Female ICR mice aged 6 to 8 weeks and male ICR mice aged 8 to 10 weeks were obtained from Orient Bio Co., Ltd. (Seoul, Korea). All procedures involving mice were conducted in accordance with guidelines approved by the Institutional Animal Care and Use Committee of Konkuk University (IACUC approval number: KU22181). The mice were housed in a controlled environment at 22°C±1°C with a 12-h light-dark cycle.

To induce superovulation in female mice, intraperitoneal injections of pregnant mare serum gonadotropin (5 IU) were administered, followed by human chorionic gonadotropin (hCG; 5 IU) 48 h later. Fifteen hours after hCG injection, cumulus-oocyte complexes (COC) were collected from the oviduct and transferred to a modified human tubal fluid (mHTF) medium supplemented with 0.625 mM glutathione (GSH; Sigma-Aldrich, St. Louis, MO, USA). Spermatozoa were collected from the caudal epididymis of male mice and incubated in mHTF medium containing 0.4 mM methyl-β-cyclodextrin (MBCD)-polyvinyl alcohol (PVA) for at least an hour to induce sperm capacitation. Capacitated sperm was then added to the GSH-mHTF medium containing COCs. Two hours after insemination, the fertilized embryos were washed multiple times and cultured in EmbryoMax KSOM Mouse Embryo media (Merck Millipore, Burlington, MA, USA) at 37°C under 5% CO₂.

Each inhibitor was added and incubated for 3 days, and dimethyl sulfoxide (DMSO; Sigma-Aldrich, USA) was used as a vehicle for inhibitors in the control group. Embryos were transferred to the KSOM medium containing either DMSO or a specific inhibitor at 4 h post-insemination (hpi) and cultured for 8 h (12 hpi). The following inhibitors were used in the present study: 5 or 100 nM trichostatin A (TSA; Sigma-Aldrich, USA) for inhibition of histone deacetylase 2 (HDAC2), 20 nM supinoxin (RX-5902; MedChemExpress, Monmouth Junction, NJ, USA) for DDX5 inhibition, and 100 μM 5,6-dichlorobenzimidazole 1-β-D-ribofuranoside (DRB; Sigma-Aldrich, USA) and 1 μM triptolide (TRP; Sigma-Aldrich, USA) for inhibition of transcription. After inhibitor treatment, the embryos were fixed with 4% paraformaldehyde in a solution containing 0.01% PVA in phosphate-buffered saline (PBS) for 15 min. They were then washed thrice with a washing buffer (0.05% Tween 20-0.01% PVA-PBS) for 10 min each. The embryos were permeabilized with 0.5% Triton X-100-PVA in PBS for 20 min, followed by three washes. The embryos were treated with 4 N HCl for 15 min to denature the DNA and then neutralized with 100 mM Tris-HCl (pH 8.5) for 20 min. Subsequently, the embryos were blocked by incubation with 5% bovine serum albumin (BSA)-PVA/PBS for 2 h. Primary antibodies, including anti-S9.6 antibodies (1:200, MABE1095; Sigma-Aldrich, USA) and anti-phospho-histone H2A.X (1:300, 2577; Cell Signaling Technology, Danvers, MA, USA), were diluted in 1% BSA-PVA/PBS and incubated with the embryos at 4°C overnight. After three washes, the embryos were incubated with secondary antibodies, including donkey anti-mouse IgG Alexa Fluor 488 conjugate (1:400; Life Technologies, Carlsbad, CA, USA) and donkey anti-rabbit IgG Alexa Fluor 546 conjugate (1:400; Life Technologies, USA), for 1 h in the dark. DNA was counterstained with 2 μM To-PRO3 Iodine (Thermo Fisher Scientific, Waltham, MA, USA) for 20 min. The stained embryos were washed several times and mounted on slides using a small drop of VECTASHIELD Antifade Mounting Medium (Vector Laboratories, Burlingame, CA, USA).

A publicly available bulk RNA-seq and low-input ribosome profiling (called LiRibo-seq) dataset [39] was used to examine the patterns of protein synthesis in mouse oocyte, 1-cell and 2-cell embryos. The dataset (GSE169632) was downloaded and reanalyzed, and the fragments per kilobase of transcripts per million mapped reads (FPKM) values of actively translated mRNAs were determined using DESeq2 tools. RNAPII ChIP-seq dataset (GSE135457) analyzed in oocyte, 1-cell and 2-cell embryos were used to identify the actively transcribed genes [5]. From the intersection of two independent IP-mass spectrometry (IP-MS) studies [21,22], a set of 205 proteins known to interact with R-loops was selected for further analysis based on their synthesis levels in the LiRibo-seq dataset.

All images were captured using a laser confocal microscope (LSM 8000; Zeiss, Oberkochen, Germany). For 1-cell stage embryos, the middle optical section of Z-stack images containing the maternal and paternal pronuclei was selected for fluorescence intensity quantification. The fluorescence intensity of the S9.6 signal within the pronuclei was measured using the ImageJ software (National Institute of Health, Bethesda, MD, USA). Quantification was performed using the following approach: the total intensity of the pronucleus was calculated by multiplying the mean intensity of the pronucleus region by the area of the pronucleus and then subtracting the product of the mean intensity of the cytoplasmic region by the area of the pronucleus.

Statistical analyses were performed using Prism software (GraphPad Inc., Boston, MA, USA). A two-tailed unpaired t-test with Welch’s correction was used to compare the two groups. One-way analysis of variance followed by Tukey’s test was performed for comparisons among more than three groups. The results are presented as mean±standard error of the mean or mean±standard deviation. Statistical significance is indicated as follows: * p<0.05, ** p<0.01, *** p<0.001, NS, not significant.

To identify potential regulators of R-loop homeostasis in oocyte and early embryos, we analyzed publicly available LiRibo-seq (GSE169632) [39] and RNAPII ChIP-seq datasets (GSE135457) generated from mouse oocyte, 1-cell and 2-cell embryos [5]. From these datasets, 205 proteins that were previously identified as R-loop-interacting proteins [21,22] were selected (Figure 1A; Table 1), and their dynamic changes of gene translation from oocytes to 2-cell stage embryos were investigated. As depicted in Figure 1A, the levels of transcripts undergoing translation differed significantly from those detected in bulk RNA-seq, indicating a specific regulation of translation during early development. Hierarchical clustering analysis of the 205 proteins revealed four distinct clusters: cluster 1 (C1, 120 genes) displayed active translation at the 2-cell stage, C2 (39 genes) exhibited active translation at the oocyte stage, C3 (12 genes) showed active translation at both the oocyte and 1-cell stages, and C4 (24 genes) displayed active translation at the 1-cell stage. Additionally, enrichment of RNAPII at the transcription start sites (TSSs) was observed for most of the 205 genes, yet the RNAPII enrichment was greatly diminished upon DRB treatment (Figure 1A). Based on the translation profiles of these genes, we selected two candidate proteins known for their roles in RNA unwinding, DNA damage repair, and histone modification from the gene clusters (Figure 1B) and employed chemical inhibitors to investigate their functions in the regulation of R-loop metabolism during the ZGA period. The inhibitors used in this study were TSA, which inhibits HDAC2 (a gene from cluster C2), and supinoxin, which inhibits DDX5 (a gene from cluster C4). Interestingly, other class I HDAC genes displayed a distinct pattern of translation in the stages (Figure 1B). Gene ontology (GO) term analysis indicated that the genes within these four clusters were mainly associated with RNA processing and splicing, suggesting their involvement in RNA biology (Figure 1C).

In our previous study, we established a correlation between zygotic R-loop dynamics and gene transcription. To validate the finding, additional experiments were performed using DRB (Figure 2A, 2B), a transcription elongation inhibitor, and TRP (Figure 2C, 2D), a transcription initiation inhibitor. After in vitro fertilization, 1-cell embryos were treated with either TRP or DRB. Consistently, the transcriptional inhibition with either inhibitor significantly reduced R-loop formation in both maternal (M) and paternal (P) pronuclei (PN) (Figure 2B, 2D).

Given that the level of translated HDAC2 was high in and diminished in 1-cell and 2-cell embryos (Figure 1A), and HDAC2 function has been implicated in the ZGA [40], 1-cell embryos were treated with TSA to determine the role of HDAC2 in regulating R-loop formation during minor ZGA. The TSA treatment had little or no effect on the R-loop levels in both the maternal and paternal pronuclei of 1-cell embryos (Figure 3A, 3B). We also found that the levels of histone H3K27ac were slightly increased by both low and high concentrations of TSA (Figure 3A, 3C).

Next, to investigate the function of RNA helicases in R-loop formation during minor ZGA of mouse zygotes, 1-cell stage embryos were treated with supinoxin (RX5902), an RNA helicase DDX5 inhibitor. The DDX5 inhibition [23] led to a notable reduction in the number of R-loop foci observed in 1-cell embryos (Figure 4A, 4B), whereas its inhibition elevated γH2AX level in both maternal and paternal pronuclei (Figure 4A, 4C). These results provide compelling evidence that RNA helicases, specifically DDX5, play critical roles in the regulation of R-loop metabolism in 1-cell embryos during minor ZGA.