Reproductive deterioration in high-producing dairy cows has caused substantial economic loss to the dairy cattle industry (Lucy, Reference Lucy2007). Two of the key factors decreasing fertility of dairy cow are low fertilization rate and early embryonic mortality (Royal et al. Reference Royal, Mann and Flint2000; Sheldon et al. Reference Sheldon, Wathes and Dobson2006), two abnormalities that occur during pre-implantation embryo development. Although genetic factors affect reproductive performance (Shook, Reference Shook2006), identification of specific genes has been a challenge, probably due to the low accuracy of fertility data collected in the field and to the low heritability of these traits (VanRaden et al. Reference VanRaden, Sanders, Tooker, Miller, Norman, Kuhn and Wiggans2004). Thus, understanding the genetic regulation of bovine pre-implantation development and identifying associated biomarkers are becoming progressively essential to improve dairy cattle fertility.
To investigate the relationship between genetic factors and pre-implantation embryo development in cattle, an in-vitro fertilization (IVF) system was created in our laboratory. This system enables us to identify genetic factors affecting fertilization and early embryonic development at both the maternal and embryo levels. At the maternal level, our IVF system has been used to uncover associations of cows’ genotypes with fertilization success and blastocyst rate of embryos produced from these cows (Khatib et al. Reference Khatib, Maltecca, Monson, Schutzkus, Wang and Rutledge2008a, Reference Khatib, Monson, Schutzkus, Kohl, Rosa and Rutledgeb, Reference Khatib, Huang, Mikheil, Schutzkus and Monson2009a, Reference Khatib, Huang, Wang, Tran, Bindrim, Schutzkus, Monson and Yandellb; Driver et al. Reference Driver, Huang, Gajic, Monson, Rosa and Khatib2009; Wang et al. Reference Wang, Schutzkus, Huang, Rosa and Khatib2009; Huang et al. Reference Huang, Kirkpatrick, Rosa and Khatib2010a). At the embryo level, differential gene expression between normal blastocysts and degenerate embryos has been investigated by applying RNA-Seq (Huang & Khatib, Reference Huang and Khatib2010), microarray expression (Huang et al. Reference Huang, Yandell and Khatib2010b) and candidate gene and pathway analyses (Laporta et al. Reference Laporta, Driver and Khatib2011; Zhang et al. Reference Zhang, Penagaricano, Driver, Chen and Khatib2011).
The transforming growth factor-β (TGF-β) signalling pathway has long been acknowledged for signal transduction and other intracellular activities, such as cell division, differentiation, migration, apoptosis and transformation (Santibanez et al. Reference Santibanez, Quintanilla and Bernabeu2011). In addition, several studies have suggested involvement of the TGF-β signalling cascade and its components in pre-implantation embryo development as well as ovarian function (Shimasaki et al. Reference Shimasaki, Moore, Otsuka and Erickson2004; Zhang et al. Reference Zhang, Yang and Wu2007). For example, Assou et al. (Reference Assou, Boumela, Haouzi, Anahory, Dechaud, De Vos and Hamamah2010) recently reported that known members of TGF-β pathway showed dynamic changes in gene expression profiles during the three early stages of human embryonic development, including oocytes, day-3 embryos, and human embryonic stem cells on day 7. Also, results from the mouse showed that multiple bone morphogenetic proteins (BMPs) and SMAD6 in the TGF-β pathway were expressed in a stage-specific pattern and were developmentally regulated in oocytes and pre-implantation embryos (Wang et al. Reference Wang, Piotrowska, Ciemerych, Milenkovic, Scott, Davis and Zernicka-Goetz2004). BMPs and GDF9, also members of TGF-β, have been reported to be crucial regulators of folliculogenesis in mouse models (Trombly et al. Reference Trombly, Woodruff and Mayo2009; Otsuka et al. Reference Otsuka, McTavish and Shimasaki2011).
It is worth mentioning that most reported data on TGF-β pathway genes have been generated in the mouse model, and there is little information on other species such as cattle. Interestingly, in a recent transcriptomic study of the bovine IVF system, significance analysis of microarrays (SAM) identified 67 transcripts differentially expressed between blastocysts and degenerative embryos, of which 33 showed at least a two-fold difference (Huang et al. Reference Huang, Yandell and Khatib2010b). To further identify signalling pathways associated with embryonic development, Gene Set Enrichment Analysis (GSEA) and gene ontology (GO) enrichment analysis were carried out. The TGF-β pathway was found to be up-regulated in degenerate embryos compared with blastocysts using microarrays (Huang et al. Reference Huang, Yandell and Khatib2010b). However, several genes from this pathway were not included in the microarray analysis. Given that the TGF-β signalling pathway has a crucial role in ovarian and embryonic development, we hypothesized that correct balance of expression of genes from this pathway is needed for proper pre-implantation development of IVF embryos. Also, there is limited information on the extent of contributions of maternal and embryonic genomes to the survival of the developing embryo. Therefore, the present study aimed to investigate the expression profiles of TGF-β genes in degenerate embryos and normal blastocysts and to evaluate the association between maternal genotypes of these genes and fertility traits such as fertilization success and blastocyst rate.
Materials and Methods
Two experiments were performed to assess involvement of the TGF-β pathway genes in embryo development and fertility traits. In the first experiment, we compared the expression profiles of the TGF-β pathway genes in two populations of embryos differing in their morphology and development. The most significant differentially expressed genes between the embryo groups were tested in the second experiment for genomic association with fertility traits in a cow population.
Expression profiles of TGF-β pathway genes in cattle embryos
Embryo production and morphological classification
Oocytes were aspirated from ovaries obtained from a local slaughterhouse and underwent maturation until they were combined with frozen-thawed semen. The procedures of in-vitro fertilization and subsequent embryo culture were as described in Khatib et al. (Reference Khatib, Maltecca, Monson, Schutzkus, Wang and Rutledge2008a, Reference Khatib, Monson, Schutzkus, Kohl, Rosa and Rutledgeb). Embryos that showed signs of cellular compaction by day 5 of culture (morula stage) were further cultured until day 8. Embryos failing to show signs of compaction were excluded from further analysis. Embryos that exhibited a clear inner cell mass and a fluid-filled cavity (blastocoele) on day 8 were classified as blastocysts, and those with abnormal blastocyst formation and morula-phenotype were classified as degenerates. Embryos randomly collected from each morphological group (n=20) were pooled and preserved in RNA later (Ambion, Austin, TX). Three sets (blastocysts and degenerates) of embryo pools were used in the RNA expression analysis, in which two sets of biological replicate pools were produced from one sire, and one set of embryos was produced from a second sire.
Real-time RT-PCR quantification
Total RNA was extracted from pools of embryos using RNaqueous Micro (Ambion) and quality controlled by a RNA6000 PicoChip (Agilent Technologies, Santa Clara CA, USA). The mRNA amplified by MessageAmp II (Ambion) was converted to cDNA using the iScript cDNA synthesis kit (Bio-Rad Laboratories, Hercules, CA, USA). Dilution of cDNA was used as template in qRT-PCR with the iQ SYBR Green Supermix kit (Bio-Rad Laboratories). The reference gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was selected as an endogenous control as described by Huang et al. (Reference Huang, Yandell and Khatib2010b). In brief, the housekeeping genes ribosomal protein large P0 (RPLPO), actin, beta (ACTB) and GAPDH were tested for expression stability across embryo samples examined for quantitative gene expression. The expression of GAPDH embryo was markedly invariable across embryo samples. Primers for 25 genes from the TGF-β pathway (Table S1) [Table S1–S3 are available online as Supplementary Material on Cambridge Journals Online (http://journals.cambridge.org)] were designed to amplify fragments that span at least one intron to avoid genomic DNA contamination using the Beacon Designer software (Premier Biosoft International, Palo Alto, CA, USA). Each sample was tested in quadruplicate. The relative quantification of gene expression was performed using 2−ΔΔCt method (Livak & Schmittgen, Reference Livak and Schmittgen2001).
Gene selection
In a previous study, the GO analysis revealed up-regulation of the TGF-β pathway in degenerate embryos compared with blastocysts using microarray expression analysis (Huang et al. Reference Huang, Yandell and Khatib2010b). The dataset is available at GEO with the accession number GSE24936. The genes INHBA, DCN, ID3, THBS2, TGFBR2, PPP2R1A, THBS4, BMP2, BMP4, PITX2, SMAD2, RPS6KB2, ACVR1, BMPR1A, BMPR1B and TGFBR1 were found to be present in the microarray. Although not represented in the array, the genes BMP3, SMAD1, SMAD6, LEFTY, ACVR1B, and ACVR1C were selected for qRT-PCR because they are part of the TGF-β pathway. The genes GDF9, BMP15, and BMPER were selected because they are members of the TGF-β family.
Association study of TGF-β pathway genes with fertilization and blastocyst rates
To further evaluate effects of the differentially expressed genes on fertilization rate and blastocyst rate, maternal genotypes were tested for association with these traits. The genes BMP4, THBS2, and ID3 were selected because they showed the most significant fold differences in expression between blastocysts and degenerates.
Phenotypic data
Oocytes were collected from a total of 496 ovaries obtained from 496 Holstein cows and fertilized by semen from 12 Holstein bulls. For each cow, fertilization rate was defined as the number of cleaved embryos at day 2 post-fertilization divided by the total number of fertilized oocytes collected from one ovary. Blastocyst rate was defined as the number of embryos that reached the blastocyst stage (day 8) out of the total number of cultured embryos. A total of 7865 fertilizations were performed and a total of 5270 embryos were produced to generate fertilization and blastocyst rate data.
Polymorphism identification and genotyping
DNA was isolated from ovaries (n=496) using standard phenol/chloroform protocols. DNA concentrations were measured using an Ultraspec 2100 spectrophotometer (Amersham Biosciences, Piscataway, NJ, USA). For single nucleotide polymorphism (SNP) identification, one DNA pool was constructed from 20 random ovary samples containing equal amounts of DNA from each (25 ng/μl). The DNA pool was amplified by 12 sets of primers designed from the exons of the three candidate genes (Table S1). Amplification was performed in a 25-μl reaction volume, which included 25 ng genomic DNA, 50 ng each primer, 200 μm each dNTP, 2·5 μl 10×PCR buffer (Promega, Madison, WI, USA), and 0·5 μ Taq DNA polymerase (Promega). The PCR products of the pooled DNA samples were sequenced using BigDye terminator (Applied Biosystems, Foster City, CA, USA) and SNPs were identified by visually inspecting sequence traces. For genotyping, three sets of primers (Table S1) were designed in ID3, BMP4 and THBS2. The PCR products of SNP rs109818980 (ID3), rs109778173 (BMP4), and rs110619673 (THBS2) were digested with the restriction enzymes FauI, HinfI and TaiI, respectively and electrophoresed on a 2·0% agarose gel. The patterns of genotypes obtained for each restriction enzyme are presented in Table S2.
Statistical analysis
For expression analysis, normalized gene expression values (ΔCt) were analysed using a general linear model as follows:
where y ijk is the normalized gene expression value (ΔCt) of sample k from pool j fertilized by bull i; μ represents an overall mean for the trait considered; p j is the random effect of pool j; b i is the fixed effect of bull i; embryo ijk is the fixed effect of the embryo type; and e ijk represents the residual. Association between the normalized gene expression and the type of embryo was tested using a likelihood ratio test by comparing model (1) to a reduced model without the embryo effect. The mean and the range of the fold change for each gene were calculated as 2−ΔΔCt using the estimated ΔΔCt value ±se. The analysis was performed by the LME4 package in R software.
The association of SNP genotypes with fertilization or blastocyst rate was tested using the following mixed linear model,
where y ijk represents the fertilization rate or embryo survival rate of oocyte k from ovary i fertilized by bull j; μ represents an overall mean for the trait considered; o i is the random effect of the ith ovary from which oocytes were harvested; b j represents the random effect of the sire used in the fertilization; SNP ijk represents the fixed effect of the genotype for the SNP considered; and e ijk represents the residuals. Ovary and bull variables were assumed to be uncorrelated. Association between fertilization rate or embryo survival rate and SNP genotype was analysed by the MIXED procedure of SAS (9.0).
Results
Two separate and complementary experiments were done in this study to investigate the roles of TGF-β genes in fertility traits in cattle. In the first experiment, we assessed and compared expression profiles of these genes in degenerate embryos that do not make a complete transition to blastocysts vs. embryos that reach the blastocyst stage in a timely manner. In the second experiment, we tested the effects of the dams’ genotypes on the fertilization and blastocyst rates of their embryos.
Differential expression of the TGF-β pathway genes
A total of 25 genes from the TGF-β pathway were evaluated for their expression patterns in blastocysts and degenerate embryos; 14 genes were expressed and 11 genes were not detectable in embryos examined (Fig. 1). Figure 2 shows differential expression of the 14 expressed genes in embryos. Expression of all examined genes, except for ACVR1, was higher in degenerate embryos than in blastocysts. The mRNA expression levels of the following genes were significantly increased in degenerate embryos: DNA-binding protein inhibitor 3 (ID3), thrombospondin-2 (THBS2), bone morphogenetic protein 4 (BMP4), growth differentiation factor-9 (GDF9), BMP binding endothelial regulator (BMPER), and decorin (DCN) (Table S3). Moreover, lesser fold differences in expression, but still statistically significant, were observed for SMAD family member 2 (SMAD2), thrombospondin-4 (THBS4), protein phosphatase 2, regulatory subunit A, alpha (PPP 2R1A) and ribosomal protein S6 kinase, 70-kDa polypeptide 2 (RPS6KB2). However, the effect of embryo group was not statistically significant for the expression levels of BMP receptor, type 1A (BMPR1A), paired-like homeodomain 2 (PITX2), inhibin, beta A (INHBA) and activin A receptor, type 1 (ACVR1) with fold change ranging from 1·16 to 2·64. Expression of BMP2, BMP3, BMP15, BMPR1B, SMAD1, SMAD6, TGF-βR2, TGF-βR1, ACVR1B, ACVR1C, and LEFTY2 was not detected in cattle embryos.
Fig. 1. Genes of TGF-β pathway investigated in the current study. 1Genes that showed differential expression between the embryo groups. 2Genes expressed in both blastocysts and degenerate embryos, but did not show significant differential expression. 3Genes not detectable in pre-implantation embryos using qRT-PCR.
Fig. 2. Differential expression of genes in the TGF-β pathway is represented as mean±maximum and minimum fold changes in degenerate embryos vs. blastocysts using qRT-PCR. Up-regulation in degenerative embryos or blastocysts is shown by bars above or below the x-axis, respectively. The qRT-PCR was performed in three sets of biological replicates of blastocysts and degenerate embryos. *P⩽0·05; **P⩽0·01; ***P⩽0·001.
Association of differentially expressed genes with fertilization rate and blastocyst rate
Genes with the most significant differential expression between embryo types (ID3, GDF9, BMP4, and THBS2) were further investigated for association analysis of cows’ genotype with fertilization and blastocyst rates. Using the pooled DNA sequencing method in the ovary/cow population, SNPs were identified in ID3, BMP4, and THBS2. No SNPs were detected in GDF9. The SNPs, rs109818980, rs109778173 and rs110619673, are located in the 3′-untranslated region (3′-UTR) of ID3, the coding region (CDS) of BMP4, and the 3′-UTR of THBS2, respectively. SNPs were in Hardy–Weinberg equilibrium (Table 1). Estimates of the three genotypic classes in each SNP for blastocyst and fertilization rate and relevant P values are given in Table 1. Analysis of SNP rs109818980 in ID3 revealed a significant association with fertilization rate (P=0·029). Oocytes from genotypes TT ovaries had 5·2 and 5·3% lower fertilization rates than those from TC and CC ovaries, respectively (Table 1). Blastocyst rate was significantly associated with SNP rs109778173 of BMP4 (P=0·006), whereas the association with fertilization rate was not statistically significant (P=0·095). Embryos produced from genotype TT cows showed 10·5 and 16·1% higher blastocyst rates than GG and GT cows, respectively (Table 1). For SNP rs110619673 of THBS2, no significant associations were found with the examined traits (Table 1).
Table 1. P-values of the Hardy-Weinberg equilibrium (HWE) test, estimate of blastocyst rate (±se), and estimate of fertilization rate (±se) for the ovary SNP genotypes in ID3, BMP4, and THBS2
† In parenthesis is the number of cows genotyped
Discussion
In the current study, we explored involvement of the TGF-β pathway in development of pre-implantation bovine embryos at both the embryonic and maternal levels. At the embryo level, we found that some genes of the TGF-β pathway were up-regulated in the growth-arrested embryos compared with normal blastocysts. At the maternal level, two genes, found differentially-expressed in embryos, showed significant association with fertilization and blastocyst rates.
Little is known about the expression pattern of the TGF-β genes and their functions in the developing bovine embryo. Interestingly, the 14 genes that were expressed in pre-implantation embryos represent all components of the signalling cascade of TGF-β pathway including ligands (BMP4, GDF9, and INHBA); receptors (BMPR1A and ACVR1); SMAD proteins (SMAD2); upstream regulators (THBS2, THBS4, DCN); and downstream regulators (ID3, BMPER, PPP 2R1A, RPS6KB2, PITX2). The signalling process of this pathway necessitates coordination of gene regulation among the different members of the pathway. For example, the activation of latent TGF-β requires binding of the thrombospondin-1 (THBS1) to the TGF-β precursor complex (Murphy-Ullrich & Poczatek, Reference Murphy-Ullrich and Poczatek2000). Protein phosphatase 2 (PPP2) and ribosomal protein S6 kinase (RPS6K) are key regulators implicated in the phosphorylation of receptor and SMADs in the TGF-β signalling cascade (Zolnierowicz, Reference Zolnierowicz2000; Fenton & Gout, Reference Fenton and Gout2010). ID proteins are direct targets of BMP and TGF-β signalling, which serve as essential mediators in biological responses in the downstream pathway (Miyazono & Miyazawa, Reference Miyazono and Miyazawa2002). This cell distribution and coordination of gene expression among the TGF-β genes testifies to the significance of this pathway in embryo development.
Furthermore, expression of many members of the TGF-β superfamily in the endometrium suggests a pivotal role for these genes in differentiation of the endometrium and the implantation process (Jones et al. Reference Jones, Stoikos, Findlay and Salamonsen2006). Consequently, expression of TGF-β genes in pre-implantation bovine embryos suggests an important role for these genes in the embryo-uterus connection. Indeed, it has been reported that blastocysts express TGF-β proteins that induce apoptosis of endometrial epithelial cells during implantation (Jones et al. Reference Jones, Stoikos, Findlay and Salamonsen2006). In addition, Chow et al. reported the presence of mRNAs and proteins of TGF-β receptors in the oviduct and uterus from day 1 to day 4 of pregnancy, which in turn suggests important roles of TGF-β signalling in the interaction between pre-implantation embryos and the reproductive tract (Chow et al. Reference Chow, Lee, Chan and Yeung2001). Collectively, expression of TGF-β signalling genes in the bovine embryo suggests an important role for the TGF-β pathway in the pre-implantation stage of bovine embryo.
In a previous study, we reported a significant differential expression of the TGFβR3 gene, a member of the TGF-β signalling pathway, in pre-implantation embryos (Huang et al. Reference Huang, Yandell and Khatib2010b). To further investigate the role in blastocyst formation of the TGF-β genes, we examined their expression profiles in normally developed blastocysts compared with growth-arrested embryos produced from the same parents and cultured at the same laboratory conditions. Interestingly, of the 14 expressed genes, 10 showed statistically significant expression differences between the embryo types, of which all were found to be up-regulated in the degenerate embryos. Although it is unknown at this point whether the differential expression is causative to the morphological degeneration or a result of this degeneration, there is accumulating evidence that balanced gene expression is likely to be essential for early embryogenesis (Rodriguez-Zas et al. Reference Rodriguez-Zas, Schellander and Lewin2008).
Members of BMP signalling subfamily, such as BMP4, BMPER, BMPR1A and ID3 were found to be highly expressed in degenerate embryos compared with blastocysts (Fig. 2). These results are consistent with earlier studies that reported their function in maintaining pluripotency in the inner cell mass of bovine blastocysts (Pant & Keefer, Reference Pant and Keefer2009). Of particular interest are the findings of Koide et al. (Reference Koide, Kiyota, Tonganunt, Pinkaew, Liu, Kato, Hutadilok-Towatana, Phongdara and Fujise2009) who demonstrated that over-activity of BMP4 signalling led to excessive apoptosis in early mammalian embryo development. Also, La Rosa et al. (Reference La Rosa, Camargo, Pereira, Fernandez-Martin, Paz and Salamone2011) reported that supplementation of BMP4 to culture medium of IVF embryos decreased blastocyst production, and concluded that a balanced BMP signalling activity is required for proper pre-implantation development of cattle embryos.
The ID proteins function as key regulators of development by stimulating and maintaining proliferation and preventing premature differentiation (Yokota & Mori, Reference Yokota and Mori2002). Given that ID expression can be regulated by other members of the TGF-β pathway such as BMPs (Hogg et al. Reference Hogg, Etherington, Young, McNeilly and Duncan2010), the up-regulation of ID3 in degenerate embryos (Fig. 2) provides strong evidence of the role of these genes in early embryonic development. BMPER is a BMP-binding endothelial regulator and has been reported to modulate BMP4 signalling in endothelial cell differentiation and angiogenesis (Moser et al. Reference Moser, Binder, Wu, Aitsebaomo, Ren, Bode, Bautch, Conlon and Patterson2003; Heinke et al. Reference Heinke, Wehofsits, Zhou, Zoeller, Baar, Helbing, Laib, Augustin, Bode, Patterson and Moser2008). Taken together, we conclude that altered expression of the BMP signalling subfamily in pre-implantation bovine embryos can lead to or be used as a marker for abnormal embryonic development.
Although the roles in bovine embryo development of other differentially expressed genes are unknown, they have critical functions in the TGF-β signalling. For example, SMAD2 belongs to the SMAD family of proteins, which are transducers of TGF-β signalling from the cell surface to the nucleus and transcription factors mediating expression of target genes in the TGF-β cascade (Heldin et al. Reference Heldin, Miyazono and ten Dijke1997; Massague et al. Reference Massague, Seoane and Wotton2005). SMAD proteins are required for pluripotency maintenance of the inner cell mass in mouse blastocysts (James et al. Reference James, Levine, Besser and Hemmati-Brivanlou2005). Thus, considering the altered expression of these regulators in pre-implantation embryos observed in this study and their roles in the activation and signalling of TGF-β pathway components, we propose that they have significant functions in regulating proper development of bovine embryos.
To explore the impact of the TGF-β signalling pathway on early IVF embryo development and hence fertility traits, four genes that showed the most significant expression differences between embryo groups were tested for SNP association with fertilization and blastocyst rates. The association analysis was done using an IVF experimental system that has been recently developed to identify genes affecting fertilization and embryo development in cattle (Khatib et al. Reference Khatib, Maltecca, Monson, Schutzkus, Wang and Rutledge2008a, Reference Khatib, Monson, Schutzkus, Kohl, Rosa and Rutledgeb, Reference Khatib, Huang, Mikheil, Schutzkus and Monson2009a, Reference Khatib, Huang, Wang, Tran, Bindrim, Schutzkus, Monson and Yandellb; Driver et al. Reference Driver, Huang, Gajic, Monson, Rosa and Khatib2009; Wang et al. Reference Wang, Schutzkus, Huang, Rosa and Khatib2009; Huang et al. Reference Huang, Kirkpatrick, Rosa and Khatib2010a; Laporta et al. Reference Laporta, Driver and Khatib2011). Allelic variants of THBS2 did not show any significant association with fertility traits in our IVF system.
A significant association was observed between ID3 maternal genotypes and fertilization rate. This result is consistent with a previous genome-wide association study, in which a SNP associated with fertilization rate was located within 50 kb distance of ID3 (Huang et al. Reference Huang, Kirkpatrick, Rosa and Khatib2010a). Although the molecular regulation of fertilization success is not fully understood, maternal genome activity and oocyte quality appear to have critical roles in embryogenesis (Stitzel & Seydoux, Reference Stitzel and Seydoux2007; Marteil et al. Reference Marteil, Richard-Parpaillon and Kubiak2009). Recently, Hogg et al. (Reference Hogg, Etherington, Young, McNeilly and Duncan2010) observed that four ID isoforms (ID1-4) were expressed across ovine ovarian follicle development and possibly regulated by TGF-β signalling via SMADs. Consequently, the authors suggested mechanistic roles of the ID proteins in mammalian oocyte development (Hogg et al. Reference Hogg, Etherington, Young, McNeilly and Duncan2010). Furthermore, ID proteins are key regulators for many cellular processes, such as cell proliferation, differentiation and cell cycle progression, which in turn are required for oocyte maturation, oocyte-to-embryo transition, and embryogenesis (Norton, Reference Norton2000; Stitzel & Seydoux, Reference Stitzel and Seydoux2007). Together, we propose that ID3 affects fertilization rate through maternal genome effects that control oocyte quality and oocyte-to-embryo transition.
Maternal genotypes of BMP4 were found to be significantly associated with blastocyst rate, suggesting its role in controlling intrinsic oocyte competence and development up to the blastocyst stage. Indeed, it has been acknowledged that blastocyst yield can be affected by intrinsic oocyte quality (Rizos et al. Reference Rizos, Ward, Duffy, Boland and Lonergan2002), and the involvement of BMP4 in ovarian function has been extensively reported (Shimasaki et al. Reference Shimasaki, Zachow, Li, Kim, Iemura, Ueno, Sampath, Chang and Erickson1999). The spatio-temporal expression of BMP4 signalling in follicle development has been broadly observed across different species including human, rat, bovine, swine and zebrafish (Nilsson & Skinner, Reference Nilsson and Skinner2003; Shimizu et al. Reference Shimizu, Yokoo, Miyake, Sasada and Sato2004; Fatehi et al. Reference Fatehi, van den Hurk, Colenbrander, Daemen, van Tol, Monteiro, Roelen and Bevers2005; Li & Ge, Reference Li and Ge2011; Tanwar & McFarlane, Reference Tanwar and McFarlane2011). Functional studies have also shown that BMP4 suppresses bovine granulose cell apoptosis and promotes follicle survival and development in rats (Nilsson & Skinner, Reference Nilsson and Skinner2003; Kayamori et al. Reference Kayamori, Kosaka, Miyamoto and Shimizu2009). Collectively, the differential expression of BMP4 in embryos and the significant association of maternal genotypes with blastocyst rate indicate that this gene could regulate pre-implantation embryo development, not only through the embryo proper, but also through the maternal genome.
In summary, the findings of this study indicate that a proper gene expression level of the TGF-β genes is required for normal IVF blastocyst development and that TGF-β signalling is likely to affect pre-implantation development of IVF embryos. There is limited information on the contributions of the genomes of dams and embryos to the development and survival of pre-implantation embryos. Results of the present study indicate that, for some genes, both embryonic and maternal genomes are required to ensure proper development. Although the mechanism by which the TGF-β signalling pathway regulates early embryonic development remains unknown, potential genetic markers such as ID3 and BMP4 have been identified and can be used to improve reproductive performance of cattle.
The study was supported by USDA Hatch grant WIS-142-PRJ17PH from the University of Wisconsin-Madison. The authors thank Francisco Peñagaricano for the statistical analysis, Jenifer Cruickshank for critical review of the manuscript, and Ashley Driver, Ricky Monson and John Parrish for embryo production and culture.
