and may be functionally involved in bovine early embryonic development. Considering the evidence that multiple BMPs and BMP signaling components are expressed in bovine embryos and follistatin binding affinity to BMP signaling ligands, we hypothesized that BMP signaling is functionally required for bovine early embryonic development and linked to the stage-specific embryotropic actions of follistatin.

To test this hypothesis, we first measured the activity of BMP signaling during early stages of bovine embryos and then used two complimentary approaches to inhibit BMP signaling in bovine embryos. We also examined the stage specific (days 1–3 and 4–7) embryotropic effect of follistatin on blastocyst production, cell number (total, TE), and mRNA/protein abundance of ICM and TE cell lineage markers in the absence of BMP signaling. To further investigate if follistatin has any direct effect on BMP signaling regulation, we also analyzed SMAD1/5 phosphorylation status in embryos treated with or without follistatin.

Materials and methods

Ethic statement

Oocytes used for in vitro experiments in this study were harvested from ovaries collected at a local slaughterhouse in the state of Michigan which does not require approval of the IACUC.

All chemicals and reagents were purchased from Sigma-Aldrich (St. Louis, MO) unless otherwise stated. Investigations using experimental animals or subjects were conducted in accordance with the SSR specific guidelines and standards.

In vitro oocyte maturation, fertilization, and embryo production

Oocytes used for all the experiments in this study were obtained from ovaries collected from a local slaughterhouse. Oocyte aspiration, in vitro oocyte maturation, fertilization, and embryo production were performed as described previously [14]. Briefly, cumulus oocyte complexes (COCs) were aspirated from follicles 3–7 mm in diameter using an 18-gauge needle connected to a 50-ml conical tube by applying 50 mm Hg of negative pressure using a vacuum pump (Cook, Australia). Intact COCs with homogenous oocyte cytoplasm and at least three layers of cumulus cells were selected and washed three times in HEPES-buffered HECM (HH) medium. Washed COCs were then matured in TCM-199 supplemented with 10% FBS (Hyclone, Logan UT), 1 μg/ml estradiol-17β, 1 IU/ml FSH, and 5 IU/ml LH (Sioux Biochemical, Sioux Center, IA), 2.3 mM of sodium pyruvate, and 25 μg/ml of gentamicin sulfate. About 50 COCs/well were incubated in four-well plates (Nunc) at 38.5 °C under 5% CO₂ in air with maximum humidity for 22–24 h.

For in vitro fertilization (IVF), COCs matured for 24 h were washed three times in fertilization medium. After washing, 50 COCs were co-incubated with percoll purified motile spermatozoa at 2 × 10⁶/ml concentration in fertilization medium for 16–18 h at 38.5 °C under 5% CO₂ in air with maximum humidity. The day of fertilization was defined as day 0. After fertilization, presumptive zygotes were completely stripped of cumulus cells and associated spermatozoa by vortexing and washing three times in HH media and twice in IVC1 medium, which comprised of potassium simplex optimization medium and 0.3% BSA. Zygotes were first cultured from days 1 to 3 (pre-compaction period) in 400 μl of IVC1 medium layered with light mineral oil (50 zygotes per well). On day 3, 8–16-cell embryos were separated, washed three times, and cultured in IVC2 medium (KSOM + 0.3% BSA + 10% FBS) from days 4 to 7 (peri-/post-compaction period). The culture was carried out in an atmosphere of 5% CO₂ in air with maximum humidity at 38.5 °C.

Analysis of BMP signaling activity during early embryonic development

To examine BMP signaling activity, SMAD1/5 phosphorylation status was determined at different stages of early embryo development. After IVF, presumptive zygotes were cultured in four-well plates (40 embryos/well) and embryos were collected at 20-h post insemination (hpi) for 1-cell (C) stage, 33 hpi for 2C, 44 hpi for 4C, 72 hpi for 8–16 cell, and morula and blastocyst were collected on 5 and 7 days post insemination (dpi), respectively. Embryos from each stage of development were subjected to western blot (WB) analysis for pSMAD1/5 or BMP4 (20 embryos/stage, n = 3 replicates per protein) expression, and immunostaining analysis (n = 3 replicates per protein).

Cloning of SMAD6 into lentiviral plasmid

The strategies for SMAD6 lentiviral (LV) expression system construction are shown in Supplementary Figure S2. Bovine SMAD6 Open Reading Frame (ORF) was custom synthesized and cloned into pUC57 vector by GeneScript. Further, cloned SMAD6 ORF was amplified using following primer: upper primer, 5′GCTTCTAGAGCCACCATGTTCAGGTCCAAACG3′, with the addition of XbaI site (boldface) and kozak sequence (italics, underline) and lower primer, 5′GCGGATCCTCTGTGGTTGTTGAGAAGGATCTCG3′, with the addition of BamH1 site (boldface). The amplified PCR product with kozak sequence and intact start codon was subcloned in to the XbaI and BamH1 sites of the pCDH-EF1-MCS-T2A-copGFP LV vector purchased from System Biosciences (Mountain View, CA). During this step, SMAD6 ORF was placed downstream of constitutive promoter EF1 and directly linked to copGFP via self-cleaving T2A peptide to monitor the transgene expression in early embryo. After cloning, the LV vector was transformed into Stabl3 bacterial cells and isolated plasmid was subjected for sequencing and restriction digestion to validate the sequence integrity of the reconstructed SMAD6 LV expression vector. A HIV-based LV vector (pWPT-EF1-GFP) validated in our lab to express GFP in bovine embryo was used as scramble control for LV injection.

Lentiviral transduction and collection of transgenic embryos

Concentrated LV particle (~1 × 10e8 TU/ml) carrying SMAD6 ORF and scramble control GFP vector were prepared by Biomedical Research Core Facilities at the University of Michigan and stored at –80 °C. LV particles (100X) were injected directly into the perivitelline space of presumptive zygotes and those embryos were cultured in IVC1 media along with uninjected and scramble control injected (IC) embryos (n = 40 embryos/treatment/well). On day 3 post-insemination, 8–16-cell embryos were counted and transferred into IVC2 media. On day 7, blastocyst rate was determined based on the percentage of presumptive zygotes that reached the blastocyst stage (n = 3 replicates) and embryo were analyzed under fluorescent microscope to check the GFP expression. Embryos positive for GFP fluorescence on day 7 post insemination were collected for real-time PCR to analyze the SMAD6 and ID3 (BMP target gene) mRNA expression (n = 3 replicates) and WB analysis to measure the protein abundance of SMAD6 and phosphorylated form of SMAD1/5 (n = 3 replicates). As required, embryos were pooled from two wells of a particular treatment to obtain sufficient material for downstream mRNA and protein analysis.

Follistatin, inhibitors, and BMP4 treatments and embryo collections

A specific pharmacological inhibitor LDN193189 (LDN) (Selleckchem, # S2618) and BMP binding protein, Noggin (R&D system, #6057-NG) were used to inhibit BMP signaling during peri-/post-compaction period of early embryonic development. To evaluate the optimum dose of inhibitors for SMAD1/5 phosphorylation, 8–16-cell embryos were cultured in IVC2 media in the presence of different doses of LDN (1, 10, and 100 nM) and Noggin (1, 10, and 100 ng/ml) with diluent control (n = 30 embryos/treatment/well). After 48 h, embryos were collected for WB analysis to detect SMAD1/5 phosphorylation level (n = 3 replicates) and real-time PCR analysis for BMP regulated genes (ID1, ID2, and ID3) (n = 3 replicates).

To analyze the days 4–7 follistatin embryotropic effects in the absence of BMP signaling, 8–16-cell embryos were cultured with diluent (control), 10 ng/ml recombinant human follistatin 300 (FST) (R&D system, #669-FO-025), 100 nM LDN or FST + LDN (n = 30 embryos/treatment; n = 4 replicates). We also used another BMP signaling inhibitor Noggin (100 ng/ml) in place of LDN in similar experimental setting to validate the effects of BMP signaling inhibition on bovine early embryo development. Further, to investigate if peri-/post-compaction (days 4–7) BMP signaling inhibition has any effect on days 1–3 follistatin embryotropic actions, IVF produced presumptive zygotes were cultured in the presence or absence of 10 ng/ml (maximal stimulatory dose) of follistatin during days 1–3 of embryo culture, then in presence or absence of 100 nM LDN from days 4 to 7 (dose determined as described above) (n = 50 zygotes/treatment/well). At day 7, the percentage of embryos that developed to blastocyst were determined in all treatment groups and blastocysts were collected to examined total and TE cell number (n = 3 replicates), gene expression analysis (n = 3 replicates), and WB analysis for TE and ICM markers. Embryos were pooled from two wells of a particular treatment to obtain required number of embryos for downstream immunostaining and expression analysis. The effect of follistatin treatment on SMAD1/5 phosphorylation status was analyzed by culturing presumptive zygotes in the presence and absence of follistatin (10 ng/ml) from days 1 to 3 and then in follistatin free IVC2 medium up to day 7 post-insemination. Embryos were collected at 8–16 cell, morula and blastocyst stage of development and subjected for WB analysis for pSMAD1/5 expression (n = 4 replicates). To analyze the BMP4 supplementation effect on SMAD1/5 phosphorylation, zygotes were cultured from days 1 to 7 in the absence and presence of 100 ng/ml BMP4 and day 7 embryos were collected to analyze the SMAD1/5 phosphorylation using WB (n = 3 replicates).

Quantitative reverse transcription-PCR analysis

Total RNA was isolated from five embryos/treatment group and eluted in 15 μl nuclease-free water using RNeasy micro kit (Qiagen; # 74004) following the manufacturer instructions. Isolated RNA was then converted in to cDNA using an iScript cDNA Synthesis Kit (BioRad; #1708841). Expression of all the target genes transcripts in this study was quantified using quantitative reverse transcription-PCR (qRT-PCR) analysis as described previously with minor modifications [12]. Briefly, after completion of cDNA synthesis, reactions were diluted three times with nuclease free water qRT-PCR was performed using 2 μl of diluted cDNA per 25 μl of reaction volume containing SsoAdvanced Universal SYBR Green Supermix (Biorad; #1725270) and 5 pmol of each primer. Reaction for each sample was performed in duplicate in C-1000 thermal cycler and CFX-96 Real time system (Bio-Rad Laboratories). The relative expression for target genes was calculated using 2^−∆∆Ct method and normalized with RPS18 as an endogenous control and a calibrator. Accession numbers and primers sequences for all the target genes used in this study are listed in Supplementary Table S1.

Western blot analysis

Expressions of all the target proteins in embryos were detected by WB analysis according to our previously published protocol [14]. Briefly, 10 embryos (unless stated otherwise) lysates were separated on 4–20% Mini-Protean TGX Precast gels (Bio-Rad) and transferred to polyvinylidene fluoride membranes (Millipore; #IPVH00010). After transfer, membranes were incubated for 1 h in blocking buffer [1X TBST (Tris-buffered saline pH 7.4 with 0.1% Tween 20) with 3% BSA] and probed with primary antibodies for pSMAD1/5^{(Ser 463/465)} (Santa Cruz; #SC-12353), tSMAD1/5 (Santa Cruz; #SC-6031-R), SMAD6 (Santa Cruz; #SC-13048), CDX2 (Abcam; #ab76541), and BMP4 (Millipore; #MAB1049) at 1:2000 and ACTIN (Millipore; #MAB1501) at 1:5000 dilution in blocking buffer at 4 °C overnight. Membranes were then washed 3–4 times with 1X TBST and further incubated for 1 h with corresponding secondary antibodies HRP anti-rabbit IgG (Cell Signaling; #7074S) or anti-mouse-IgG (Thermo Scientific; #PA1–74421) at 1:5000 dilutions in blocking buffer at room temperature. The membranes were washed 3 × 4 min in 1X TBST and then immunoreactive protein bands were visualized by SuperSignal west Dura Chemiluminescent substrate (Thermo Scientific; #34076) and image were captured using myECL Imager (Thermo Scientific). ImageJ software (http://imagej.nih.gov/ij/) was used to perform quantitative densitometry analysis and target protein band intensity was normalized to ACTIN and fold change was determined with respect to controls set to 1.0.

Immunostaining

Immunofluorescent staining was performed to determine the pSMAD1/5 expression localization at different stages of bovine early embryonic development and CDX2 positive cell to calculate TE cell number using the protocol described elsewhere [24]. After fixing, permeabilization and blocking, total five embryos per treatment or developmental stage were incubated for 18–24 h with primary antibodies for pSMAD1/5^{(Ser 463/465)} (Santa Cruz, rabbit monoclonal; #SC-12353) or CDX2 (abcam, rabbit monoclonal; #ab76541) at 1:200 dilution at 4 °C. After washing three times, samples were incubated with corresponding secondary antibody (Alexa Fluor 488 goat anti-rabbit IgG, abcam; # ab150077). Embryos were washed three times and mounted on glass side with anti-fade solution containing DAPI (Vector Laboratories; #H-1200) and visualize under epifluorescent microscope. A number of total cells were determined based on DAPI, and TE cell based on CDX2 positive staining in embryos collected at day 7 from all the treatment groups. Manufacturer detail and RRID number of antibodies used in WB and immunostaining are provided in Supplementary Table S2.

Statistical analysis

All the experimental data were obtained from three or more independent biological replicates. Data were analyzed in SAS using one-way ANOVA followed by Fisher Protected Least Significant Difference test to detect differences between means. Percent data of embryo development were calculated based on the number of zygotes cultured and was arc-sine transformed prior to analysis. Gene expression and WB data are presented as fold change, and statistical analysis was performed on the ΔΔCt values to determine differences in mRNA expression. All data are presented as untransformed mean ± SEM. Variables that are statistically significant (P < 0.05) are represented by different letters on the bar graph.

Results

BMP signaling is active during bovine early embryonic development

Results demonstrated that diffuse p-SMAD1/5 immunoreactivity is present in both the cytoplasm and nucleus from the 1- to 8-cell stage. At subsequent stages, p-SMAD1/5 is localized predominantly in nucleus from the 16 cell to blastocyst stage (Figure 1A). In addition, WB analysis revealed that p-SMAD1/5 is significantly increased at 8–16-cell stage and persists through day 7 blastocyst (Figure 1B). Collectively, these results suggest that BMP signaling is highly active during the peri-/post-compaction stage (8–16 cell to day 7 blastocyst) of bovine embryonic development.

Figure 1.

BMP signaling activity during early bovine embryo development. Pre-implantation bovine embryos were collected at 1-cell (1C), 2-cell (2C), 4-cell (4C), 8–16 cell (8–16C), day 5-morula (d5-Mor), and day 7-blastocyst (d7-Blast) stages of development and analyzed for localization and abundance of phosphorylated (active) form of SMAD1/5 (pSMAD1/5) transcription factor. (A) pSMAD1/5 Immunostaining (green) counterstained with 4′,6-diamidino-2-phenylindole (DAPI) for nuclear staining (blue). Negative control (Neg) embryos were incubated in the absence of pSMAD1/5 antibody (n = 5 embryos/stage; n = 3 replicates). (B) WB analysis for pSMAD1/5 abundance at different stages of embryo development (n = 20 embryos/stage, n = 3 replicates). Representative blot and quantification of pSMAD1/5 signals from each group are shown here. WB data were normalized relative to the abundance of ACTIN as a loading control and expressed as mean ± SEM. Values with different superscripts indicate statistically significant differences at P < 0.05.

Validation of lentiviral and pharmacological approaches to inhibit BMP signaling during peri-/post-compaction stage of bovine embryo

Two complimentary approaches were used to inhibit BMP signaling in the days 4–7 pre-implantation embryo. The first approach involved LV-mediated overexpression of SMAD6. In preliminary experiments, we showed that delivery of concentrated LV particle under the zona yielded transgenic bovine embryo with high (>90%) efficiency. As shown in Supplementary Figure S3, bovine embryos injected with control LV particle displayed strong GFP fluorescence after 8-cell stage indicating that LV-mediated transgene expression in bovine embryo can be achieved only after embryonic genome activation that occurs at 8–16-cell stage in cattle [25]. This finding allowed us to use LV approach for BMP signaling inhibition during peri-/post-compaction stage (8–16 cell to day 7 blastocyst) of bovine embryo development. We observed high transcript and protein abundance of SMAD6 in day 7 blastocyst derived from embryos injected with SMAD6 LV particles compared to injected control (IC) and uninjected (UI) bovine embryos (Figure 2C and D, P < 0.05). Significantly reduced expression of ID3 (P < 0.05), BMP target gene, was observed in blastocyst derived from SMAD6 LV (SD6LV) injected embryos compared to IC demonstrating the SMAD6 inhibitory effect on BMP signaling in bovine embryo (Figure 2E. However, no statistically significant change was observed in SMAD1/5 phosphorylation status in embryos derived from SMAD6 LV injected, IC, or UI group on day 7 post-insemination (data not shown).

Figure 2.

Effects of SMAD6 overexpression on BMP signaling and blastocyst development. (A) Map of the two LV vectors used to drive the transgene expression from EF1⍺ promoter in bovine embryos. SD6LV construct contained SMAD6-T2A-copGFP fusion gene downstream of EF1⍺ promoter with an extended 2A peptide linker (T2A) inserted in-frame to separate proteins after translation. Basic control LV construct (IC) contained GFP only. (B) GFP expression in blastocyst developed from bovine embryos that were uninjected (UI), or transduced with subzonal zygote injection of LV particles carrying pWPT-GFP (IC) or EF1⍺-SD6-copGFP (SD6LV) vector at presumptive zygotic stage. Red arrows indicate embryos with complete absence or mosaic expression of transgene. (C) Quantitative real-time PCR analysis of SMAD6 mRNA expression in blastocyst derived from uninjected, IC, and SD6LV injected embryos (n = 5 embryos/treatment, n = 3 replicates). Expression was normalized with RPS18 as an endogenous control and presented as fold induction relative to basal expression level in uninjected group. (D) WB analysis of SMAD6 protein abundance in GFP sorted blastocyst from IC and SD6LV injected bovine embryos (n = 20 embryos/treatment, n = 3 replicates). SMAD6 expression was normalized to ACTIN abundance. (E) Effect of inhibitory SMAD6 overexpression on BMP signaling targeted ID3 gene expression in day 7 blastocyst derived from IC and SD6LV injected bovine embryos (n = 5 embryos/treatment, n = 3 replicates). (F) Percent of embryos that developed to blastocyst stage from UI, IC, and SD6LV injected bovine embryos (n = 3 replicates). All data values are presented as mean ± SEM. Different superscripts letters denote P < 0.05.

In a second approach, the small molecule BMP antagonist LDN193189 and Noggin, the BMP binding protein, were used to inhibit BMP signaling activity during peri-/post-compaction stage (8–16 cell to day 7 blastocyst) of embryo development. Validation of these BMP signaling inhibitors was performed by WB analysis of their ability to inhibit SMAD1/5 phosphorylation level (phospho-/total-SMAD1/5) in bovine embryos. Results demonstrated a statistically significant (P < 0.05) decrease in SMAD1/5 phosphorylation level in response to both inhibitor treatments (Supplementary Figure S4A), which not only suggest that both inhibitors are functionally active, but also validate the specificity of pSMAD1/5 and tSMAD1/5 antibodies used. To further analyze the specificity of LDN193189 inhibitor, we examined the mRNA expression of known BMP target genes (ID1, ID2, and ID3) [26] and TGF-beta target gene (CTGF) [20] in day 5 embryos after 48 h of treatment with LDN193189. Results show that LDN inhibitor effectively downregulated (P < 0.05) the transcript abundance of ID genes (Supplementary Figure S4B), but has no effect on expression of CTGF gene transcript compared to the untreated control (Supplementary Figure S4C), suggesting that LDN inhibitor is functionally active and specific to the BMP signaling.

Inhibition of peri-/post-compaction (days 4–7) BMP signaling perturbs early embryo development and block days 4–7 follistatin embryotropic actions

SMAD6 overexpression in bovine embryos demonstrated no effect on 8–16 cell and day 5 morula production but significantly (P < 0.05) reduced the blastocyst production compared to scramble injected and uninjected control embryos (Figure 2F). SMAD6 lentivirus injected embryos that reached the blastocyst stage displayed either no or mosaic expression of GFP (expressed as SMAD6 fusion protein) indicating a lack of ubiquitous and stable overexpression of SMAD6 transgene may be needed to develop to the blastocyst stage in these embryos (Figure 2B). To further investigate the role of SMAD1/5 signaling in bovine embryo development and days 4–7 follistatin-mediated embryotropic action, 8–16-cell embryos were treated with diluent control, FST days 4–7, LDN days 4–7, and LDN + FST (days 4–7) and the effects of treatments on blastocyst production, total, and TE cell number on day 7 were analyzed (n = 50 zygotes/treatments, n = 6 replicates). We found that LDN treatment during days 4–7 showed statistically significantly reduction (P < 0.05) in blastocyst production and decreased the number of total and TE cells in resulting day 7 blastocyst compared to diluent treated control (Figure 3A–C). As reported previously [14], days 4–7 follistatin treatment significantly (P < 0.05) enhanced the day 7 blastocyst production and their cell number (total, TE), but this embryotropic effect of follistatin was completely blocked in the presence of LDN supplementation (Figure 3A–C). We observed similar decrease in blastocyst production and total cell number when LDN was replaced with another type of BMP signaling inhibitor Noggin (Supplementary Figure S5A and B), which further confirms that BMP signaling is required for early embryo development and days 4–7 follistatin-mediated embryotropic actions in cattle.

Figure 3.

LDN193189 (BMP signaling inhibitor) effect on blastocyst production and total and TE cell number in the presence and absence of follistatin. In vitro produced 8–16-cell bovine embryos were cultured with diluent control (control), 10 ng/ml recombinant human follistatin (FST) (F(days 4–7)), 100 nM LDN 193189 (L(days 4–7)), and LDN + FST (L + F (days 4–7)). (A) Effect of LDN treatment on the percentage of embryos developing to the blastocyst stage upon BMP signaling inhibition in the absence and presence of follistatin (n = 30 embryos/treatment, n = 4 replicates). (B and C) Blastocyst total and trophectoderm (TE) cell number in embryos treated with or without LDN inhibitor in the presence and absence of follistatin (n = 5 embryos/replicate, n = 3 replicates). Total and TE cell number were determined by DAPI nuclear staining and CDX2 (TE cell markers) immunostaining, respectively. Data are expressed as mean ± SEM. Values with different superscripts across treatment indicate statistically significant differences (P < 0.05).

We further examined if BMP signaling inhibition effect on blastocyst TE cell number was mediated via modulation of CDX2, a lineage-specific transcription factor, in the TE. Effects of BMP signaling inhibitor LDN193189 on transcript and protein abundance of CDX2 were analyzed in similar experimental design as described above. Gene expression and WB data demonstrated that BMP signaling inhibition (LDN days 4–7) significantly (P < 0.05) reduced CDX2 mRNA and protein abundance compared to control, and blocked the days 4–7 FST induced CDX2 mRNA and protein expression (Figure 4A, P < 0.05). Expression analysis of BMP4, trophoblast regulator, at different stage of bovine embryos demonstrated its consistent presence from presumptive zygotes to blastocyst stage of embryo and increase (P < 0.05) the SMAD1/5 phosphorylation in blastocyst upon exogenous supplementation during embryo culture (Supplementary Figure S6A and B).

Figure 4.

Effects of LDN193189 mediated BMP signaling inhibition on blastocyst CDX2 mRNA and protein abundance in follistatin treated and untreated embryos. Bovine embryos at 8–16-cell stage were cultured from days 4 to 7 with diluent control (control), 10 ng/ml recombinant human follistatin (FST) (F(days 4–7)), 100 nM LDN 193189 (L(days 4–7)), and LDN + FST (L + F (days 4–7)) (n = 30 embryos/treatment/well). (A) Real-time PCR analysis of CDX2 mRNA expression in blastocyst embryos resulted from all the treatment group (n = pool of five embryos/treatment, n = 3 replicates). Expression was normalized relative to the RPS18 abundance as a housekeeping gene. (B) WB data analysis for CDX2 protein abundance in blastocyst from all the treatment group (n = pool of 10 embryos/treatment, n = 4 replicates). Data were normalized to total ACTIN abundance as endogenous loading control. All the values are presented as mean ± SEM. Different letter across the treatment indicate P < 0.05.

Follistatin days 1–3 embryotropic actions do not require peri-/post-compaction (days 4–7) BMP signaling

We performed further experiments to investigate if peri-/post-compaction BMP signaling (days 4–7) is required for days 1–3 follistatin actions in promoting blastocyst development, blastocyst (total and TE) cell number, and CDX2 mRNA and protein abundance [13]. Results showed that blastocyst development, total, and TE cell number were significantly (P < 0.05) reduced by LDN inhibition of SMAD1/5 signaling during days 4–7 of embryo culture, whereas they were increased by days 1–3 follistatin treatment compared to controls. However, follistatin pretreatment (days 1–3) totally blocked the subsequent negative effect of SMAD1/5 signaling inhibition on blastocyst development and blastocyst total and TE cell number (Figure 5A–C). Further analysis of gene expression and WB data demonstrated that days 1–3 follistatin treatment also rescues the negative effects of days 4–7 SMAD1/5 signaling inhibition on CDX2 mRNA and protein abundance (Supplementary Figure S7A and B).

Figure 5.

BMP signaling inhibitor LDN193189 (LDN) effect on days 1–3 FST stimulated blastocyst production and total and TE cell number. Presumptive zygotes were cultured from days 1 to 3 in the absence (control) and presence of 10 ng/ml recombinant human follistatin (F(days 1–3)). The 8–16-cell embryos from both groups were then isolated, washed, and cultured from days 4 to 7 in fresh media in the absence [control, F(days 1–3)] and presence [L(days 4–7), F(days 1–3) + L(days 4–7)] of 100 nM LDN treatment. (A) Percent of days 1–3 follistatin treated and untreated embryos reaching to blastocyst stage upon peri-/post-compaction (days 4–7) BMP signaling inhibition by LDN 193189 (n = 50 zygotes/treatment, n = 4 replicates). (B) Blastocyst total cell number determined by DAPI staining in all the treatment groups (n = 5 blastocyst/treatment, n = 3 replicates). (C) Trophectoderm (TE) cell number determined by CDX2 (TE cell marker) immunostaining in blastocyst resulted from all the treatment groups (n = 5 embryos/treatment, n = 3 replicates). Data are presented as mean ± SEM. Values with different superscripts across treatment indicate statistically significant differences (P < 0.05).

Follistatin days 1–3 treatment has an inhibitory effect on BMP signaling pathway

Phosphorylation of SMAD1/5 (p-SMAD1/5) was evaluated at different stages of embryo development in the presence and absence of days 1–3 follistatin treatment to further investigate whether follistatin had any regulatory effect on BMP signaling activity. Interestingly, we observed a decrease p-SMAD1/5 abundance at 8–16 cell and day 5 morula (P < 0.05) but not day 7 blastocyst stage of embryos derived from days 1 to 3 follistatin treatment (Figure 6). These results suggest that days 1–3 follistatin treatment inhibits SMAD1/5 signaling during pre-compaction period and has a post-treatment inhibitory effect up to the morula stage of embryo development.

Figure 6.

Follistatin days 1–3 treatment effect on phosphorylated (active) form of SMAD1/5 (pSMAD1/5) abundance in early bovine embryos. Presumptive zygotes were cultured from days 1 to 3 with or without 10 ng/ml recombinant human follistatin (n = 50 zygotes/treatment, n = 4 replicates). On day 3, 8–16-cell embryos were isolated, washed, and cultured in fresh media without follistatin till day 7. Embryos were collected at 8–16-cell stage, day 5 morula, and day 7 blastocyst and subjected for WB analysis with antibody specific to phosphorylated form of SMAD1/5 transcription factors (n = pool of 10 embryos/replicate, n = 4 replicates). Representative blot of pSMAD1/5 abundance at 8–16-cell stage (A), day 5 morula (B), and day 7 blastocyst (C) are shown. Expressions of pSMAD1/5 were normalized relative to the abundance of ACTIN as an endogenous control or loading control and presented as mean ± SEM. Values with different superscripts across treatments denote statistically significant differences (P < 0.05).

Discussion

BMPs, the largest subfamily of the TGF-beta superfamily, signal through binding to types I and II receptors and activating their receptor kinase activity to phosphorylate SMAD1/5 transcription factor, which complex with SMAD4 and move to the nucleus to induce BMP target gene transcription [27]. BMP signaling has been shown to play important roles in mammalian development, mainly during early post-implantation embryo development in regulating gastrulation, anterior–posterior patterning [28–32] and germ-cell differentiation [33, 34]. Several studies carried out in polyovulatory mammals have demonstrated that BMP signaling also regulates several aspects of pre-implantation embryo development including cell cleavage, extra-embryonic cell lineage determination, and trophoblast differentiation [35–39]. However, the role of BMP signaling during early embryonic development in mono-ovulatory species like cattle and human has not been thoroughly investigated. In cattle, transcriptomic analysis of early embryos has shown that ligands, receptors, and signaling components of BMP signaling pathway are dynamically regulated during oocyte maturation and early embryogenesis [21, 22, 40]. However, the presence of these transcripts does not mean that BMP signaling is active during bovine early embryonic development. Importantly, in the present study nuclear localization of pSMAD1/5 from 8–16 cell to the blastocyst stage was observed suggesting that SMAD-dependent BMP signaling becomes active at the time of embryonic genome activation (8–16 cell) and could be involved in developmental events such as embryonic genome activation, cell polarization, compaction, blastocyst formation, and first cell lineage determination in cattle. Results of WB analysis demonstrated that phosphorylated form of SMAD1/5, produced by BMP activated receptor kinase, is highly abundant during this period, which further confirms the high activity of BMP signaling during peri-/post-compaction stages of embryo development in cattle.

To investigate if active BMP signaling during the peri-/post-compaction period is important for bovine early embryo development, we used LV-mediated SMAD6 overexpression and specific pharmacological inhibitors (LDN193189 and Noggin) to inhibit BMP signaling. Embryos injected with SMAD6 lentivirus showed reduced expression of BMP signaling targeted genes ID3, but had no effect on SMAD1/5 phosphorylation level on day 7 post-insemination. Since SMAD 6 is an inhibitory SMAD and acts by disrupting the SMAD1–SMAD4 complex required for BMP signaling activity, change in SMAD1/5 phosphorylation level is not required for SMAD6-mediated BMP signaling inhbition. [41]. LDN193189 selectively target BMP type I receptor kinase to inhibit SMAD1/5 phosphorylation level, whereas Noggin blocks the ligand binding site of both (BMP types I and II) receptors [42, 43]. Results from using two independent inhibitors and LV approach provide consistent support for the requirement of peri-/post-compaction BMP signaling in blastocyst development and cell lineage determination in cattle. LDN-mediated inhibition of SMAD1/5 phosphorylation level impaired the 8–16-cell development to the blastocyst stage and decreased the number of total and TE cell in resulting day 7 blastocysts. A previous study demonstrated a similar effect of BMP signaling inhibition on mouse pre-implantation embryo development and showed that effects on reduced blastocyst cell numbers were mediated through the inhibition of BMP target genes (ID1, ID2, and ID3) known to regulate cell cleavage [37]. This mechanism could be conserved in bovine embryos as we also observed reduced expression of ID1, ID2, and ID3 genes in embryos treated with the LDN inhibitor of BMP signaling. In addition, inhibition of BMP type I receptor kinase activity has been reported to inhibit cell division in early mouse embryo, and cell growth and migration in human cancer stem cells [44]. These findings suggest that BMP signaling could play an important role in cell division during the peri-/post-compaction period of bovine embryo development.

Molecular understanding of events leading to blastocyst TE and ICM cell lineage specification could help to identify the causes of early embryonic failure in cattle. Our results demonstrated that inhibition of BMP (SMAD1/5) signaling causes a statistically significant reduction in TE cell number and CDX2 (TE cell marker) expression in day 7 blastocysts. SMAD1/5 is a transcription factor that can interact with evolutionary conserved enhancer region of CDX2 and alter its expression [45], suggesting that BMP signaling is required for CDX2 expression and development of TE cell during early embryo development in cattle. BMP4 is a well-known regulator of trophoblast cell lineage determination in mouse and human [45]. The presence of BMP4 at different stages of bovine embryonic development and its ability to induce BMP signaling in blastocysts treated with BMP4 further suggests its endogenous production from bovine embryos and the functional role of BMP signaling in bovine TE cell lineage determination. Our previous study has also shown that transcripts of BMPs 2, 3, 7, and 10 isoforms are produced by bovine embryo in stage specific manner and could regulate the early embryo development if present in the culture media [22].

We recently reported pre-compaction (1 cell to 8–16 cell, days 1–3 pi) and peri-/post-compaction (8–16 cell to day 7 blastocyst, days 4–7 pi) stage specific follistatin actions in embryo culture on promoting blastocyst production, total and TE cell numbers, and CDX2 mRNA and protein abundance in cattle [14]. What remained unclear, however, was how follistatin was facilitating these embryotropic effects, including the molecular mechanisms responsible. Follistatin is a secretory glycoprotein, which is known to bind activin and select BMPs ligands of TGF beta superfamily in extracellular milieu and hypothesized to function via modulating activin regulated SMAD2/3 pathway and/or BMPs activated SMAD1/5 signaling. Our previous studies demonstrated that days 1–3 follistatin effects were lost in the absence of SMAD4, a common SMAD required for both SMAD2/3 and SMAD1/5 signaling. Considering the role of SMAD4 in SMAD1/5 signaling in follistatin actions and presence of active BMP signaling observed in the present study, we further investigated if BMP signaling activated after embryonic genome activation (8–16 cell to day 7 blastocyst) has any functional role in follistatin days 1–3 and/or days 4–7 stage specific embryotropic effects. Pharmacological inhibition of active BMP signaling completely blocked the positive effects of days 4–7 follistatin treatment but had no effect on days 1–3 follistatin stimulatory actions on blastocyst production, TE cell number, and CDX2 expression suggesting requirement of BMP signaling for peri–post-compaction follistatin actions in bovine embryo culture.

Though the positive effects of days 1–3 follistatin treatment on embryo culture were not influenced by BMP signaling inhibition after embryonic genome activation (8–16 cell to day 7 blastocyst), this does not preclude the possibility of follistatin regulation of BMP signaling before or near the onset of embryonic genome activation at 8-cell stage of bovine embryo [25], which is day 3 of in vitro embryo culture. Therefore, we analyzed the BMP signaling activity at different stages of embryo development in the presence and absence of days 1–3 follistatin treatment and observed reduced activity of BMP signaling at 8–16 cell and day 5 morula stage embryos derived from days 1 to 3 follistatin treatment. Follistatin is a BMP antagonist and has ability to inhibit SMAD1/5 signaling activity by binding to different BMP ligands such as BMPs-2, -4, -7, and their receptors [46]. Our previous studies demonstrated that multiple BMP ligands and receptors are also expressed and dynamically regulated during early stage of bovine embryos development [22]. Moreover, BMP4 supplementation during early stage of embryo culture has shown negative effects on bovine embryo production in vitro [23]. These findings suggest that days 1–3 follistatin embryotropic effects could be mediated via binding to BMPs ligands (e.g., BMP4) and neutralizing their signaling activity and adverse effects on bovine embryo development. This could also explain the observed effect of days 1–3 follistatin-mediated negative regulation of BMP signaling in bovine embryos analyzed in the present study. However, further studies would be needed to investigate the role of individual BMPs ligands in days 1–3 follistatin-mediated embryotropic effects.

In summary, using two independent approaches, we have uncovered the activity and functional role of BMP signaling in regulation of bovine early embryonic development. Moreover, we found that active BMP signaling from 8–16 cell to day 7 blastocyst is required for peri-/post-compaction, but not pre-compaction follistatin actions in promoting bovine embryo development suggesting that follistatin use different mechanisms to cause its stage specific embryotropic effects. We have also observed a direct effect of days 1–3 follistatin treatment on BMP signaling activity at 8–16 cell and day 5 morula stages of bovine embryos, which suggests that while days 1–3 follistatin actions may not be affected by BMP signaling inhibition during days 4–7, they may involve BMP signaling to cause the observed stimulatory effects on embryo development. We described a successful overexpression of SMAD6 ORF after 8–16 cell to inhibit BMP signaling, which demonstrates the utility and relevance of our LV approach for future studies to understand the functional role of embryonic genes and signaling pathways activated after embryonic genome activation in cattle. Results from this study not only help to understand the molecular regulation of early embryonic development in cattle, but also could have a potential implication in improving Assisted reproductive technology (ART) media and designing better strategies for infertility treatment in humans.