The histone chaperone CAF-1 safeguards somatic cell identity

Journal name:
Nature
Volume:
528,
Pages:
218–224
Date published:
DOI:
doi:10.1038/nature15749
Received
Accepted
Published online

Abstract

Cellular differentiation involves profound remodelling of chromatic landscapes, yet the mechanisms by which somatic cell identity is subsequently maintained remain incompletely understood. To further elucidate regulatory pathways that safeguard the somatic state, we performed two comprehensive RNA interference (RNAi) screens targeting chromatin factors during transcription-factor-mediated reprogramming of mouse fibroblasts to induced pluripotent stem cells (iPS cells). Subunits of the chromatin assembly factor-1 (CAF-1) complex, including Chaf1a and Chaf1b, emerged as the most prominent hits from both screens, followed by modulators of lysine sumoylation and heterochromatin maintenance. Optimal modulation of both CAF-1 and transcription factor levels increased reprogramming efficiency by several orders of magnitude and facilitated iPS cell formation in as little as 4 days. Mechanistically, CAF-1 suppression led to a more accessible chromatin structure at enhancer elements early during reprogramming. These changes were accompanied by a decrease in somatic heterochromatin domains, increased binding of Sox2 to pluripotency-specific targets and activation of associated genes. Notably, suppression of CAF-1 also enhanced the direct conversion of B cells into macrophages and fibroblasts into neurons. Together, our findings reveal the histone chaperone CAF-1 to be a novel regulator of somatic cell identity during transcription-factor-induced cell-fate transitions and provide a potential strategy to modulate cellular plasticity in a regenerative setting.

At a glance

Figures

  1. Arrayed and multiplexed shRNAmiR screening strategies to identify suppressors of reprogramming.
    Figure 1: Arrayed and multiplexed shRNAmiR screening strategies to identify suppressors of reprogramming.

    a, b, Schematic of arrayed (a) and multiplexed (b) RNAi screens. Dox, doxycycline. c, Results from arrayed screen, depicting average reprogramming efficiency ratios of two biological replicates normalized to Renilla (Ren.713) shRNA control. d, Heatmap depicting enrichment of selected shRNAs (shown in rows, ordered by gene symbol) over all 96 replicates (columns). e, Scatter plot representing sum score of enriched shRNAs across all replicates. f, Western blot analysis confirming shRNA suppression of CAF-1 p150 (Chaf1a), CAF-1 p60 (Chaf1b) and Ube2i at day 3 of reprogramming (see Supplementary Fig. 1 for full scans). g, Validation of hits from multiplex screen. Values are the mean from biological triplicates; error bars indicate standard deviation (*P < 0.05; **P < 0.01; ***P < 0.001).

  2. CAF-1 suppression accelerates reprogramming and yields developmentally competent iPS cells.
    Figure 2: CAF-1 suppression accelerates reprogramming and yields developmentally competent iPS cells.

    a, Generation of iPS-cell-derived chimaeras using indicated shRNAmiRs. b, Combinatorial RNAi studies involving pairs of shRNAmiRs sequentially transduced 5 and 1 day before OKSM induction. Reprogramming efficiencies are shown as ratio of Oct4–GFP+ to Oct4–GFP cells at day 11 relative to an empty vector control. c, Flow cytometry plots of representative samples used for b. d, e, Time-course analysis of Oct4–GFP (d) and Nanog (e) expression during reprogramming of MEFs transduced with indicated shRNAmiRs. f, Establishment of transgene-independent Oct4−GFP+ iPS cells in the presence of indicated shRNAmiRs (colour code as in d). Samples were induced with doxycycline for indicated number of days and analysed at day 13. g, Expression dynamics of reprogramming markers Epcam and Oct4–tomato after 4 and 6 days of OKSM expression (media supplemented with 2i, ascorbate and Dot1l inhibitor). h, Alkaline phosphatase (AP)-positive iPS cell colonies scored at day 11 after 4 or 6 days of OKSM expression (representative example from two biological replicates and three technical replicates).

  3. Enhanced reprogramming depends on optimal CAF-1 and OKSM dosage.
    Figure 3: Enhanced reprogramming depends on optimal CAF-1 and OKSM dosage.

    a, Comparison of reprogramming efficiency upon Chaf1a knockdown using MEFs carrying one or two copies of Col1a1::tetOP-OKSM and R26-M2rtTA. Colonies were scored at day 10 following 6 days of OKSM induction and 4 days of culture in the absence of doxycyline. b, Quantification of data shown in a. Values are the mean of biological triplicates; error bars indicate standard deviation. c, d, Effect of CAF-1 suppression on reprogramming efficiency when directly infecting MEFs with lentiviral vectors achieving medium (c) or high (d) OKSM expression levels, as determined by flow cytometry for Oct4–GFP at day 11. Values are the mean from biological triplicates; error bars indicate standard deviation. e, Influence of duration and degree of Chaf1a suppression on reprogramming potential of MEFs carrying doxycycline-inducible shRNA cassette (top), as determined by immunocytochemistry for Nanog at day 9. Data points represent single experiment. f, Comparison of reprogramming efficiencies when using shRNAs or sgRNAs targeting Chaf1a, as determined by flow cytometry for Oct4–GFP after 7 days of doxycycline exposure and 4 days of doxycycline-independent growth. Values are the mean from biological triplicates; error bars indicate standard deviation.

  4. CAF-1 suppression enhances reprogramming in different cell conversion systems.
    Figure 4: CAF-1 suppression enhances reprogramming in different cell conversion systems.

    a, Reprogramming of fetal haematopoietic stem and progenitor cells (HSP cells) into iPS cells. b, Flow cytometric analysis of Pecam expression during HSP cell reprogramming using indicated shRNAs. c, Quantification of data shown in b. Values represent fold-change expression differences between experimental and control samples using geometric mean. Data were obtained from one experiment using two different Chaf1a shRNAs. d, Transdifferentiation of MEFs into induced neurons. e, Representative image of Map2+ induced neurons after 13 days of transdifferentiation. Scale bars, 100 μM. f, Quantification of transdifferentiation efficiency (n = 5 independent experiments; values are mean ± standard deviation; unpaired t-test; **P = 0.0075). g, Transdifferentiation of pre-B cells into macrophages. h, Activation of macrophage markers Cd14 and Mac1 in representative samples at indicated time points. i, Cd14 and Mac1 expression levels in indicated samples (values represent fold-change expression differences between experimental and control samples using geometric mean; n = 2 independent viral transductions).

  5. CAF-1 suppression affects chromatin dynamics and facilitates activation of pluripotency genes during iPS cell generation.
    Figure 5: CAF-1 suppression affects chromatin dynamics and facilitates activation of pluripotency genes during iPS cell generation.

    a, ATAC-seq analysis of ES-cell-specific enhancers and promoters at day 3 of reprogramming. Shown are merged data for Chaf1a.164 and Chaf1a.2120 shRNA-infected cells (P > 0.5 and P < 10−15 between Chaf1a and Renilla shRNAs for promoters and enhancers, respectively; n denotes number of examined promoter and enhancer elements). b, Sox2 ChIP-seq analysis of ES-cell-specific enhancers and promoters at day 3 of reprogramming using weak (shRNA no. 2120) and strong (shRNA no. 164) Chaf1a hairpin (P < 1 × 10−15 for both shRNAs; see a for definition of n). c, Representative ATAC-seq and Sox2 ChIP-seq peaks at the Sall1 super-enhancer (y axis: tag density profiles). d, H3K9me3 ChIP-seq analysis of reprogramming-resistant regions (RRRs)29 after 0 and 3 days of OKSM expression. Heatmap shows all RRRs (rows); box plots show individual RRRs between day 0 and 3 in Chaf1a knockdown cells (P < 0.05 for both shRNAs). e, Chromatin accessibility at days 0, 3 and 6 for genes that become transcriptionally upregulated in Chaf1a shRNA-treated cells by day 6 (*P < 0.05; **P < 0.01). f, Chromatin in vivo assay (CiA) to directly measure effect of CAF-1 suppression on transcriptional activity of endogenous Oct4 locus in fibroblasts upon overexpression of Gal4–VP16 fusion protein targeted to the Oct4 promoter. g, Summary and model. TF, transcription factor; Pol II, RNA polymerase II.

  6. Validation of hits from chromatin-focused shRNA screens.
    Extended Data Fig. 1: Validation of hits from chromatin-focused shRNA screens.

    a, Quantitative RT–PCR analysis to confirm suppression of Chaf1a and Chaf1b expression with miR-30-based vectors from arrayed screen. Sh Chaf1a pool, sh Chaf1b pool and sh CAF-1 pool denote pools of shRNAs targeting either Chaf1a, Chaf1b or both. b, Western blot analysis to confirm knockdown of CAF-1 components using the top-scoring miR-30-based shRNAs from arrayed screen (see Supplementary Fig. 1 for full scans). c, Quantification of data shown in Fig. 1f. d, Quantitative RT–PCR analysis confirming knockdown with top-scoring miR-E-based shRNAmiRs targeting Chaf1a, Chaf1b or Ube2i from the multiplexed screen. Error bars show s.d. from biological triplicates. RNA and protein were extracted from reprogrammable MEFs 72 h after doxycycline induction in panels ad. e, Suppression of CAF-1 components, Ube2i and Setdb2 enhances reprogramming in the presence or absence of ascorbic acid (AA) as well as in serum replacement media containing LIF (SR-LIF). Oct4–GFP+ cells were scored by flow cytometry on day 11 after 7 days of OKSM induction and 4 days of transgene-independent growth. Error bars show s.d. from biological triplicates. f, Number of doxycycline-independent, alkaline phosphatase (AP)-positive colonies emerging two weeks after plating 10,000 reprogrammable MEFs carrying shRNA vectors against indicated targets and cultured in serum replacement media containing 2i (SR-2i), n = 1 experiment. g, Effect of suppressing SUMO E2 ligase Ube2i, E1 ligases Sae1 and Uba2 on iPS cell formation. Shown is fraction of Oct4–GFP+ cells at day 11 (7 days of OKSM induction, 4 days of transgene-independent growth). Error bars depict s.d. from biological triplicates.

  7. Germline transmission of iPS cells, genetic interaction of shRNA hits and effect of CAF-1 or Ube2i suppression on reprogramming dynamics.
    Extended Data Fig. 2: Germline transmission of iPS cells, genetic interaction of shRNA hits and effect of CAF-1 or Ube2i suppression on reprogramming dynamics.

    a, Germline transmission of agouti chimaeras generated from iPS cells using doxycycline-inducible shRNA vectors targeting Chaf1a, Chaf1b or Ube2i. Germline transmission was determined by scoring for agouti coat colour offspring upon breeding chimaeras with albino females. Germline transmission was observed in 8/8, 4/4 and 6/8 cases for Chaf1a iPS-cell-derived chimaeras, in 7/7, 4/4, 7/7 and 9/9 cases for Chaf1b iPS-cell-derived chimaeras, and in 5/5, 7/7 and 5/5 cases for Ube2i iPS-cell-derived chimaeras. b, Table summarizing effects of co-suppressing pairs of targets on emergence of Oct4–GFP+ cells, shown as the ratio of Oct4–GFP+ to Oct4–GFP cells relative to an empty vector control. Experiment equivalent to Fig. 2b except that second shRNAs were transduced two days after induction of reprogramming. c, Representative FACS plots showing effects of Chaf1a/b or Ube2i suppression on emergence of Oct4–GFP+ cells at days 7, 9 and 11 of OKSM expression. Histogram plots show fraction of Nanog+ cells within Oct4–GFP+ cells.

  8. Effect of CAF-1 suppression on OKSM levels and cellular growth, and shRNA rescue experiment.
    Extended Data Fig. 3: Effect of CAF-1 suppression on OKSM levels and cellular growth, and shRNA rescue experiment.

    a, Quantitative RT–PCR for transgenic OKSM expression using reprogrammable MEFs transduced with indicated shRNA vectors. Error bars show s.d. from biological triplicates. b, RNA-seq analysis of OKSM transgene expression in reprogrammable MEFs transduced with Renilla and Chaf1a shRNAs and exposed to doxycycline for 0, 3 or 6 days. Error bars indicate s.d. from biological triplicates. c, Western blot analysis for Sox2 and Tbp (loading control) in reprogrammable MEFs transduced with shRNA vectors targeting Renilla (Ren.713) or different CAF-1 components and exposed to doxycycline for 3 days (see Supplementary Fig. 1 for full scans). The same membrane was probed with anti-CAF-1 p150 and anti-CAF-1 p60 antibody to confirm knockdown (data not shown). d, Rescue experiment to demonstrate specificity of Chaf1b.367 shRNA vector. Reprogrammable MEFs carrying Oct4–tomato knock-in reporter were infected with lentiviral vectors expressing either EGFP or human CAF-1 p60 (CHAF1B) before transducing cells with Renilla or Cha1fb.367 shRNAs and applying doxycycline for 6 days. Colonies were counted at day 11. Note that CAF-1 p60 overexpression attenuates enhanced reprogramming elicited by Chaf1b suppression. e, f, Competitive proliferation assay between shRNA vector-infected and non-infected reprogrammable cells using indicated shRNAs in the presence or absence of doxycycline (OKSM expression). Note that CAF-1 suppression does not substantially affect the proliferation potential of reprogrammable MEFs after 1–3 days of doxycycline (OKSM) induction while it impairs the long-term growth potential of uninduced MEFs. Data were normalized to cell counts in ‘no OKSM’ condition for e and ‘day 2’ time point for f. Error bars show s.d. from biological triplicates.

  9. Confirmation of CAF-1 reprogramming phenotype with alternative transgenic and non-transgenic vector systems.
    Extended Data Fig. 4: Confirmation of CAF-1 reprogramming phenotype with alternative transgenic and non-transgenic vector systems.

    a, Alkaline phosphatase (AP)-positive, transgene-independent iPS cell colonies at day 14 following transduction of R26-M2rtTA MEFs with tetO-STEMCCA lentiviral OKSM expression vector and either Chaf1a.164 or Ren.713 shRNA vectors and treatment with high (2 μg ml−1) or low (0.2 μg ml−1) doses of doxycycline for 10 days. b, Quantification of data shown in a. Experiment was performed at 3 different plating densities (n = 1 experiment per density), representative data are shown. c, Comparison of reprogramming efficiencies between Col1a1::tetOP-OKSM; R26-M2rtTA reprogrammable MEFs and wild-type MEFs infected directly with OKSM-expressing lentiviral vectors containing either a strong Ef1a full-length promoter (Ef1a-OKSM long) or a weaker truncated promoter (Ef1a-OKSM short). TRE3G-OKSM is a lentiviral vector with a strong promoter, whose activity is downregulated over time upon infection of CAGS-rtTA3 transgenic MEFs (see below). Error bars show s.d. from biological triplicates. d, Quantitative RT–PCR data showing variability in OKSM expression levels over time using different vector systems. Cells were analysed after 3 and 6 days of infection (lentiviral vectors) or doxycycline exposure (reprogrammable MEFs). Error bars show s.d. from biological triplicates. OGR MEF, transgenic MEFs carrying Oct4–GFP and CAGS-rtTA3 alleles. e, Quantification of Oct4 protein levels by intracellular flow cytometry (top) and cellular granularity/complexity by side scatter (SSC) analysis of indicated samples (bottom). Error bars show s.d. from biological triplicates.

  10. Effects of CAF-1 dose on NIH3T3 growth and reprogramming potential.
    Extended Data Fig. 5: Effects of CAF-1 dose on NIH3T3 growth and reprogramming potential.

    a, Competitive proliferation assay to determine effect of indicated Chaf1a and Chaf1b shRNA vectors on long-term growth potential of immortalized NIH3T3 cell line. Cells were infected with indicated constructs and the fraction of shRNA vector-positive cells was measured by flow cytometry at different time points. Data were normalized to cell counts at day 2 post-transduction. Rpa3.455, validated control shRNA targeting the broadly essential replication protein A3. Error bars show s.d. from biological triplicates. b, Histogram plots of MEFs harbouring R26-M2rtTA allele and either Col1a1::tetOP-miR30-tRFP-Ren.713 or Col1a1::tetOP-miR30-tRFP-Chaf1a.164 shRNA knock-in allele after transduction with pHAGE (Ef1a-OKSM) lentiviral vector and exposure of cells to different doses of doxycycline for 2, 4 and 6 days. Low doses of doxycycline (0.2 µg ml−1) result in lower expression of the shRNA miR cassettes than high doses of doxycycline (2 µg ml−1). c, Quantification of data shown in b using the geometric mean (n = 1 experiment for 3 indicated time points). d, Reprogramming efficiency of Col1a1::tetOP-miR30-tRFP-Chaf1a.164; R26-M2rtTA MEFs infected with pHAGE (Ef1a-OKSM) vector and induced with high (2 μg ml−1) or low (0.2 μg ml−1) doses of doxycycline for indicated number of days before scoring for Nanog+ iPS cells by immunocytochemistry on day 9. e, Classification of CRISPR/Cas9-induced mutations by sequence analysis of representative iPS cell clones (wt, wild type; indel, insertion/deletion; fs, frameshift; *, point mutation). f, Western blot analysis for CAF-1 subunits p150 and p60 in 6 representative iPS cell clones after CRISPR/Cas9-induced modifications of the Chaf1a locus (see Supplementary Fig. 1 for full scans). Wt/wt samples show unmodified wild-type control samples.

  11. Effect of CAF-1 suppression on HSP cell reprogramming and transdifferentiation.
    Extended Data Fig. 6: Effect of CAF-1 suppression on HSP cell reprogramming and transdifferentiation.

    a, Gating strategy for determining Pecam+ fraction (shaded area) in panel b; data identical to Fig. 4c. b, Quantification of the fraction of Pecam+ cells at day 4 and day 6 of reprogramming. Data obtained from one experiment using two different Chaf1 shRNAs. c, Transgene dependence assay during the reprogramming of haematopoietic stem and progenitor cells (HSP cells) into iPS cells in the presence of Chaf1a or Renilla shRNAs. Doxycycline pulses were given for 3 or 6 days and alkaline phosphatase (AP)-positive colonies were scored at day 10. d, Quantitative RT–PCR analysis of Chaf1a expression to confirm knockdown after 3 days of doxycycline induction, that is, coexpression of shRNAmiR and Ascl1 (n = 4 independent infections of the same Col1a1::tetOP-Chaf1a.164 shRNA MEF line; mean value ± s.d.). e, Gating strategy for determining Cd14+ and Mac1+ fractions (shaded area) shown in f; data identical to Fig. 4g. Positive gates were based on untreated (0 h) control cells. f, Quantification of the fraction of Cd14+ and Mac1+ cells at 0, 24 and 48 h of transdifferentiation using indicated CAF-1 shRNA or empty control vector (n = 2 independent infections; rep, replicate). g, Quantitative RT–PCR analysis of Chaf1a and Chaf1b expression to confirm knockdown in transduced pre-B cell line before induction of transdifferentiation (kd/ctrl, knockdown/empty vector control; n = 1 experiment, representative of 2 independent infections).

  12. CAF-1 suppression promotes chromatin accessibility at enhancer elements.
    Extended Data Fig. 7: CAF-1 suppression promotes chromatin accessibility at enhancer elements.

    a, Experimental outline and assays (SONO-seq, ATAC-seq, Sox2 ChIP-seq, H3K9me3 ChIP-seq, microarrays and RNA-seq) to dissect effect of CAF-1 suppression on chromatin accessibility, transcription factor binding, heterochromatin patterns and gene expression. Assays were performed either in early reprogramming intermediates (day 3) or throughout the reprogramming time course (ATAC-seq and gene expression). b, SONO-seq analysis of CAF-1 knockdown and control cells at day 3 of reprogramming to determine accessible chromatin regions across promoters (n = 5,513) and ES-cell-specific enhancers (n = 14,265). CAF-1 shRNA vectors Chaf1a.164, Chaf1a.2120, Chaf1b.365 and Chaf1b.1221 were pooled for this experiment. c, ATAC-seq peak distribution across different genomic features. Shown is classification of peaks that are gained in CAF-1 knockdown cells compared to Renilla. d, ATAC-seq analysis of Chaf1a and Renilla control cells at day 3 of reprogramming to measure global chromatin accessibility over pluripotency-specific super-enhancer elements. ATAC-seq data from Chaf1a.164 shRNA- and Chaf1a.2120 shRNA-transduced cells were merged for this analysis. e, ATAC-seq accessibility maps at super-enhancer elements associated with the Sall4 locus. Shaded grey bars highlight more accessible sites in Chaf1a knockdown samples at days 3 and 6 of reprogramming compared to Renilla shRNA controls. f, ATAC-seq analysis of Chaf1a and control cells at day 3 of reprogramming to measure global chromatin accessibility over lineage-specific super-enhancer elements43 (C2C12, myoblast cell line; proB, progenitor B cells; Th, T helper cells). The n denotes number of examined enhancer elements for each cell type. ATAC-seq data from Chaf1a.164 shRNA- and Chaf1a.2120 shRNA-transduced cells were merged for this analysis.

  13. CAF-1 suppression facilitates Sox2 binding to chromatin.
    Extended Data Fig. 8: CAF-1 suppression facilitates Sox2 binding to chromatin.

    a, Sox2 ChIP-seq enrichment across pluripotency-specific super-enhancer elements at day 3 of reprogramming in the presence of indicated shRNA vectors. b, Venn diagram depicting shared and unique Sox2 targets in Chaf1a and Renilla knockdown cells. c, Bar graph shows the number and fraction of ES-cell-specific Sox2 targets (blue colour) among Sox2-bound sites that are unique to Chaf1a or Renilla knockdown cells at day 3 of OKSM expression. d, Sox2 ChIP-seq analysis of Chaf1a and control shRNA-infected cells at day 3 of reprogramming to determine enrichment of Sox2 binding across lineage-specific super-enhancer elements43 (C2C12, myoblast cell line; proB, progenitor B cells; Th, T helper cells; P value <10−15 for all comparisons between Chaf1a knockdown cells and control).

  14. CAF-1 suppression induces specific depletion of H3K9me3 at somatic heterochromatin domains.
    Extended Data Fig. 9: CAF-1 suppression induces specific depletion of H3K9me3 at somatic heterochromatin domains.

    a, Scatter plots comparing H3K9me3 enrichment nearby ATAC-seq sensitive and super-enhancer regions between control (Ren.713) and Chaf1a knockdown cells (Chaf1a.164 and Chaf1a.2122) at day 3 of reprogramming. Values reflect normalized H3K9me3 ChIP signal (IP/input) for 5-kb genomic regions overlapping ATAC-seq sensitive regions (red), super-enhancer regions (orange) and regions within 50-kb upstream and downstream of super-enhancers (black). b, Scatter plots comparing H3K9me3 enrichment over transposable element (TE) families in control and Chaf1a knockdown cells at day 3 of reprogramming. Values reflect normalized H3K9me3 ChIP-seq signal (IP/input) over families of TEs in the mouse genome. c, Heatmap shows the relative changes (z-normalized) of TE family expression as estimated by RNA sequencing in control and Chaf1a knockdown cells at day 0, 3 and 6 of reprogramming. Data are clustered using the k-means algorithm. d, Cumulative histogram showing the relative fraction of reprogramming-resistant regions (RRRs)29 (x axis) that display negative or positive enrichment (fold change) of average H3K9me3 signal at day 3 of reprogramming in control and Chaf1a knockdown cells. Note that more RRRs exhibit depletion of H3K9me3 in Chaf1a knockdown samples. e, H3K9me3 ChIP-seq analysis of RRRs after 0 and 3 days of reprogramming. Box plots depict representative RRRs on chromosome 7 (P < 0.05 for both shRNAs). See also Fig. 5d. f, Histogram plot showing activation of UAS–Oct4–GFP transgene upon suppression of Chaf1b (shRNA+ line) in the presence of Gal4–VP16 fusion protein. See Fig. 5f for quantification.

Accession codes

Primary accessions

Gene Expression Omnibus

References

  1. Lee, T. I. & Young, R. A. Transcriptional regulation and its misregulation in disease. Cell 152, 12371251 (2013)
  2. Takahashi, K. & Yamanaka, S. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126, 663676 (2006)
  3. Vierbuchen, T. & Wernig, M. Molecular roadblocks for cellular reprogramming. Mol. Cell 47, 827838 (2012)
  4. Yang, C. S., Chang, K. Y. & Rana, T. M. Genome-wide functional analysis reveals factors needed at the transition steps of induced reprogramming. Cell Rep. 8, 327337 (2014)
  5. Mikkelsen, T. S. et al. Dissecting direct reprogramming through integrative genomic analysis. Nature 454, 4955 10.1038/nature07056 (2008)
  6. Onder, T. T. et al. Chromatin-modifying enzymes as modulators of reprogramming. Nature 483, 598602 (2012)
  7. Rais, Y. et al. Deterministic direct reprogramming of somatic cells to pluripotency. Nature 502, 6570 (2013)
  8. Dejosez, M., Ura, H., Brandt, V. L. & Zwaka, T. P. Safeguards for cell cooperation in mouse embryogenesis shown by genome-wide cheater screen. Science 341, 15111514 (2013)
  9. Qin, H. et al. Systematic identification of barriers to human iPSC generation. Cell 158, 449461 (2014)
  10. Stadtfeld, M., Maherali, N., Borkent, M. & Hochedlinger, K. A reprogrammable mouse strain from gene-targeted embryonic stem cells. Nature Methods 7, 5355 (2010)
  11. Zuber, J. et al. RNAi screen identifies Brd4 as a therapeutic target in acute myeloid leukaemia. Nature 478, 524528 (2011)
  12. Fellmann, C. et al. An optimized microRNA backbone for effective single-copy RNAi. Cell Rep. 5, 17041713 (2013)
  13. Smith, S. & Stillman, B. Purification and characterization of CAF-I, a human cell factor required for chromatin assembly during DNA replication in vitro. Cell 58, 1525 (1989)
  14. Chen, J. et al. H3K9 methylation is a barrier during somatic cell reprogramming into iPSCs. Nature Genet. 45, 3442 (2013)
  15. Soufi, A., Donahue, G. & Zaret, K. S. Facilitators and impediments of the pluripotency reprogramming factors’ initial engagement with the genome. Cell 151, 9941004 (2012)
  16. Sridharan, R. et al. Proteomic and genomic approaches reveal critical functions of H3K9 methylation and heterochromatin protein-1γ in reprogramming to pluripotency. Nature Cell Biol. 15, 872882 (2013)
  17. Krizhanovsky, V. & Lowe, S. W. Stem cells: the promises and perils of p53. Nature 460, 10851086 (2009)
  18. Polo, J. M. et al. A molecular roadmap of reprogramming somatic cells into iPS cells. Cell 151, 16171632 (2012)
  19. Quivy, J. P., Gerard, A., Cook, A. J., Roche, D. & Almouzni, G. The HP1-p150/CAF-1 interaction is required for pericentric heterochromatin replication and S-phase progression in mouse cells. Nature Struct. Mol. Biol. 15, 972979 (2008)
  20. Hoek, M. & Stillman, B. Chromatin assembly factor 1 is essential and couples chromatin assembly to DNA replication in vivo. Proc. Natl Acad. Sci. USA 100, 1218312188 (2003)
  21. Houlard, M. et al. CAF-1 is essential for heterochromatin organization in pluripotent embryonic cells. PLoS Genet. 2, e181 (2006)
  22. Ye, X. et al. Defective S phase chromatin assembly causes DNA damage, activation of the S phase checkpoint, and S phase arrest. Mol. Cell 11, 341351 (2003)
  23. Rolef Ben-Shahar, T. et al. Two fundamentally distinct PCNA interaction peptides contribute to chromatin assembly factor 1 function. Mol. Cell. Biol. 29, 63536365 (2009)
  24. Chanda, S. et al. Generation of induced neuronal cells by the single reprogramming factor ASCL1. Stem Cell Rep. 3, 282296 (2014)
  25. Bussmann, L. H. et al. A robust and highly efficient immune cell reprogramming system. Cell Stem Cell 5, 554566 (2009)
  26. Auerbach, R. K. et al. Mapping accessible chromatin regions using Sono-seq. Proc. Natl Acad. Sci. USA 106, 1492614931 (2009)
  27. Buenrostro, J. D., Giresi, P. G., Zaba, L. C., Chang, H. Y. & Greenleaf, W. J. Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin, DNA-binding proteins and nucleosome position. Nature Methods 10, 12131218 (2013)
  28. Huang, H. et al. Drosophila CAF-1 regulates HP1-mediated epigenetic silencing and pericentric heterochromatin stability. J. Cell Sci. 123, 28532861 (2010)
  29. Matoba, S. et al. Embryonic development following somatic cell nuclear transfer impeded by persisting histone methylation. Cell 159, 884895 (2014)
  30. Hathaway, N. A. et al. Dynamics and memory of heterochromatin in living cells. Cell 149, 14471460 (2012)
  31. Ishiuchi, T. et al. Early embryonic-like cells are induced by downregulating replication-dependent chromatin assembly. Nature Struct. Mol. Biol. (2015)
  32. Ray-Gallet, D. et al. Dynamics of histone H3 deposition in vivo reveal a nucleosome gap-filling mechanism for H3.3 to maintain chromatin integrity. Mol. Cell 44, 928941 (2011)
  33. Jullien, J. et al. HIRA dependent H3.3 deposition is required for transcriptional reprogramming following nuclear transfer to Xenopus oocytes. Epigenetics Chromatin 5, 17 (2012)
  34. Wen, D., Banaszynski, L. A., Rosenwaks, Z., Allis, C. D. & Rafii, S. H3.3 replacement facilitates epigenetic reprogramming of donor nuclei in somatic cell nuclear transfer embryos. Nucleus 5, 369375 (2014)
  35. Premsrirut, P. K. et al. A rapid and scalable system for studying gene function in mice using conditional RNA interference. Cell 145, 145158 (2011)
  36. Zuber, J. et al. Toolkit for evaluating genes required for proliferation and survival using tetracycline-regulated RNAi. Nature Biotechnol. 29, 7983 (2011)
  37. Lengner, C. J. et al. Oct4 expression is not required for mouse somatic stem cell self-renewal. Cell Stem Cell 1, 403415 (2007)
  38. Bernstein, B. E. et al. Genomic maps and comparative analysis of histone modifications in human and mouse. Cell 120, 169181 (2005)
  39. Langmead, B., Trapnell, C., Pop, M. & Salzberg, S. L. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol. 10, R25 (2009)
  40. Kharchenko, P. V., Tolstorukov, M. Y. & Park, P. J. Design and analysis of ChIP-seq experiments for DNA-binding proteins. Nature Biotechnol. 26, 13511359 (2008)
  41. Shen, Y. et al. A map of the cis-regulatory sequences in the mouse genome. Nature 488, 116120 (2012)
  42. Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows–Wheeler transform. Bioinformatics 25, 17541760 (2009)
  43. Whyte, W. A. et al. Master transcription factors and mediator establish super-enhancers at key cell identity genes. Cell 153, 307319 (2013)
  44. Sabo, P. J. et al. Discovery of functional noncoding elements by digital analysis of chromatin structure. Proc. Natl Acad. Sci. USA 101, 1683716842 (2004)
  45. Ross-Innes, C. S. et al. Differential oestrogen receptor binding is associated with clinical outcome in breast cancer. Nature 481, 389393 10.1038/nature10730 (2012)
  46. Robinson, M. D., McCarthy, D. J. & Smyth, G. K. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139140 (2010)
  47. Marson, A. et al. Connecting microRNA genes to the core transcriptional regulatory circuitry of embryonic stem cells. Cell 134, 521533 (2008)
  48. Quinlan, A. R. & Hall, I. M. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics 26, 841842 (2010)
  49. Pezic, D., Manakov, S. A., Sachidanandam, R. & Aravin, A. A. piRNA pathway targets active LINE1 elements to establish the repressive H3K9me3 mark in germ cells. Genes Dev. 28, 14101428 (2014)
  50. Smyth, G. K. Linear models and empirical Bayes methods for assessing differential expression in microarray experiments. Stat. Appl. Genet. Mol. Biol. 3, 125 (2004)
  51. Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl Acad. Sci. USA 102, 1554515550 (2005)
  52. Dowen, J. M. et al. Control of cell identity genes occurs in insulated neighborhoods in mammalian chromosomes. Cell 159, 374387 (2014)
  53. Davis, M. P., van Dongen, S., Abreu-Goodger, C., Bartonicek, N. & Enright, A. J. Kraken: a set of tools for quality control and analysis of high-throughput sequence data. Methods 63, 4149 (2013)
  54. Anders, S. & Huber, W. Differential expression analysis for sequence count data. Genome Biol. 11, R106 (2010)

Download references

Author information

  1. These authors contributed equally to this work.

    • Sihem Cheloufi &
    • Ulrich Elling
  2. These authors jointly supervised this work.

    • Johannes Zuber &
    • Konrad Hochedlinger

Affiliations

  1. Department of Molecular Biology, Cancer Center and Center for Regenerative Medicine, Massachusetts General Hospital, Boston, Massachusetts 02114, USA

    • Sihem Cheloufi,
    • Daniel J. Wesche,
    • Nadezhda Abazova,
    • Max Hogue,
    • Justin Brumbaugh &
    • Konrad Hochedlinger
  2. Department of Stem Cell and Regenerative Biology and Harvard Stem Cell Institute, Cambridge, Massachusetts 02138, USA

    • Sihem Cheloufi,
    • Danielle Tenen,
    • Daniel J. Wesche,
    • Nadezhda Abazova,
    • Max Hogue,
    • Justin Brumbaugh,
    • John Rinn &
    • Konrad Hochedlinger
  3. Howard Hughes Medical Institute, Chevy Chase, Maryland 20815, USA

    • Sihem Cheloufi,
    • Daniel J. Wesche,
    • Nadezhda Abazova,
    • Max Hogue,
    • Justin Brumbaugh,
    • Howard Y. Chang,
    • Scott W. Lowe &
    • Konrad Hochedlinger
  4. Institute of Molecular Biotechnology of the Austrian Academy of Sciences (IMBA), Vienna Biocenter (VBC), A-1030 Vienna, Austria

    • Ulrich Elling,
    • Maria Hubmann,
    • Daniel Wenzel,
    • Marietta Zinner,
    • Oliver Bell &
    • Josef M. Penninger
  5. Research Institute of Molecular Pathology (IMP), Vienna Biocenter (VBC), A-1030 Vienna, Austria

    • Barbara Hopfgartner,
    • Philipp Rathert,
    • Julian Jude,
    • Michaela Fellner &
    • Johannes Zuber
  6. Department of Biomedical Informatics, Harvard Medical School, Boston, Massachusetts 02115, USA

    • Youngsook L. Jung,
    • Francesco Ferrari &
    • Peter J. Park
  7. Division of Genetics, Brigham and Women’s Hospital, Boston, Massachusetts 02115, USA

    • Youngsook L. Jung,
    • Francesco Ferrari &
    • Peter J. Park
  8. Department of Cell Biology, Harvard Medical School, Boston, Massachusetts 02115, USA

    • Jernej Murn,
    • Aimee I. Badeaux,
    • Andres Blanco &
    • Yang Shi
  9. Division of Newborn Medicine, Boston Children’s Hospital, Boston, Massachusetts 02115, USA

    • Jernej Murn,
    • Aimee I. Badeaux,
    • Andres Blanco &
    • Yang Shi
  10. California Institute of Technology, Division of Biology and Biological Engineering, Pasadena, California 91125, USA

    • Maria Ninova &
    • Alexei Aravin
  11. Institute for Stem Cell Biology and Regenerative Medicine, Department of Pathology and Department of Bioengineering, Stanford University, Stanford, California 94305, USA

    • Cheen Euong Ang &
    • Marius Wernig
  12. Broad Institute of Massachusetts Institute of Technology and Harvard, Cambridge, Massachusetts 02142, USA

    • Danielle Tenen &
    • John Rinn
  13. Memorial Sloan Kettering Cancer Center, New York, New York 10065, USA

    • Nilgun Tasdemir &
    • Scott W. Lowe
  14. The Helen L. and Martin S. Kimmel Center for Biology and Medicine, Skirball Institute of Biomolecular Medicine, Department of Cell Biology, NYU School of Medicine, New York, New York 10016, USA

    • Simon E. Vidal &
    • Matthias Stadtfeld
  15. Center for Personal Dynamic Regulomes and Program in Epithelial Biology, Stanford University School of Medicine, Stanford, California 94305, USA

    • Howard Y. Chang
  16. Centre de Recherche, Institut Curie, 75248 Paris, France

    • Genevieve Almouzni

Contributions

Y. L. Jung and B. Hopfgartner contributed equally to this work. S.C., K.H., U.E. and J.Z. designed primary screens, analysed and interpreted data. S.C., J.M. and N.A. performed the arrayed screen and S.C. conducted follow-up cell biology and chromatin studies. U.E. and B.H. performed the multiplexed screen. U.E. performed validation experiments, genetic interaction assays and cell biology experiments with support from B.H., M.H. and D.W. Human reprogramming experiments were performed by S.C. and J.B.; N.T. and S.W.L. assisted in the generation of inducible Col1a1::tetOP-Chaf1a shRNA cell lines. S.C., A.I.B., A.B. and Y.S. performed B-cell to macrophage conversion experiments. C.E.A. and M.W. conducted MEF to induced neuron transdifferentiation experiments. Y.L.J., M.N., A.A., F.F. and P.J.P. performed bioinformatics analyses. M.H. and U.E. conducted the CiA assay with support from O.B. D.J.W. assisted with the SONO-seq experiments and H.Y.C. helped with the ATAC-seq assay. J.M., M.H. and M.Z. assisted with western blot and chromatin studies. D.T. and J.R. conducted ChIP experiments and library construction. M.S. and S.E.V. provided secondary Oct4–tdTomato MEFs. J.Z. and S.W.L. provided the arrayed library. J.Z. and P.R. designed the extended chromatin library. M.F., J.J. and B.H. generated lentiviral vectors and RNAi reagents. J.M.P. and G.A. provided intellectual support and mentoring. K.H., S.C., J.Z. and U.E. wrote the paper with input from all co-authors.

Competing financial interests

The authors declare no competing financial interests.

Corresponding authors

Correspondence to:

All SONO-seq, ATAC-seq, ChIP-seq, RNA-seq and microarray data have been deposited in the Gene Expression Omnibus database under accession number GSE66534.

Author details

Extended data figures and tables

Extended Data Figures

  1. Extended Data Figure 1: Validation of hits from chromatin-focused shRNA screens. (285 KB)

    a, Quantitative RT–PCR analysis to confirm suppression of Chaf1a and Chaf1b expression with miR-30-based vectors from arrayed screen. Sh Chaf1a pool, sh Chaf1b pool and sh CAF-1 pool denote pools of shRNAs targeting either Chaf1a, Chaf1b or both. b, Western blot analysis to confirm knockdown of CAF-1 components using the top-scoring miR-30-based shRNAs from arrayed screen (see Supplementary Fig. 1 for full scans). c, Quantification of data shown in Fig. 1f. d, Quantitative RT–PCR analysis confirming knockdown with top-scoring miR-E-based shRNAmiRs targeting Chaf1a, Chaf1b or Ube2i from the multiplexed screen. Error bars show s.d. from biological triplicates. RNA and protein were extracted from reprogrammable MEFs 72 h after doxycycline induction in panels ad. e, Suppression of CAF-1 components, Ube2i and Setdb2 enhances reprogramming in the presence or absence of ascorbic acid (AA) as well as in serum replacement media containing LIF (SR-LIF). Oct4–GFP+ cells were scored by flow cytometry on day 11 after 7 days of OKSM induction and 4 days of transgene-independent growth. Error bars show s.d. from biological triplicates. f, Number of doxycycline-independent, alkaline phosphatase (AP)-positive colonies emerging two weeks after plating 10,000 reprogrammable MEFs carrying shRNA vectors against indicated targets and cultured in serum replacement media containing 2i (SR-2i), n = 1 experiment. g, Effect of suppressing SUMO E2 ligase Ube2i, E1 ligases Sae1 and Uba2 on iPS cell formation. Shown is fraction of Oct4–GFP+ cells at day 11 (7 days of OKSM induction, 4 days of transgene-independent growth). Error bars depict s.d. from biological triplicates.

  2. Extended Data Figure 2: Germline transmission of iPS cells, genetic interaction of shRNA hits and effect of CAF-1 or Ube2i suppression on reprogramming dynamics. (475 KB)

    a, Germline transmission of agouti chimaeras generated from iPS cells using doxycycline-inducible shRNA vectors targeting Chaf1a, Chaf1b or Ube2i. Germline transmission was determined by scoring for agouti coat colour offspring upon breeding chimaeras with albino females. Germline transmission was observed in 8/8, 4/4 and 6/8 cases for Chaf1a iPS-cell-derived chimaeras, in 7/7, 4/4, 7/7 and 9/9 cases for Chaf1b iPS-cell-derived chimaeras, and in 5/5, 7/7 and 5/5 cases for Ube2i iPS-cell-derived chimaeras. b, Table summarizing effects of co-suppressing pairs of targets on emergence of Oct4–GFP+ cells, shown as the ratio of Oct4–GFP+ to Oct4–GFP cells relative to an empty vector control. Experiment equivalent to Fig. 2b except that second shRNAs were transduced two days after induction of reprogramming. c, Representative FACS plots showing effects of Chaf1a/b or Ube2i suppression on emergence of Oct4–GFP+ cells at days 7, 9 and 11 of OKSM expression. Histogram plots show fraction of Nanog+ cells within Oct4–GFP+ cells.

  3. Extended Data Figure 3: Effect of CAF-1 suppression on OKSM levels and cellular growth, and shRNA rescue experiment. (243 KB)

    a, Quantitative RT–PCR for transgenic OKSM expression using reprogrammable MEFs transduced with indicated shRNA vectors. Error bars show s.d. from biological triplicates. b, RNA-seq analysis of OKSM transgene expression in reprogrammable MEFs transduced with Renilla and Chaf1a shRNAs and exposed to doxycycline for 0, 3 or 6 days. Error bars indicate s.d. from biological triplicates. c, Western blot analysis for Sox2 and Tbp (loading control) in reprogrammable MEFs transduced with shRNA vectors targeting Renilla (Ren.713) or different CAF-1 components and exposed to doxycycline for 3 days (see Supplementary Fig. 1 for full scans). The same membrane was probed with anti-CAF-1 p150 and anti-CAF-1 p60 antibody to confirm knockdown (data not shown). d, Rescue experiment to demonstrate specificity of Chaf1b.367 shRNA vector. Reprogrammable MEFs carrying Oct4–tomato knock-in reporter were infected with lentiviral vectors expressing either EGFP or human CAF-1 p60 (CHAF1B) before transducing cells with Renilla or Cha1fb.367 shRNAs and applying doxycycline for 6 days. Colonies were counted at day 11. Note that CAF-1 p60 overexpression attenuates enhanced reprogramming elicited by Chaf1b suppression. e, f, Competitive proliferation assay between shRNA vector-infected and non-infected reprogrammable cells using indicated shRNAs in the presence or absence of doxycycline (OKSM expression). Note that CAF-1 suppression does not substantially affect the proliferation potential of reprogrammable MEFs after 1–3 days of doxycycline (OKSM) induction while it impairs the long-term growth potential of uninduced MEFs. Data were normalized to cell counts in ‘no OKSM’ condition for e and ‘day 2’ time point for f. Error bars show s.d. from biological triplicates.

  4. Extended Data Figure 4: Confirmation of CAF-1 reprogramming phenotype with alternative transgenic and non-transgenic vector systems. (299 KB)

    a, Alkaline phosphatase (AP)-positive, transgene-independent iPS cell colonies at day 14 following transduction of R26-M2rtTA MEFs with tetO-STEMCCA lentiviral OKSM expression vector and either Chaf1a.164 or Ren.713 shRNA vectors and treatment with high (2 μg ml−1) or low (0.2 μg ml−1) doses of doxycycline for 10 days. b, Quantification of data shown in a. Experiment was performed at 3 different plating densities (n = 1 experiment per density), representative data are shown. c, Comparison of reprogramming efficiencies between Col1a1::tetOP-OKSM; R26-M2rtTA reprogrammable MEFs and wild-type MEFs infected directly with OKSM-expressing lentiviral vectors containing either a strong Ef1a full-length promoter (Ef1a-OKSM long) or a weaker truncated promoter (Ef1a-OKSM short). TRE3G-OKSM is a lentiviral vector with a strong promoter, whose activity is downregulated over time upon infection of CAGS-rtTA3 transgenic MEFs (see below). Error bars show s.d. from biological triplicates. d, Quantitative RT–PCR data showing variability in OKSM expression levels over time using different vector systems. Cells were analysed after 3 and 6 days of infection (lentiviral vectors) or doxycycline exposure (reprogrammable MEFs). Error bars show s.d. from biological triplicates. OGR MEF, transgenic MEFs carrying Oct4–GFP and CAGS-rtTA3 alleles. e, Quantification of Oct4 protein levels by intracellular flow cytometry (top) and cellular granularity/complexity by side scatter (SSC) analysis of indicated samples (bottom). Error bars show s.d. from biological triplicates.

  5. Extended Data Figure 5: Effects of CAF-1 dose on NIH3T3 growth and reprogramming potential. (474 KB)

    a, Competitive proliferation assay to determine effect of indicated Chaf1a and Chaf1b shRNA vectors on long-term growth potential of immortalized NIH3T3 cell line. Cells were infected with indicated constructs and the fraction of shRNA vector-positive cells was measured by flow cytometry at different time points. Data were normalized to cell counts at day 2 post-transduction. Rpa3.455, validated control shRNA targeting the broadly essential replication protein A3. Error bars show s.d. from biological triplicates. b, Histogram plots of MEFs harbouring R26-M2rtTA allele and either Col1a1::tetOP-miR30-tRFP-Ren.713 or Col1a1::tetOP-miR30-tRFP-Chaf1a.164 shRNA knock-in allele after transduction with pHAGE (Ef1a-OKSM) lentiviral vector and exposure of cells to different doses of doxycycline for 2, 4 and 6 days. Low doses of doxycycline (0.2 µg ml−1) result in lower expression of the shRNA miR cassettes than high doses of doxycycline (2 µg ml−1). c, Quantification of data shown in b using the geometric mean (n = 1 experiment for 3 indicated time points). d, Reprogramming efficiency of Col1a1::tetOP-miR30-tRFP-Chaf1a.164; R26-M2rtTA MEFs infected with pHAGE (Ef1a-OKSM) vector and induced with high (2 μg ml−1) or low (0.2 μg ml−1) doses of doxycycline for indicated number of days before scoring for Nanog+ iPS cells by immunocytochemistry on day 9. e, Classification of CRISPR/Cas9-induced mutations by sequence analysis of representative iPS cell clones (wt, wild type; indel, insertion/deletion; fs, frameshift; *, point mutation). f, Western blot analysis for CAF-1 subunits p150 and p60 in 6 representative iPS cell clones after CRISPR/Cas9-induced modifications of the Chaf1a locus (see Supplementary Fig. 1 for full scans). Wt/wt samples show unmodified wild-type control samples.

  6. Extended Data Figure 6: Effect of CAF-1 suppression on HSP cell reprogramming and transdifferentiation. (327 KB)

    a, Gating strategy for determining Pecam+ fraction (shaded area) in panel b; data identical to Fig. 4c. b, Quantification of the fraction of Pecam+ cells at day 4 and day 6 of reprogramming. Data obtained from one experiment using two different Chaf1 shRNAs. c, Transgene dependence assay during the reprogramming of haematopoietic stem and progenitor cells (HSP cells) into iPS cells in the presence of Chaf1a or Renilla shRNAs. Doxycycline pulses were given for 3 or 6 days and alkaline phosphatase (AP)-positive colonies were scored at day 10. d, Quantitative RT–PCR analysis of Chaf1a expression to confirm knockdown after 3 days of doxycycline induction, that is, coexpression of shRNAmiR and Ascl1 (n = 4 independent infections of the same Col1a1::tetOP-Chaf1a.164 shRNA MEF line; mean value ± s.d.). e, Gating strategy for determining Cd14+ and Mac1+ fractions (shaded area) shown in f; data identical to Fig. 4g. Positive gates were based on untreated (0 h) control cells. f, Quantification of the fraction of Cd14+ and Mac1+ cells at 0, 24 and 48 h of transdifferentiation using indicated CAF-1 shRNA or empty control vector (n = 2 independent infections; rep, replicate). g, Quantitative RT–PCR analysis of Chaf1a and Chaf1b expression to confirm knockdown in transduced pre-B cell line before induction of transdifferentiation (kd/ctrl, knockdown/empty vector control; n = 1 experiment, representative of 2 independent infections).

  7. Extended Data Figure 7: CAF-1 suppression promotes chromatin accessibility at enhancer elements. (273 KB)

    a, Experimental outline and assays (SONO-seq, ATAC-seq, Sox2 ChIP-seq, H3K9me3 ChIP-seq, microarrays and RNA-seq) to dissect effect of CAF-1 suppression on chromatin accessibility, transcription factor binding, heterochromatin patterns and gene expression. Assays were performed either in early reprogramming intermediates (day 3) or throughout the reprogramming time course (ATAC-seq and gene expression). b, SONO-seq analysis of CAF-1 knockdown and control cells at day 3 of reprogramming to determine accessible chromatin regions across promoters (n = 5,513) and ES-cell-specific enhancers (n = 14,265). CAF-1 shRNA vectors Chaf1a.164, Chaf1a.2120, Chaf1b.365 and Chaf1b.1221 were pooled for this experiment. c, ATAC-seq peak distribution across different genomic features. Shown is classification of peaks that are gained in CAF-1 knockdown cells compared to Renilla. d, ATAC-seq analysis of Chaf1a and Renilla control cells at day 3 of reprogramming to measure global chromatin accessibility over pluripotency-specific super-enhancer elements. ATAC-seq data from Chaf1a.164 shRNA- and Chaf1a.2120 shRNA-transduced cells were merged for this analysis. e, ATAC-seq accessibility maps at super-enhancer elements associated with the Sall4 locus. Shaded grey bars highlight more accessible sites in Chaf1a knockdown samples at days 3 and 6 of reprogramming compared to Renilla shRNA controls. f, ATAC-seq analysis of Chaf1a and control cells at day 3 of reprogramming to measure global chromatin accessibility over lineage-specific super-enhancer elements43 (C2C12, myoblast cell line; proB, progenitor B cells; Th, T helper cells). The n denotes number of examined enhancer elements for each cell type. ATAC-seq data from Chaf1a.164 shRNA- and Chaf1a.2120 shRNA-transduced cells were merged for this analysis.

  8. Extended Data Figure 8: CAF-1 suppression facilitates Sox2 binding to chromatin. (179 KB)

    a, Sox2 ChIP-seq enrichment across pluripotency-specific super-enhancer elements at day 3 of reprogramming in the presence of indicated shRNA vectors. b, Venn diagram depicting shared and unique Sox2 targets in Chaf1a and Renilla knockdown cells. c, Bar graph shows the number and fraction of ES-cell-specific Sox2 targets (blue colour) among Sox2-bound sites that are unique to Chaf1a or Renilla knockdown cells at day 3 of OKSM expression. d, Sox2 ChIP-seq analysis of Chaf1a and control shRNA-infected cells at day 3 of reprogramming to determine enrichment of Sox2 binding across lineage-specific super-enhancer elements43 (C2C12, myoblast cell line; proB, progenitor B cells; Th, T helper cells; P value <10−15 for all comparisons between Chaf1a knockdown cells and control).

  9. Extended Data Figure 9: CAF-1 suppression induces specific depletion of H3K9me3 at somatic heterochromatin domains. (317 KB)

    a, Scatter plots comparing H3K9me3 enrichment nearby ATAC-seq sensitive and super-enhancer regions between control (Ren.713) and Chaf1a knockdown cells (Chaf1a.164 and Chaf1a.2122) at day 3 of reprogramming. Values reflect normalized H3K9me3 ChIP signal (IP/input) for 5-kb genomic regions overlapping ATAC-seq sensitive regions (red), super-enhancer regions (orange) and regions within 50-kb upstream and downstream of super-enhancers (black). b, Scatter plots comparing H3K9me3 enrichment over transposable element (TE) families in control and Chaf1a knockdown cells at day 3 of reprogramming. Values reflect normalized H3K9me3 ChIP-seq signal (IP/input) over families of TEs in the mouse genome. c, Heatmap shows the relative changes (z-normalized) of TE family expression as estimated by RNA sequencing in control and Chaf1a knockdown cells at day 0, 3 and 6 of reprogramming. Data are clustered using the k-means algorithm. d, Cumulative histogram showing the relative fraction of reprogramming-resistant regions (RRRs)29 (x axis) that display negative or positive enrichment (fold change) of average H3K9me3 signal at day 3 of reprogramming in control and Chaf1a knockdown cells. Note that more RRRs exhibit depletion of H3K9me3 in Chaf1a knockdown samples. e, H3K9me3 ChIP-seq analysis of RRRs after 0 and 3 days of reprogramming. Box plots depict representative RRRs on chromosome 7 (P < 0.05 for both shRNAs). See also Fig. 5d. f, Histogram plot showing activation of UAS–Oct4–GFP transgene upon suppression of Chaf1b (shRNA+ line) in the presence of Gal4–VP16 fusion protein. See Fig. 5f for quantification.

Supplementary information

PDF files

  1. Supplementary Figure 1 (1.3 MB)

    This file contains full scans of Western Blot data for Figures 1f and Extended Data Figures 1b, 3c and 5f.

Excel files

  1. Supplementary Table 1 (130 KB)

    This file contains Primary data from the arrayed shRNAmiR screen including all 1,075 tested shRNAmiRs, their sequences, corresponding scores for each biological replicate (FC replicate A/B) and their average (FC avg), which are plotted.

  2. Supplementary Table 2 (2.8 MB)

    This file contains primary data from the multiplexed shRNAmiR screen. Sheet 1 includes all 5,049 tested shRNAmiRs, their sequences and corresponding average scores for each time point. Sheet 2 shows raw data as normalized reads for all biological replicates (3-A01-D12, 48 replicates with OKSM induction after 3 days of G418 selection; 6-A01-D12, 48 replicates with OKSM induction after 3 days of G418 selection). Fold-change (FC) ratios for each shRNA were calculated by dividing the normalized reads of each shRNA by the normalized reads in MEFs 3 days after viral transduction. The mean of all FC values for the day 3 and day 6 time point were used to calculate the average FC for both time points (AvgFC_D3/6). In addition, FC ratios were used to calculate scores for each shRNA in each replicate (default score=0; score=1 if FC>1, score=3 if FC>10), which were summed up over all replicates separately for the day 3 and day 6 time point (Score_D3/6).

  3. Supplementary Table 3 (12 KB)

    This file shows accessible super-enhancers in reprogramming intermediates upon CAF-1 suppression. Genomic coordinates of significantly more accessible super-enhancer (SE) sites in CAF-1 knockdown samples compared to Renilla control including log2 fold-enrichment scores (log2FC) and p-values for ATAC-seq data with indication of neighboring genes.

  4. Supplementary Table 4 (10 KB)

    This file shows correlation analysis between Sox2 binding sites and chromatin accessibility at super-enhancer sites in control and CAF-1 knockdown cells expressing OKSM. Sheet 1 shows Sox2 binding site counts in reprogramming intermediates (day 3) containing Chaf1a or Renilla shRNAs (unique and shared sites), ESC-specific sites, overlap with super-enhancers and positive correlation with ATAC-seq data. Sheet 2 depicts unique Sox2-bound super-enhancer sites in Chaf1a knockdown cells that are also more accessible by ATAC-Seq analysis, including genomic coordinates, neighboring genes and log2 fold-enrichment for both ChIP-seq and ATAC-seq data.

  5. Supplementary Table 5 (9.1 KB)

    This file shows significantly up-regulated genes in CAF-1 shRNA transduced MEFs relative to Renilla control by day 6 of reprogramming. Sheet1 contains genes that are either (1) significantly up-regulated at day 6 compared to day 3 in CAF-1 knockdown cells relative to Renilla control as determined by expression array data, or (2) significantly up-regulated at day 6 in CAF-1 shRNA transduced MEFs compared to Renilla control by RNA-seq analysis.

Additional data