Rbm14 maintains the integrity of genomic DNA during early mouse embryogenesis via mediating alternative splicing

Abstract Objective In this study, we generated an Rbm14 knockout mouse model to explore its functions during early mouse embryogenesis. Materials and methods The Rbm14 knockout mouse model was generated by a combination of clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9 and microinjection techniques. The developmental defects of the knockout embryos were characterized by histological analyses. The accumulation of DNA damage in mouse embryonic stem cells (ESCs) was detected by γH2AX staining and comet assay. The altered mRNA splicing of DNA damage response (DDR)‐related genes was detected by RNA‐Seq analysis and confirmed by semi‐quantitative PCR. The interaction of RBM14 with alternative splicing‐related genes was detected by immunoprecipitation‐mass spectra (IP‐MS) and confirmed by co‐immunoprecipitation (Co‐IP). Results Rbm14 knockout in mice results in apoptosis and cell proliferation defects in early post‐implantation epiblast cells, leading to gastrulation disruption and embryonic lethality. FACS and immunostaining demonstrate accumulation of DNA damage in Rbm14 knockout ES cells. We also identified altered splicing of DDR‐related genes in the knockout mouse ESCs by RNA‐Seq, indicating that RBM14‐mediated alternative splicing is required for the maintenance of genome integrity during early mouse embryogenesis. Conclusions Our work reveals that Rbm14 plays an essential role in the maintenance of genome integrity during early mouse embryonic development by regulating alternative splicing of DDR‐related genes.

were characterized by histological analyses. The accumulation of DNA damage in mouse embryonic stem cells (ESCs) was detected by γH2AX staining and comet assay. The altered mRNA splicing of DNA damage response (DDR)-related genes was detected by RNA-Seq analysis and confirmed by semi-quantitative PCR. The interaction of RBM14 with alternative splicing-related genes was detected by immunoprecipitation-mass spectra (IP-MS) and confirmed by co-immunoprecipitation (Co-IP).
Results: Rbm14 knockout in mice results in apoptosis and cell proliferation defects in early post-implantation epiblast cells, leading to gastrulation disruption and embryonic lethality. FACS and immunostaining demonstrate accumulation of DNA damage in Rbm14 knockout ES cells. We also identified altered splicing of DDR-related genes in the knockout mouse ESCs by RNA-Seq, indicating that RBM14-mediated alternative splicing is required for the maintenance of genome integrity during early mouse embryogenesis.
Conclusions: Our work reveals that Rbm14 plays an essential role in the maintenance of genome integrity during early mouse embryonic development by regulating alternative splicing of DDR-related genes.

| INTRODUC TI ON
RNA binding proteins (RBPs) are a diverse protein family that is designated by their ability to bind to single or double strand RNAs. RNA transcripts are recognized and covered by RBPs as soon as they are synthetized and form ribonucleoprotein (RNP) complexes. 1,2 RBPs are reported to be involved in various RNA metabolic processes, including transcription, 3 editing, 4 splicing, 5,6 transport 7 and translation. 8 RBPs recognize RNA through specific amino acid sequences, such as the RNA recognition motif (RRM), arginine-rich motif (ARM), K homology domain (KHD) and arginine-glycine-glycine (RGG) box. 9 As one of the largest subgroups of single strand RNA binding proteins in eukaryotes, 10 RRM family proteins are reported to be involved in multiple cellular functions and diseases, such as germ cell development, 11 senescence 12 and malignancy. 13 Many RRM proteins have been reported to be implicated in mammalian embryonic development. For example, RNA binding motif protein 15 (RBM15), also known as OTT1, is highly expressed in various tissue types. The germ line deletion of Rbm15 gene in mice results in defects in placental trophoblast development and placental vascular branching morphogenesis, thus leads to embryonic lethality beyond E9.5. 14 While a conditional deletion of Rbm15 within the hematopoietic compartment blocks B-cell development. 15 Another RRM family protein, RBM20, is a tissue-specific pre-mRNA splicing factor that is highly expressed in human heart. RBM20 mediates the alternative splicing of specific mRNA variants of many genes in cardiac muscles. 6 Mutations in Rbm20 gene are reported to be related with dilated cardiomyopathy in humans. 16 RBM14 is an RBP that contains two RRMs in the N-terminus and a prion-like domain (PLD) in the C terminus. 17 With its ability to interact with both RNAs and proteins, RBM14 acts as a multifunctional protein in eukaryotic cells and is reported to be implicated in many aspects of cellular processes such as transcription coactivation, 18 alternative splicing, 19 spindle assembly, 20 DNA repair 21 and cell differentiation. 22 Recently, we had reported that RBM14 participates in pluripotency maintenance and mesoderm development of mouse embryonic stem cells (ESCs). 23 However, the in vivo function of RBM14 in mammalian embryogenesis remains unclear. In this study, we investigated the role of RBM14 in mouse embryonic development using a Rbm14 knockout mouse model generated through clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9. 24 Depletion of RBM14 causes DNA damage accumulation and cell proliferation defects, thus leads to the arrest of embryogenesis during gastrulation. Further studies demonstrate that RBM14 plays a vital role in the maintenance of genome integrity during early mouse embryogenesis through regulating alternative splicing of genes associated with DNA damage response (DDR).

| Animals
The ICR mice were purchased from the Beijing Vital River Laboratory Animal Center. All mice were housed under specific pathogen-free (SPF) conditions in the animal facilities of the Institute of Zoology, Chinese Academy of Sciences. All animal experiments were approved by the Committee on Animal Care at the Institute of Zoology, Chinese Academy of Sciences. All institutional and national guidelines for the care and use of laboratory animals were followed.

| Generation of the Rbm14 knockout allele
The construction of the T7-sgRNA plasmid was performed as described previously. 25 The single stranded sgRNA oligonucleotides synthesized by the Beijing Genomics Institute (BGI) were annealed to form 5′ and 3′ overhangs, which are complementary to the sticky ends of BbsI-digested T7-gRNA backbone. The sequence of the oligonucleotides is listed in Table S1. The T7-sgRNA and T7-Cas9 plasmids were linearized using the endonuclease to serve as templates for in vitro transcription with the HiScribe T7 In Vitro Transcription Kit (New England Biolabs), following the manufacturer's instructions.
The Cas9 mRNA and sgRNA were introduced into the fertilized eggs by microinjection. Each egg was injected with 2.5 ng of sgRNA and 10 ng of Cas9 mRNA. The injected embryos were cultured in vitro for 2 hours before implantation into the oviducts of the pseudopregnant female mice. Full-term pups were obtained by natural labour 19.5 days post coitum (dpc). The genotype of the pups was determined using the tail tip DNA and primers F and R (sequences listed in Table S1). A Δ68 knockout allele that caused premature termination of RBM14 translation was identified in a male founder mouse.

| Micromorphological observation
Early postimplantation embryos at E5.5, E6.5 and E7.5 stages obtained after interbreeding of Rbm14 +/− male and female mice were dissected from the uteruses of the pregnant female mice. The size and morphology of the embryos were observed and imaged using a Leica orthographic microscope. The embryos were subjected to genotyping using primers F and R.

| Haematoxylin and eosin staining
Haematoxylin and eosin (H&E) staining were performed as described previously. 25 Briefly, E7.5 stage embryos obtained after interbreeding of Rbm14 +/− male and female mice were dissected and fixed with 4% paraformaldehyde. The fixed embryos were dehydrated in an ethanol series (70%, 80%, 90% and 100% ethanol) and embedded in paraffin. The embryos were cut into 3 μm sections from the paraffin-embedded blocks using a Leica slicing machine (Leica Biosystems) and mounted on poly-d-lysine-coated glass slides (Zhong Shan Golding Bridge Biotechnology). The slides were heated at 65°C for 2 hours and immersed in xylene to remove the paraffin. After rehydration in an ethanol series (100%, 100%, 90%, 80% and 70% ethanol), the sections were stained with H&E following standard methods and imaged using a Leica Aperio VERSA 8 microscope (Leica Biosystems).

| Immunofluorescent staining
Immunofluorescent staining was performed as described previously. 25 For pre-implantation embryos, the embryos obtained after interbreeding of Rbm14 +/− male and female mice were flushed from the uteruses of the pregnant female mice and fixed with 4% paraformaldehyde. For ESCs, the cells were plated on 15-mm round coverslips and cultured for one or two days and fixed with 4% paraformaldehyde. For postimplantation embryos, the embryos obtained after interbreeding of Rbm14 +/− male and female mice were dissected from the uterus of the pregnant female mice, fixed, dehydrated and embedded in paraffin followed by sectioning. The sections were then dewaxed and rehydrated before staining. The fixed embryos, cells or rehydrated sections were subjected to immunostaining with appropriate primary antibodies and corresponding secondary antibodies following standard methods. The fluorescent images of the samples were captured using a Carl Zeiss LSM 780 confocal system. The primary antibodies and appropriate fluorophore (FITC, Cy3 and Cy5)-conjugated secondary antibodies used in this study are listed in Table S2.

| Cell proliferation assay
The ESCs at the logarithmic growth phase were trypsinized using 0.25% Trypsin-EDTA (Gibco) and plated on 48-well plates at a density of 5000 cells/well and cultured for 1-5 days. Cell proliferation was examined using the Cell Counting Kit-8 (CCK-8) kit (Solarbio), following the manufacturer's instructions. Briefly, 1/10 volume of CCK-8 reagent was added into the culture medium in each well and incubated for 2 hours. The absorbance of each well at 450 nm was measured using a BioTek spectrophotometer.

| Western blotting
Western blotting was performed as described previously. 25 Briefly, ESCs were lysed with RIPA lysis buffer (0.5% NP-40, 50 mmol/L Tris-HCl, 150 mmol/L NaCl, 1 mmol/L EDTA, pH = 8) supplemented with proteinase and phosphatase inhibitors (Thermo Fisher). The cell lysates were subjected to sodium dodecyl sulphate gel electrophoresis (SDS-PAGE) at 110 V for 2 hours. The proteins on the gel were transferred to a nitrocellulose membrane (Thermo Fisher).
The membrane was then blocked with 5% bovine serum albumin (BSA) prepared in PBS. The membrane was probed with appropriate primary antibodies and corresponding secondary antibodies. The proteins were detected in a LI-COR (ODYSSEY CLx) exposing equipment. Antibodies used in this assay are listed in Table S2.

| Apoptosis assay
The apoptosis rate of ESCs was examined using the FITC Annexin V

| Cell cycle analysis
The cell cycle of the ESCs was analysed using the BD Pharmingen

| Single cell gel electrophoresis assay
Single cell gel electrophoresis assay was performed using the

Immunoprecipitation (IP) was performed using the Pierce Crosslink
Magnetic IP/Co-IP Kit (Thermo Fisher), following the manufacturer's instructions with minor modifications. Briefly, 10 μg of anti-GFP antibody (Abcam, ab290) or IgG antibody was bound to 50 μL of Pierce Protein A/G Magnetic Beads. The coupled antibodies were crosslinked to the magnetic beads using the DSS cross-linker. The ESCs (1 × 10 7 ) were lysed with 500 μL of ice-cold IP lysis/wash buffer and incubated with the antibody-bead mixture at 4°C overnight.
The beads were collected using a magnetic stand and washed with IP lysis/wash buffer for five times followed by elution with 100 µL of elution buffer. The eluted proteins were subjected to mass spectrometry or western blot analysis.

| PAR-CLIP-Seq and bioinformatic analysis
PAR-CLIP experiments were conducted as described. 26 Briefly, Gfp-Rbm14 mESCs were pulsed with 100 μmol/L 4-thiourdine (4SU, Sigma) overnight. After irradiation with UV at 365 nm, RBM14/RNA complexes were immunoprecipitated with the anti-GFP antibody and protein A magnetic beads (Thermo). After RNase T1 treatment, the immunoprecipitates were resolved on 4%-12% LDS gel with Mops buffer (Life Science Tech). RNA on the gel was visualized with SYBR green I staining on the PhosphorImager (GE). RNA bound to the RBM14 protein was extracted from the gel, freed from protein by digestion with proteinase K, purified, converted into cDNA and sequenced by Illumina X Ten. The PAR-CLIP-Seq data were analysed as described in previous paper. 27 In brief, The PAR-CLIP-Seq reads of the RBM14 protein were trimmed with the Cutadapt software (version 2.4). 28 Reads that were less than 18 nt in length or contained an ambiguous nucleotide were discarded. The remained reads were aligned to the mouse genome (mm10), with parameters recommended by PARalyzer (v1.5) 29 by the Bowtie algorithm (version 1.2.3). 30 Clusters were obtained by PARalyzer (v1.5) with the default parameters.

| Quantitative PCR and semi-quantitative PCR
RNA was reverse transcribed into cDNA using the Reverse Transcription System (Promega), following the manufacturer's instructions. Briefly, 1 μg RNA were mixed with 1 μL random primers and 1 μL oligo primers and then added with RNase-free ddH 2 O up to 13.5 μL, then the RNA was annealed by incubation at 70°C for 5 minutes followed by incubation on ice for 2 minutes. After then 1 μL MLV reverse transcriptase, 4 μL 5× reaction buffer, 0.5 μL RNasin and 1 μL dNTP were added in the annealed RNA. RNA was reverse transcribed in 42°C for 1 hour followed by heat inactivation at 85°C for 5 minutes. For quantitative PCR, 0.2 μL cDNA were used as templates for PCR using SYBR Green PCR Master Mix (Toyobo) on an Agilent real-time fluorescence quantitative PCR instrument (Mx3000P). The formula −2 ΔΔCT was used to calculate the relative expression of Rbm14. For semi-quantitative PCR, 1 μL cDNA was then used as templates for PCR using the Vazyme 2 × PCR Master Mix. The PCR products were then analysed using the agarose gel electrophoresis system (Bio-Rad). Primers for quantitative PCR and semi-quantitative PCR were listed in Table S1.

| Statistical analysis
All statistical analyses (unless stated otherwise) were performed using the R package for statistical computing. For experimental data quantification, Students' t-test was applied using the GraphPad Prism 6 software. The error bar represents the standard error of mean (SEM) (unless stated otherwise). The difference was considered statistically significant when the P value was <.05.

| Rbm14 knockout leads to embryonic lethality in mice
To investigate the role of Rbm14 in mouse embryonic development, we generated a Rbm14 knockout allele through CRISPR/ Cas9-mediated non-homologous end joining (NHEJ) ( Figure S1A).
The Cas9 mRNA and a single guide RNA (sgRNA) targeting the first exon of Rbm14 just downstream of the start codon were introduced into the zygote-stage mouse embryos through microinjection. Finally, a male founder mouse harbouring a Rbm14 knockout allele was obtained. The Rbm14 knockout allele contains a 68 bp deletion in the target site and results in frame shift and premature termination of translation at the N-terminus of the RBM14 protein.
The founder mouse was crossed with wild-type mice (Rbm14 +/+ ) to generate Rbm14 +/− heterozygous pups, the genotypes of the pups were detected by polymerase chain reaction (PCR, Figure S1B) and confirmed by Sanger sequencing ( Figure S1C). The Rbm14 +/− mice were fertile and showed no abnormalities when compared to their Rbm14 +/+ counterparts. Next, we investigated the role of Rbm14 in mouse embryonic development using various mating strategies: ♀  Immunostaining for NANOG and OCT4 revealed no substantial defects in the pre-implantation blastocyst stage embryos with the knockout of Rbm14 ( Figure S2B). Genotype analysis of the embryos revealed that the proportion of embryos with different genotypes was in accordance with the Mendelian ratio (20.7% for Rbm14 +/+ to 55.2% for Rbm14 +/− to 24.1% for Rbm14 −/− , Figure S2C). Our findings indicated that although Rbm14 was highly expressed during the preimplantation stage, Rbm14 knockout could support the development of mouse embryos, at least, from zygotes to blastocysts.

| Rbm14 knockout inhibits gastrulation during early embryonic development in mice
We then dissected embryos at E5.5, E6.5 and E7.5 by mating Rbm14 +/− male and female mice and observed abnormalities in size and morphology in the Rbm14 knockout embryos as early as E5.5 ( Figure 1A and Figure S3A

| Rbm14 knockout induces apoptosis and cell cycle arrest in the post-implantation epiblast
The reduced size and loss of cell density in the Rbm14 knock-

| Rbm14 knockout causes apoptosis and cell cycle arrest in mouse embryonic stem cells
To further investigate the underlying mechanisms for the requirement of Rbm14 in early mouse embryonic development, we derived ESCs from the E3.5 Rbm14 +/+ and Rbm14 −/− blastula stage embryos.
The expression of pluripotency-related genes, such as Nanog and Oct4 were evaluated in these embryos by immunostaining. Our results revealed that both Rbm14 +/+ and Rbm14 −/− ESCs expressed these pluripotent genes in the protein level ( Figure 3A). We also confirmed the depletion of RBM14 protein in the Rbm14 knockout ESCs by western blotting ( Figure 3B). Although no obvious developmental defect was observed in the pre-implantation Rbm14 −/− embryo as described above, we detected decreased proliferation rate in the Rbm14 −/− ESCs ( Figure 3C). We then analysed the apoptosis of Rbm14 −/− ESCs. We detected markedly elevated apoptosis in the Rbm14 −/− ESCs ( Figure 3D,F). Moreover, cell cycle analysis using 5-ethynyl-2′-deoxyuridine (EdU) labelling revealed that the G1 phase was delayed in the Rbm14 −/− ESCs ( Figure 3E,G,H). These results demonstrated that the knockout of Rbm14 results in apoptosis and cell cycle arrest in the mouse ESCs, which was consistent with the proliferation defects observed in the epiblast of the Rbm14 −/− early mouse embryo.

| Rbm14 is required for maintenance of genomic DNA integrity
For proliferating cells, DNA damage is an intense stress that may cause cell death and cell cycle arrest. 31,32 To investigate whether

| Rbm14 regulates alternative splicing of DDRrelated genes
Previous studies have reported that Rbm14 may participate in NHEJ DNA repair by interacting with Ku80 in glioblastoma (GBM) and HeLa cells. 21,22 To further investigate the role of Rbm14 in DDR in ESCs and in early embryonic development, we performed

Immunoprecipitation-Mass Spectra (IP-MS) with a GFP tagged
Rbm14 (Gfp-Rbm14) ESC cell line ( Figure S4A) and identified 172 proteins that interacted with RBM14 protein ( Figure 5A). Gene ontology (GO) term analysis revealed that most of the proteins interacting with RBM14 were enriched in RNA processing related processes, particularly in "RNA splicing" (Figure 5B), indicating that Rbm14 was involved in regulating mRNA alternative splicing. We then confirmed the interaction of RBM14 protein with splicing associated proteins such as SF3B1 and SF3B3 by co-immunoprecipitation (Co-IP). Interestingly, we detected no interaction of RBM14 with Ku80 ( Figure S4B), indicating that RBM14 participated in DDR of ESCs in quite a different way from that in HeLa cells. We then performed RNA sequencing for Rbm14 +/+ and Rbm14 −/− ESCs and analysed the altered splicing events. We detected 330 alternative splicing events under the threshold of 10% percent spliced in (PSI) changes ( Figure   S4C). GO analysis showed that many of the alternatively spliced genes were enriched in the term "Cellular response to DNA damage stimulus" (Figure S4D). To identify the direct interaction of RBM14 with the transcripts of alternatively spliced genes, we performed PAR-CLIP-Seq and obtained 4871 genes whose transcripts were bound by RBM14. Two independent replicates were conducted with a large overlap between the RBM14-binding transcripts. Of all the 214 genes with alternatively spliced exons, 48.6% have RBM14 binding peaks on their mRNA ( Figure 5C). GO analysis showed that genes having RBM14 binding peaks on their mRNA were enriched in terms such as "Cellular response to DNA damage stimulus" and "DNA repair", indicating that RBM14 might directly bind to the transcripts of DDR-associated genes and regulate their splicing process ( Figure 5D). We then confirmed the altered splicing of genes related to DDR, such as Hnrnpk, Syce2 and Fance in Rbm14 knockout ESCs by semi-quantitative PCR. We also found that all these genes harboured RBM14 binding peaks on the adjacent exons or introns of the alternatively spliced exons ( Figure 5E-G). These results demonstrated that Rbm14 was involved in genomic DNA maintenance by regulating the alternative splicing of DDR-related genes. . Two independent replicates were used and identified 172 proteins interacting with RBM14. B, GO analysis for proteins interacting with RBM14 shows enriched RNA processing associated terms. C, Venn diagram showing the identified numbers of genes with alternatively spliced exons and with RBM14 binding peaks in the PAR-CLIP-Seq assay. Two independent replicates were conducted. Of all the 214 genes with alternatively spliced exons, 48.6% have RBM14 binding peaks on their mRNA. D, GO analysis for genes having RBM14 binding peaks on their mRNA. E, Representative sashimi plots displaying RNA-Seq reads coverage and the RBM14 binding sites (indicated by blue rectangular blocks) of the alternatively spliced genes associated with DDR. Alternatively, spliced regions are indicated by semitransparent blocks. F, Calculated PSI value of the alternatively spliced genes associated with DDR in both wild type and Rbm14 knockout embryonic stem cells (ESCs). Data are shown as mean ± SEM. Three cell lines for each genotype were sequenced and analysed (n = 3). *P < .05, **P < .01, ***P < .001, ****P < .0001, Student's t-test. G, Semi quantitative-PCR confirms the altered splicing of DDR associated genes. PSI, percent spliced in many proteins that differ in their structures and functions. 33 A genome-wide analysis of the alternative isoform expression revealed that alternative splicing and isoform-specific expression were frequently observed during the development of early mouse embryo. 34 Previous studies have reported the implication of alternative splicing mediated by RRM proteins in early embryonic development. [35][36][37] RBM14 was originally reported to be a general nuclear coactivator that interacts with thyroid hormone receptor-binding protein (TRBP) to regulate the activation of gene transcription. 38 Further studies revealed roles of Rbm14 in preserving the mitotic spindle integrity 20 and regulation of DNA virus-mediated innate immune response. 39 However, few studies have evaluated the role of Rbm14 in alternative splicing regulation. 19 In this study, we recognized the interaction of RBM14 with core splicing factors such as SF3B1 and SF3B3 ( Figure 5B and Figure S4B).
Additionally, we demonstrated that Rbm14 has a role in the maintenance of genome integrity by regulating the alternative splicing of DDR-related genes ( Figure S4C). The altered splicing of these genes results in cell apoptosis and cell cycle arrest in both mouse embryos and ESCs, and thus inhibits gastrulation during early mouse embryogenesis.
Our previous work reported that knockout of Rbm14 inhibited the expression of pluripotency-associated genes such as Oct4 and Nanog and thus influenced pluripotency maintenance of mESCs.
RNA-Seq also demonstrated decreased expression of genes associated with TGF-beta and Wnt pathways. 23 Interestingly, in this present study, we performed immunostaining and detected no substantial suppression of Oct4 and Nanog expression in Rbm14 knockout mESCs, at least during the first 5-7 passages. This contradiction might be due to the fact that distinct pathways are implicated in pluripotency maintenance of mESCs in different culture conditions. In our previous work, when mESCs were cultured in the serum + LIF system, the LIF/STAT3 signalling and the TGF-beta signalling were essential for the pluripotency maintenance of mESCs. [40][41][42][43] While in this work, mESCs were maintained in the ground state in the 2i + LIF condition, in which a cocktail of two inhibitors for the MEK1/2 and GSK3 pathways (2i), has been proved to be sufficient to maintain mESCs with full pluripotency and LIF supplementation strengthens the robustness. 44 In this condition, the TGF-beta pathway has been shown to be less active. 45 DNA damage is commonly associated with cell proliferation defects such as cell apoptosis and cell cycle arrest. 31 may also regulate the cell proliferation and embryonic development by directly regulating the alternative splicing of related genes. This is because we also detected alternative splicing of genes related to transcription regulation, cell proliferation and embryonic development in the Rbm14 knockout mouse ESCs.
Recent studies have reported the involvement of Rbm14 in NHEJ by recruiting XRCC4 and XLF through Ku80 in HeLa and glioblastoma cells. 21,22 In our work, however, we detected no interaction between RBM14 and Ku80 or other NHEJ repair factors ( Figure   S4B). This may be due to the species specificity of Rbm14 as all these previous studies were conducted in human cells.
In summary, we identified a pivotal role of Rbm14 in the regulation of alternative splicing of DDR-related genes and established that Rbm14 is required for the maintenance of genome integrity during early stage development of mouse embryos. Our study revealed a novel role of RBPs in the regulation of early embryogenesis in mice.

ACK N OWLED G EM ENTS
The

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest. for the sequencing data reported in this paper is CRA001790.