Maternal DDB1 regulates apoptosis and lineage differentiation in porcine preimplantation embryos

Keywords: blastocyst, cellular apoptosis, DDB1, embryo, lineage differentiation, maternal-effect gene, porcine, preimplantation development.

Mammalian early embryonic development is a precise and complex biological process that includes a series of morphological (e.g. cleavage or compaction), cytological (e.g. polarisation or lineage differentiation), and molecular biological (e.g. maternal regulation or epigenetic modification) events before implantation. These events are controlled by the accurate temporal and spatial expression of genes (Hamatani et al. 2004; Wang et al. 2021). Maternal-effect genes (MEGs) are maternal origin genes that play a critical role in modulating early embryonic development such as embryonic genome, cleavage, and embryonic cell lineage establishment (Li et al. 2010; Zhang and Smith 2015). An impairment in the expression of MEGs can lead to abnormal embryo development, impairing cleavage or compaction (Wu et al. 2003; de Vries et al. 2004; Xu et al. 2015), blastocyst formation (Kim et al. 2016; Cao et al. 2019), and trophectoderm (TE) and inner cell mass (ICM) differentiation (Wu et al. 2010; Cockburn et al. 2013).

TE and ICM are the major components of a blastocyst and they are both formed in the blastocyst after the first lineage differentiation. TE is a type of squamous epithelium derived from the polar cells produced by an asymmetric division during the 8-cell to the morula stage (Saini and Yamanaka 2018). It is located on the outer edge of the embryo and mediates implantation and placentation (Maître 2017). Several MEGs play a critical role in the formation and differentiation of TE cells. For example, maternal E-cadherin-deficient embryos can have a delayed cell division (de Vries et al. 2004) and an E-cadherin null mutant embryo failed to form an intact TE cell (Larue et al. 1994). ICM takes an inner position within the embryo, which develops the fetus and the extra-embryonic yolk sac (Saini and Yamanaka 2018). During mammalian embryonic development, it is well-known that the Hippo/YAP signalling cascade plays a central role in the TE/ICM lineage specification (Chazaud and Yamanaka 2016).

Damage-specific DNA binding protein 1 (DDB1) is one of the subunits of the damaged DNA binding protein complex that is involved in the nucleotide excision repair pathway (the other component is DDB2) (Chu and Yang 2008). DDB1 is an adapter protein between Cullin 4A (CUL4A) and CUL4-associated factors (DCAFs) that acts as a component of the CUL4A-RING ubiquitin E3 ligases (CRL4) complex to target substrates for ubiquitination (Iovine et al. 2011). DDB1 is required for meiotic resumption and meiosis I progression in mice oocytes (Yu et al. 2015a). Oocyte-specific Ddb1 knockout mice demonstrated oocyte loss and female infertility due to increased apoptosis and poor responses to ovulation signals of the granulosa cells (Yu et al. 2013, 2015b). The deletion of maternal DDB1 in oocytes contributes to embryonic deaths before the 8-cell stage (Yu et al. 2013). Although Ddb1 has been reported to be a MEG in mice, with a role in oocyte meiosis maturation, whether it is a conservative MEG among species – as well as its function in preimplantation embryonic development of other species – is still unknown.

In this study, we identified DDB1 as a MEG during porcine preimplantation embryonic development. We further explored the dynamic expression pattern of DDB1 in porcine oocytes and preimplantation embryos. In addition to this novel finding, we also assessed the molecular function of DDB1 using RNA interference and RNA-seq analysis methods. Our work provides evidence supporting the critical role of DDB1 in cellular apoptosis, lineage differentiation, and the pluripotency maintenance of blastocyst formations in porcine models.

All chemicals were purchased from Sigma-Aldrich (Shanghai, China) unless specific statement.

Porcine ovaries were collected from a local slaughterhouse and transported into the lab for processing within 2 h of collection. The cumulus-oocyte complexes (COCs) with 3–6 mm diameter were used for maturation culture. Briefly, the COCs were randomly placed in four-well Petri dishes. Eighty oocytes were cultured in 400 μL IVM medium [Medium-199 supplemented with 10% fetal bovine serum, 10% porcine follicular fluid, 10 ng/mL epidermal growth factors, 10 IU/mL equine chorionic gonadotropin, 10 IU/mL human chorionic gonadotropin, 0.1 mg/mL l-Cysteine (Ding et al. 2017)] per well, and covered with mineral oil. After 42–44 h culture (38.5°C, 5% CO₂, saturated humidity), the cumulus cells were removed using 1 mg/mL hyaluronidase. The oocytes with normal morphology and first polar body (pb1) were identified as mature oocytes and used for the subsequent experiments. The subsequent experiments were replicated at least three times.

The matured oocytes PA, IVF and embryo culture were performed as previous described (Ding et al. 2017). Briefly, matured oocytes (metaphase II stage, MII) were activated with a single direct current pulse of 1.56 kV/cm for 80 μs in activation medium (0.3 M mannitol supplemented with 0.1 mM CaCl₂, 0.1 mM MgCl₂, and 0.01% polyvinyl alcohol) using cell fusion apparatus (CF-150B, BLS, Hungary). Then, the oocytes were further incubated with assisting activation medium [porcine zygote medium 3 (PZM-3 was prepared as previous described (Yoshioka et al. 2002)) supplemented with 10 μg/mL cycloheximide and 10 μg/mL Cytochalasin B] for 4 h at 38.5°C with 5% CO₂ and saturated humidity.

For the IVF, fresh Duroc boar semen was collected from a local pig breeding farm (Hefei Antai, China) and transported into the lab for processing within 2 h of collection. A fertilisation medium droplet (55 μL) was made by adding 5.5 μL of resuspended fresh boar sperm solution (1 × 10⁶ cells/mL) into 49.5 μL fertilisation medium (modified tris-buffered medium (components are listed in Supplementary Table S1) plus 2 mg/mL BSA and 2 mM caffeine). A total of 15 MII oocytes were co-incubated with sperm in the fertilisation droplet for 6 h in the incubator (38.5°C, 5% CO₂, saturated humidity). After washing out the surrounding sperm, the presumptive zygotes were processed into embryo culture. A pool of 15 zygotes were cultured in one 50 μL PZM-3 droplet, covered with mineral oil and placed in the incubator (38.5°C with 5% CO₂ and saturated humidity). The cultured embryos were assessed and collected at the following hours: pronucleus (PN, 16 h); 2-cell (2C, 32 h); 4-cell (4C, 44 h); 8-cell (8C, 82 h); morula (MO, 104 h); blastocyst-day 5 (BL, 120 h); blastocyst-day 6 (BL, 144 h); blastocyst-day 7 (BL, 168 h).

The embryos were treated with 25 μg/mL α-amanitin (RNA polymerase II inhibitor) for 30 min at the 2C, 4C, and 8C stages. The embryos in the control groups were treated with solvent (dimethylsulfoxide). 2C embryos from treatment and control groups were washed with PZM-3 three times and cultured and then were collected at 4C for RT-qPCR. 4C embryos from treatment and control groups were washed with PZM-3 three times and cultured and then collected at 8C for RT-qPCR. 8C embryos from treatment and control groups were washed with PZM-3 three times and cultured and then collected at MO for RT-qPCR. All experiments were replicated three times.

DDB1 expression was knocked down in MII oocytes by cytoplasm microinjecting two small interfering RNA species (siRNA sequences are listed in Table S2). A non-specific siRNA was used as a negative control (GenePharma, Shanghai, China). Briefly, MII oocytes were placed in an operating medium (TCM199 with 2% fetal bovine serum and 7.5 μg/mL Cytochalasin B), and ∼10 pL of the siRNA solution (25 μM) was cytoplasm microinjected per cell with a micromanipulator system (Olympus, Japan). After 30 min of recovery in PZM-3 media, the oocytes were processed into PA or IVF. FAM-siRNA (green fluorescence) was used to validate injection efficiency. Recovered oocytes with tightness, margin smooth and good refraction of cytoplasm were considered survival (Fig. S1a). The non-treated (control), RNase-free water injection (sham water), and non-specific siRNA injection (NC-siRNA) groups were set as controls. There were more than 20 oocytes per group and the experiment was replicated three times.

More than 20 oocytes and 10 embryos were used for total RNA extraction (74034, Qiagen, Germany). After that, total RNA was used for reverse transcription (205311, Qiagen, Germany). Quantitative real time PCR (StepOnePlus, Applied Biosystems, USA) was used to measure target gene expression. The amplification reaction consisted of template denaturation and polymerase activation at 95°C for 10 min, followed by 40 cycles of denaturation at 95°C for 15 s, annealing at 60°C for 60 s. and 72°C extension for 30 s. The genes expression level was normalised against H2AFZ (Lee et al. 2017) and presented as relative fold change to the controls (n = 3). Three independent experiments and qPCR reactions were performed. All primers are listed in Table S3.

Blastocyst apoptotic cells were stained using an In Situ Cell Death Detection Kit (12156792910, Roche, USA). Briefly, blastocyst (day 7) embryos were fixed with 4% paraformaldehyde for 15 min and permeabilised with 1% Triton X-100 for 30 min. After that, the blastocyst embryos were incubated with TUNEL reaction solution for 1 h at 37°C, then with 10 μg/mL Hoechst 33342 for 15 min. Finally, the apoptotic cells were imaged using an inverted fluorescence microscope (Olympus, Japan) and quantified with Image J software (ImageJ 1.52a, NIH, USA). Five blastocysts were quantified in each group, and the experiment was replicated three times.

CDX2 protein was stained for immunofluorescence as previously described (Ding et al. 2017; Gao et al. 2020). Briefly, blastocyst (day 7) embryos were fixed (4% paraformaldehyde, 15 min) and permeabilised (1% Triton X-100, 30 min), then blocked (2% BSA, 2 h, RT). The embryos were then followed by incubating with anti-CDX2 antibody (CDX2, AM392, Biogenex, USA) overnight at 4°C. After washing, the embryos were incubated with secondary antibody (1:200, goat anti-mouse IgG, Alexa Fluor 488, A11029, Invitrogen, USA) for 2 h at 37°C. After being counterstained with propidium iodide or 4,6-diamidino-2-phenylindole dihydrochloride for 10 min, the stained embryos were imaged using an inverted fluorescence microscope (Olympus, Japan) and quantified using Image J software (ImageJ 1.52a, NIH, USA).

The non-specific siRNA injected (NC) and DDB1-siRNA injected (DDB1-KD) porcine morulae were collected for RNA sequencing (five morulae per group, n = 3). Briefly, the morulae RNA was extracted (TRK1001, LC Science, Houston, TX) and the cDNA library was constructed (TruSeq Nano DNA LT Library Preparation Kit, FC-121-4001, Illumina, USA). Then the cDNA was fragmented by dsDNA Fragmentase (NEB, M0348S, USA) and paired-end sequencing was performed on an Illumina Novaseq™ 6000 (LC-bio, China) using Illumina paired-end RNA-seq approach. A total of 57 G bp of cleaned, paired-end reads were produced and mapped to the Sus scrofa reference genome. StringTie was used to determine mRNA expression levels by calculating the fragments per kilobase of exon model per million mapped fragments (FPKM). The differentially expressed genes were screened with log₂ fold change > 1 or log₂ fold change < −1 and with a statistical significance (P-value < 0.05) by the R package. Gene ontology (GO) analysis was performed using http://geneontology.org.

At least three replicates were performed for all experiments. The statistical analyses were performed using a student’s t-test or a one-way ANOVA (SPSS 17.0, USA). The graphs were constructed by GraphPad Prism (GraphPad software 5.01, USA). A P value of < 0.05 was considered to be statistically significant.

The mRNA expression levels of DDB1 in oocytes and preimplantation embryos were measured by RT-qPCR. The expression level of the DDB1 transcript level was highest in oocyte at the GV stage and was significantly higher than the MII stage and embryos in different developmental stage (Fig. 1a). After fertilisation, the DDB1 mRNA abundance peaked at the stage of 2C, followed by 4C, PN, MO, BL, and was lowest in 8C (Fig. 1a).

Fig. 1.  Characterisation of DDB1 in porcine oocytes and preimplantation embryos. (a) DDB1 mRNA was analysed in oocytes and preimplantation embryos by RT-qPCR at different developmental stages (at least 20 oocytes and 10 embryos per pool, n = 3). (b) Relative abundance of DDB1 mRNA in 4C, 8C, and MO embryos derived from α-amanitin treated 2C, 4C, and 8C embryos, respectively, determined by RT-qPCR (10 embryos per pool, n = 3). (c) RT-qPCR analysis of DDB1 mRNA levels in non-injected (control), RNase-free water injection (sham water), negative control siRNA (NC-siRNA), and DDB1-siRNA injected 4C or 8C embryos (10 embryos per pool, n = 3). D1 and D2 correspond to two different DDB1 siRNA sequences; D1 + D2, cocktail of D1 and D2 (1:1). All data are presented as mean ± s.e.m. and different letters on the column indicate significant differences (P < 0.05).

Porcine zygotic genome activation occurs at the 4-cell stage (Cao et al. 2014). α-amanitin, an inhibitor of RNA polymerase II, has been commonly used to block zygote mRNA synthesis during preimplantation stage (Uh and Lee 2017). We treated the 2C, 4C, and 8C embryos with α-amanitin to inhibit zygote mRNA produce. The DDB1 mRNA expression level did not change between the treatment and the control groups (Fig. 1b), suggesting that DDB1 is one of the maternal origin transcripts in porcine early embryos.

We designed two siRNA species (DDB1-siRNA1, D1; DDB1-siRNA2, D2) to target the porcine DDB1 sequence for the DDB1 knockdown. 25 μM siRNA was determined to be the most appropriate concentration (Fig. S1d). As shown in Fig. 1c, the endogenous DDB1 mRNA was significantly reduced in the 4C embryos (D1, 78.21%; D2, 79.65%; D1 + D2, 84.49%) and the 8C embryos (D1, 80.43%; D2, 81.58%; D1 + D2, 91.46%). A DDB1-siRNA cocktail was used for the subsequent knockdown experiments. The injection efficiency of FAM-siRNA was 97.75% and survival rate of MII oocytes after injection was 95.75% (Fig. S1b, c).

We knocked down DDB1 in IVF embryos to explore the effect of DDB1 on preimplantation development. As can be seen in Fig. 2a, b, the DDB1-deficient IVF embryos had an impaired development ability as the ratios of 8C, MO, and BL were all significantly lower than the controls. The DDB1-deficient blastocysts had a significantly decreased total cell number and TE cell number (Fig. 2c, d). Although the DDB1-deficient embryos had no change in ICM cell population, the ICM/TE significantly increased compared with the controls (Fig. 2d). Similar results were detected in the DDB1-deficient PA embryos (Fig. S2).

Fig. 2.  Effect of DDB1 knockdown on in vitro development of embryos derived from IVF. (a) Representative brightfield images of embryos at different developmental stages. Scale bar: 100 μm. (b) Developmental rate of embryos at different stages. The number of 2C embryos was regarded as the initial data and the embryos of all other stages were compared to it (at least 20 oocytes per treatment, n = 4). Data are shown as mean ± s.e.m. and different letters indicate significant difference (P < 0.05). (c) Representative fluorescence staining images of DDB1 knockdown and the control. Nuclei (left panel, red), CDX2 (middle panel, green), and merged images between the DNA and CDX2 (right panel). Scale bar: 100 μm. (d) Detection of the quality of blastocysts by an analysis of the number of total cells, TE cells, ICM cells, and the ratio of ICM to TE (at least eight blastocysts per group, n = 3). Data are presented as mean ± s.e.m. and different letters indicate significant difference (P < 0.05).

Due to the decreased number of TE cells in the DDB1-deficient blastocysts, TUNEL staining was applied to establish whether DDB1 ablation caused cell apoptosis. As shown in Fig. 3, there was a greater number of apoptotic cells and a higher apoptotic index in the DDB1-deficient blastocysts relative to the control group.

Fig. 3.  Effect of DDB1 deficiency on cell apoptosis in blastocysts. (a) Apoptosis of blastocysts was detected by TUNEL staining and representative images of the DDB1-KD groups and the control groups are shown. DNA (top panel, blue), TUNEL (middle panel, red), merged images between the DNA and TUNEL (bottom panel). Scale bar: 100 μm. (b) The number of TUNEL positive cells was statistically analysed in the DDB1-KD and control groups, respectively (at least 10 blastocysts per group, n = 3). (c) The apoptotic index (ratio of apoptotic cells to total cells) was statistically analysed in the DDB1-KD and control groups. Data are presented as the mean ± s.e.m. and the different letters indicate significant difference (P < 0.05).

An RNA-seq analysis was used for exploring the molecular difference in the DBB1-deficient embryos (Fig. 4a). All five replicates in the same group had a high correlation (DDB1-KD groups, R = 0.81∼0.96; NC groups, R = 0.88∼0.97) (Fig. S3a, b).

Fig. 4.  Transcriptome analysis of the morula embryos deficient of DDB1. (a) Schematic design of the morula selected for RNA-seq (three morulae per group, n = 5). (b) Volcano plot showing all DEGs between the DDB1-KD and NC groups. (c) Validation of selected downregulated (top panel) and upregulated genes (bottom panel) in the DDB1-KD and the control groups by RT-qPCR (10 morulae per pool, n = 3). Data are shown as mean ± s.e.m. and the different letters indicate significant difference (P < 0.05). (d) Heatmaps showing partial DEGs at the morula between the DDB1-KD and NC groups (P < 0.05, log₂ fold change > 0.8 or < −0.8). (e) GO analysis exhibiting the most enriched functional categories by the DEGs.

Our result demonstrated that 236 DEGs accounted for the transcriptome changes between the DDB1-deficient and control groups. There were 131 upregulated and 105 downregulated genes in the DDB1-deficient morula (Fig. 4b, d).

A total of 19 differentially expressed genes (DEGs) (FC > 2 or <0.5, P < 0.05) were validated by RT-qPCR (UP: EZH1, JAM3, PHF1, PPP1CB, BMP15, KPNA7, MAPK9, CD3E, PIWIL2, FKBP6; DOWN: YAP1 (also known as YAP), SH2B3, ANXA1, CSF1R, IL1B, ERBB4, FRMD4B, DAB2, TCIRG1) (Fig. 4c) and their expression patterns were consistent with the RNA-seq result.

A gene ontology (GO) analysis was performed on the DEGs and three function categories (the biological process, cellular component, and molecular function) were used for the classification (Fig. 4e). Several blastocyst genesis key genes were found to be enriched in the biological process such as YAP1 (Cao et al. 2019), JAM3 (Su et al. 2012), KPNA7 (Tejomurtula et al. 2009), and PIWIL2 (Roovers et al. 2015).

The Hippo signalling pathway plays an essential role in the establishment of TE-specific features (Saini and Yamanaka 2018). Our results demonstrated a significantly downregulated YAP1 in the DDB1-deficient morula (Fig. 4c). The expression level of TEAD4, CDX2, and GATA3 – the downstream effectors of YAP – were all significantly decreased in the DDB1-deficient blastocysts compared to the control groups (Fig. 5a). Immunofluorescent staining confirmed the reduced protein CDX2 level in the DDB1-deficient blastocysts (Fig. 5b, c). The expression of the ICM-specific genes OCT4 and NANOG decreased, while SOX2 increased in the DDB1-deficient blastocysts (Fig. 5a).

Fig. 5.  DDB1 deficiency disturbs the expression of pluripotency and the differentiation of associated genes in the blastocysts. (a) RT-qPCR analysis of the pluripotency and differentiation-associated gene mRNA abundance in the NC and DDB1-KD groups (five blastocysts per pool, n = 3). Data are shown as mean ± s.e.m. and the different letters indicate significant difference (P < 0.05). (b) Representative fluorescence staining images of the nuclei and CDX2 protein. Nuclei (left panel, blue), CDX2 (middle panel, green), and merged images between the DNA and CDX2 (right panel). Scale bar: 100 μm. (c) The average values of the fluorescence intensities of CDX2 were assessed in the NC and DDB1-KD groups by densitometry (at least five blastocysts per treatment; n = 3 replicates). **P < 0.01.

DDB1 is a multifunctional gene and plays a role in embryonic stem cell differentiation, oocyte maturation, and early embryonic development (Yu et al. 2013, 2015a; Gao et al. 2015). In this study, we demonstrated that DDB1 was a transcript of maternal origin in the porcine model and was consistently expressed in the oocytes and embryos. Our generated DDB1-deficient embryos showed an impaired 8C, MO, and BL ratio as well as a significantly decreased TE population associated with increased apoptotic cells in blastocysts. The RNA-seq analysis showed a changed gene expression profile in the DDB1-deficient embryos indicating that DDB1 could regulate porcine embryo development by modulating the differentiation of the embryonic cell lineages and pluripotency. Our work highlights the critical function of DDB1 in porcine embryonic development, which is of benefit to embryonic development-related research.

MEGs are not all highly conserved among mammal species. For example, Linker histone H1FOO as a maternal gene is essential for the morula–blastocyst transition in bovine early embryonic development (Li et al. 2021); however, the knockdown of H1foo did not affect the preimplantation development to the blastocyst stage for mice (Funaya et al. 2018). Ddb1 was first reported in mice, which is a maternal factor required for zygotic genome activation, and a Ddb1-deficient embryo died before the 8-cell stage (Yu et al. 2013). The expression pattern of Ddb1 in mice oocytes and embryos was unclear. In this study, we reported the expression pattern of DDB1 in porcine oocytes and embryos. A DDB1-deficient porcine embryo could develop to the blastocyst stage but the quality of the resultant blastocyst was poor. These results suggested that the function of maternal DDB1 was less conserved in preimplantation embryonic development between the mouse and porcine models.

Apoptosis is a critical factor that correlated with embryo developmental arrest in bovine and porcine studies (Mateusen et al. 2005; Antunes et al. 2010). In our study, we found that DDB1 depletion stimulated apoptosis in the developing porcine embryos. DDB1 takes part in the nucleotide excision repair pathway (Chu and Yang 2008). The loss of DDB1 led to transcriptional p53 pathway activation, cell cycle deregulation, and an increase in apoptosis in zebrafish embryos (Hu et al. 2015). The deletion of DDB1 in hematopoietic stem cells has been shown to induce p53 pathway activation leading to DNA damage and rapid apoptosis (Gao et al. 2015). Our RNA-seq result showed that myeloid leukaemia factor 1 (MLF1) was upregulated in the DDB1-KD morula (Fig. 4d). MLF1 is a negative regulator of cell cycle progression functioning upstream of the p53, which induced p53-dependent cell cycle arrest in mice embryonic fibroblasts (Yoneda-Kato et al. 2005). Overexpressed MLF1 has been shown to increase apoptosis in neonatal rat cardiomyocytes (Rangrez et al. 2017). Therefore, MLF1 may be a potential regulator of increased cellular apoptosis in DDB1-deficient porcine embryos, however further work is required.

The morula is a pivotal period for blastocyst formation, controlled by a significant amount of differentially expressed genes between the morula and blastocyst stage (Hsu et al. 2012). In this study, the most changed genes in the DDB1-deficient morula were found to be associated with signal transduction, cell adhesion, and stem cell differentiation including YAP1 (also known as YAP). The Hippo/YAP signalling cascade is central to cell fate specification processes indicating a TE/ICM lineage specification (Sasaki 2015; Chazaud and Yamanaka 2016). In mice, the polarisation of the outer cells promoted YAP nuclear localisation through the regulation of the Hippo signalling pathway (Nishioka et al. 2009; Saini and Yamanaka 2018). Non-phosphorylated YAP entered the nucleus and interacted with TEAD4 to promote the expression of downstream genes such as CDX2 and GATA3, which are related to trophoblast differentiation (Ralston et al. 2010). The formation of the TE lineage depends on the transcription factor CDX2. The deletion of CDX2 has been shown to result in a failure to form TE and impair the TE function (Strumpf et al. 2005; Goissis and Cibelli 2014; Bou et al. 2017; Liu et al. 2018). Our result showed downregulated gene and protein levels of CDX2 and GATA3 and significantly reduced TE population, indicating that CDX2-mediated TE was reduced in the DDB1-deficient porcine embryos.

Pluripotency is one property for ICM cells that generates all somatic lineages and germlines (Nichols and Smith 2012). OCT4, NANOG, and SOX2 are key transcription factors that maintain the pluripotency of ICM cells (Boyer et al. 2005). OCT4 and NANOG are highly expressed in embryonic stem cells (ESCs) and have been shown to trigger differentiation when repressed (Pan and Thomson 2007; Shi and Jin 2010). In contrast, an increase in SOX2 can induce the differentiation of ESCs (Kopp et al. 2008). Our results showed a decreased OCT4 and NANOG and an increased SOX2 expression in the DDB1-deficient blastocysts (Fig. 5a), suggesting an essential role of DDB1 in ICM pluripotency maintenance.

In conclusion, our study demonstrated the maternal origin of DDB1 in a porcine model. A lack of DDB1 enhanced cellular apoptosis and impaired TE formation and pluripotency maintenance in porcine blastocysts. The decreased expression of CDX2 was the potential molecular mechanism of the decreased TE population in DDB1-deficient porcine blastocysts. Further studies are required to explore the function of DDB1 in embryonic development.

Supplementary material is available online.

The data that support the findings of this study are openly available in Gene Expression Omnibus (GEO) database (accession number GSE189720).

The authors declare no conflicts of interest with regard to the study.

This research was funded by University Natural Science Research Project of Anhui Province (KJ2019A0517) and The First Affiliated Hospital of Bengbu Medical College Science Fund for Outstanding Young Scholars (2019byyfyyq10).