SUPPLEMENRY INFORMION doi:.38/nature In vivo nucleosome mapping D4+ Lymphocytes radient-based and I-bead cell sorting D8+ Lymphocytes ranulocytes Lyse the cells Isolate and sequence mononucleosome cores Micrococcal nuclease Supplementary Figure. Schematic depiction of in vivo nucleosome mapping experiment. lood cells were isolated from a human donor blood and sorted into populations representing D4+ -cells, D8+ -cells and granulocytes. Nuclear chromatin was released by crushing the cells, followed by Micrococcal nuclease treatment. Mononucleosome fraction was isolated by gel electrophoresis and sequenced to high depth using SOLiD platform. WWW.NURE.OM/NURE
RESERH SUPPLEMENRY INFORMION nucleosome reconstitution experiment Recombinant purified histones Reconstituted histones Purified nucleosomes H H H3 H4 Dialysis Sheared human DN, (.6-.7 Kb) High salt solution, nuclesome / 85 bps MNase Library DN Unprotected DN is removed Nucleosomes occupy sequence-determined positions Supplementary Figure. Schematic representation of in vitro reconstitution experiment. Recombinant histones were assembled to produce the histone octamer particles. Human genomic DN was sheared to a range of.6-.7 Kb and combined with octamers at a ratio of one octamer per 85 bps of DN. he salt was gradually dialyzed away and unbound DN was removed by Micrococcal nuclease treatment. Nucleosome-bound DN was purified and sequenced on the SOLiD platform. WWW.NURE.OM/NURE
SUPPLEMENRY INFORMION RESERH Distances Reads Distogram calculation + - Phasogram calculation Phases (+) Reads (+) Reads (-) Phases (-) Supplementary Figure 3. Distograms and phasograms. () Schematic depiction of the distogram calculation. lue arcs represent recorded distances between nucleosome reads that map on opposite strands. Distance frequencies are represented as a histogram (distogram, see Fig. - of the main text). Distograms are used to reveal the existence of consistently positioned nucleosomes in the main data. () Schematic depiction of the phasogram calculation. lue arcs represent recorded phases between the nucleosome reads mapping on the same strand of the reference genome. Phase frequencies are represented as a histogram (phasogram, see Fig -D). Phasograms are used to reveal the existance of consistently spaced nucleosomes forming regular arrays. WWW.NURE.OM/NURE 3
RESERH SUPPLEMENRY INFORMION % positioned nucleosomes 5 % positioned nucleosomes ell ell Mapped cores.. Positioning.5 stringency Dyad Dyad.5 Dyad Dyad Supplementary Figure 4. Schematic depiction of the nucleosome positioning stringency metric. t the sites containing perfrectly positioning nucleosomes (panel ) the stringency values are. (% positioning), and at the sites containing two mutually exclusive nucleosome positions which are utilized with 5% frequency across cells (panel ), the stringency values are.5 (5% positioning frequency at each of the two sites). Nucleosome dyad positions are identified as the local maxima of the stringency profile (green arrows). 4 WWW.NURE.OM/NURE
SUPPLEMENRY INFORMION RESERH Positioned dyads.8.4. lobal in vivo positioning Difference enomic DN ranulocyte dyads M....3.4.5 Positioning stringency Fraction of genome covered.8.6.4. Difference enomic DN ranulocyte cores.....3.4.5.6 Positioning stringency Positioned dyads 3 M M M lobal in vitro positioning enomic DN dyads Difference...4.6.8 Positioning stringency Fraction of genome covered.8.6.4.. enomic DN cores Difference...4.6.8 Positioning stringency Supplementary Figure 5. enome-wide positioning of nucleosomes. () lobal in vivo nucleosome positioning of granulocytes. In both panels, X axis represents a range of positioning stringency cutoffs. In the left panel, Y axis represents the number of positioned dyads at a given positioning stringency cutoff. he red curve represents granulocyte data, the blue curve represents genomic DN control matched to the number of granulocyte reads, the green curve represents the difference curve that provides the number of statistically positioned dyads at a given stringency cutoff. In the right panel, Y axis represents the fraction of the genome covered by 47 bp nucleosome cores centered at the dyad positions exceeding a given stringency. he red curve represents granulocyte nucleosome data, the blue curve represents genomic control matching the granulocyte data read number, and the green curve represents the difference between granulocytes and control curves and gives the fraction of the genome covered by statistically positioned nucleosomes. () lobal in vitro nucleosome positioning. he data are plotted as in () using in vitro data and control matching the read number of the in vitro data set. WWW.NURE.OM/NURE 5
RESERH SUPPLEMENRY INFORMION Normalized nucleosome frequency....9 Nucleosome frequency within genes -..- D4+ -cells ranulocytes - -3 3-5 ene RPKM bins >5 Supplementary Figure 6. ssociation between transcriptional levels and measured nucleosome occupancy. X axis represents gene expression values binned according to their RPKM values. Y axis represents normalized frequencies of observed nucleosome coverage within the regions occupied by genes in each bin. 6 WWW.NURE.OM/NURE
SUPPLEMENRY INFORMION RESERH Oligomer preferences of nucleosome Normalized dyad frequency 3... Elements 3... In vivo Elements 5-5 5 5 Distance from the element, bp 5-5 5 5 Distance from the element, bp Supplementary Figure 7. () Signatures of rotational positioning of in vitro nucleosomes. Shown are preferences relative to most dimers and trimers composed of s and s. X axis represents a distance from a given oligomer to a dyad inferred from mapped sequence reads. Y axis represents the frequency of dyads at a given distance normalized to the expected frequency. bp-spaced peaks represent helical rotational preferences of oligomers relative to nucleosome surface. () Signatures of rotational positioning of in vivo granulocyte nucleosomes against the same panel of oligomers. WWW.NURE.OM/NURE 7
RESERH SUPPLEMENRY INFORMION nucleosome dyad distribution around gene promoters (expression groups from D4+ -cells) Normalized dyad frequency..5..5..5 3. Highly expressed Moderately expressed Low expressed Not expressed 5 5 Distance relative to SS, bps Supplementary Figure 8. Sequence-encoded nucleosome organization around SS. Plotted are frequencies of in vitro nucleosome dyads around promoters of genes binned according to their expression levels in D4+ -cells. X axis represents the distances relative to the SS (left of zero is away from the gene). Y-axis represents frequencies of nucleosome dyads normalized to the genome-wide average. Each of the 4 gene bins is represented by a line of a corresponding color displayed in the legend. 8 WWW.NURE.OM/NURE
SUPPLEMENRY INFORMION RESERH Dyad instances, fold enrichment.5..5. 4 6 In vivo F ranulocytes D4+ -cells Distance from the F binding site, bp Supplementary Figure 9. Nucleosome organization around F binding sites. () Schematic depiction of nucleosome organization inferred from the data. he blue ovals represent in vivo nucleosome positions, the green square represents binding of F protein which is flanked by two well-positioned nuclesoomes. he orange oval represents preferred position of nucleosomes in vitro. () Dyad frequencies around F binding site. inidng sites were aligned so that position represents coordinate of F binding inferred from F data in D4+ -cells. X-axis represents 4 Kbp window around F binding site, Y-axis represents normlized frequencies of dyads across the regions. he red curve represents smoothed frequency of nucleosome dyads from granulocytes, the blue curve represents smoothed nucleosome dyad frequency in D4+ -cells. () Dyad frequencies in the in vitro reconstitution data around F binding sites. WWW.NURE.OM/NURE 9
kb plus ladder 5 5 5 units of MNase 5 bp bp Nucleosomes 5bp ladder kb plus ladder 5 5 5 5 units of MNase D bp bp kb plus ladder ranulocytes kb plus ladder bp bp in vitro nucleosome MNase digested 5bp ladder 5 5 5 5 5 units of MNase D8+ -cells kb plus ladder D4+ -cells Supplementary Figure kb plus ladder RESERH SUPPLEMENRY INFORMION bp bp Supplementary Figures. Isolation of nucleosome-bound DN. garose gels of nucleosome-bound DN after micrococcal-treatment in D4+ -cells (), D8+ -cells (), ranulocytes (), and in vitro reconstituted nucleosomes (D). ands isolated for sequencing are marked by red rectangles. W W W. N U R E. O M / N U R E
SUPPLEMENRY INFORMION RESERH MNase control ranulocytes D4+ -cells D D8+ -cells Pile Pile Pile Pile 5-4 -3 - - 3 5-4 -3 - - 3 5-4 -3 - - 3 5-4 -3 - - 3 5-4 -3 - - 3 5-4 -3 - - 3 5-4 -3 - - 3 5-4 -3 - - 3 5-4 -3 - - 3 5-4 -3 - - 3 5-4 -3 - - 3 5-4 -3 - - 3 E F stringency >.5 stringency >.7 Pile 5-4 -3 - - 3 Pile 5-4 -3 - - 3 Pile 5-4 -3 - - 3 5-4 -3 - - 3 5-4 -3 - - 3 5-4 -3 - - 3 5-4 -3 - - 3 5-4 -3 - - 3 5-4 -3 - - 3 Supplementary Figure. Micrococcal nuclease sequence bias anaysis. Shown are Weblogos (rooks et al 4) across sites cleaved by micrococcal nuclease in the control data (), in vivo nucleosome data (-D), and in vitro nucleosome data (E-). We examined sites containing nucleosomes of increasing positioning strength (Pile, sites with or more read starts on the same strand; Pile3, sites with 3 or more read starts; Pile5, sites with 5 or more reads starts). For each subset, we aligned start positions and plotted nucleotide frequency at corresponding sites, with representing the first sequenced base of the fragments. For the sites containing positioned in vitro nucleosomes (stringency >.5 and >.7), we plotted nucleotide frequencies from overlapping nucleosome fragments. WWW.NURE.OM/NURE
RESERH SUPPLEMENRY INFORMION Dyad counts 7 5 3 Duke DnaseI (m878) Nucleosome dyads (D4+ -cells) 6 3 5 read start counts Distance from F site, bp Supplementary Figure. hromatin structure around F sites. We plotted Dnase I cutting frequency (brown) and dyad frequencies (blue) around F binding sites. Dnase I cleavage frequency is represented by plotting frequency of 5 ends from Dnase I sequence reads using Duke Dnase-seq protocol (Song and rawford, ) in the lymphoblastoid cell line. Peaks of Dnase I are in strong counter-phase with dyads, representing cleavage sites localizing within the nucleosome linker DN. In addition, a strong peak of Dnase I can be seen between the F binding site and the first wel-positioned nuclesome. WWW.NURE.OM/NURE
SUPPLEMENRY INFORMION RESERH Phasograms of nucleosomes from Schones, et al, ell 8 Phase counts 5.4 M 5. M 5. M -cell nucleosomes (resting D4+ -cells) -cell nucleosomes (activated D4+ -cells) 4.7 M 4.5 M 4.8 M 4.3 M 5 5 5 3 Resting D4+ -cells Phase, bp ctivated D4+ -cells Peak coordinate, bp 4 6 8 Phase =.9 bp Std. Err = 3. bp djusted R =.999-7 p-value = 3.8 x Peak coordinate, bp 4 6 8 Phase =. bp Std. Err =.8 bp djusted R =.999-7 p-value =.3 x 3 4 5 6 Peak count 3 4 5 6 Peak count Supplementary Figure 3. Nucleosome spacing in resting and activated -cells. () Phasograms of nucleosomes in resting and activated -cells (Schones et al, 8). Nucleosome spacing was estimated using a linear fit to peak positions in the corresponding phasograms. () Spacing was estimated to be.9 bps in resting -cells, and (). bps in activated -cells. hese results provide independent replication of phasing estimates in D4+ and D8+ cells (Fig. D). WWW.NURE.OM/NURE 3