Citrullination is the post-translational conversion of an arginine residue within a protein to the non-coded amino acid citrulline1. This modification leads to the loss of a positive charge and reduction in hydrogen-bonding ability. It is carried out by a small family of tissue-specific vertebrate enzymes called peptidylarginine deiminases (PADIs)2 and is associated with the development of diverse pathological states such as autoimmunity, cancer, neurodegenerative disorders, prion diseases and thrombosis2,3. Nevertheless, the physiological functions of citrullination remain ill-defined, although citrullination of core histones has been linked to transcriptional regulation and the DNA damage response4,5,6,7,8. PADI4 (also called PAD4 or PADV), the only PADI with a nuclear localization signal9, was previously shown to act in myeloid cells where it mediates profound chromatin decondensation during the innate immune response to infection10. Here we show that the expression and enzymatic activity of Padi4 are also induced under conditions of ground-state pluripotency and during reprogramming in mouse. Padi4 is part of the pluripotency transcriptional network, binding to regulatory elements of key stem-cell genes and activating their expression. Its inhibition lowers the percentage of pluripotent cells in the early mouse embryo and significantly reduces reprogramming efficiency. Using an unbiased proteomic approach we identify linker histone H1 variants, which are involved in the generation of compact chromatin11, as novel PADI4 substrates. Citrullination of a single arginine residue within the DNA-binding site of H1 results in its displacement from chromatin and global chromatin decondensation. Together, these results uncover a role for citrullination in the regulation of pluripotency and provide new mechanistic insights into how citrullination regulates chromatin compaction.
Pluripotent cells have the capacity to self-renew and differentiate into all somatic and germ-cell lineages and, hence, possess therapeutic potential for a multitude of medical conditions. Their generation by reprogramming of differentiated somatic cells has been achieved by nuclear transfer, cell fusion and transduction of transcription factors, such as Oct4 (also called Pou5f1), Sox2, Klf4 and c-Myc12. Pluripotent cells have a distinctly open chromatin structure that is essential for unrestricted developmental potential13,14, and reprogramming involves an almost complete epigenetic resetting of somatic cells13. The ability of PADI4-mediated histone citrullination to induce chromatin decondensation in neutrophils10 prompted us to ask whether it can have a role in pluripotency, where chromatin decondensation is also necessary. To investigate this we first assessed the expression of Padi4 in the mouse embryonic stem-cell line ES Oct4-GIP (ES), the mouse neural stem-cell line NSO4G and in induced pluripotent stem (iPS) cells derived from NSO4G (see Methods). Padi4 is expressed in pluripotent ES and iPS cells but not multipotent neural stem cells (Fig. 1a). Culture of ES cells in 2i/LIF medium establishes a ground state of pluripotency15. This leads to the downregulation of lineage-specific markers and the upregulation of pluripotency factors, as well as rapid induction of Padi4 expression (Fig. 1b). The pattern of Padi4 expression follows closely that of Nanog, an essential transcription factor for the transition to ground-state pluripotency16 (Fig. 1a, b). Whereas other Padi genes are expressed in pluripotent cells, Padi4 is the only one for which expression clearly associates with naive pluripotency (Extended Data Fig. 1a, b). Citrullination of histone H3 (H3Cit), a modification shown previously to be carried out specifically by PADI4 (ref. 17), is detectable in ES and iPS cells (Extended Data Fig. 1c), indicating that Padi4 is also enzymatically active. H3 and global citrullination are undetectable in NSO4G cells (Extended Data Fig. 1c, d).
To determine the kinetics of Padi4 activation during the establishment of pluripotency, we examined RNA and protein samples collected daily during the course of reprogramming of NSO4G into iPS cells16. Padi4 is rapidly induced in NSO4G after transduction of reprogramming factors but only becomes active to citrullinate H3 after introduction of 2i/LIF, closely following the onset of Nanog expression (Fig. 1c). These observations strongly suggested that Padi4 activity is associated with ground-state pluripotency and prompted us to examine whether Padi4 is part of the pluripotency transcriptional network.
First, we asked whether the reprogramming factors regulate Padi4 expression, using the ZHBTc4.1 and 2TS22C cell lines where Oct4 and Sox2, respectively, can be deleted acutely in response to doxycycline treatment (see Methods). Deletion of Oct4, but not Sox2, led to a decrease in Padi4 messenger RNA levels (Extended Data Fig. 1e). Furthermore, whereas Oct4 and Klf4 occupy the Padi4 promoter in ES, but not NSO4G, cells, Sox2 is bound in both cell types (Extended Data Fig. 1f). To understand the effects of Padi4 on transcriptional regulation in pluripotent cells, we analysed the transcriptome of ES cells upon overexpression of human PADI4 and inhibition of endogenous Padi4. Several key pluripotency genes are upregulated in response to PADI4 overexpression (Fig. 1d, Extended Data Fig. 2a and Supplementary Table 2), including Klf2, Tcl1, Tcfap2c and Kit. Tcl1 was previously identified as the only regulator of self-renewal upregulated in ground-state pluripotency15, and overexpression of Tcl1 or Tcfap2c positively influences reprogramming18. Gene Ontology (GO) analysis of this data set indicates an enrichment of genes involved in stem-cell development and maintenance (Fig. 1e). In addition, knockdown of Padi4 in mouse ES cells leads to decreased expression of Tcl1 and Nanog, which is rescued by exogenous expression of RNA interference (RNAi)-resistant human PADI4 (Fig. 1f and Extended Data Fig. 2b, c). These genes are under the control of PADI4 enzymatic activity, as treatment with the chemical inhibitor Cl-amidine, which disrupts citrullination by PADI4 (ref. 19), downregulates their expression (Fig. 1g). Chromatin immunoprecipitation (ChIP) analysis indicated that H3Cit is present on regulatory regions of Tcl1 and Nanog in ES and iPS cells, but not NSO4G (Fig. 1h and Extended Data Fig. 3a). Accordingly, exogenously expressed human PADI4 localizes to and is enzymatically active on these regions, as well as regulatory regions of Klf2 and Kit in ES cells (Extended Data Fig. 3b, c). In contrast to PADI4 overexpression, treatment of ES cells with Cl-amidine led to upregulation of differentiation markers such as Prickle1, Epha1 and Wnt8a and downregulation of pluripotency markers such as Klf5 (Extended Data Fig. 4a, b and Supplementary Table 3), in addition to Nanog and Tcl1 (Fig. 1g). GO analysis of this data set indicated enrichment in genes involved in cell differentiation (Extended Data Fig. 4c). Pou5f1, Klf4, Sox2 and c-Myc were not affected by PADI4 modulation (Extended Data Fig. 2a and Supplementary Tables 2 and 3). Cumulatively, the above results place Padi4 within the pluripotency transcriptional network, indicating that it acts downstream of some of the cardinal reprogramming factors to regulate a specific subset of pluripotency genes.
Prompted by the above observations, we investigated whether Padi4 is necessary for pluripotency, as assessed during reprogramming (Extended Data Fig. 5a) and in the early stages of embryo development. NSO4G cells express a green fluorescent protein (GFP) reporter under the control of the Oct4 regulatory sequences, which is activated on acquisition of pluripotency16, allowing us to trace reprogrammed cells. Knockdown of Padi4 in NSO4G-derived pre-iPS cells impaired the ability of the cells to establish H3Cit upon switch to 2i/LIF medium and led to a clear reduction in reprogramming (Fig. 2a, Extended Data Fig. 5b–e and Supplementary Video 1). Consistent with this finding, levels of Tcl1 and Nanog were not elevated on reprogramming to the same extent as in control cells (Fig. 2b and Extended Data Fig. 5f). Cl-amidine treatment led to a marked reduction of reprogramming efficiency and H3Cit (Fig. 2c and Extended Data Fig. 5g–i), indicating that the catalytic activity of Padi4 is important for the induction of pluripotency.
Padi4 expression and H3Cit are detected in the early embryo20,21 and Padi4-null mice are born in lower numbers than would be expected by Mendelian inheritance22, indicating that Padi4 loss affects embryonic development. To assess the role of Padi4 in early development, we cultured mouse embryos in Cl-amidine-containing medium from the 2-cell stage and throughout pre-implantation development (see Methods and Extended Data Fig. 6a–c). Using 200 μM Cl-amidine resulted in a complete developmental arrest of the embryos at the 8-cell stage (Extended Data Fig. 6a). We therefore used the maximum dose of Cl-amidine that reduced H3Cit (Extended Data Fig. 6b, c) but did not induce arrest (10 μM). This led to a reduced percentage of pluripotent Nanog-positive epiblast cells and an increased percentage of differentiated trophectoderm cells at the blastocyst stage (Fig. 2d, e and Extended Data Fig. 6d, e). Time-course analyses of the cleavage patterns and cell-fate decisions in early embryos showed that Cl-amidine increased the number of symmetric cell divisions at the expense of asymmetrical divisions at the 8–16- and 16–32-cell transitions (Fig. 2f, g). This resulted in 16-cell-stage embryos with fewer inner cells (destined for pluripotency) and greater numbers of outer cells (destined for differentiation into trophectoderm; reviewed in ref. 23) (Extended Data Fig. 6f). Treatment with another Padi4 inhibitor, Thr-Asp-F-amidine (TDFA)24, but not the HDAC inhibitor trichostatin A (TSA), had similar effects (Extended Data Figs 7 and 8). These results indicate that Padi4 activity also promotes the maintenance of pluripotent cells in the early mouse embryo.
To elucidate the molecular mechanisms by which Padi4 regulates pluripotency, we aimed to identify PADI4 substrates in the chromatin fraction of mouse ES cells using stable isotope labelling of amino acids in cell culture (SILAC) (Fig. 3a and Extended Data Fig. 9a, b). Among the identified PADI4 substrates were Atrx, Dnmt3b, Trim28 and variants of linker histone H1 (Fig. 3b–e, Extended Data Figs 9c–f and 10a, b, and Supplementary Table 4), all of which can have an impact on pluripotency. Histone H1 stabilizes the nucleosome and facilitates chromatin condensation, a state that is less permissive to processes that require access to the DNA, such as transcription11. The identified citrullinated H1 peptides correspond to, and are common between, variants H1.2, H1.3 and H1.4 (Fig. 3d, e), whereas an additional peptide corresponds to the same residue in H1.5 (Extended Data Fig. 10a, b). In ES cells, H1.2, H1.3 and H1.4 are required for chromatin compaction25, whereas their depletion leads to increased expression of pluripotency genes such as Nanog and stalls them in a self-renewal state with impaired differentiation capability26. Notably, H1 is more loosely bound to chromatin in ES cells than in differentiated cells14 and its genomic localization in cancer cells was shown to anti-correlate with that of PADI4 (ref. 7). Mass spectrometric analysis accounted for all arginine residues within H1 but indicated that Arg 54 is the only site citrullinated by PADI4 (Fig. 3e and Extended Data Fig. 9c). Indeed, we found that although H1.2 is citrullinated in ES cells (by endogenous Padi4, and significantly increased on PADI4 overexpression), it is refractory to modification when Arg 54 is mutated (Fig. 3f). Similar results were obtained in in vitro citrullination assays (Extended Data Fig. 10c). H1 Arg 54 lies within the globular domain of H1 (Extended Data Fig. 10d), which is highly conserved among the linker histone family and is necessary for interaction with nucleosomal DNA27,28. To test whether H1 Arg 54 is necessary for binding of H1 to nucleosomes, we mutated and assessed it in nucleosome-binding assays. Figure 3g shows that an R54A mutant, which mimics the charge change that accompanies citrullination, is impaired for nucleosome binding. An R54K mutant, which retains the positive charge, is impaired to a lesser extent (Fig. 3g), indicating that H1R54 is important for electrostatic interactions between H1.2 and the nucleosome. The above results open up the possibility that PADI4 may affect chromatin compaction in pluripotent cells. To test this hypothesis, we first assessed whether citrullination by ectopic PADI4 can lead to decondensation of differentiated cell chromatin. Recombinant PADI4 was added to permeabilized and stabilized differentiated C2C12 mouse myoblast nuclei (Fig. 4a). This protocol ensures stabilization of the nuclear component while allowing the free diffusion of non-chromatin-bound nuclear proteins into the extranuclear fraction, and their collection by washing. Incubation with active PADI4 (Extended Data Fig. 10e, f) leads to the eviction of H1 from the chromatin and its diffusion out of the permeabilized nucleus (Fig. 4b). The evicted H1 is citrullinated on Arg 54, as determined by mass spectrometry (Extended Data Fig. 10g, h). Consistent with this, PADI4-treated cells showed evidence of decondensed chromatin, as determined by nuclear swelling, diffuse 4′,6-diamidino-2-phenylindole (DAPI) staining and increased sensitivity to micrococcal nuclease (Fig. 4c, d and Extended Data Fig. 10i). Similar results were observed when PADI4 was overexpressed in C2C12 cells (Fig. 4e) or NSO4G cells (data not shown). To monitor whether Padi4 can affect H1 binding on pluripotent cell chromatin, we performed ChIP-qPCR (quantitative PCR) analyses of H1.2 on the regulatory regions of Tcl1 and Nanog and found that it is stabilized on Padi4 knockdown (Fig. 4f). The ability of Padi4 to disrupt the binding of H1 to nucleosomal DNA provides a novel mechanistic example of how citrullination regulates protein function and chromatin condensation.
The work described here identifies citrullination of chromatin components by Padi4 as a feature of pluripotency (Fig. 4g), in addition to its previously described role in the myeloid lineage. One of the reasons for the restricted expression pattern of Padi4 may be the requirement for an open chromatin state in these cell types. The selective expression characteristics and the inducible nature of the catalytic activity of Padi4 indicate that it is under tight spatial and temporal regulation, giving it a unique status among chromatin-modifying enzymes. As such, inappropriate Padi4 activity may have deleterious consequences, which may explain its activation in cancers of varying origin during progression to malignancy29. Indeed, citrullination is a common feature of several unrelated diseases, suggesting that strict regulation is probably a requirement for the physiological function of all PADIs. During review of this manuscript, a study suggested that PADI2, thought to be mainly cytoplasmic, can also citrullinate histones and lead to transcriptional activation30. This opens up the possibility that other PADIs may mediate nuclear events in specific contexts, including in pluripotent cells. Further research into the function and targets of PADIs is likely to illuminate the aetiology of several pathologies.
Information regarding cell lines, antibodies, plasmids and chemical inhibitors used in this study, as well as detailed protocols for reprogramming and cell culture, overexpression and knockdown, gene expression analysis (qRT–PCR and microarray), chromatin immunoprecipitation, embryo collection, culture, immunofluorescence staining and analysis, immunoblot analysis, SILAC labelling, mass spectrometry and identification of citrullinated peptides and proteins, native purification of wild-type and mutant linker histone H1.2, H1–GFP and nucleosome pull-down assays, purification of recombinant PADI4–GST and in vitro deimination assays, treatment of permeabilized cells with recombinant PADI4, and micrococcal nuclease digestion are provided in the Methods.
NSO4G neural stem cells31 were cultured in RHB-A media (Stem Cell Sciences), supplemented with penicillin/streptomycin (Life Technologies) and 10 ng ml−1 bFGF and EGF (PeproTech). ES Oct4–GIP cells32 and E14 ES cells were cultured in GMEM supplemented with 10% fetal calf serum (FCS) for ES cells (Biosera), 0.1 mM non-essential amino acids, penicillin/streptomycin, 2 mM l-glutamine, 1 mM sodium pyruvate, 0.1 mM β-mercaptoethanol and 106 units l−1 leukaemia inhibitory factor (LIF) (ESGRO, Millipore), or in 2i/LIF media, based on GMEM and containing 10% knockout serum replacement (Life Technologies), 1% FCS for ES cells (Biosera), 0.1 mM non-essential amino acids, penicillin/streptomycin, l-glutamine, sodium pyruvate, 0.1 mM β-mercaptoethanol, 1 μM PD0325901 (AxonMedChem), 3 μM CHIR99021 (AxonMedChem) and 106 units l−1 LIF (ESGRO, Millipore). Plat-E packaging cells were grown in DMEM media (Life Technologies) supplemented with 10% FCS, 1 μg ml−1 puromycin, 10 μg ml−1 blasticidin and penicillin/streptomycin. iPS cells were maintained in 2i/LIF media. 1 μg ml−1 puromycin was added to iPS and ES Oct4–GIP cultures during expansion. ZHBTc4.1 (ref. 33) and 2TS22C (ref. 34) ES cell lines were expanded in ES cell media and treated with 1 μg ml−1 doxycycline for 48 h before RNA extraction and qRT–PCR analysis. Complete knockdown of Oct4 in the ZHBTc4.1 cell line and of Sox2 in the 2TS22C ES cell line was confirmed by western blot analysis (data not shown). All cells were grown at 37 °C with 5% (Biostation) or 7.5% CO2.
Overexpression of human PADI4 or shRNA against mouse Padi4 in mouse ES cells
Human PADI4 was inserted into ES E14 cells using the piggyBac transposon system35. The Gateway system was used to clone human PADI4 into the piggyBac vector using the following primers: PADI4_AttB1_F, 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTTCACCATGGCCCAGGGGACATTGATCCG-3′; PADI4_AttB2_R, 5′-GGGGACCACTTTGTACAAGAAAGCTGGGTCTCAGGGCACCATGTTCCACC-3′.
pB-CAG-Ctrl or pB-CAG-hPADI4 vectors (1 μg) were transfected with piggyBac transposase (pPBase) expression vector, pCAGPBase (2 μg) , by nucleofection according to the manufacturer’s instructions (Lonza). ES E14 cells constitutively expressing the hygromycin resistance gene and human PADI4 were selected and expanded in media containing 200 μg ml−1 hygromycin.
For mouse Padi4 knockdown experiments, ES E14 cells were transfected with Lipofectamine 2000 (Life Technologies) or by nucleofection with pRFP-C-RS HuSH shRNA RFP vectors (Origene) containing either the scrambled shRNA cassette TR30015 (ctrl), mouse Padi4 targeting shRNAs FI516326 (shRNA 1) or FI516328 (shRNA 2), or with Mission RNAi pLKO.1-puro vectors (Sigma), containing either the non-targeting shRNA SHC002 (ctrl) or mouse Padi4 targeting shRNA TRCN000101833 (shRNA 3). Where applicable, ES cell lines were generated after selection with 1 μg ml−1 puromycin.
Reprogramming was performed as described previously36,37. For retroviral supernatant preparation, 9 μg of pMXs-Oct4, pMX-Klf4 and pMXs-c-Myc were transfected with FuGENE 6 into 1–2 × 106 Plat-E cells in separate 10 cm dishes. After 24 h incubation, the media was replaced with DMEM + 10% FCS and penicillin/streptomycin. Virus-containing supernatants from Plat-E cultures were filtered through a 0.22-μm cellulose acetate filter, mixed in equal ratios, and 4 μg ml−1 polybrene was added. 2 ml of the final viral mix was then added to previously plated 1.2 × 105 NSO4G cells, in 6-well plates coated with gelatin. After 1 day, the media was replaced with NSO4G cell culture medium. After 3 days incubation, the media was changed to ES cell serum-containing media (see above). At this stage, pre-iPS colony formation was evident, but none of the colonies was positive for GFP expression (GFP+). After 2 days and to complete reprogramming, medium was replaced with 2i/LIF. Cells were maintained in 2i/LIF for 8 days, with media change every 2 days. Oct4–GFP+ colonies were counted at day 7 in 5–9 selected fields per well, either at the microscope or after time-lapse image acquisition on the Biostation CT (Nikon). The percentage of GFP+ cells was determined by flow cytometry at day 8, using a Dako Cytomation CyAN ADP high-performance cytometer and Summit software, as described previously36,37. Statistical significance was determined by two-tailed unpaired t-test. For the time course experiments, cells were collected with 350 μl RLT buffer (Qiagen) or 2× Laemmli buffer, from individual wells in consecutive days and after a PBS wash.
Cl-amidine (200 μM) was added at the time of media exchange to 2i/LIF and replenished every 2 days. For Padi4 knockdown experiments, pre-iPS cells were maintained in ES-cell-serum-containing media and cell lines were generated after transfection with lipofectamine 2000 with the Mission RNAi pLKO.1-puro vectors, containing either a non-targeting shRNA (SHC002) or mouse Padi4 targeting shRNAs TRCN000101833 (shRNA 3) or TRCN000101834 (shRNA 4). After puromycin selection, control and Padi4 knockdown pre-iPS cell lines were generated. Notably, puromycin treatment abolished all non-transfected pre-iPS cells, and control and shPadi4 pre-iPS cell lines were Oct4–GFP negative (Fig. 2a). For reprogramming experiments, 1 × 105 cells were plated in individual wells (6-well plates) in triplicate and without puromycin, and media was changed to 2i/LIF after day 1. Cells were maintained in 2i/LIF for 8 days, with media change every 2 days, and assessed for Oct4–GFP as previously described.
qRT–PCR primer design
GenBank and Ensembl cDNA sequences were used to design gene-specific primers in Primer 3 (ref. 38) or in the Universal ProbeLibrary Assay Design Center (Roche Applied Science). The specificity of PCR primers was determined via the in-Silico PCR (UCSC Genome Browser) and Primer-BLAST (NCBI) web-based tools. Oligonucleotides were obtained from Sigma. Primer sequences can be found in Supplementary Table 1.
Total RNA was isolated from ES cells, NSO4G and cells during reprogramming, using the RNeasy extraction kit with in-column DNase treatment (Qiagen). Total RNA was reverse transcribed with the High-Capacity cDNA Reverse Transcription kit for 1 h (Applied Biosystems). Samples were aliquoted equally into positive and negative (RT−) reactions. Before qPCR analysis, samples were diluted five- or tenfold with DNase/RNase free dH2O (Ambion).
qPCR reactions were performed in duplicate or triplicate for each sample. Each PCR reaction had a final volume of 10–20 μl and 2.5–5 μl of diluted cDNA or ChIP DNA. RT− samples were assayed to discount genomic DNA amplification. Fast SYBR green Master Mix or TaqMan Fast Universal PCR Master Mix (Applied Biosystems) was used according to the manufacturer’s instructions. A melting curve was obtained for each PCR product after each run, to confirm that the SYBR green signal corresponded to a unique and specific amplicon. Random PCR products were also run in a 2–3% agarose gel to verify the size of the amplicon. Standard curves were generated for each real-time PCR run using serial threefold dilutions of a sample containing the sequence of interest. Their plots were used to convert Ct values (number of PCR cycles needed for a given template to be amplified to an established fluorescence threshold) into arbitrary quantities of initial template per sample. Expression levels were then obtained by dividing the quantity by the value of housekeeping genes, such as ubiquitin (UbC). UbC assays were run every time samples were frozen/thawed. Statistical analysis was performed in Prism 6 using one-way ANOVA analysis of variance with Holm-Sidak’s multiple comparisons test or two-tailed unpaired t-test.
Gene expression analysis
Mouse WG-6 Expression BeadChip microarrays (Illumina) were processed at the Cambridge Genomic Services, Department of Pathology, University of Cambridge. Three biological replicates were assayed for each condition. Illumina microarray probes were matched to gene identifiers according to the re-annotation of the microarray platform39. For both the PADI4 overexpression and the Cl-amidine inhibition experiments, normalization was performed using the lumi40 R package. Limma41 was used for differential expression analysis, with Benjamini–Hochberg (FDR) adjusted P-values <0.05 considered significant. Gene ontology (GO) enrichment analysis was performed using GOstats42, and results adjusted for multiple testing using the Benjamini–Hochberg procedure (FDR).
For immunoblot analysis, cell monolayers or pellets were re-suspended in 2× Laemmli buffer, incubated for 5 min at 95 °C and passed 10 times through a 21G needle to shear genomic DNA. In the case of trichloroacetic acid precipitated proteins, pellets were re-suspended in buffer and boiled as above. Proteins were separated by SDS–PAGE, transferred to nitrocellulose membrane (Millipore) using wet transfer and incubated in blocking solution (5% BSA in TBS containing 0.1% Tween) for 1 h at room temperature. Membranes were incubated with primary antibody at 4 °C overnight and appropriate HRP-conjugated secondary antibody for 2 h at room temperature. Membranes were then incubated for enhanced chemiluminescence (ECLH; GE Healthcare) and proteins were detected by exposure to X-ray film. Primary antibodies, diluted in blocking solution, were used against citrullinated histone H3 (anti-H3Cit, Abcam, ab5103 at 1:50,000 dilution), unmodified histone H3 (anti-H3, Abcam, ab10799 at 1:2,000), linker histone H1 (anti-H1, Santa Cruz Biotechnology, sc-34464 at 1:200), GFP (anti-GFP, Abcam, ab290 at 1:5,000) and Gapdh (anti-Gapdh, Abcam, ab9485 at 1:2,500). Citrulline-containing proteins were modified on the membrane and detected using the anti-modified citrulline detection kit (Millipore) as per manufacturer’s instructions.
ChIP-IT Express (Active Motif) was used according to the supplier’s recommendations. Cells were crosslinked using 1% formaldehyde for 10 min at room temperature. Formaldehyde was quenched by a 5-min incubation with glycine, cells were rinsed twice with cold PBS, collected by scraping and pelleted at 2,500 r.p.m. for 10 min at 4 °C. Frozen pelleted cells were thawed and re-suspended in lysis buffer, rotated for 30 min at 4 °C, dounced and centrifuged at 5,000 r.p.m. for 10 min at 4 °C. Pelleted nuclei were re-suspended in shearing buffer. Chromatin was then sonicated using a Bioruptor 200 (Diagenode), high frequency, 0.5 min/0.5 min, for 10 min twice. Sonicated chromatin was analysed in 1% agarose gel, to confirm efficient sonication. Input was collected for further analysis. 5–15 µg of chromatin was incubated with 2 µg of rabbit IgG as control (Abcam, ab6742), anti-PADI4 (Sigma, P4749), anti-Klf4 (R&D Systems, AF3158), anti-Oct4 (Santa Cruz Biotechnology, sc-5279X (mouse) or sc-8628X (goat)), anti-Sox2 (Santa Cruz Biotechnology, sc-17320), anti-H3Cit (Abcam, ab5103), anti-H2A (Abcam, ab18255), anti-H3K4me3 (Millipore, 17-614), anti-RNA polymerase II (Millipore, clone CTD4H8, 05-623) or anti-H1.2 (anti-H1C, a gift from A. Skoultchi) for 1 h at 4 °C and subsequently with protein G magnetic beads. After overnight immunoprecipitation at 4 °C, beads were washed three times with ChIP buffer 1 and two times with ChIP buffer 2. After elution and reverse crosslink (95 °C for 15 min), samples were treated with proteinase K for 1 h. Purified DNA and 1% input were analysed by Fast SYBR green Master Mix qPCR, using serial fourfold dilutions of the concentrated input for standard curves and triplicates per sample. Occupancy is plotted as fold enrichment over input and after subtracting background signal from the beads. Primers are listed in Supplementary Table 1.
Embryo collection and culture
Embryos were collected into M2 medium (including 4 mg ml−1 BSA) from superovulated C57BL/6 × CBA females mated with C57BL/6 × CBA or H2B–EGFP43 males as previously described44. Embryos were cultured in KSOM (including 4 mg ml−1 BSA) under paraffin oil in 5% CO2 at 37.5 °C. PADI4 inhibitors (Cl-amidine, 10 μM or 200 μM, and TDFA, 100 μM) and HDAC inhibitor (TSA, Sigma, 10 nM) were added to final KSOM from 2-cell-stage embryos (44 h after hCG) onwards. The concentration for inhibitors was determined by titration, and was set as the lowest dose that leads to inhibition of enzyme activity and allowed embryonic development to blastocyst stage. No randomization was performed and there was no blinding. Animals were maintained in the Animal Facility of the Gurdon Institute at a 12:12 light cycle and provided with food and water ad libitum. All experiments were conducted in compliance with Home Office regulations.
Immunofluorescence staining and analysis of embryos
Embryos were fixed in 4% PFA for Nanog, Sox17, Cdx2 and H3K9 acetylated staining, and in methanol for H3Cit staining. Immunofluorescence staining was carried out as described previously45. Primary antibodies used were as follows: rabbit anti-H3Cit (Abcam, 1:100), rabbit anti-H3K9ac (Upstate, 1:100), rabbit anti-Nanog (Cosmo Bio Co., 1:200), goat anti-Sox17 (R&D systems, 1:200), mouse anti-Cdx2 (Cdx2-88) (Biogenex, 1:200). Secondary antibodies used were AlexaFluor 568-conjugated donkey anti-goat, AlexaFluor 488-conjugated donkey anti-mouse and AlexaFluor 633-conjugated donkey anti-rabbit, and AlexaFluor 488-conjugated anti-rabbit and Texas red-conjugated goat anti-rabbit (Invitrogen), at 1:400. Collection of 4-cell embryos and E4.5 blastocysts was performed 56 h and 106 h after hCG, respectively. Confocal microscopy was performed and images analysed using a 40/1.4 NA oil DIC Plan-Apochromatic lens on an inverted Zeiss 510 Meta confocal microscope. Confocal sections were taken every 2 μm through the whole embryo. To measure the fluorescence levels of H3Cit objectively, individual cells were outlined manually in Image J, and the intensity of the fluorescent signal was recorded for each z-stack (three measurements per nucleus). Only cells in the same z-stacks were compared. Statistical analysis was performed with two-tailed unpaired t-test or Mann–Whitney U-test (for non-normal distributions). Fluorescence and DIC z-stacks of embryos from late 2-cell (52 h after hCG) to blastocyst stage were collected on 15 focal planes every 15 min for 72 h of continuous embryo culture. The images were processed as described previously44. All cells were followed in 4D using SIMI Biocell software (http://www.simi.com/en/products/cell-research/simi-biocell.html/)46 as previously described44.
SILAC labelling and mass spectrometry
ES cells were labelled in culture using the Mouse Embryonic Stem Cell SILAC Protein Quantitation kit (Pierce) for at least six passages. Extracted proteins were re-suspended in Laemmli sample buffer, and resolved on a 4–20% SDS–PAGE (NuPAGE, Life Technologies). The gel was stained with Coomassie blue, cut into 20 slices and processed for mass spectrometric analysis using standard in gel procedure47. Briefly, cysteines were reduced with dithiothreitol (DTT), alkylated using chloroacetamide (CAA)48, and finally the proteins were digested overnight with endoproteinase LysC and loaded onto C18 StageTips before mass spectrometric analysis.
All MS experiments were performed on a nanoscale HPLC system (EASY-nLC from Thermo Scientific) connected to a hybrid LTQ–Orbitrap Velos49 equipped with a nanoelectrospray source (Thermo Scientific). Each peptide sample was auto-sampled and separated on a 15-cm analytical column (75-cm inner diameter) in-house packed with 3 μm C18 beads (Reprosil Pur-AQ, Dr. Maisch) with a 2 h gradient from 5% to 40% acetonitrile in 0.5% acetic acid. The effluent from the HPLC was directly electrosprayed into the mass spectrometer.
The MS instrument was operated in data-dependent mode to switch between full-scan MS and MS/MS acquisition automatically. Survey full-scan MS spectra (from m/z 300–1,700) were acquired in the Orbitrap mass analyser with resolution R = 30,000 at m/z 400 (after accumulation to a ‘target value’ of 1,000,000 in the linear ion trap). The ten most intense peptide ions with charge states ≥2 were subsequently isolated to a target value of 50,000 using predictive automatic gain control (pAGC) and fragmented by higher-energy collisional dissociation (HCD) in the octopole collision cell using normalized collision energy of 40%. The ion selection threshold was set to 5,000 counts for HCD and the maximum allowed ion accumulation times was set to 500 ms for full scans and 250 ms for HCD. All HCD fragment ion spectra were recorded in the Orbitrap mass analyser with a resolution of 7,500 at m/z 400. For all full-scan measurements, a lock-mass ion from ambient air (m/z 445.120025) was used for internal calibration when present, as described50.
Identification of peptides and proteins
Mass spectrometry data analysis was performed with the MaxQuant software suite (version 18.104.22.168) as described51 supported by Andromeda (http://www.maxquant.org) as the database search engine for peptide identifications52. We followed the step-by-step protocol of the MaxQuant software suite53 to generate MS/MS peak lists that were filtered to contain at most six peaks per 100 Da interval and searched by Andromeda against a concatenated target/decoy54 (forward and reversed) version of the Uniprot human database version (70.101 forward protein entries). Protein sequences of common contaminants such as human keratins and proteases used were added to the database. The initial mass tolerance in MS mode was set to 7 p.p.m. and MS/MS mass tolerance was set to 20 p.p.m. Cysteine carbamidomethylation was searched as a fixed modification, whereas protein N-acetylation, oxidized methionine, deamidation of asparagine and glutamine, and citrullination of arginines were searched as variable modifications. A maximum of two mis-cleavages was allowed whereas we required strict LysC specificity. Peptide assignments were statistically evaluated in a Bayesian model on the basis of sequence length and Andromeda score. We only accepted peptides and proteins with a false discovery rate of less than 1%, estimated on the basis of the number of accepted reverse hits.
Native purification of wild-type and mutant linker histone H1.2
Linker histone H1.2 was expressed from a pET 28b(+) vector, provided by the laboratory of R. Schneider. The vector was mutated using Quickchange site-directed mutagenesis protocol (Stratagene) for the expression of R54A and R54K mutants. After transformation into expression strain BL21(DE3)-RIL, pre-cultures were diluted 1:100 into 3 l of LB medium containing kanamycin and camptothecin and grown at 37 °C, 200 r.p.m. At an optical density at 600 nm of 0.6, 0.2 mM IPTG was added and the bacteria grown for another 2 h under the same conditions. Bacterial pellets were collected by centrifugation (15 min, 4,000g, 4 °C), frozen in liquid nitrogen and stored at −80 °C. Linker histone expression was assessed by SDS–PAGE and Coomassie blue stain. Pellets were thawed on ice, re-suspended in 25 ml lysis buffer (20 mM HEPES, pH 7.6; 100 mM NaCl; 5 mM EDTA; 1× complete protease inhibitors (Roche); 1 mM DTT; 0.2 mM PMSF) per litre of culture and lysed through an Avestin high-pressure homogenizer. Lysates were cleared by centrifugation (20,000g, 40 min, 4 °C) and 4 M ammonium sulphate added slowly to the supernatant to a final concentration of 2 M. The mixture was incubated on an end-to-end rotator at 4 °C for 20 min before the precipitated proteins were removed by centrifugation (20,000g, 40 min, 4 °C). The supernatant was filtered through a 0.2 μm syringe filter and loaded onto a 20 ml HiPrep 16/10 Phenyl FF (low sub) hydrophobic interaction column (Amersham Biosciences), equilibrated in HEMG buffer (20 mM HEPES, pH 7.6; 50 μM EDTA; 6.25 mM MgCl2; 0.5 mM DTT; 10% glycerol) containing 2 M ammonium sulphate. The sample was injected via a 50-ml super loop at a flow rate of 0.5 ml min−1, while the flow through, which contained H1.2, was collected in 5-ml fractions. Ten column volumes of HEMG buffer were used to flush the remaining H1.2 from the column. Fractions were analysed by SDS–PAGE and H1.2-containing fractions were pooled and dialysed overnight into HEG buffer (20 mM HEPES, pH 7.6; 50 μM EDTA; 10% glycerol) containing 100 mM NaCl, using 6–8 kDa MWCO dialysis tubing. The protein solution was centrifuged (20,000g, 40 min, 4 °C) to remove precipitated proteins and filtered as above. The solution was loaded onto a 6 ml Resource S column, pre-equilibrated with HEG buffer, and proteins eluted with a gradient of 100 mM to 1 M NaCl-containing HEG buffer (0.5 ml min−1, 20CV) in 500 μl fractions. Protein fractions were analysed for the presence of H1.2, pooled as above and dialysed in HEG buffer with 100 mM NaCl using dialysis tubing with 10 kDa MWCO. Samples were filtered as above and loaded onto a Mono S column (Amersham Biosciences) equilibrated with HEG buffer containing 100 ml NaCl. 500 μl fractions were collected with a gradient of 100 mM to 1 M NaCl-containing HEG buffer (0.5 ml min−1, 20CV) and analysed on a 15% SDS–PAGE. Fractions containing H1.2 were pooled, dialysed as above into 50% glycerol and stored at −20 °C.
H1–GFP and nucleosome pull-down assays
GFP-tagged linker histone H1.2 (H1.GFP) was expressed in ES cells stably overexpressing human PADI4 or control vector by transfection (Lipofectamine 2000) of a pEGFP-H1.2 vector, containing either the wild-type or the R54A mutant sequence, or with empty pEGFP vector. Transfection efficiency was assessed by visualization of GFP using fluorescence microscopy and determined to be >90%. 24 h after transfection, cells were collected and H1–GFP was pulled down using GFP-TRAP conjugated magnetic beads (Chromotek) as per the manufacturer’s instructions, and subjected to immunoblot analysis.
For in vitro nucleosome pull down, nucleosomes containing purified and refolded histone octamers and biotinylated 601 DNA were assembled as described previously55 and immobilized on Dynabeads Streptavidin MyOne T1 (Invitrogen) for 2 h at 4 °C. Nucleosome-loaded beads in varying amounts were incubated on an end-to-end rotator with wild-type or mutant linker histone H1.2 in binding buffer (20 mM HEPES, pH 7.9; 150 mM NaCl; 0.2 mM EDTA, 20% glycerol; 0.1% NP40; 1 mM DTT, and complete protease inhibitors (Roche)) for 4 h at 4 °C. After five washes in binding buffer, the beads were collected and bound proteins were eluted in sample buffer and subjected to immunoblot analysis.
Purification of recombinant PADI4–GST and in vitro deimination assays
Recombinant human PADI4–GST was expressed from pGEX6p constructs in LBKan media, induced with 0.1 mM IPTG at 25 °C, purified using glutathione-sepharose resin, eluted using a 25 mM glutathione solution and stored in 50% glycerol at −20 °C. In vitro deimination of linker histone H1.2 was carried out in deimination buffer (50 mM HEPES, pH 7.5; 2 mM DTT; in the presence or absence of 10 mM CaCl2) at 37 °C for 1 h, using the active enzyme. Samples were re-suspended in sample buffer for immunoblot analysis.
Treatment of permeabilized cells with recombinant PADI4
Cells were re-suspended in a cold solution of 80 μg ml−1 digitonin in SuNaSP56, mixed gently but thoroughly and incubated on ice for 3 min. Complete permeabilization was checked by Trypan blue uptake of sub-aliquots of the suspension and the reaction was stopped by addition of SuNaSP containing 3% BSA. Permeabilized cells were then pre-incubated in wash buffer (20 mM HEPES-KOH, 75 mM KCl, 1.5 mM MgCl2, 5 mM CaCl2) with 0.2% Triton X-100 for 30 min at 37 °C. Nuclei were washed in wash buffer and incubated in reaction buffer (20 mM HEPES-KOH, 75 mM KCl, 1.5 mM MgCl2, 5 mM CaCl2, 2 mM DTT) with or without recombinant PADI4 for 1.5 h at 37 °C. The nuclei were washed several times in wash buffer to remove excess PADI4 and released proteins that were not stably bound within the nucleus. The ‘Washes’ fraction contains all proteins that are no longer stably bound, including recombinant human PADI4. The ‘Pellet’ contains all other nuclear-retained proteins. The wash supernatants were concentrated by trichloroacetic acid protein precipitation and subjected to immunoblot analysis. The washed nuclei were washed a further time in PBS for immunoblot analysis or in MNase buffer and used for DNA compaction assays.
Micrococcal nuclease digestion
Cell pellets were washed twice and re-suspended in micrococcal nuclease (MNase) buffer (15 mM Tris, pH 7.5; 15 mM NaCl; 60 mM KCl; 0.34 M sucrose; 0.5 mM spermidine; 0.15 mM spermine; 0.25 mM PMSF; 0.1% β-mercaptoethanol). 1 mM of CaCl2 was added and the suspension was divided into equal aliquots, which were kept on ice. MNase was added to a final concentration of 0.5 U ml−1 and samples were incubated at 25 °C in a heat block for varying amounts of time. The reaction was stopped by addition of 0.5 M EDTA and 0.5 M EGTA and DNA purified using a QIAGEN PCR purification kit. DNA was quantified and 1,500 ng of each sample was loaded on a 1.5% agarose gel containing ethidium bromide.
Vossenaar, E. R., Zendman, A. J., van Venrooij, W. J. & Pruijn, G. J. PAD, a growing family of citrullinating enzymes: genes, features and involvement in disease. BioEssays 25, 1106–1118 (2003)
Wang, S. & Wang, Y. Peptidylarginine deiminases in citrullination, gene regulation, health and pathogenesis. Biochim. Biophys. Acta 1829, 1126–1135 (2013)
Martinod, K. et al. Neutrophil histone modification by peptidylarginine deiminase 4 is critical for deep vein thrombosis in mice. Proc. Natl Acad. Sci. USA 110, 8674–8679 (2013)
Hagiwara, T., Nakashima, K., Hirano, H., Senshu, T. & Yamada, M. Deimination of arginine residues in nucleophosmin/B23 and histones in HL-60 granulocytes. Biochem. Biophys. Res. Commun. 290, 979–983 (2002)
Cuthbert, G. L. et al. Histone deimination antagonizes arginine methylation. Cell 118, 545–553 (2004)
Wang, Y. et al. Human PAD4 regulates histone arginine methylation levels via demethylimination. Science 306, 279–283 (2004)
Zhang, X. et al. Genome-wide analysis reveals PADI4 cooperates with Elk-1 to activate c-Fos expression in breast cancer cells. PLoS Genet. 7, e1002112 (2011)
Tanikawa, C. et al. Regulation of histone modification and chromatin structure by the p53-PADI4 pathway. Nature Commun. 3, 676 (2012)
Asaga, H., Nakashima, K., Senshu, T., Ishigami, A. & Yamada, M. Immunocytochemical localization of peptidylarginine deiminase in human eosinophils and neutrophils. J. Leukoc. Biol. 70, 46–51 (2001)
Neeli, I., Khan, S. N. & Radic, M. Histone deimination as a response to inflammatory stimuli in neutrophils. J. Immunol. 180, 1895–1902 (2008)
Buttinelli, M., Panetta, G., Rhodes, D. & Travers, A. The role of histone H1 in chromatin condensation and transcriptional repression. Genetica 106, 117–124 (1999)
Yamanaka, S. & Blau, H. M. Nuclear reprogramming to a pluripotent state by three approaches. Nature 465, 704–712 (2010)
Gaspar-Maia, A., Alajem, A., Meshorer, E. & Ramalho-Santos, M. Open chromatin in pluripotency and reprogramming. Nature Rev. Mol. Cell Biol. 12, 36–47 (2011)
Meshorer, E. et al. Hyperdynamic plasticity of chromatin proteins in pluripotent embryonic stem cells. Dev. Cell 10, 105–116 (2006)
Marks, H. et al. The transcriptional and epigenomic foundations of ground state pluripotency. Cell 149, 590–604 (2012)
Theunissen, T. W. et al. Nanog overcomes reprogramming barriers and induces pluripotency in minimal conditions. Curr. Biol. 21, 65–71 (2011)
Darrah, E., Rosen, A., Giles, J. T. & Andrade, F. Peptidylarginine deiminase 2, 3 and 4 have distinct specificities against cellular substrates: novel insights into autoantigen selection in rheumatoid arthritis. Ann. Rheum. Dis. 71, 92–98 (2012)
Polo, J. M. et al. A molecular roadmap of reprogramming somatic cells into iPS cells. Cell 151, 1617–1632 (2012)
Luo, Y. et al. Inhibitors and inactivators of protein arginine deiminase 4: functional and structural characterization. Biochemistry 45, 11727–11736 (2006)
Brahmajosyula, M. & Miyake, M. Localization and expression of peptidylarginine deiminase 4 (PAD4) in mammalian oocytes and preimplantation embryos. Zygote 21, 314–324 (2013)
Kan, R. et al. Potential role for PADI-mediated histone citrullination in preimplantation development. BMC Dev. Biol. 12, 19 (2012)
Li, P. et al. PAD4 is essential for antibacterial innate immunity mediated by neutrophil extracellular traps. J. Exp. Med. 207, 1853–1862 (2010)
Zernicka-Goetz, M., Morris, S. A. & Bruce, A. W. Making a firm decision: multifaceted regulation of cell fate in the early mouse embryo. Nature Rev. Genet. 10, 467–477 (2009)
Jones, J. E. et al. Synthesis and screening of a haloacetamidine containing library to identify PAD4 selective inhibitors. ACS Chem. Biol. 7, 160–165 (2012)
Fan, Y. et al. Histone H1 depletion in mammals alters global chromatin structure but causes specific changes in gene regulation. Cell 123, 1199–1212 (2005)
Zhang, Y. et al. Histone H1 depletion impairs embryonic stem cell differentiation. PLoS Genet. 8, e1002691 (2012)
Goytisolo, F. A. et al. Identification of two DNA-binding sites on the globular domain of histone H5. EMBO J. 15, 3421–3429 (1996)
Brown, D. T., Izard, T. & Misteli, T. Mapping the interaction surface of linker histone H1(0) with the nucleosome of native chromatin in vivo. Nature Struct. Mol. Biol. 13, 250–255 (2006)
Chang, X. et al. Increased PADI4 expression in blood and tissues of patients with malignant tumors. BMC Cancer 9, 40 (2009)
Zhang, X. et al. Peptidylarginine deiminase 2-catalyzed histone H3 arginine 26 citrullination facilitates estrogen receptor α target gene activation. Proc. Natl Acad. Sci. USA 109, 13331–13336 (2012)
Silva, J., Chambers, I., Pollard, S. & Smith, A. Nanog promotes transfer of pluripotency after cell fusion. Nature 441, 997–1001 (2006)
Ying, Q. L., Nichols, J., Evans, E. P. & Smith, A. G. Changing potency by spontaneous fusion. Nature 416, 545–548 (2002)
Niwa, H., Miyazaki, J. & Smith, A. G. Quantitative expression of Oct-3/4 defines differentiation, dedifferentiation or self-renewal of ES cells. Nature Genet. 24, 372–376 (2000)
Masui, S. et al. Pluripotency governed by Sox2 via regulation of Oct3/4 expression in mouse embryonic stem cells. Nature Cell Biol. 9, 625–635 (2007)
Ding, S. et al. Efficient transposition of the piggyBac (PB) transposon in mammalian cells and mice. Cell 122, 473–483 (2005)
Silva, J. et al. Nanog is the gateway to the pluripotent ground state. Cell 138, 722–737 (2009)
Theunissen, T. W. et al. Nanog overcomes reprogramming barriers and induces pluripotency in minimal conditions. Curr. Biol. 21, 65–71 (2011)
Rozen, S. & Skaletsky, H. Primer3 on the WWW for general users and for biologist programmers. Methods Mol. Biol. 132, 365–386 (2000)
Barbosa-Morais, N. L. et al. A re-annotation pipeline for Illumina BeadArrays: improving the interpretation of gene expression data. Nucleic Acids Res. 38, e17 (2010)
Du, P., Kibbe, W. A. & Lin, S. M. lumi: a pipeline for processing Illumina microarray. Bioinformatics 24, 1547–1548 (2008)
Smyth, G. K. In Bioinformatics and Computational Biology Solutions using R and Bioconductor (eds Carey, V., Gentleman, R., Dudoit, S., Irizarry, R. & Huber, W. ) 397–420 (Springer, 2005)
Falcon, S. & Gentleman, R. Using GOstats to test gene lists for GO term association. Bioinformatics 23, 257–258 (2007)
Hadjantonakis, A. K. & Papaioannou, V. E. Dynamic in vivo imaging and cell tracking using a histone fluorescent protein fusion in mice. BMC Biotechnol. 4, 33 (2004)
Bischoff, M., Parfitt, D. E. & Zernicka-Goetz, M. Formation of the embryonic-abembryonic axis of the mouse blastocyst: relationships between orientation of early cleavage divisions and pattern of symmetric/asymmetric divisions. Development 135, 953–962 (2008)
Jedrusik, A. et al. Role of Cdx2 and cell polarity in cell allocation and specification of trophectoderm and inner cell mass in the mouse embryo. Genes Dev. 22, 2692–2706 (2008)
Schnabel, R., Hutter, H., Moerman, D. & Schnabel, H. Assessing normal embryogenesis in Caenorhabditis elegans using a 4D microscope: variability of development and regional specification. Dev. Biol. 184, 234–265 (1997)
Shevchenko, A., Tomas, H., Havlis, J., Olsen, J. V. & Mann, M. In-gel digestion for mass spectrometric characterization of proteins and proteomes. Nature Protocols 1, 2856–2860 (2006)
Nielsen, M. L. et al. Iodoacetamide-induced artifact mimics ubiquitination in mass spectrometry. Nature Methods 5, 459–460 (2008)
Olsen, J. V. et al. A dual pressure linear ion trap Orbitrap instrument with very high sequencing speed. Mol. Cell. Proteomics 8, 2759–2769 (2009)
Olsen, J. V. et al. Parts per million mass accuracy on an Orbitrap mass spectrometer via lock mass injection into a C-trap. Mol. Cell. Proteomics 4, 2010–2021 (2005)
Cox, J. & Mann, M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nature Biotechnol. 26, 1367–1372 (2008)
Cox, J. et al. Andromeda: a peptide search engine integrated into the MaxQuant environment. J. Proteome Res. 10, 1794–1805 (2011)
Cox, J. et al. A practical guide to the MaxQuant computational platform for SILAC-based quantitative proteomics. Nature Protoc. 4, 698–705 (2009)
Elias, J. E. & Gygi, S. P. Target-decoy search strategy for increased confidence in large-scale protein identifications by mass spectrometry. Nature Methods 4, 207–214 (2007)
Bartke, T. et al. Nucleosome-interacting proteins regulated by DNA and histone methylation. Cell 143, 470–484 (2010)
Halley-Stott, R. P. et al. Mammalian nuclear transplantation to germinal vesicle stage Xenopus oocytes—a method for quantitative transcriptional reprogramming. Methods 51, 56–65 (2010)
This work was funded by programme grants from Cancer Research UK (T.K.) and EMBL (P.B., R.L.). R.P.H.-S. and J.B.G. are supported by the Medical Research Council (G1001690) and the Wellcome Trust. G.C.-B. was funded by EMBO (Long-Term Post-Doctoral Fellowship), European Union (FP7 Marie Curie Intra-European Fellowship for Career Development) and Swedish Research Council. M.A.C. was funded by an EMBO Long-Term Post-Doctoral Fellowship and a Human Frontier Science Programme Long-Term Post-Doctoral Fellowship. C.S.O. was supported by FAPESP (Foundation for Research Support of the State of São Paulo) and mouse embryo work was supported by the Wellcome Trust programme grant to M.Z.-G. M.L.N. was partly supported by the Novo Nordisk Foundation Center for Protein Research, the Lundbeck Foundation, and by and the European Commission’s 7th Framework Programme HEALTH-F7-2010-242129/SYBOSS. K.A.M. was funded by NIH grant AI099728. We would like to thank S. Lestari, A. Cook and C. Hill for technical assistance; P. Thompson for providing the TDFA compound; GSK Epinova for Cl-amidine; T. Bartke for the gift of histone octamers and help with nucleosome pull-down assays; A. Finch for help with FPLC chromatography; A. Jedrusik for help with embryo work; R. Walker at the Flow Cytometry Core Facility at Wellcome Trust Centre for Stem Cell Research, University of Cambridge and T. Theunissen for help with the flow cytometry; and members of the Kouzarides laboratory for critical discussions of the work. 2TS22C cells were provided by H. Niwa. The ChIP grade H1.2 antibody was a gift from A. Skoultchi.
T.K. is a co-founder of Abcam.
Extended data figures and tables
Extended Data Figure 1 Citrullination and Padi expression profiles in ES, NSO4G and iPS cells; regulation of Padi4 by pluripotency factors in ES cells.
a, Transcript levels for Padi1, Padi2 and Padi3 in ES, NSO4G (NSC) and iPS cells, as assessed by qRT–PCR. Padi6 was undetectable in all three cell types. Expression is normalized to endogenous levels of ubiquitin (UbC). Error bars represent the standard error of the mean of three biological replicates. b, Transcript levels of Padi1, Padi2 and Padi3 in ES cells on switch to 2i containing medium for one passage, as assessed by qRT–PCR. Padi6 was undetectable in both conditions. Expression normalized to UbC. Error bars represent the standard error of the mean of three biological replicates. c, Immunoblot analysis of H3Cit levels in ES, NSO4G and iPS cells. Total H3 is presented as a loading control. d, Immunoblot analysis of total citrullination levels in ES, NSO4G and iPS cells, using an antibody against modified citrulline (ModCit), which recognizes peptidylcitrulline irrespective of amino acid sequence. Total H3 is presented as a loading control. e, ZHBTc4.1 and 2TS22C ES cell lines were treated with 1 μg ml−1 doxycycline for 48 h, resulting in depletion of Oct4 or Sox2 (data not shown). Padi4 mRNA was significantly reduced on Oct4, but not Sox2 knockdown, as assessed by qPCR. Error bars represent standard error of the mean of four biological replicates. f, ChIP-qPCR for Oct4, Sox2, Klf4, RNA polymerase II (PolII), H3K4me3 and H2A on the promoter of Padi4 in mouse ES and NSO4G cells. For each cell condition, the signal is presented as fold enrichment over input and after subtracting background signal from the beads. Error bars represent the standard deviation of three technical qPCR replicates. Asterisks denote difference with ES cells (a) or ES media (b), and 0 h time point (e); NS, not significant, *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001 by ANOVA (a) or t-test (b, e).
Extended Data Figure 2 PADI4 overexpression or knockdown in ES cells modulates expression of pluripotency genes.
a, Validation of selected targets from the PADI4 overexpression microarray data set by qRT–PCR. Expression of Pou5f1, Sox2, Klf4 and c-Myc is not affected by PADI4 overexpression. Expression is normalized to UbC. Error bars are presented as standard error of the mean of three biological replicates. b, Transcript levels of mouse Padi4 and human PADI4 in mouse ES cells after transient knockdown with Padi4 or control (ctrl) shRNA, and overexpression of human PADI4 or control vector (pPB ctrl), as assessed by qRT–PCR. Expression normalized to UbC. Error bars represent the standard error of the mean of three biological replicates. c, Transcript levels of mouse Padi4, Tcl1 and Nanog in a mouse ES cell clone stably expressing Padi4 or control (ctrl) shRNA, as assessed by qRT–PCR. Expression is normalized to UbC. Error bars represent the standard error of the mean of three biological replicates. Asterisks denote difference with ctrl (a, b, c) and between samples (b); NS, not significant, *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001 by ANOVA (b) or t-test (a, c).
a, Representative ChIP-qPCR for H2A on regulatory regions of Tcl1 and Nanog in mouse ES, NSO4G and iPS cells (corresponding to Fig. 1h). For each cell condition, the signal is presented as fold enrichment over input and after subtracting background signal from the beads. Error bars represent the standard deviation of three technical qPCR replicates. b, ChIP-qPCR for human PADI4 on regulatory regions of Tcl1, Nanog, Klf2 and Kit, which are upregulated by human PADI4 overexpression, in mouse ES cells stably expressing human PADI4. For each cell condition, the signal is presented as fold enrichment over input and after subtracting background signal from the beads. Enhancer regions for Kit located +3.4 kb and +12 kb downstream of transcription termination site. Error bars represent the standard deviation of three technical qPCR replicates. c, Representative ChIP-qPCR for H3Cit on regulatory regions of Tcl1 and Nanog in mouse ES cells stably expressing human PADI4 and treated with 200 μM Cl-amidine for 48 h. For each cell condition, the signal is presented as fold enrichment over input and after subtracting background signal from the beads. Error bars represent the standard deviation of three technical qPCR replicates.
a, Heat map of the top 70 genes that showed differential expression after Padi4 inhibition in ES cells by with 200 μM Cl-amidine for 48 h, as determined by microarray analysis. Displayed values are normalized log intensities, minus the mean expression of the gene across the two samples. Hierarchical clustering based on correlation. b, Validation of selected targets from the above microarray data set by qRT–PCR. Expression is normalized to UbC. Error bars presented as standard error of the mean of three biological replicates. Asterisks denote difference with Ctrl; *P ≤ 0.05, **P ≤ 0.01, ****P ≤ 0.0001 by t-test. c, Gene Ontology for Biological Process (GOBP) analysis for the most regulated gene categories within the microarray data set of Cl-amidine treatment in mouse ES cells. P value is corrected for multiple testing using Benjamini–Hochberg FDR.
a, Scheme of reprogramming of neural stem cells to pluripotent state. NSO4G cells were retrovirally transduced with Oct4, Klf4 and c-Myc. After 6 days, partially reprogrammed pre-iPS cells arose. For shRNA experiments, pre-iPS cells were stably transfected with control or Padi4 shRNA and then full reprogramming was performed in the presence of 2i/LIF media for 8 days. For Padi4 enzymatic inhibition, pre-iPS cells were immediately changed to 2i/LIF media in the presence of 200 μM Cl-amidine for 8 days. b, Quantification of flow cytometry analysis for the assessment of Oct4–GFP reporter expression in a reprogramming assay using pre-iPS cells stably expressing Padi4 shRNA 4 and control shRNA. Error bars represent standard error of the mean of triplicate samples within a representative from four reprogramming experiments. c, Quantification of Oct4–GFP-positive colonies in the reprogramming assay where pre-iPS cells were Padi4 shRNA 4 versus control (see Fig. 2a), after time-lapse image acquisition with Biostation CT. Error bars represent standard error of the mean of triplicate samples within a representative reprogramming experiment. See Supplementary Video 1 for time-lapse video. d, Immunoblot analysis of H3Cit in pre-iPS cells treated with 2i, after Padi4 knockdown (Padi4 shRNA 4) versus control cells (ctrl shRNA). 2i-containing medium was added on day 2. Gapdh presented as loading control. e, Quantification of flow cytometry analysis for the assessment of Oct4–GFP reporter expression in a reprogramming assay using pre-iPS cells stably expressing Padi4 shRNA 3 and control shRNA. Error bars represent standard error of the mean of triplicate samples. f, qRT–PCR analysis for the expression of Tcl, Nanog and Padi4 mRNAs at the end of the above reprogramming assay (e). Error bars represent standard error of the mean of triplicate samples. g, Quantification of flow cytometry analysis for the assessment of Oct4–GFP reporter expression in a reprogramming assay were treated with 200 μM Cl-amidine. Error bars represent standard error of the mean of triplicate samples within a representative from three reprogramming experiments. h, Quantification of Oct4–GFP-positive colonies in the reprogramming assay where pre-iPS cells were treated with 200 μM Cl-amidine (see Fig. 2c) after time-lapse image acquisition with Biostation CT. Error bars represent standard error of the mean of triplicate samples within a representative reprogramming experiment. i, Immunoblot analysis for the presence of H3Cit at the end of the above reprogramming assay (g). Total histone H3 presented as loading control. Asterisks denote difference with control; NS, not significant, *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001 by t-test.
a, Embryos at 2-cell stage were treated with 200 μM Cl-amidine and snapshots were taken at E3.0, E3.5 and E4.0. 200 μM Cl-amidine embryos arrested at 8-cell stage, whereas control embryos continued development to form blastocysts. Phase contrast images are shown. b, Embryos at 2-cell stage were treated with 10 μM Cl-amidine for 12 h, fixed and stained for H3Cit at the 4-cell stage. Phase contrast, H3Cit (white) and Hoechst 33342 (blue) images are shown. Scale bar, 20 μm. c, Embryos at E3.5 were treated with 10 μM Cl-amidine for 24 h, fixed and stained for H3Cit at E4.5. H3Cit (green) and Hoechst 33342 (blue) images are shown. Bar represents 20 μm. d, Table with quantifications of lineage commitment in E4.5 blastocysts treated with 10 μM Cl-amidine from the 2-cell stage. Asterisks denote difference with control, unpaired t-test; *P < 0.05; n = 3 (50 embryos). e, Embryos were cultured in medium supplemented with 10 μM Cl-amidine from 2-cell stage and through pre-implantation development. E4.5 blastocysts were fixed and stained for Sox17 (primitive endoderm marker, red), Cdx2 (trophectoderm marker, green) and Hoechst 33342 (blue). Scale bar, 20 μm. f, Time-lapse analysis of distribution of inner and outer cells at the 8–16-cell transition, on culturing of embryos with medium containing 10 μM Cl-amidine from 2-cell stage. Error bars represent standard error of the mean. Statistical significance was determined by unpaired t-test or Mann–Whitney U-test upon non-normal distribution. Asterisks denote difference with control; *P ≤ 0.05.
a, Embryos at 2-cell stage were treated with 100 μM TDFA for 12 h and fixed and stained for H3Cit at 4-cell stage. H3Cit and Hoechst 33342 images are shown. b, Table representing the percentage of cells committed to each embryonic lineage in E4.5 blastocysts on treatment of embryos at 2-cell stage with 100 μM TDFA. Scale bars represent mean percentage (±s.e.m.). Asterisks denote difference with control, Mann–Whitney U-test, *P < 0.05; n = 3 (60 embryos). c, Embryos at 2-cell stage were treated with 100 μM TDFA and fixed at embryonic day E4.5. Phase contrast, Nanog (green), Sox17 (purple), Cdx2 (red) and Hoechst 33342 (blue) images are shown.
a, Embryos at 2-cell stage were treated with 10 nM TSA for 12 h, and fixed and stained for H3K9ac at 4-cell stage. H3K9ac and Hoechst 33342 images are shown. b, Table representing the percentage of cells committed to each embryonic lineage in E4.5 blastocysts on treatment of embryos at 2-cell stage with 10 nM TSA. Scale bars represent mean percentage (±s.e.m.). Asterisks denote difference with control, unpaired t-test; *P < 0.05; n = 2 (32 embryos). c, Embryos at 2-cell stage were treated with 10 nM TSA and fixed at embryonic day E4.5. Phase contrast, Nanog (green), Sox17 (purple), Cdx2 (red) and Hoechst 33342 (blue) images are shown.
a, Histogram demonstrating the mass accuracies of all fragment ion masses used for identifying citrullinated peptides in our HCD MS/MS spectra. >490,000 y- and b-ion masses are depicted. The average absolute mass accuracy for all of these fragment ions is 3.97 p.p.m. b, Scatter plot representing SILAC ratios in ES cells cultured in 13C6 l-lysine (HEAVY) and LIGHT medium separately, to assess the extent and quality of SILAC labelling. No significant outliers are observed, indicating equal labelling. c, Peptide coverage of histone H1 by LC-MS analysis. Detected peptides are highlighted in light green and cover >60% of H1. Whereas all arginine residues of histone H1 (highlighted in dark green) were accounted for by the analysis, Arg 54 was the only one found citrullinated. d, Fragmentation spectra of the unmodified LysC peptide ERSGVSLAALKK surrounding Arg 54 of H1.2 (unmodified counterpart of citrullinated peptide depicted in Fig. 3d). The y and b series indicate fragments at amide bonds of the peptide. e, Fragment ion table (expected and observed masses for detected y and b ions) for the identified H1R54 citrullination of peptide ERSGVSLAALKK on histone H1.2 (as shown in Fig. 3d). All measured fragment ions were detected with mass accuracies <10 p.p.m., unambiguously identifying that the detected peptide sequence harbours a citrullination at position Arg 54. f, Theoretical and measured b- and y-ion fragment masses for the corresponding unmodified and heavy SILAC labelled H1.2 peptide, as presented in d above.
Extended Data Figure 10 Mass spectrometry spectra for citrullinated H1.5; PADI4 treatment of differentiated nuclei leads to H1 citrullination and chromatin decompaction.
a, MS spectrum of histone H1.5 in a SILAC proteomic screen for identification of PADI4 substrates. Linker histone H1.5 is deiminated by PADI4, as identified by a highly increased SILAC ratio of the heavy labelled identified peptide (marked by a red dot). b, Fragmentation spectra of the doubly charged LysC peptide ERGGVSLPALK surrounding Arg 54 of H1.5. The y and b series indicate fragments at amide bonds of the peptide, unambiguously verifying the citrullinated peptide. c, Mutation of Arg 54 renders histone H1.2 refractory to deimination. Immunoblot analysis of recombinant histone H1.2 using an antibody that detects all deimination events (ModCit). Wild-type and Arg 54-mutant H1.2 were treated with recombinant PADI4, in the presence of activating calcium. Only wild-type H1.2 can be deiminated, indicating that Arg 54 is the only substrate of PADI4 in H1.2. No-calcium reactions presented as negative controls. Total H1.2 presented as loading control. d, Schematic representation of the position of Arg 54 within the globular domain linker histone H1.2. e, Immunoblot analysis of the ‘Pellet’ fraction of C2C12 permeabilized cells treated with recombinant PADI4. Presence of H3Cit species indicates PADI4 activity. Total H3 is presented as a control for equal use of starting material in the two experimental conditions. f, Immunofluorescence analysis of C2C12 nuclei after treatment with recombinant PADI4. Presence of H3Cit species indicates PADI4 activity. DNA is visualized by staining with DAPI. g, Fragmentation spectra of the citrullination site Arg 54 on the evicted H1.2 peptide ERSGVSLAALK (corresponding to Fig. 4b). The evicted histone H1 is citrulinated at Arg 54. h, Theoretical and measured b- and y-ion fragment masses for the citrullinated H1.2 peptide (peptide sequence ERSGVSLAALK) evicted after treatment of C2C12 cells with recombinant human PADI4 (corresponding to Fig. 4b). i, Micrococcal nuclease digestion of C2C12 nuclei after treatment with recombinant PADI4, as described in Fig. 4a. M, size marker.
The file contains the primer sequences. (XLSX 12 kb)
This file contains the complete microarray data for PADI4 over-expression in ES cells. (XLSX 0 kb)
This file contains complete microarray data for PADI4 inhibition in ES cells. (XLSX 7756 kb)
This file contains the complete Mass Spec dataset. (XLSX 31 kb)
Phase contrast and fluorescence time-lapse videos for the assessment of Oct4-GFP reporter expression after reprogramming of pre-iPS cells stably expressing Padi4 and Ctrl shRNAs. (MOV 1463 kb)
About this article
Cite this article
Christophorou, M., Castelo-Branco, G., Halley-Stott, R. et al. Citrullination regulates pluripotency and histone H1 binding to chromatin. Nature 507, 104–108 (2014). https://doi.org/10.1038/nature12942
Biochimica et Biophysica Acta (BBA) - Reviews on Cancer (2021)
Nature Communications (2021)
Molecular Cell (2021)
Epigenetic regulation of replication origin assembly: A role for histone H1 and chromatin remodeling factors