Positive coactivator PC4 shows dynamic nucleolar distribution required for rDNA transcription and protein synthesis

Abstract

The nucleolus is the site of rDNA transcription and ribosome biogenesis. Alterations in nucleolar function and architecture correlate with drastic heterochromatin rearrangement and global changes in gene expression. However, the precise mechanism that connects nucleolar function to heterochromatin organization and transcription is yet unknown. Here, we report that the RNA polymerase II (RNA pol II) transactivator and chromatin condenser, Positive Coactivator 4 (PC4), is a bona fide nucleolar protein. PC4 showed dynamic nucleolar accumulation, which is critical for rDNA transcription. The lysine acetyltransferase, KAT5 (Tip60) acetylates PC4 at K35, which facilitates nucleolar release of PC4 and concomitated inhibition of rDNA transcription. By employing PC4 mutant, which is defective in nucleolar accumulation, we found that nucleolar PC4 is crucial for RNA pol I-mediated rDNA transcription. To validate this significant novel role of PC4, in the context of nucleolus organization and function, at the organismal level, we looked into B cell-specific conditional knockout of Sub1 encoding PC4 in mice, which revealed that indeed the rDNA transcription and protein synthesis in B cells are severely repressed in the absence of PC4. Furthermore, PC4 CKO B cells were associated with the loss of H3K9me3-marked heterochromatin foci but not global H3K9me3 levels. LC–MS/MS analysis of the H3K9me3 chromatin complexes revealed that most non-histone heterochromatin proteins were reduced or absent in the constitutive heterochromatin of PC4 CKO B cells. These findings establish PC4 as a critical functional component of nucleolus for rDNA transcription.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12964-025-02238-4.

Introduction

Chromatin is compartmentalized into different functional spatial domains such as the actively transcribed euchromatin and the condensed transcriptionally inactive heterochromatin. Heterochromatin drives the three-dimensional arrangement of chromatin in the cells namely the lamina-associated domains (LADs) and nucleolus associated domains (NADs), where the heterochromatin is positioned at the periphery of the nucleus or nucleolus, respectively [1]. The nucleolus is a membrane-less nuclear compartment formed around tandem ribosomal DNA (rDNA) repeats known as nucleolar organizer regions (NORs). It is the site of RNA polymerase (pol) I-mediated rDNA transcription and ribosome biogenesis, which are essential for cell viability. The heterochromatin and nucleolus, though appear to be distinct components of the nucleus, are intimately connected both structurally and functionally. Maintenance of heterochromatin is crucial for nucleolar architecture and function. Demethylation on histone H3 lysine 9 position (H3K9) resulted in R-loop formation at rDNA loci and nucleolar distortion [2]. Loss of H3K9 methylation in Drosophila due to Su(var)3–9 mutation led to nucleolar disorganization [3]. Deficiency of Suv39h1/2 and HP1α/β, which are critical for heterochromatin formation driven by H3K9 methylation, resulted in nucleolar defects and reduction in ribosomal RNA biosynthesis in mammalian cells [4]. Conversely, the inhibition of RNA pol I transcription resulted in the 3D rearrangement of heterochromatin, culminating in altered gene expression networks [5]. Depletion of nucleolar protein NPM1 triggered the rearrangement of perinucleolar heterochromatin and nucleolar deformation [6]. Overall, the disturbance of either the heterochromatin or nucleolar function seems to cause distortion or disorganization in the other, leading to changes in gene expression networks.

Positive Coactivator 4 (PC4) is a small, highly conserved, and abundant nuclear protein that has been implicated in multiple cellular processes, including transcription [7–11], DNA repair [12, 13], chromatin compaction [14–16], neurogenesis [17], and autophagy [18]. These functional diversities of PC4 are achieved through post-translational modifications, such as acetylation and phosphorylation based on the cellular signals [15, 19–21]. PC4 interacts directly with RNA pol II and III to promote their transcriptional activity and to enhance the recruitment of other transcription factors to the promoter region of target genes [10, 19, 22, 23]. By generating B cell-specific knockout of Sub1 gene encoding PC4, we previously reported that PC4 is critical for regulating lineage-specific gene regulatory networks (GRN) during B cell differentiation and the B cells devoid of PC4 exhibited compromised plasma cell differentiation [23].

Despite its essential roles in the cell, not much is known about how the diverse nuclear functions of PC4 are regulated and linked. Here we report a previously unknown function of PC4 in increasing RNA pol I transcription in the nucleolus. Loss of PC4 resulted in attenuated protein synthesis. The absence of PC4 resulted in abrogation of H3K9me3-marked heterochromatin assembly and distortion of nuclear morphology. Collectively, our data revealed an unexpected role for PC4 in the nucleolus, the disruption of which leads to disorganizations of rDNA gene expression and constitutive heterochromatin.

Results

The transcription coactivator PC4 is a nucleolar protein

In our previous study, we explored how PC4 regulated gene expression networks in mature B cells. We isolated the PC4 complex from splenic B cell nuclei to identify the PC4 interactome by LC–MS/MS. By analyzing the PC4 interactome, we identified 128 proteins as PC4-interacting proteins [23] of which approximately 27% were related to nucleolar function (Fig. 1A, Table S1), suggesting a function of PC4 at the nucleolus. Supporting this idea, a recent ATAC-Seq study suggested a role of PC4 at the rDNA loci [15]. Therefore, to further elucidate the connection between PC4 and nucleolus, we tagged PC4 with GFP to visualize the subcellular localization of PC4 in HeLa cells. We found that PC4-GFP was localized in the nucleolus in addition to nucleocytoplasm, colocalizing with the nucleolar protein NOP56 (Fig. 1B). Many nucleolar proteins are known to be dislocated upon inhibition of rDNA transcription [24] or cellular stresses like DNA damage. We observed that PC4, similarly to NOP56, diffused out of the nucleolus upon treatment with Actinomycin D (ActD) at a lower concentration, which preferentially inhibits RNA pol I [25] (Fig. 1B, S1A). We also observed that upon treatment with the DNA damage agent, Bleocin, PC4 localization at the nucleolus was greatly reduced (Fig. S1B). These observations supported the idea that PC4 is a nucleolar protein and plays a role concerning nucleolar functions and agreed with previous nucleolar proteome studies that identified PC4 in the datasets as a potential nucleolar protein [24].

Fig. 1.

The RNA pol II coactivator, PC4 is a nucleolar protein. A. A Venn Diagram depicting the proportion of nucleolar proteins present in the PC4 interactome identified previously [23]. B. Immunofluorescence microscopy to show the localization of PC4 at the nucleolus and its efflux from the nucleolus upon 0.05 μg/ml actinomycin D (ActD) treatment for 2 h in HeLa cells. PC4 tagged with GFP, NOP56, and DNA were stained in green, red, and blue, respectively in the merged images. Line plots across the yellow arrows were plotted to show the colocalization between PC4 and NOP56 nucleolar protein. The nucleolar regions are shaded in blue. Scale bar = 10 μm

Tip60-mediated acetylation on lysine 35 is crucial for PC4 dynamics

The dynamics of many nucleolar proteins are regulated by post-translational modifications, including lysine acetylation [26, 27]. A previously published acetylome study revealed that PC4 is acetylated on two lysine residues, lysine 35 (K35) and lysine 68 (K68) [28] (Fig. 2A). To test whether lysine acetylation may be involved in the nuclear dynamics of PC4, HeLa cells were treated with the HDAC inhibitor, sodium butyrate (Na-butyrate). This led to the release of PC4 from the nucleolus (Figs. 2B, C). Since the result suggested the control of PC4 nuclear dynamics by acetylation, we tried to locate the target site of the acetylation. Based on our hypothesis that acetylation within nucleolar localization signal (NoLS) might govern the PC4 nuclear dynamics, we predicted a putative NoLS in PC4 using the Nucleolar localization sequence Detector (NoD) prediction tool [29, 30] and obtained the region spanning from residue 17 to 43 as the putative NoLS (Fig. 2A). Since K35 was located within the putative nucleolus localization signal (NoLS), we speculated that the lysine acetylation on K35 may be responsible for the flux of PC4 from the nucleolus (Fig. 2D). The lysine residue at the 35th position on PC4 was substituted by arginine to create an acetylation-defective mutant (K35R). We found that the PC4-K35R mutant appeared to exit the nucleolus slower than the WT counterpart upon ActD treatment (Fig. 2D, E, S1A) and was unaffected by Na-butyrate treatment (Figs. 2D, E). These data indicated that K35 acetylation is crucial for the outward flux of PC4 from the nucleolus.

Fig. 2.

PC4 sub-nuclear dynamics is controlled by lysine acetylation on K35. A Schematic of PC4 domain architecture with the position of the putative nucleolar localization and K35 indicated. B Wildtype PC4 dynamics under 0.05 μg/ml ActD for 2 h or 5 μM sodium butyrate (Na-butyrate) treatment for 12 h. C Graphical representation of the comparison between the nucleolar fluorescence intensities of PC4 control cells (n = 127 nucleoli in 70 cells), and ActD (n = 69 nucleoli in 46 cells) and Na-butyrate (n = 102 nucleoli in 58 cells) treatment respectively. D PC4 K35R dynamics under 0.05 μg/ml ActD for 2 h and 5 μM Na-butyrate treatment for 12 h. E Graphical representation of the comparison between the nucleolar fluorescence intensities of PC4 K35R control cells (n = 112 nucleoli in 78 cells), and ActD (n = 99 nucleoli in 53 cells) and Na-butyrate (n = 95 nucleoli in 40 cells) treatment respectively. F PC4 dynamics under ActD treatment and the overexpression of Tip60 and dominant-negative mutant Tip60 (mutTip60). G Graphical representation of the comparison between the nucleolar fluorescence intensities of PC4-GFP cells under ActD treatment either expressing Tip60 (n = 78 nucleoli in 44 cells), or mutTip60 (n = 97 nucleoli in 76 cells). Representative images from N = 3 independent biological repeats for B and D, and N = 2 independent biological repeats for F. PC4 tagged with GFP, NOP56, and DNA were stained in green, red, and blue, respectively in the merged images. Scale bar = 10 μm. One-way ANOVA with Dunnett’s multiple comparison test was performed for C and E. Unpaired t-test was performed for G. Significant P values are mentioned in the figures

After establishing the importance of lysine acetylation on PC4 dynamics, we wanted to identify the lysine acetyltransferase (KAT) responsible. To achieve this, Hela cells were treated with three kinds of small molecule inhibitors targeting the three major nuclear KAT family: p300/CBP family inhibitor, C646, PCAF inhibitor, Embelin, and Tip60/Moz inhibitor, NU9056. Cells were pre-treated with the KAT inhibitor followed by ActD treatment for 2 h. Cells treated with C646 or Embelin showed a predominant nuclear PC4 localization and moved out of the nucleolus, similarly to the cells treated with ActD alone (Fig. S2A, B), suggesting that p300/CBP or PCAF is not the enzyme acetylating PC4 on K35. In contrast, when treated with the Tip60 inhibitor NU9056 and ActD, the cells exhibited prolonged retention in the nucleolus compared to the nucleolar protein NOP56 (Fig. S2A). To further confirm the role of Tip60 in PC4 localization, a dominant negative catalytic mutant of Tip60 [31] was overexpressed in HeLa cells. Compared to the cells expressing the WT Tip60, PC4 in the cells expressing the mutant Tip60 demonstrated longer nucleolar retention (Fig. 2F, G). Cells expressing PC4 K35R mutant protein, also exhibited longer nucleolar retention upon co-expression with Tip60 and ActD treatment (Figure S2C). These results together suggest that the acetylation-dependent release of PC4 from nucleolus is regulated by Tip60.

Nucleolar PC4 is important for RNA Pol I-mediated transcription

PC4 is a multifunctional protein; however, there are no known functions of PC4 in the nucleolus. Since PC4 is a known transcription coactivator, we hypothesized that PC4 may also be a coactivator for RNA pol I-mediated transcription. To dissect the functions of PC4 in the nucleolus, we created additional mutant derivatives of lysine residues, aiming to identify mutations that would inhibit its nucleolar accumulation. Among them, two double mutants of PC4 abolished nucleolar or nuclear accumulation. PC4 K23A and K24A (nucleolar localization defective mutant or NoLD) exhibited a predominant nuclear localization with clear nucleolar exclusion (Fig. 3A). PC4 K26A, R27A mutant (nuclear localization defective mutant or NLD) was exclusively cytoplasmic (Fig. S3A). To test the role of PC4 in rDNA transcription, we overexpressed PC4, NoLD, and NLD mutants, and quantified the levels of 47S pre-rRNA through qPCR. We used cells transfected with empty vector and those treated with ActD as controls (Fig. 3B). As expected, ActD abolished rDNA transcription. The cells treated with Na-butyrate showed a significant reduction in 47S pre-rRNA transcription, suggesting that the presence of PC4 in the nucleolus was important for RNA pol I-mediated transcription. Corroborating this observation, we found that PC4 overexpression increased rDNA transcription whereas the NoLD mutant reduced rDNA transcription, suggesting that the NoLD mutant acted as a dominant negative mutant. The NLD mutant did not show any effect on the levels of 47S pre-rRNA (Fig. 3B), suggesting that NoLD sequestered some PC4-interacting nucleolar proteins in nuclei. To further confirm the role of PC4 as an RNA pol I coactivator, chromatin IP was performed followed by qPCR using primers across distinct elements of the rDNA gene, namely the promoter, 18S, 5.8S, 28S, and the intergenic spacer (IGS) (Fig. 3C). For this experiment, we established cells stably expressing FLAG-PC4. As expected, anti-FLAG IP enriched the rDNA promoter, while the enrichment was greatly diminished upon RNA pol I inhibition using ActD. These results suggest that PC4 is required for efficient transcription of rDNA genes by RNA pol I. Corroborating these results, it was observed that PC4 knockdown in HeLa cells reduced global protein translation (Fig. S3B) whereas cells overexpressing WT and PC4 K35R exhibited higher translation as judged by puromycin labeling of nascent peptides (Fig. 3D), suggesting that PC4 can influence global translation through the increase of RNA pol I transcription. Unexpectedly, overexpression of PC4 NoLD mutant did not cause a clear reduction in translation.

Fig. 3.

Nucleolar PC4 is important for RNA pol I transcription and protein synthesis. A Schematic of PC4 domain architecture with the position of the putative nucleolar localization and K23,K24 and K26,R27 indicated. B Pre-rRNA qPCR results comparing WT PC4, ActD, Na-butyrate, and PC4 localization mutants. Statistical analysis performed was one-way ANOVA with Dunnett’s multiple comparisons test. N = 3. C ChIP-qPCR results showing the enrichment of WT PC4 on the rDNA loci. Statistical analysis performed was two-way ANOVA with multiple comparisons test. ****, p < 0.0001, N = 2. D Western blotting analysis of puromycin labelling assay on Hela cells overexpressing PC4, PC4 K35R, NoLD, and NLD PC4 mutants N = 2. E Schematic depicting the experimental procedure performed on mouse B cells. F Pre-rRNA qPCR results comparing WT PC4 to PC4 CKO B cells. N = 5 WT mice for and 6 PC4 CKO mice. G Western blotting analysis of puromycin labelling assay on LPS-activated WT and PC4 CKO B cells. N = 4 mice for each group. Unpaired t-test statistical test was performed

We next examined whether PC4 absence would lead to reduced protein synthesis. Because we reported previously that PC4 plays important role in B cells by generating conditional knockout (CKO) of Sub1 encoding PC4 in these cells [23], B cells WT and PC4 CKO mice (Fig. 3E). We found that the 47S pre-rRNA transcription was significantly reduced in the PC4 CKO B cells (Fig. 3F). We performed a puromycin labeling assay to determine the levels of the nascent peptide being translated. Substantiating the qPCR data, we found that there was a drastic decrease in the levels of puromycin-labeled peptides in the PC4 CKO lanes versus the WT (Fig. 3G). These results further support the notion that PC4 is direct or indirect facilitator of transcription by RNA pol I and the loss of PC4 culminates in the abrogation of protein synthesis.

Prolonged loss of PC4 results in altered chromatin function

Recent studies have highlighted the crosstalk between the nucleolus and heterochromatin. Nucleolar stress interfering with RNA pol I transcription results in reorganization of heterochromatin and altered gene expression programs [5]. In our previous study, we observed a loss of H3K9me3-marked foci in PC4 CKO B cells [23]. Together with the previous observation, the present results prompted us to further explore the potential role of PC4 in heterochromatin maintenance and nuclear function. Consistent with our previous study [23], we observed that H3K9me3 heterochromatin foci were absent in PC4 CKO B cells (Fig. S4A); however, the global levels of H3K9me3 histone mark remained unaltered (Fig. S4B). Intrigued by this, we set out to understand the fundamental differences in the H3K9me3 constitutive heterochromatin in the presence and absence of PC4. Since only the H3K9me3 foci were altered, we reasoned that the composition of the constitutive heterochromatin may be altered in the absence of PC4. To test this hypothesis, we immunoprecipitated H3K9me3-marked chromatin and subjected it to LC–MS/MS analysis (Fig. S4C). The abundance of canonical histones and linker histones were comparable in both WT and the PC4 CKO H3K9me3 IP (Fig. 4A), indicating successful immunoprecipitations of chromatin fragments from these cells. Strikingly however, histone variants, especially H2A variants known to be present in heterochromatin, such as H2A.Z, H2A.V, H2A.J, and macroH2A were reduced or absent in the H3K9me3 chromatin in PC4 CKO B cells (Fig. 4A). Among linker histone H1, H1t was also reduced.

Fig. 4.

Prolonged loss of PC4 results in altered chromatin function. A Graphical representation of the mean protein score and protein coverage of core histones, histone variants, and linker histone proteins present in the H3K9me3 IP from WT and PC4 CKO B cells. Data from 3 independent biological repeats. B Gene Ontology analysis of the protein hits found exclusively in the WT H3K9me3 IP. C Gene Ontology analysis of the common protein hits found in the PC4 CKO H3K9me3 IP. D Graphical representation of the mean protein score and protein coverage of non-histone heterochromatin proteins present in the H3K9me3 IP from WT and PC4 CKO B cells. Representative data from 2 independent biological repeats. E Electron micrographs depicting the nuclear morphology in WT and PC4 CKO B cells. F Illustration depicting the lysine acetylation-dependent subnuclear shuttling of PC4 protein which is important for RNA pol I-mediated transcription and subsequently heterochromatin maintenance

Gene Ontology (GO) analysis was performed on the interactome of H3K9me3 chromatin. For the WT interactome, proteins present exclusively in the WT complexes (74 proteins) were considered for GO analysis (Fig. S4D). As expected, chromatin silencing was the most enriched pathway. In addition, pathways associated with nucleolar function were also enriched such as “GO:0006412 ~ translation”, “GO:0000027 ~ ribosomal large subunit assembly”, “GO:0042254 ~ ribosome biogenesis”, and “GO:0000183 ~ chromatin silencing at rDNA” (Fig. 4B). This result suggested that H3K9me3 chromatin was closely located with nucleoli in WT B cells. In contrast, however, the GO analysis of the common proteins present in the PC4 CKO complex (30 proteins) were devoid of heterochromatin and nucleolar function and are enriched in pathways that relate to RNA processing and gene expression (Fig. 4C), suggesting a failure of proper gene silencing by the H3K9me3-deposited chromatin. To gain an in depth understanding of the difference between the interactome of WT and PC4 CKO H3K9me3 chromatin, we plotted the protein score and coverage of common non-histone chromatin-associated proteins. Reflecting the GO analyses it was observed that many chromatin-associated proteins or heterochromatin-associated proteins were greatly reduced or absent in the PC4 CKO H3K9me3 complex (Fig. 4D); known regulators of heterochromatin, including Cbx5 and Suv39h1, were reduced in the PC4 CKO. We found an enrichment in nucleolar proteins involved in RNA Pol I transcription (such as UBF, TCOF1, NOLC1), and ribosomal biosynthesis in the WT complex over the PC4 CKO (Fig. S4E).

Heterochromatin maintains the rigidity and morphology of the nucleus [32]; therefore, we wondered if the loss of heterochromatin foci would exhibit phenotypic changes in the PC4 CKO B cell nucleus. Validating our GO analyses data, electron microscopy observation revealed that the loss of heterochromatin foci in PC4 CKO B cells resulted in the deformation of the nuclei, including invagination of the nuclear membrane and lobular structure (Fig. 4E).

Discussion

In this manuscript, we demonstrate the interdependency between nucleolar function and heterochromatin. We show that the transcription coactivator PC4 is required for efficient transcription of rDNA genes by RNA pol I in addition to RNA pol II and III, and as an extension, is also crucial for ribosome biogenesis and protein synthesis. These observations would suggest that PC4 may play a pivotal role in deciding gene expression profiles under various internal and environmental cues, especially stress conditions like DNA damage where global protein synthesis is dampened [33, 34].

Mechanistically, we have discovered that the subnuclear localization of PC4 is significantly influenced by Tip60-mediated acetylation. This finding, along with other studies, underscores the vital importance of post-translational modifications, particularly acetylation and phosphorylation, in determining the functional segregation of PC4. Previous investigations have indicated that these modifications are frequently mutually exclusive [21], possibly hinting at the functional finetuning of this important transcriptional coactivator (Fig. 4F). The direct enzyme–substrate relationship of Tip60 and PC4 is yet to be established. It also remains unclear how acetylated PC4 is excluded from the nucleolus. Considering that acetylation of lysine neutralizes the basic residue, PC4 acetylation may suppress the electrostatic interaction of PC4 with rDNA or other components of the nucleolus. Recently, dynamic localization of histone deacetylase SIRT1 has been suggested to regulate epigenetic modification [35]. Upon DNA damage, SIRT1 is mobilized from the nucleolus to DNA damage sites for DNA repair. However, this response is considered to abrogate the acetylation status of rDNA chromatin, leading to deregulation of nucleolar heterochromatin and hence accelerated aging. It is an interesting possibility that the dynamic regulation of nucleolar PC4 may reflect a new mode of nucleolar-nuclear signaling.

Our results suggest that PC4 stimulates rDNA gene transcription by RNA pol I by binding to the promoter region. Initially recognized as a coactivator for RNA pol II, PC4 was found to engage with activators and components of the pol II basal transcription apparatus [7, 36–38]. It enhances activator-dependent transcription, inducing transcription initiation in a phosphorylation-dependent manner [38]. Moreover, PC4 facilitates RNA pol III recycling and is enriched on RNA pol III gene promoters [9, 10]. As an extension, PC4 may activate rDNA transcription by RNA pol I through the interaction with RNA pol I transcription machinery. We have found TCOF1, a protein important for RNA pol I transcription initiation, present in the PC4 interactome analysis (Fig. 1A and Table S1). In addition, we also observe an enrichment of RNA Pol I transcription coactivators such as TCOF1, UBF, and NOLC1, in the WT H3K9me3 complex from B cells over the PC4 CKO complex (Fig. S4E). However, it remains to be examined whether PC4 functions as a coactivator of RNA pol I by interacting with pol I itself or its basal transcription factors. Therefore, it will be important to test whether PC4 can regulate RNA Pol I-mediated transcription through the stabilization of the transcription initiation complex.

The observed alteration in the composition of constitutive heterochromatin in PC4 CKO B cells is consistent with the well-established association between the nucleolus and heterochromatin functionality [4–6]. The extended absence of PC4 led to the disappearance of heterochromatin foci (Fig. S4A). Importantly, many of the proteins important for heterochromatin-mediated gene silencing were lost from chromatin regions with H3K9me3 in the absence of PC4 (Fig. 4D). Considering its function in chromatin compaction [14, 15], PC4 loss on heterochromatin may directly impact the protein interactions on chromatin and gene silencing. However, it may potentially be due to the disruption of nucleolar function. Proteins involved in RNA pol I-mediated transcription were reduced in the PC4 CKO complex, possibly indicative of the reduced and altered nucleolar function. While we cannot dismiss the direct impact of PC4 loss on heterochromatin, it is important to consider that the dysfunction of nucleoli in the absence of PC4 could also have an impact on the overall heterochromatin landscape. The phenotypic changes PC4-deficient B cells [23] may reflect not only altered heterochromatin but also the defective nucleolar function which leads to a severe reduction in the protein synthesis capacity.

Materials and methods

Mice

B1-8 hi:Mb1-cre:Sub1f/f mice were generated as described previously [23] and maintained in pathogen-free conditions in accordance with guidelines approved by the Institute for Animal Experimentation Committee at Tohoku University Graduate School of Medicine. The experiments were performed using age-matched mice between 8–16 weeks of age.

Cell culture

HeLa cells were cultured in Dulbecco's Modified Eagle’s Medium (D5796, DMEM-high glucose, Sigma-Aldrich) supplemented with 10% fetal bovine serum (FBS) (172012, Sigma-Aldrich), and 100 U/ml penicillin-100 μg/ml streptomycin (15140–122, penicillin–streptomycin, Thermo Fisher Scientific).

B cell isolation and in vitro activation

Splenic B cells were isolated using a B cell isolation kit (130–090–862, Miltenyi Biotec) and LS columns (130–042–401, Miltenyi Biotec). For in vitro B cell activation, purified splenic B cells were cultured in RPMI-1640 (Sigma) media supplemented with 10% FBS, 10 mM HEPES, 1 mM Na-Pyruvate, 0.1 mM non-essential amino acids, 100 mg/mL penicillin, 10 U/mL streptomycin, 50 mM β-mercaptoethanol, and 20 ug/mL LPS.

Plasmids

MSCV-FLAG-PC4-IRES-GFP was described previously [23]. The PC4 construct was subcloned into the EGFP-N1 vector using the XhoI/EcoRI sites. The mutations were introduced into the PC4 sequence through the site-directed mutagenesis method by amplifying the plasmid with primers containing the desired mutagenized sequences, followed by DpnI restriction digestion and transformation into DH5⍺ competent E. coli cells. The incorporated mutations were confirmed by Sanger’s sequencing. Primers used are listed in Table S1.

Antibodies and chemicals

The antibodies used in this study are: anti-PC4 (sc-166280, Santa Cruz), anti-actin (GTX-109639; GeneTex), anti-puromycin (clone 12D10, MABE343, Merck), anti-H3K9 me3 (homemade), anti-H3K9 me3 (EPR16601, Abcam), anti-NOP56 (GTX130973, GeneTex), anti-DDDDK-tag (M185-3L, MBL), HRP linked anti-mouse IgG (NA931, GE Healthcare) and HRP-linked anti-rabbit IgG (NA934, GE Healthcare) antibodies. The reagents and chemicals used in this study are as follows: puromycin (P8833, Puromycin dihydrochloride, Sigma-Aldrich), embelin (E1406, Sigma), NU9056 (ab255734, Abcam), C646 (382113, Merck) and Lipopolysaccharides (LPS) from Escherichia coli O111:B4 (L2630, Sigma-Aldrich).

RNA interference

Knockdown was performed using Lipofectamine® RNAiMAX transfection reagent (13778150; Thermo Fisher Scientific) according to the manufacturer’s instructions. siRNA Stealth RNAi duplexes against SUB1 (PC4) or control were purchased from Thermo Fisher Scientific.

siRNAs used in this study:

• siControl: 5’- AGGAGGAUAUUACCAUGAAGAAGAC-3’

• siPC4-1: 5’-UGAGGUACGUUAGUGUUCGCGAUUU-3’

• siPC4-2: 5’-GCUGAAGGAACAGAUUUCUGACAUU-3’

• siPC4-3: 5’-GGUGAGACUUCGAGAGCCCUGUCAU-3’

RT-qPCR

Total RNA was extracted using an RNeasy Plus Mini kit (74134, Qiagen) following the manufacturer’s instructions. cDNA was synthesized from 1 µg of total RNA using the High-Capacity cDNA Reverse Transcription Kit (4374966, Thermo Fisher Scientific) with random primers. Quantitative PCR (qPCR) was performed using the hot start ‘FastStart Essential DNA Green Master (06924204001, Roche) in a LightCycler® 96 instrument (Roche). Primers used are listed in Table S2.

PC4 stable expression cell lines

Hela cells were transfected with either the pEGP-N1, pEGFP-N1-PC4, or mutant PC4 plasmid using GeneJuice® Transfection Reagent (70967, Merck) following the manufacturer’s protocol. After 24 h, cells were selected in 400 μg/ml G418 (509290, Calbiochem). The surviving cell population was grown and maintained under selection.

Puromycin assay

HeLa cells stably expressing empty vector, WT PC4, or mutant PC4 proteins were grown to 70 to 80% confluency and pulse-labeled with 20 μg/ml puromycin for 30 min at 37 °C. Isolated splenic B cells were activated for 24 h in 20 μg/ml LPS, as described earlier, and pulse-labeled with 20 μg/ml puromycin for 30 min at 37 °C. Cells were lysed in lysis buffer composed of 50 mM Tris–HCl pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% NP-40, 0.5% Na-deoxycholate, 0.1% SDS, and protease inhibitor cocktail (04693132001, Roche). The lysates were separated on a 12% SDS-PAGE. The gel was cut at the 25 kDa marker, the upper part was transferred onto a 0.45-μm PVDF membrane as described elsewhere [39], and the lower portion of the gel was stained in Coomassie Brilliant Blue as the loading control. Intensity measurements were performed using the ImageJ software (https://imagej.nih.gov/ij/index.html).

Immunofluorescence microscopy

HeLa cells were grown on a coverslip for 12–24 h before transfection with pEGFP-N1 empty vector, WT PC4, or mutant PC4 expressing plasmids using GeneJuice® Transfection Reagent (70967, Merck). After 24 h, cells were fixed in 4% paraformaldehyde, permeabilized with 0.5% Triton X-100, and blocked with 5% FBS at 37 °C. The samples were incubated with primary antibodies: anti-H3K9 me3 (EPR16601, Abcam) or anti-NOP56 (GTX130973, GeneTex) antibody diluted in 1% FBS at 37 °C for 1 h, and then with a Goat anti-Rabbit IgG H&L (Alexa Fluor® 594) secondary antibody (ab150080, Abcam). Nuclei were counterstained with Hoechst. The coverslips were mounted on VECTASHIELD® Antifade Mounting Medium (H-1000–10, Vector Laboratories). The samples were visualized under the Zeiss LSM780, confocal microscope system. Data obtained were analyzed using the ImageJ software (https://imagej.nih.gov/ij/index.html).

Chromatin immunoprecipitation (ChIP)-qPCR

ChIP was performed as described previously [40]. Chromatin from splenic B cells was isolated and sonicated to obtain DNA fragments ranging from 100 to 500 base pairs. Chromatin fragments were immunoprecipitated with anti-DDDDK-tag mAb-Magnetic Beads (M185-11R, MBL). After crosslink reversal, the enriched DNA sequences were assessed by real-time quantitative PCR. Primers used are listed in Table S2.

H3K9me3 Complex purification and LC–MS/MS analysis

B1-8 hi splenic B cells were crosslinked with 1% formaldehyde. Nuclear extracts were prepared following the ChIP protocol. After pre-clearing with protein A/G beads (Invitrogen) at 4 °C for 1 h, samples were incubated with anti-H3K9 me3 conjugated with protein A/G beads overnight. Beads were washed three times and eluted with SDS sample buffer containing 120 mM Tris (pH 6.8), 4% SDS, and 0.2 M DTT. The complex was subjected to SDS-PAGE on 5%–20% gradient gel (HOG-0520–13; Oriental Instruments Co, Ltd), and stained with Coomassie Brilliant Blue. The immobilized samples were excised into gel pieces and subjected to in-gel reduction using 10 mM DTT in 25 mM ammonium bicarbonate. Alkylation of the cysteine residues was performed with 55 mM acrylamide in 25 mM ammonium bicarbonate, and digestion was carried out with trypsin (V5280, Trypsin Gold, Promega). The tryptic peptides were analyzed in an Orbitrap Fusion mass spectrometer connected with an Easy-nLC 1000 HPLC (Thermo Fisher Scientific) and the MASCOT search engine (Matrix Science) as previously described [41]. Components of the H3K9me3 complex were determined with a ≥ fivefold protein score of test IP over control IgG, and ≥ 2 unique peptides. Three independent biological repeats of the H3K9me3 complex IP were performed.

Gene ontology analysis

GO and KEGG analysis were performed using DAVID v2021 online software (https://david.ncifcrf.gov/home.jsp). Enriched pathways with FDR ≤ 0.05 were plotted. Venn Diagram was plotted using Venny2.1.0 online software (https://bioinfogp.cnb.csic.es/tools/venny/). Two biological repeats each, from WT and PC4 CKO B cell samples, of the H3K9me3 complex IP with the highest number of protein overlaps were used for plotting the Venn diagram and the subsequent analyses.

Statistics and data reproducibility

The statistical test performed between two samples is an unpaired t-test. For comparison between multiple samples, one-way or two-way ANOVA with Dunnett’s multiple comparison test were used. Graphical representation and statistical analyses have been performed in GraphPad Prism 9.3.1. The number of independent biological repeats are mentioned in the figure legends.

Authors’ contributions

S.K. conceptualized the research, carried out the main experiments, analyzed and interpreted data, prepared figures and wrote the manuscript. K.O. conceptualized the research, generated, maintained and providePC4-deficient mice and guided and supported experiments. H.S. carried out proteomics analyses. M.M. carried out bioinformatics analyses of omics data. M.A. supported protein synthesis assay. T.I. provided Tip60 derivatives and experimental conditions related to them. T.K.K. generated PC4 floxed mice, conceptualized the research, provided unpublished findings on PC4 to tune the research, and edited the manuscript. K.I. conceptualized, guided and organized the research, and edited the manuscript and figures. All authors reviewed the manuscript.

Funding

This work was supported in part by Grants-in-Aid from the Japan Society for the Promotion of Science 25H01020, 22H00443, 20 KK0176, and 18H04021 (to K. I.), 16H01295 and 23 K27362 (K.O.), 23K24156 (M.M.), 24K03083 (T.I.) and a Research Grant in the Natural Sciences from the Mitsubishi Foundation (to K. I.). S. K. was supported by fellowship from Takeda Science Foundation.

Data availability

Data on proteomics measurement using LC–MS/MS is available from the corresponding author upon request.

Declarations

Competing interests

The authors declare no competing interests.