Neuroepithelial cell competition triggers loss of cellular juvenescence

Cell competition is a cell–cell interaction mechanism which maintains tissue homeostasis through selective elimination of unfit cells. During early brain development, cells are eliminated through apoptosis. How cells are selected to undergo elimination remains unclear. Here we aimed to identify a role for cell competition in the elimination of suboptimal cells using an in vitro neuroepithelial model. Cell competition was observed when neural progenitor HypoE-N1 cells expressing RASV12 were surrounded by normal cells in the co-culture. The elimination through apoptosis was observed by cellular changes of RASV12 cells with rounding/fragmented morphology, by SYTOX blue-positivity, and by expression of apoptotic markers active caspase-3 and cleaved PARP. In this model, expression of juvenility-associated genes Srsf7 and Ezh2 were suppressed under cell-competitive conditions. Srsf7 depletion led to loss of cellular juvenescence characterized by suppression of Ezh2, cell growth impairment and enhancement of senescence-associated proteins. The cell bodies of eliminated cells were engulfed by the surrounding cells through phagocytosis. Our data indicates that neuroepithelial cell competition may have an important role for maintaining homeostasis in the neuroepithelium by eliminating suboptimal cells through loss of cellular juvenescence.

Well-characterized inducers of cell competition include minute 22 , Bone morphogenetic proteins (BMPs) 1 , Myc 2 , and Ras 20,23 . The discovery of HRAS (Harvey rat sarcoma virus) in 1964 was isolated from the passage of Moloney type-C virus in rats 24 . High expression of activated HRAS oncogene has been proposed to induce senescence 25 . An oncogene-induced senescence (OIS) has been reported by induction of an oncogenic form of RAS in normal human fibroblasts 26 . Cellular juvenescence is characterized by the ability of cells to grow, differentiate and resist premature senescence 27 . We previously identified juvenility-associated genes (JAGs) expressed in brain 28 , cardiomyocytes and hepatocytes 29 . JAGs are genes selectively highly expressed in juvenile cells. Srsf7, Serine-Arginine-rich family of pre-mRNA splicing factors, which constitutes a part of the spliceosome, is one of the JAGs 29 .
In this communication, to investigate the role of cell competition in the developing brain, we have established an in vitro model system. In mice, formation of a neural plate begins at E7.5. Neural plate is an earliest indication of neural development. At this stage, cephalization begins with the presence of enlarged head folds and formation of neural plate from the ectoderm, which converts into neuroectoderm. Based on this developmental stage, we have established a heterogeneous neuroepithelial model in which neural progenitor HypoE-N1 cells were cultured as a monolayer. Heterogeneity was introduced by induction of RAS V12 in a small proportion of the cells. We investigated the molecular mechanisms in neuroepithelial cell competition by which juvenescence-losing cells are sensed and eliminated from the neuroepithelium. In this study, we clarify a role of cell competition as a quality-control mechanism to orchestrate homeostasis in the brain.

Results
Establishment of the neuroepithelial cell competition model. To investigate the phenomenon of cell competition in the developing mammalian central nervous system, we established a cellular model for neuroepithelium using HypoE-N1 cells, a neural progenitor cell line derived from mouse fetal hypothalamus. We performed conditioned medium assays and indirect co-culture assays to exclude the possibility of indirect effects. We confirmed that neuroepithelial cell competition requires direct cell-cell contact, and was not mediated by soluble factors produced during competitive co-culture ( Supplementary Fig. S1). In this system, we co-cultured normal HypoE-N1 cells with doxycycline-inducible RAS V12 -EGFP-transfected HypoE-N1 cells (hereinafter referred to as RAS V12 cells) at a 100:1 ratio to create competitive conditions. For the non-competitive condition, we co-cultured RAS V12 cells with RAS V12 cells labeled with the fluorescent dye CMTPX (Fig. 1a). The expression of RAS V12 upon doxycycline treatment was validated by western blot analysis. RAS V12 was expressed stably up to 72 h with doxycycline treatment (Fig. 1b). SYTOX blue staining showed RAS V12 cells in a competitive condition are positive, indicating they are dying (Fig. 1c). The percentage of SYTOX blue positive in RAS V12 cells in competitive condition were significantly higher compared to normal cells. However, in the non-competitive control condition, both RAS V12 and RAS V12 (CMTPX) cells showed lower frequency of SYTOX blue staining with no significant difference between them (Fig. 1d).
Next, we evaluated the effects of co-culture on the growth rate of cells. The changes in RAS V12 cell numbers over time were compared between competitive and non-competitive co-cultures. We found competitive co-cultures showed reduction in RAS V12 cell numbers after 24 h compared to RAS V12 (CMTPX) in the non-competitive condition (Fig. 1e). We then recorded the phenomenon occurring in the competitive co-culture. Time-lapse video revealed that RAS V12 cells surrounded by normal HypoE-N1 cells showed morphological changes involving rounding-up and eventually fragmentation (Fig. 1f, Supplementary Video S1). Collectively, our data suggest introduction of a small number of RAS V12 cells induced heterogeneity to the neuroepithelium and this established a model for cell competition with the loser RAS V12 cells being eliminated through non-cell autonomous cell death.

Neuroepithelial cell competition occurs by non-cell autonomous induction of apoptosis.
We next investigated the mechanism of RAS V12 cell death from competitive co-culture. During cell competition, most of the loser cells will be eliminated by apoptosis 30 . Thus, we asked whether RAS V12 cells in competitive coculture were also eliminated by apoptosis. Immunostaining with apoptosis markers active caspase-3 and cleaved PARP displayed positive signals in RAS V12 cells surrounded by normal cells in the competitive co-cultures. By contrast, RAS V12 (CMTPX) cells in non-competitive co-cultures did not show positive signals for either apoptosis marker (Fig. 2a,c). Quantification for active caspase-3 and cleaved PARP-positive signals showed RAS V12 cells were significantly higher compared to normal cells in the competitive condition (Fig. 2b,d).
During apoptosis, cytochrome-c is released from mitochondria 31 . Healthy cells displayed punctate patterns of cytochrome-c staining that co-localized with mitochondria. In contrast, cells undergoing apoptosis showed a diffuse pattern of cytochrome-c. RAS V12 cells surrounded by normal cells in competitive co-culture displayed diffuse cytochrome-c staining that did not co-localize with mitochondria (Fig. 2e). We next tried to block apoptosis by using the apoptosis inhibitor ZVAD in the co-cultures. ZVAD treatment significantly increased the number of RAS V12 cells in the competitive co-culture compared to those without ZVAD treatment (Fig. 2f).
A previous study reported involvement of the p38 MAPK pathway during cell competition 32 . Immunofluorescence staining for phospho-p38 in the competitive condition indicated p38 was activated in RAS V12 cells ( Supplementary Fig. S2). To answer whether apoptosis was induced through the MAPK pathway, we inhibited p38 activation in the competitive co-cultures. The percentage of SYTOX blue positive cells was significantly lower in the presence of p38 inhibitor (Fig. 2g). These results indicated that RAS V12 cells in competitive co-culture underwent cell death through non-cell autonomous apoptosis.    www.nature.com/scientificreports/ petitive condition. First, we investigated protein expression of JAGs in cells by immunofluorescence analysis. We observed as early as 16 h of competitive condition, Srsf7 expression was lost in RAS V12 cells surrounded by normal cells (Fig. 3a). In the non-competitive condition, the percentage of Srsf7 positivity among RAS V12 and RAS V12 (CMTPX) cells was not significantly different, but in competitive co-cultures, the percentage of Srsf7 positive cells was significantly decreased in RAS V12 cells compared to normal cells (Fig. 3b). Western blot analysis of non-competitive condition cells showed Srsf7 expression was slightly reduced with RAS V12 induction (Fig. 3c). These results suggest competitive co-culture caused acute loss of Srsf7 in RAS V12 cells. We next explored how Srsf7 loss was faster in RAS V12 cells during the competitive condition. SR proteins have been reported to undergo post-translational modifications by phosphorylation, acetylation, methylation, ubiquitylation and sumoylation 33 . A previous study reported that SRSF5, one of the SR family members, was ubiquitylated by Smurf1 and undergoes proteasome-mediated degradation 34 . To assess whether Srsf7 is degraded through the ubiquitin-proteasome pathway, we conducted ubiquitylation assays. We found that the ubiquitin immunoreactive smear was significantly enhanced with MG132 treatment in HA-Ub transfected cells (Fig. 3d). To further analyze the loss of Srsf7 by proteasome-mediated degradation in the competitive condition, we treated the cocultures with MG132 and quantified the number of Srsf7 positive in RAS V12 cells in the absence and presence of MG132 (Fig. 3e). The percentage of Srsf7 positive in RAS V12 cells with MG132 treatment was higher compared to the non-treated group (Fig. 3f). These observations indicated that during competitive co-culture, Srsf7 in RAS V12 cells was degraded through the ubiquitin-proteasome pathway. To investigate if restoring Srsf7 expression is sufficient to prevent RAS V12 cell elimination, we overexpressed Srsf7 in RAS V12 cells and co-cultured them with normal cells. Srsf7 overexpression in RAS V12 cells (HA-Srsf7) significantly increased the number of RAS V12 cells compared to control treatment (TdTomato, Fig. 3g). In addition, Srsf7 overexpression in RAS V12 cells (HA-Srsf7) significantly reduced the frequency of caspase-3 positive cells (Fig. 3h), which verifies that Srsf7 antagonizes the execution of cell competition. Collectively, these results revealed cell competition triggers the acute loss of Srsf7 in RAS V12 cells through proteasome-mediated degradation.

Loss of Srsf7 suppresses Ezh2 expression in the loser cells.
To investigate how the expression of other proteins change, we measured other JAGs protein expression after 24 h of co-culture. Competitive cocultures displayed acute loss of Ezh2 in RAS V12 cells (Fig. 4a). The percentage of Ezh2-expressing RAS V12 cells in competitive condition was significantly lower compared to normal cells (Fig. 4b). This observation showed a similar pattern with Srsf7 expression in non-competitive and competitive condition. Western blot analysis revealed Ezh2 expression in non-competitive condition was slightly downregulated with RAS V12 induction up to 72 h (Fig. 4c). Next, we did immunofluorescence staining analysis for cytochrome-c to investigate whether Ezh2 loser cells underwent apoptosis. We found that both Ezh2 positive and negative RAS V12 cells showed punctate staining, which suggests that Ezh2 loss occurs before apoptosis induction (Fig. 4d). Thus, our analyses revealed Srsf7 loss induced Ezh2 suppression in loser cells and was a primary mechanism in neuroepithelial cell competition.

Srsf7 depletion induces loss of cellular juvenescence through Ezh2.
To elucidate the role of Srsf7 in cell competition, we depleted Srsf7 in HypoE-N1 cells using gene-specific siRNA transfection. Knockdown efficiency was validated with quantitative PCR analysis (Fig. 5a). Srsf7 depletion significantly impaired cell growth (Fig. 5b,c). We had previously identified Srsf7 and Ezh2 as JAGs 27 . Ezh2 mRNA expression was downregulated upon Srsf7 knockdown (Fig. 5d), which suggests Srsf7 regulates Ezh2 transcription. Immunostaining showed Ezh2 expression was markedly reduced in Srsf7-depleted cells (Fig. 5e). This finding was corroborated by western blot analysis which showed Ezh2 (one of the JAGs) protein levels were suppressed in the Srsf7-depleted cells, while the expression of gamma H2AX (senescence-associated protein) was increased (Fig. 5f). To analyze how cellular juvenescence was regulated by Ezh2, we tested the consequences of Ezh2 depletion on HypoE-N1 cells. siRNA-mediated knockdown of Ezh2 led to increased expression of senescence-associated proteins P16 and gamma H2AX (Fig. 5g). To further test whether Srsf7 suppression was sufficient to induce cell competition, we co-cultured Srsf7-knocked down in normal HypoE-N1 cells with normal HypoE-N1 cells. We found the number of Srsf7-knocked down cells was reduced compared to control siRNA transfected cells. Furthermore, we observed a significant increase of caspase-3 positive in Srsf7-knocked down cells ( Supplementary Fig. S3).
These findings indicate Srsf7 depletion in loser RAS V12 cells induces loss of cellular juvenescence through Ezh2 suppression and is sufficient to induce cell competition.
Neuroepithelial cell competition removes loser cell corpora via phagocytosis. The phagocytic elimination of cells undergoing apoptosis is a conserved mechanism among multicellular organisms. Defects in phagocytosis processes can lead to structural and functional errors in morphogenesis and tissue homeostasis 35 .
We therefore sought to identify the removal of apoptosis-undergoing RAS V12 cells in the competitive co-culture.
In competitive co-culture, we found RAS V12 cells had been engulfed by normal cells (Fig. 6a). To confirm that RAS V12 cells were phagocytized by normal cells, we labeled the phagosome with CD68 antibody. We found RAS V12 cells were co-localized with CD68 in normal cells which suggested active phagocytosis (Fig. 6b). Next, we quantified phagocytosis by counting co-localization of RAS V12 cells with CD68, and our results showed phagocytosis was being promoted in competitive co-culture (Fig. 6c). To demonstrate that HypoE-N1 cells possess phagocytic capacity, we used a FITC-dextran assay and a phagocytosis assay of cells induced to undergo apoptosis by ultraviolet (UV) exposure. In the cells incubated with FITC-dextran, we observed uptake of FITCdextran into the cells by positive green signals (Fig. 6d). Flow-cytometric analysis of cells incubated with FITCdextran at 37 °C showed positive signal (Fig. 6e,f). In the phagocytosis assay, cultures of HypoE-N1 cells were stained with red-conjugated phalloidin, a cytoskeletal marker, before co-incubating with UV-induced apop- www.nature.com/scientificreports/ totic RAS V12 cells. Confocal microscopy showed cytoplasmic localization of RAS V12 in normal HypoE-N1 cells (Fig. 6g). Analyses of z-stack images revealed that RAS V12 cells were phagocytized by normal HypoE-N1 cells, and orthogonal projection view confirmed that RAS V12 apoptotic cells were internalized (Fig. 6h). These observations indicate HypoE-N1 cells have phagocytic capacity, and they can remove apoptotic RAS V12 cells from competitive co-culture through phagocytosis.

Discussion
Cell competition is an emerging field for developmental biology. The mechanistic relevance of cell competition addresses unresolved questions during development, for example, organ size and quality control mechanisms to eliminate unfit cells 36 . Despite increasing numbers of cell competition models for development in different organs, there had not been a valid model for cell competition in developing brain. Findings of endogenous cell competition occurring in vivo during epiblast formation 1,37 had prompted us to investigate the role of cell competition in early brain development. Previous studies demonstrated cell competition in cell lines grown under routine culture conditions, and these have been used as in vitro models with different cell strains including human osteosarcoma U2OS 14 , canine epithelial MDCK 20 and embryonic fibroblasts NIH3T3 38 . By utilizing HypoE-N1, a neural progenitor cells (NPCs) line, we established an in vitro model system for cell competition in the brain and explored the molecular mechanisms for loser cell elimination. During brain development, mitotic events and environmental insults lead to somatic mutations which can produce suboptimal cells that can compromise tissue integrity. Competitive interactions between cells have clear benefits in a multicellular organism 39 . RAS oncogene is activated in 30% of human cancers 40   www.nature.com/scientificreports/ discovered to be activated in glioblastoma multiforme in adults 41 . We proposed that cell competition act as a quality control mechanism in the developing brain to ensure proper tissue growth and homeostasis maintenance. Studies have reported the potential role of cell competition to eliminate RasV12-transformed cells from epithelia through cell death or apical extrusion 20,42 . RasV12-transformed cells in epithelia trigger non-cell autonomous changes in both types of cells 43 . In this study, we found RAS V12 cells demonstrated loss of cellular juvenescence which resulted in the important selective elimination of RAS V12 cells through a non-cell autonomous manner. This mechanism of cell competition maintains tissue integrity and acts as quality control by ensuring that cells with deleterious mutations such as RAS V12 are eliminated from neuroepithelia during brain development.
Our results showed that introduction of heterogeneity by induction of RAS V12 in a small proportion of the cells caused cell competition. Previous studies reported that elimination of loser cells during cell competition was through apoptosis 14,17,19 with the removal of apoptotic loser cells by phagocytosis 17 . These reports supported our data, which showed that in competitive condition, RAS V12 cells were recognized by the surrounding normal cells and eliminated by non-cell autonomous induction of apoptosis followed by phagocytosis by the neighbouring winner cells. Neural progenitor cells are known to be phagocytes, and their phagocytic activity can regulate adult neurogenesis 44 .
We showed in this study that the elimination of RAS V12 cells was context-dependent. This was evidenced with our finding in non-competitive condition of RAS V12 to RAS V12 (CMTPX) co-culture that apoptosis was not observed. Furthermore, we demonstrated neuroepithelial cell competition requires cell-cell contact and was not mediated by soluble factors. Thus, the apoptosis observed in RAS V12 cells was the result of cell competition in a non-cell autonomous manner. Currently, the molecular mechanism underlying loser cells elimination is unresolved. In Drosophila, expression of Flower gene can distinguish loser cells from winner cells by alternative splicing of the FweLose-A/B isoforms in loser cells. This isoform is necessary and sufficient for competitiondriven cell elimination 15 . Myc-driven cell competition in mouse epiblast demonstrated BMP signalling was decreased in loser cells 1 . Our study revealed for the first time that during neuroepithelial cell competition, acute loss of Srsf7 and Ezh2 in loser cells preceded their elimination. We hypothesized that the presence of senescent cells in the early developing brain was related to apoptosis, which has been widely reported. Our findings provide evidence that during early brain development the presence of senescent cells losing cellular juvenescence might activate cell competition resulting in cellular discrimination and removal from the neuroepithelia by apoptosis.
Proteasome regulation of the SR protein family is less understood. We hypothesized that the acute loss of Srsf7 in competitive condition observed in this study was linked to ubiquitylation and proteasome degradation. Our results suggested that Srsf7 underwent degradation via the ubiquitin-proteasome pathway. Besides the known function of Srsf7 as a splicing factor, SR proteins have also been identified as regulators of gene expression 33 . Through functional analyses of Srsf7, we demonstrated that it was indispensable for cell growth and survival. Srsf7 suppression induced loss of cellular juvenescence through Ezh2 suppression at both mRNA and protein levels. The loss of cellular juvenescence was further evidenced with enhancement of senescence-associated proteins P16 and gamma H2AX. Ezh2 suppression has been reported to rapidly elicit DNA damage and trigger the onset of senescence without loss of H3K27me3 45 . Akizu et al. reported that Ezh2 is required for neural progenitor proliferation and has been linked to a number of human diseases including some neurodegenerative disease 46 .
In summary, we have uncovered how cell competition as a quality control mechanism might maintain brain homeostasis. Mechanistically, cellular juvenescence in unwanted loser cells is deregulated during brain development. Further studies should determine whether this molecular mechanism was conserved in other mammalian tissues. The potential of Srsf7 and other JAGs to be drug targets should also be investigated. Cell competition could be exploited to conceive novel therapeutic interventions in neurological diseases that may boost elimination of unwanted and damaged cells.

Generation of inducible RAS V12 -EGFP in HypoE-N1 cells.
HypoE-N1 (Cedarlane, #CLU101), a neural progenitor cell line (NPCs) derived from mouse hypothalamus was used in this study. HypoE-N1 cells were cultured in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum. HypoE-N1 cells that express doxycycline-inducible RAS V12 -EGFP were established by serial transfection of pcDNA6/TR and pcDNA4/TO-RAS V12 -EGFP. The plasmids were gifts from Shunsuke Kon and Yasuyuki Fujita (Hokkaido University, Japan). To induce RAS V12 expression, 20 µg/ml of doxycycline (LKT laboratories, Inc) was used. Establishment of neuroepithelial cell competition model. For competitive co-culture, normal and RAS V12 were co-cultured at 100:1 ratio while for non-competitive condition, RAS V12 and RAS V12 labeled with CMTPX (red dye) were co-cultured at 100:1 ratio. Both co-culture conditions were established at > 95% confluency and incubated overnight before doxycycline treatment to induce the expression of RAS V12 -EGFP. We optimized the cell numbers plated in a 6-well plate to create a condition where RAS V12 cells were surrounded by normal cells (competitive condition), and RAS V12 (CMTPX) cells were surrounded by RAS V12 cells (noncompetitive condition).
Immunofluorescence staining. Co-culture cells were plated on glass cover slips. The mixture of cells was incubated overnight, followed by doxycycline treatment for 24-72 h. Cells were fixed in 4% paraformaldehyde (PFA) for 10 min at RT, permeabilized with 0.1% Triton X-100 for 2 min at RT, blocked with 2% FBS for 1 h, and incubated with primary antibodies at 4 °C overnight. After three 10-min washes in PBS, cells were incubated with secondary antibodies (anti-rabbit-IgG conjugated to Alexa-546 [Invitrogen, AQ17) or anti-mouse-IgG conjugated to Alexa-647 (Invitrogen, A21235)] at 1:1,000 for 1 h at RT. After three 10-min washes in PBS, coverslips with cells were mounted onto Prolong Gold DAPI (Invitrogen, P36971) on a glass slide. Images were captured by Olympus FV1000 or Leica, TCS SP8 confocal microscopes. ImageJ software was used for image analysis. Positive and negative cells were defined based on fluorescence intensity. Signals in negative controls were set as threshold and cells with fluorescence intensity higher than threshold were considered positive.
SYTOX blue. Cells were seeded in 6-well plates and induced with 20 µg/ml of doxycycline for 24 h. Culture media were then removed and replaced with media containing 5 nM SYTOX blue. Cells were incubated with SYTOX blue at 37 °C for 30 min in 5% CO 2 incubator. After incubation, the SYTOX Blue solution was removed and cell images captured using an EVOS fluorescence microscope. Numbers of SYTOX blue positive cells were counted in 10 areas and averaged.
Apoptosis assay. Dead cells that had undergone apoptosis were analyzed after 24 h doxycycline induction by immunofluorescence staining for active caspase-3 and cleaved PARP. Positive cells were counted on five independent 40X fields for each sample and averaged.
Time-lapse microscopy. Cells were co-cultured on glass-bottom culture dishes (Mat-Tek Corporation) and allowed to attach for 16 h before addition of doxycycline. After 8 h of doxycycline induction, cells were recorded at 10 min/frame interval for 24 h using an Olympus IX83 microscope. Cells were maintained at 37 °C, 5% CO 2 and 100% humidity during recording. Videos were analyzed by an Olympus CellSens software and processed with Adobe Photoshop and Image J.  www.nature.com/scientificreports/ FITC dextran uptake assay. HypoE-N1 cells were incubated with FITC-conjugated dextran (1 mg/ml) (Tokyo Chemical Industry) at 37 °C, or at 4 °C (negative control) for 2 h in culture medium. After washing with PBS supplemented with 1% BSA, cells were prepared for flow cytometric analysis by centrifugation at 300-400 g for 5 min (RT). Cell pellets were then collected and resuspended in PBS/BSA, followed by centrifugation at 300-400 g for 5 min (RT). Pellets were collected and diluted with PBS/BSA to 1 × 10 7 cells/ml or greater. The diluted cells were transferred into FACs tubes and analyzed by flow cytometry (FACSCalibur, BD Biosciences). HypoE-N1 cells positive for FITC fluorescent signal were considered as cells that had engulfed the dextran.
Phagocytosis assay. The phagocytosis assay was modified from Lu et al. 44 . Doxycycline (20 ug/ml) was added to HypoE-N1-RAS V12 -EGFP cell lines for 24 h to induce expression of EGFP-RAS V12 . RAS V12 cells were then dissociated and exposed to ultraviolet light for 15 min to induce apoptosis before incubating with preplated HypoE-N1 cells for 4 h. Cells were then washed with cold PBS to stop phagocytosis, fixed in 4% PFA for 10 min at RT, and stained with 50 µg/ml phalloidin-conjugated TRITC in PBS (Sigma-Aldrich, P1951, 30 min at RT) to identify actin. Nuclei were counterstained with Prolong Gold DAPI (Invitrogen, P36971) mounting agent. Images were visualized with an Olympus FV1000 laser confocal microscope under 100 × magnification and Z-stack images were generated with z-step size at 0.5 µm.
Quantitative RT-PCR. Total RNA was extracted with TRIzol reagent (Thermo Fisher Scientific). Extracted RNA was quantified using a NanoDrop Lite Spectrophotometer (Thermo Fisher Scientific), and reverse transcribed with high-capacity RNA-to-cDNA kit (Applied Biosystems) using a PCR thermal cycler Dice (Takara) according to the manufacturer's instructions. Synthesized cDNAs were amplified by qPCR using a LightCycler 480 SYBR Green I Master Kit on a LightCycler 480 instrument (Roche). Quality and specificity of the qPCR amplification was determined by a melting curve analysis. Normalization was made to mouse Tubb5 or Polr2a. Primers used for the qPCR are listed in Supplementary Table S1.
In vivo ubiquitylation assay. To analyze Srsf7 ubiquitylation, HypoE-N1 cells were transfected with a HA-Ubiquitin expressing plasmid, kindly gifted by Toshifumi Morimura (Shiga University of Medical Science). Transfected cells were treated with 5 µM MG132 (Cell Signaling, #2194) for 8 h before cell harvest to enhance ubiquitylation signals. For immunoprecipitation, Dynabeads Protein A (Invitrogen) were incubated with 2 µg Srsf7 and rabbit IgG antibody for 1 h at RT, and then cell lysates were added to the antibody-bead complexes for a further 16 h at 4 °C. Beads were washed with lysis buffer three times followed by elution with 1X Laemmli sample buffer and denaturation at 70 °C for 10 min. The samples were separated in an 8% SDS-PAGE gel followed by western blotting with HA and Srsf7 antibodies. www.nature.com/scientificreports/