New Results

Micro-C XL: assaying chromosome conformation at length scales from the nucleosome to the entire genome

Tsung-Han S. Hsieh, Geoffrey Fudenberg, Anton Goloborodko, Oliver J. Rando
doi: https://doi.org/10.1101/071357

ABSTRACT

Structural analysis of chromosome folding in vivo has been revolutionized by Chromosome Conformation Capture (3C) and related methods, which use proximity ligation to identify chromosomal loci in physical contact. We recently described a variant 3C technique, Micro-C, in which chromatin is fragmented to mononucleosomes using micrococcal nuclease, enabling nucleosome-resolution folding maps of the genome. Here, we describe an improved Micro-C protocol using long crosslinkers, termed Micro-C XL, which exhibits greatly increased signal to noise, and provides further insight into the folding of the yeast genome. We also find that signal to noise is much improved in Micro-C XL libraries generated from relatively insoluble chromatin as opposed to soluble material, providing a simple method to physically enrich for bona-fide long-range interactions. Micro-C XL maps of the budding and fission yeast genomes reveal both short-range chromosome fiber features such as chromosomally-interacting domains (CIDs), as well as higher-order features such as clustering of centromeres and telomeres, thereby addressing the primary discrepancy between prior Micro-C data and reported 3C and Hi-C analyses. Interestingly, comparison of chromosome folding maps of S. cerevisiae and S. pombe revealed widespread qualitative similarities, yet quantitative differences, between these distantly-related species. Micro-C XL thus provides a single assay suitable for interrogation of chromosome folding at length scales from the nucleosome to the full genome.

INTRODUCTION

The compaction and organization of the physical genome has wide-ranging consequences for genomic function1-4. In eukaryotes, the first level of genome compaction is organization into the characteristic “beads on a string” structure, with nucleosomes separated by relatively accessible linker DNA. Our understanding of this primary structure of chromatin is well-developed, with multiple crystal structures solved for the nucleosome5, 6, and a plethora of genome-wide studies that identify the positions of individual nucleosomes across the genome in various organisms, in some cases at single nucleotide-resolution7. The next step in chromosome folding remains relatively poorly-characterized; for example, the long-held belief that chromatin fibers form a helical secondary structure termed the 30 nm fiber is increasingly subject to debate8-15.

Structural analysis of chromosome folding beyond the nucleosome fiber has been revolutionized by the Chromosome Conformation Capture (3C) family of techniques2, 16. In 3C-based protocols, chromatin is first crosslinked in vivo using formaldehyde to capture physical interactions between distal regions of the genome. Chromatin is subsequently fragmented, and ligation of chromatin fragments is used to generate chimeric DNA molecules. Sequencing these molecular libraries provides a readout of genomic loci that were crosslinked to one another via protein-protein interactions. Genome-wide variants of 3C, such as Hi-C, have revealed a number of organizational features of eukaryotic genomes at increasingly fine resolutions, from the scale of full chromosomal territories, to multi-Mb active and inactive compartments, to hundred-kb contact domains (TADs), to enhancer-promoter loops17-28. While many factors impact the effective resolution of a 3C/Hi-C dataset, including sequencing depth and library complexity29, a fundamental limit to genomic resolution is the size of the fragments generated before physical interactions are captured via ligation. Since the majority of 3C-based experiments rely on restriction enzymes for fragmentation of the genome – resulting in genomic fragments that are both long relative to the nucleosome, and inhomogeneously spaced along the genome – current Hi-C datasets are limited to ∼1 kb resolution.

To improve the resolution of 3C-based techniques, we recently developed a high resolution 3C-based technique, dubbed “Micro-C”, in which fragmentation of the genome is accomplished using micrococcal nuclease (MNase) to enable mononucleosome-resolution analysis of chromosome folding17. While the improved resolution afforded by Micro-C enabled the identification of features such as chromosomally-interacting domains – “CIDs” – in budding yeast that had not previously been discernible using a restriction enzyme-based 3C technique30, known higher-order interactions such as centromere clustering were poorly recovered using our technique. In addition, although several studies have reported “gene loops” in budding yeast using 3C methods31, 32, we found no evidence for gene loops using Micro-C.

These discrepancies with the literature motivated a deeper exploration of the effects of specific protocol steps on the results of Micro-C analysis of chromosome folding. Most notably, we sought to determine whether the reliance on formaldehyde, a “zero length” crosslinker, to crosslink genomic loci to one another might limit the ability of 3C-related methods to fully interrogate chromosome structure. To investigate whether longer crosslinkers might reveal additional features of local chromatin structure, we characterized the effects of two long protein-protein crosslinkers on Micro-C maps of the budding yeast genome. A revised Micro-C protocol incorporating long crosslinkers, which we named “Micro-C XL”, not only recapitulated the local chromatin structures previously revealed by Micro-C, but also robustly recovered higher-order features such as centromere-centromere interactions. Micro-C XL thus overcomes the key technical limitation of the original Micro-C protocol, providing a single protocol for analysis of chromosomal folding from the scale of nucleosomes to the full genome. We also characterized Micro-C XL profiles in pellet and supernatant fractions of crosslinked chromatin, finding that chromatin contacts are enriched in relatively insoluble chromatin, thereby providing a simple technical approach to improve signal-to-noise in Micro-C maps. Finally, we compared Micro-C XL maps from S. cerevisiae and S. pombe, finding a general conservation of gene-scale folding behavior in these distantly-related species. Taken together, our results provide an updated Micro-C protocol for characterization of chromosome folding at all length scales, and provide additional high resolution insights into chromosome structure in two key model organisms.

RESULTS

Optimization of crosslinking conditions for Micro-C

We recently detailed a modified Hi-C19 protocol, termed Micro-C, in which micrococcal nuclease (MNase) digestion of crosslinked chromatin enables the analysis of chromosome folding at mononucleosomal resolution17. Our reported Micro-C maps robustly captured short-range interactions such as chromosomally-interacting domains (CIDs) in budding yeast, but exhibited poor recovery of higher-order features such as the centromere-centromere (CEN-CEN) and telomere-telomere (TEL-TEL) interactions that are well-known features of yeast genome organization. We reasoned that the use of the short-range crosslinker formaldehyde in 3C protocols might limit the chromosomal interactions captured due to the extreme physical proximity required for formaldehyde to crosslink two proteins.

We therefore sought to identify alternative crosslinkers or crosslinking conditions to improve Micro-C capture of CEN-CEN and other long-range interactions. Using qPCR primers designed to assay interactions either within the contact domain associated with MDJ1 or between pairs of centromeres (Fig. S1), we tested a variety of different protein-protein crosslinkers and crosslinking conditions to identify conditions that best enabled recovery of longer-range (greater than ∼1 kb) interactions. These analyses identified two protein-protein crosslinkers that appeared to more efficiently crosslink distant nucleosomes within the MDJ1 CID and to more efficiently capture CEN-CEN interactions – disuccinimidyl glutarate (DSG, a 7.7Å crosslinker), and ethylene glycol bis(succinimidyl succinate) (EGS, a 16.1Å crosslinker) (Fig. 1a-b). The improvements in signal-to-noise afforded by DSG and EGS were not observed when DSG or EGS were added prior to cell permabilization (not shown), consistent with our expectation that these molecules are too large to cross the yeast cell wall 33. In addition, improved signal-to-noise was not observed following crosslinking with higher concentrations of formaldehyde, longer incubation times, or when a second round of formaldehyde crosslinking was carried out after cell wall digestion (Fig. S1e-f), demonstrating that the improvements in the Micro-C protocol required some specific aspect of the DSG and EGS crosslinkers, rather than, say, an increase in the sheer density of crosslinks introduced into chromatin.

Fig. 1.
Fig. 1. Overview of Micro-C XL.

(a) Outline of changes to the Micro-C protocol. After budding yeast are fixed with formaldehyde, cells are permeabilized, then treated with one of several additional protein-protein crosslinkers. Crosslinked chromatin is then digested to mononucleosomes using micrococcal nuclease. End digestion and repair is used to introduce biotinylated nucleotides into mononucleosomal ends, and nucleosomes crosslinked to one another are ligated together at high dilution. Ligation products are then purified via streptavidin capture and size selection of dinucleosomesized DNA, and paired-end deep sequencing is used to characterize internucleosomal interactions genome-wide. (b) Structures of the two protein-protein crosslinkers used in Micro-C XL. (c) Example of Micro-C XL contact map for a 20 kb genomic stretch. The Micro-C XL protocol effectively recovers the chromosomally-interacting domains previously observed in Micro-C data17. (d) Plot of interaction density for all unidirectional (“IN-OUT”) read pairs, expressed as a fraction of potential pairwise interactions (per bp2) (y axis, log10), vs. genomic distance (x axis, log10) for various Micro-C protocols scaled to 109 interactions.

We incorporated each of these longer crosslinkers into an altered Micro-C protocol, which we dubbed Micro-C XL (MICROcoccal nuclease-based analysis of Chromosome folding using long X-Linkers), and then sought to identify those features of yeast chromosome folding uniquely revealed using these crosslinkers (Methods). Briefly, actively growing budding yeast cultures are crosslinked with formaldehyde alone, formaldehyde + DSG, formaldehyde + EGS, or all three crosslinkers, and resulting chromatin is fragmented to mononucleosomes using MNase digestion (Fig. 1a). Crosslinked chromatin is then treated with T4 DNA polymerase in the absence of dNTPs to promote exonuclease activity. This leaves single stranded DNA ends, which are then repaired and biotinylated upon the addition of dNTPs, including biotin-dATP and biotin-dCTP, to the T4 polymerase reaction. Following DNA ligation, ligation products are purified away from unligated mononucleosomal DNA based 1) on ligationdependent protection of biotinylated DNA from exonuclease attack, and 2) on size selection specifically of dinucleosome-sized ligation products. In practice, nucleosomal ligation products are first treated with exonuclease III to remove biotinylated nucleotides from free DNA ends, leaving biotinylated nucleotides specifically in nucleosomal ends that had been ligated to one another and thereby protected from exonuclease attack. DNA is then purified from deproteinated chromatin, and dinucleosome-sized ligation products are gel-purified away from unligated mononucleosomal DNA. Recovered DNA is then further purified on streptavidin beads to isolate only DNA carrying biotinylated nucleotides at ligation junctions that had been protected from exonuclease digestion. Purified ligation products are then used to generate deep sequencing libraries, and subject to Illumina paired-end sequencing using standard methods.

Genome-wide analysis of chromosome folding by Micro-C XL

To investigate whether adding DSG or EGS to the Micro-C protocol provided additional insights into chromosome folding, we generated genome-wide Micro-C maps for budding yeast subject to a variety of crosslinking conditions. These conditions include 1% or 3% formaldehyde (FA) alone, or FA (1% or 3%) plus DSG, EGS, or both crosslinkers. Also, as described above, we generated similar datasets in which DSG or EGS were added prior to cell wall digestion as a negative control, as these molecules are not expected to cross the cell wall in budding yeast. Below, we primarily focus on the results using 3% FA with or without DSG and EGS, but results of other crosslinking conditions are noted when relevant.

In general, all four conditions (FA, FA/DSG, FS/EGS, and FA/DSG/EGS) yielded qualitatively similar results at the scale of individual genes, with chromosomal interaction domains of varying strength covering ∼1-5 genes (Fig. 1c, S2). Although CIDs were clearly observed in all four conditions, the addition of longer crosslinkers to the Micro-C protocol resulted in improved ability to visualize these structures (Fig. S2 and see below). Importantly, as previously observed with Micro-C, we again found no evidence for a regular organization of the chromatin fiber above the nucleosomal scale, which would have manifested as a peak in interaction density at a genomic distance corresponding to the fiber size (Fig. 1d, Table S1). Moreover, compared to standard Micro-C we found that Micro-C XL exhibited substantially higher signal-to-noise (Fig. 1d, S3), consistent with the q-PCR results in Fig. S1.

Beyond recapitulating the key aspects of chromosome folding previously revealed by Micro-C, Micro-C XL resolved additional details that were not apparent in prior Micro-C maps. Most interestingly, in contrast to standard Micro-C crosslinking conditions, all three long crosslinking conditions captured very robust CEN-CEN and TEL-TEL interactions characteristic of the Rabl configuration for interphase chromosomes (Figs. 2, S4)24, 27, 30. This finding thus resolves the primary qualitative discrepancy between prior Micro-C data and known features of genomic folding in yeast, while preserving the ability of Micro-C to interrogate chromatin interactions at the 2-10 nucleosome scale.

Fig. 2.
Fig. 2. Micro-C XL robustly captures known interchromosomal interactions.

Interaction maps for Micro-C data generated using 3% formaldehyde, 3% formaldehyde+DSG, or 3% formaldehyde+EGS are shown as indicated for budding yeast chromosomes VII through X. Here, data are normalized by read depth only and displayed as log10(counts per million). Lower right panel shows Micro-C XL data for 3% formaldehyde+DSG in which crosslinked chromatin was subject to centrifugation following MNase digestion, and the pellet fraction was subject to downstream protocol steps (see also Fig. 3). For 3% FA+DSG, CEN-CEN interactions are shown with blue arrows. As previously observed, telomere-telomere interactions are only observed between a subset of chromosome arms (which do not include interactions between chromosomes VII to X) in budding yeast.

Fig. 3.
Fig. 3. Micro-C XL interactions are enriched in insoluble chromatin.

(a) Plot of interaction density (y axis, log10) vs. genomic distance (x axis, log10) for four Micro-C XL libraries, normalized as in Fig. 1d. One pair of libraries were crosslinked with 3% FA+DSG, then MNase-digested chromatin was separated into soluble and insoluble fractions by centrifugation; the same procedure was also repeated for yeast crosslinked with 3% FA+EGS. In both cases, the relatively soluble Micro-C library exhibited far lower signal to noise, with relatively rapid decay of interactions with increasing distance, compared to libraries constructed from pellet material. (b-f) Micro-C interaction maps for chromosomes VII to X are shown as in Fig. 2, for pellet (b, d, f) and supernatant (c, e) libraries as indicated. Note that the strong enrichment of CEN-CEN interactions in the pellet fraction requires long-distance crosslinkers, as it is not observed for 3% FA chromatin pellets (b).

We conclude that both DSG and EGS dramatically extend the length scale at which chromosome folding can be assayed by Micro-C, enabling analysis at scales from the local chromatin fiber to the full genome. Interestingly, this improvement can largely be ascribed to a decrease in the background levels of ligation between distant genomic regions (Fig. S3, Table S1) – the decrease in this “noise floor” seen using the Micro-C XL protocol is likely to account for the improved ability to measure relatively lowabundance signals such as CEN-CEN interactions.

Bona-fide Micro-C contacts are primarily found in relatively insoluble chromatin

We next sought to uncover whether Micro-C data are affected by fractionation of crosslinked chromatin prior to proximity ligation. This was motivated by the absence of “gene loops”31, 32 in our previous Micro-C analysis – the 3C method used in several studies of gene loops includes a step in which insoluble chromatin is pelleted and isolated prior to ligation, and it is known that different chromatin structures are likely to be present in soluble vs. insoluble crosslinked chromatin34. We therefore carried out Micro-C XL in which fragmented chromatin was centrifuged after the completion of MNase digestion to separate soluble from insoluble chromatin (Methods), and proximity ligation was carried out separately on pellet and supernatant material (Fig. 3).

Micro-C XL maps from supernatant material were extremely noisy at longer distances, and did not identify known aspects of higher-order organization. In contrast, contact maps generated from relatively insoluble chromatin had excellent signal to noise, and robustly captured CEN-CEN interactions. We conclude that this reflects either preferential precipitation of crosslinked fragments, higher ligation efficiency in the pellet, or higher specificity of ligation in the pellet; i.e. noise in the supernatant dataset is elevated due to ligations in solution between freely-moving, likely uncrosslinked, nucleosomes causing artefactual contacts in trans. These related hypotheses are of course not mutually exclusive, and may all contribute to the improved signal to noise seen in Micro-C XL maps from pellet material (Fig. S3, right panel).

We next searched for evidence of gene loops31, 32 in the dataset generated from relatively insoluble chromatin. Here, we consider a gene “loop” to be characterized by an increased contact frequency between the gene start and stop relative to other locus pairs in their vicinity (similar to previous definitions of peaks in Hi-C contact maps –28), rather than a gene-wide increase in relative contact frequency. In general, it is clear that, for any given nucleosome, raw interaction counts (either normalized only for library depth, or normalized additionally for nucleosome occupancy) decay steadily with increasing distance and do not exhibit an uptick at gene ends, which is the signature of a looping interaction35, 36 – see example in Fig. 1c, or averaged “metagenes” in Figs. S5-6. Nevertheless, with our population-average contact maps we cannot rule out a scenario where populations of various length gene loops are formed dynamically over gene bodies (as proposed for enhancer-promoter interactions –37).

Although our data thus do not support the concept of widespread end-to-end gene loops at transcribed genes, visual inspection of Micro-C XL data did reveal a small number of possible looping interactions at multi-gene scale that were apparent even in interaction counts not normalized for distance (Fig. S7). Validation and functional analysis of these apparent loops will be the subject of future studies.

Comparison of chromosome folding in S. cerevisiae and S. pombe

Although many aspects of chromosome folding are conserved between S. cerevisiae and other eukaryotes, S. cerevisiae lack several evolutionarily widespread chromatin regulatory systems, such as the H3K9me3/HP1 and H3K27me3/Polycomb systems for gene repression found in many eukaryotes. We therefore carried out Micro-C XL in the fission yeast S. pombe to ascertain the similarities and differences in chromosome folding between these distantly-related microbes, and to demonstrate the broad applicability of our methods. Key aspects of the Micro-C XL protocol proved equally important in fission yeast, as for example maps generated from relatively insoluble chromatin exhibited far less noise compared to maps based on soluble chromatin (Fig. S8). Overall, our data were well-correlated (Spearman’s r = 0.77 using 10 kb bins) with a prior Hi-C analysis of S. pombe chromatin by Mizuguchi et al (Fig. S9)24. As in budding yeast, Micro-C XL maps in S. pombe revealed frequent interactions along the diagonal, robust CEN-CEN and TEL-TEL interactions, and a depletion of interactions between centromeres and chromosome arms (Fig. 4a). We did find quantitative differences in such large-scale aspects of chromosome folding, as S. pombe chromosomes exhibited slightly stronger centromere clustering, and substantially stronger telomere clustering (Fig. S10). These differences do not appear to be a consequence of the profound cell cycle differences between budding and fission yeast (Fig. S10c-d), but could potentially be explained by any number of other features ranging from the smaller number of longer chromosomes in S. pombe, to the molecular details of interactions between pairs of H3K9-methylated nucleosomes present in this species but not in S. cerevisiae.

Fig. 4.
Fig. 4. Analysis of chromosome folding in S. pombe.

(a) Whole genome Micro-C XL interaction map for S. pombe. Here, data are shown for yeast crosslinked with 3% FA+DSG and pelleted prior to ligation. Similar results were obtained with 3% FA+EGS (not shown). Key features of this map include robust clustering of centromeres and telomeres (with the exception of the rDNA-carrying chromosome III telomeres, which were excluded from analysis based on their repetitive nature), and strong depletion of interactions between centromeres and chromosome arms. (b) Zoom-in on S. pombe Chr I:553,200-609,400, showing widespread contact domains typically associated with individual genes, but occasionally associated with blocks of ∼2-5 genes.

As prior studies of chromosome folding in S. pombe were performed with ∼10 kb kb resolution, we next turned to those aspects of chromatin structure uniquely interrogated using the enhanced resolution of Micro-C. Visual inspection of chromosome folding revealed abundant contact domains associated with ∼1-5 genes and separated by promoter regions (Fig. 4b), analogous to the CID structures in budding yeast. As in budding yeast, promoters in fission yeast acted as efficient boundaries between CIDs, and metagene analysis revealed remarkably similar behavior in both yeast species at the length scale of individual genes or promoters (Fig. 5, S11-12). In addition, plots of interaction frequency vs. distance, and distributions of contact domain length, were qualitatively similar, yet quantitatively different, in these two species; in particular, differences at short distances in the positions of interaction maxima correspond to known differences in nucleosome repeat length in these two species (Fig. S13).

Fig. 5.
Fig. 5. Comparison of chromosome folding in S. cerevisiae and S. pombe.

(a-b) For each species, intergenic regions were separated into those falling between pairs of genes oriented divergently, in tandem, or convergently, as indicated, and were aligned according to the midpoint of the nearest respective intergenic region. Data from DSG pellet libraries from S. cerevisiae (a) or S. pombe (b) are averaged for all genes in each category. Top panels show interaction counts (normalized to library read depth for coverage-corrected 200 bp-binned maps) for the three classes of intergenic region. Bottom panels show the same data, additionally corrected for the decay in interaction frequency with increasing distance, and expressed as the log2 ratio of observed interactions divided by expected interactions for a given genomic distance.

We conclude that broadly similar principles underlie chromosome folding behavior in these distantly-related fungi, with modest quantitative differences in chromosome structure that could potentially result from the interspecies differences in aspects of genomic structure including chromosome length, gene length, intron abundance, location of rDNA clusters, and nucleosome repeat length.

DISCUSSION

Here, we present an improved protocol for nucleosome-resolution mapping of chromosome folding, termed Micro-C XL. The primary technical improvements detailed here are 1) the use of additional “long-range” crosslinkers to supplement formaldehyde crosslinking, and 2) fractionation of relatively insoluble chromatin prior to nucleosome ligation and subsequent library construction. Contrary to our initial expectations, the dramatic improvement seen in apparent capture of long-range interactions using these protocols likely results not from the ability of long-range crosslinkers to bridge interacting genomic loci associated with proteins that are more than 3 Å away from one another, but rather from a decrease in the noise caused by soluble nucleosomes encountering one another in solution during the ligation reaction and causing artefactual “interactions” between unlinked nucleosomes (Fig. S3, Table S1). This hypothesis is based on the fact that DSG- and EGS-based Micro-C maps are extremely similar despite their substantial difference in crosslinking distance, as well as the finding that isolation of soluble chromatin results in greatly increased noise in Micro-C maps (Fig. 3). In addition, we note that chromatin fragments generated by restriction enzymes in typical Hi-C protocols are significantly larger than mononucleosomes, increasing the number of crosslinking opportunities per fragment and thus presumably restricting their diffusion and resultant ability to generate artefactual ligation products. We propose that this difference in fragment size/mobility accounts for the increased noise seen previously in Micro-C relative to standard Hi-C protocols. Further supporting this idea, we find that the improved protocol strongly reduces the incidence of artefactual ligation products between the nuclear genome and the mitochondrial genome, relative to the standard Micro-C protocol (Fig. S14). We note this is in general agreement with prior comparisons of in-solution versus both pellet and “in situ” Hi-C protocols28, 38, 39. Together, these considerations support the idea that the use of long crosslinkers and isolation of insoluble chromatin may be important to prevent mononucleosomes from freely diffusing prior to ligation and introducing noise into Micro-C measurements. Still, this does not rule out the additional possibility that in some cases our long crosslinkers capture nearby genomic loci for which the closest crosslinkable proteins are not in immediate physical proximity, and indeed both of these features may contribute to the improvement in data quality seen in Micro-C XL.

Chromosome structure

Validated via comparisons to prior data, our method provides insight into yeast genome folding at all length scales of interest. At larger scales, the Rabl configuration of chromosomes is seen as clustering of centromeres, and interactions between the telomeres of chromosome arms of similar length. Centromeric chromatin also shows a characteristic “X” shape resulting from the two arms of the chromosome both statistically leading away from the centromere together for some ∼20 kb, with centromeres otherwise being relatively isolated from chromosome arms. At higher resolution, genes in both budding and fission yeast are organized into chromosomallyinteracting domains (CIDs), typically spanning 1-5 genes, that are in some ways similar to the “topologically-associating domains” described in a multitude of other model organisms. Boundaries between CIDs occur at active promoters, highly-expressed genes, and tRNA genes (Fig. S15). In both budding and fission yeast, genomic regions surrounding cohesin-associated loci are relatively insulated from physically interacting with one another (Fig. S16). However, this insulation is stronger and persists over greater genomic distances in S. pombe relative to S. cerevisiae, pointing towards important differences in the role of cohesin, potentially in a cell-cycle dependent fashion, between the species. Taken together, these analyses highlight the ability of Micro-C XL to assay chromosome folding across all scales, as well as its broad applicability and future utility in comparative genomics.

Applications to other biological systems

We finally turn to considerations of the sequencing depth required for applications of Micro-C XL in organisms with larger genomes. As reviewed in Lajoie et al29, the fundamental genomic resolution of a chromosome capture dataset is set by the frequency at which the genome is fragmented prior to capture of physical interactions by ligation; beyond this lower bound to resolution, the effective genomic resolution is further influenced by sequencing depth and library complexity. The proportion of molecular byproducts in a library additionally influences the amount of sequencing required to achieve a given coverage per fragment. Given that Micro-C XL does not display a preponderance of molecular byproducts (Fig. S14), the sequencing depth required to achieve a given genomic resolution should be similar to a Hi-C protocol. Nevertheless, Micro-C XL has the capacity to analyze chromatin interactions at genomic distances smaller than currently available Hi-C protocols. Indeed, the highest-resolution studies performed to date in mammals28 utilize restriction enzymes with 4 bp target sequences to yield average fragment lengths of ∼256 bp, although due to the heterogeneous distribution of restriction sites across the genome lower-resolution (∼1 or 5 kb) binning approaches must be used in analysis of such datasets. In comparison to any individual 4-cutter, MNase digestion of chromatin to mononucleosomes results in at most ∼75% more genomic fragments (depending on the nucleosome repeat length in the tissue of interest), which in turn increases the fundamental genomic resolution of Micro-C by substantially more than ∼1.75-fold, thanks to the more even spacing of the resulting fragments.

In addition, beyond binning Micro-C data to mimic lower-resolution Hi-C, it is important to note that a wide variety of biological questions can be addressed – at high resolution – by Micro-C at much lower sequencing depth. First of all, the strength of the Micro-C protocol is its ability to interrogate chromatin fiber structure at ∼150-1000 bp resolution – there is little reason to carry out Micro-C to investigate >1 MB chromatin domains19. A key measure in this regard 40 is the decay of interaction frequency with increasing distances (see, e.g., Fig. 1d or Fig. 3a), which is an averaged measure across the entire genome and is thus extremely robust to undersequencing (Fig. S17). We anticipate that very low coverage (below 1-2 million reads) Micro-C XL maps in mammals will thus allow robust comparison of average chromatin fiber folding for, say, Polycomb-repressed genes, or for exons vs. introns, etc. In addition to using such computational averaging methods to make use of multiple instances of any given annotation, the molecular complexity of the sequencing library can also be experimentally reduced. This is commonly done in sequence-capture RNA-Seq protocols in cancer exome studies, and more recently such methods have been applied to 3C methods based either on protein capture (eg, Pol2 IP) or on capture of specific “bait” sequences41. The reduction in read depth required for such methods naturally depends on the distribution and abundance of the feature to be captured, but many proteins of interest – CTCF, cohesin, TFIIB, and others – are sparsely distributed enough to enable >100-fold reductions in the sequencing depth required for highresolution Hi-C studies. These and other considerations29, 41 must be a part of any experimental design for a 3C-based study.

Conclusion

Here, we describe a modified protocol for genome-wide analysis of 3D chromatin structure that captures aspects of chromosome folding at all scales from mononucleosome resolution up to interactions between different chromosomes. This protocol, Micro-C XL, should find broad utility in a multitude of biological systems.

METHODS

Yeast strains and culture conditions

All experiments reported here were carried out with either S. cerevisiae strain BY4741 or S. pombe strain 972 h-. BY4741 cultures were grown in YPD media at 30°C, while S. pombe cells were grown at 30°C in “Compromise Media”42, consisting of Yeast extract (1.5%), Peptone (1%), Dextrose (2%), SC Amino Acid mix (Sunrise Science) 2 grams per liter, Adenine 100 mg/L, Tryptophan 100 mg/L, and Uracil 100 mg/L.

Fixation conditions

Midlog yeast cultures were crosslinked with either 1% or 3% final concentration of formaldehyde (Sigma) for 15 minutes at 30°C, then quenched with 125 mM glycine for 5 min at room temperature. Yeast were then spheroplasted as previously described42, 43. For DSG and EGS (Thermo Fisher Scientific) crosslinking studies, spheroplasts were resuspended in a 3 mM final concentration of the crosslinker of interest in PBS, and crosslinked for 40 min at 30°C, then quenched with 125 mM glycine for 5 min at room temperature. Note that we use the protocol “Micro-C XL” for DSG and EGS-based protocols interchangeably, as the data from these protocols are nearly indistinguishable.

Separation of soluble and insoluble chromatin

Crosslinked chromatin was digested with micrococcal nuclease (MNase, Worthington) to yield > 95% mononucleosomes. After inhibition of MNase with 2 mM EGTA at 65°C, fragmented lysate was in some cases used directly for the standard Micro-C protocol, or in some experiments (Figs. 3-5) was separated into supernatant and pellet portions. Here, MNase-digested lysate was spun at 10,000 g for 5 minutes, and Micro-C was separately performed on supernatant or on the pellet fraction.

Micro-C protocol

The termini of nucleosomal DNA were dephosphorylated by Shrimp Alkaline Phosphatase and then subjected to T4 DNA polymerase for end repairing and biotin labeling by supplementing with biotin-dCTP, biotin-dATP, dTTP, and dGTP. Crosslinked chromatin was diluted to 10 ml and treated with T4 DNA ligase. After heat inactivation, chromatin was concentrated to 250 µl in an Amicon 30k spin column and treated with 100 U exonuclease III for 5 min to eliminate biotinylated ends of unligated DNA. Proteinase K was then added and incubated for 65°C overnight. DNA was purified by PCI extraction and ethanol precipitation, treated with RNase A, and ∼250–350 bp DNA was gel-purified. Purified DNA was treated with End-it, subject to A-tailing with Exo-Klenow, and ligated to Illumina adapters. Adapter-ligated DNA was purified with streptavidin beads to isolate ligated Micro-C products away from undigested dinucleosomal DNA. Streptavidin beads were then subject to ∼10–12 cycles of PCR using Illumina paired-end primers. Amplified library was purified and subject to Illumina NextSeq paired end sequencing. Detailed methods are described in Supplemental Text.

Computational analysis of Micro-C interactions

MicroC data was mapped to the sacCer3 genome using Bowtie 2.1.0 as described in44 using the hiclib library for python, publicly available at https://bitbucket.org/mirnylab/hiclib, with virtual 100bp fragments. To obtain corrected contact maps, genomic coverage was calculated by summing the total number of interactions per bin. Low coverage bins were then excluded from further analysis using a MAD-max (maximum allowed median absolute deviation) filter on genomic coverage, set to 9 median absolute deviations. Following this filtering, standalone bins were removed (ie. regions where both neighboring bins did not pass filters), and the resulting maps were then iteratively corrected to equalize genomic coverage44. Observed/expected contact maps were obtained by additionally dividing out the dependence on genomic distance, calculated empirically as the mean number of contacts at each genomic separation, using a sliding window with linearly increasing size, as previously described45. Log-log plots of contact probability P(s) versus distance were calculated using log-spaced bins with a constant step size. For average plots around genomic features, gene positions and orientations, centromere positions, and tRNA positions were obtained from the SGD (http://www.yeastgenome.org/).

Micro-C XL Protocol

I. First Crosslinking

  1. Culture 100 mL of yeast to the midlog stage, OD=0.55 o/n.

  2. Add 37% formaldehyde directly to the culture to 3% of a final concentration.

  3. Shake the culture at 210 rpm for 15 min at 30°C (FA only) or 10 min at 30°C (Dual crosslinking).

  4. Quench the crosslinking by adding 10 mL of 2.5M Glycine.

  5. Incubate for 5 min at room temperature.

  6. Centrifuge the cells at 4000 rpm for 5 min at 4°C.

  7. Pour off the medium and wash the cells in 50 mL of sterile water by vortexing.

  8. Centrifuge the cells at 4000 rpm for 5 min at 4°C.

  9. Pour off the water.

II. Permeabilize the cell wall

  1. Resuspend the cell pellet in 10 mL of Buffer Z and add 7 µL of 2-Mercaptoethanol (final 10mM).

  2. Add 250 µL Zymolyase solution (final 250 µg/mL).

  3. Shake the tube at 210 rpm for 40 min at 30°C.

  4. Centrifuge the cells at 4000 rpm for 10 min at 4°C.

  5. Aspirate the supernatant with a vacuum suction.

  6. Rinse the permeabilized cells by 5mL cold 1X PBS.

  7. Centrifuge the cells at 4000 rpm for 2 min at 4°C.

  8. Aspirate the supernatant with a vacuum suction.

III. Second Crosslinking

  1. Freshly prepare the long crosslinker stock and working solution as below:

  2. Resuspend the cells homogenously by 5 mL of working solution.

  3. Rotate the tube for 40 min at 30°C.

  4. Quench the crosslinking by adding 1 mL of 2.5M Glycine.

  5. Centrifuge the cells at 4000 rpm for 10 min at 4°C.

  6. Aspirate the supernatant with a vacuum suction.

  7. Rinse the permeabilized cells by 5 mL cold 1X PBS.

  8. Centrifuge the cells at 4000 rpm for 2 min at 4°C.

  9. Aspirate the supernatant with a vacuum suction.

  10. The crosslinked pellet can be store at -80°C for few months.

IV. Chromatin fragmentation

  1. Resuspend the cell pellet in 200 µL of MBuffer#1 (freshly complete).

  2. Add the appropriate amount of MNase to digest the chromatin to > 95% mononucleosomes.

  3. Incubate the tube for 20 min at 37°C.

  4. Add 2 mM EGTA and incubate the tube for 10 min at 65°C to stop the MNase activity.

  5. Here, you can proceed the sample depending on the experimental design: 1) Total chromatin. 2) Supernatant. 3) Pellet. Any of those fractions can be subjected to following Micro-C protocol.

V. Chromatin cleaning and concentration

1) Total chromatin

  1. Transfer the whole sample into the 0.5 mL Amicon 10K spin column.

  2. Concentrate the sample at 16000xg for at 4°C until the volume goes down to ∼50 µL.

  3. Wash / pipette the sample by 450 µL MBuffer#2.

  4. Repeat wash step 2 - 3.

  5. Concentrate the sample at 16000xg for at 4°C until the volume goes down to ∼30 µL.

  6. Add BSA to 1X final concentration.

2) Supernatant

  1. Centrifuge the tube at 16000xg for 5 min at 4°C.

  2. Collect the supernatant.

  3. Concentrate the sample by the 0.5 mL Amicon 10K spin column at 16000xg for at 4°C until the volume goes down to ∼50 µL.

  4. Wash / pipette the sample by 450 µL MBuffer#2.

  5. Repeat wash step 2 - 3.

  6. Concentrate the sample at 16000xg for at 4°C until the volume goes down to ∼30 µL.

  7. Add BSA to 1X final concentration.

3) Pellet

  1. Centrifuge the tube at 16000xg for 5 min at 4°C.

  2. Collect the pellet.

  3. Resuspend the pellet in 1 mL MBuffer#2.

  4. Centrifuge the tube at 16000xg for 5 min at 4°C.

  5. Aspirate the buffer with a vacuum suction.

  6. Repeat wash steps 3 - 5.

  7. Resuspend the pellet to 30 µL of MBuffer#2 + final 1X BSA (or NEBuffer 2.1).

VI. Repair and label the end of chromatin fragments

1) De-phosphorylation

2) End-Chewing

3) End-labeling

→ PCR machine: Incubate for 25min at 25°C → 15min at 12°C → 4°C.

→ Add EDTA (final 30mM) and heat inactivation for 20 min at 65°C.

VII. Proximity ligation and Remove unligated ends

1) Ligation

Although in our test Micro-C in “pellet” can be scaled down to 1mL ligate reaction, we suggest at least using 2.5 mL for routine experiment.

2) Remove the biotin-dNTP at unligated ends

3) Reverse crosslinking

→ Add 20X proteinase K to 1X final concentration.

→ Incubate for overnight at 55°C.

VIII. Dinucleosomal DNA purification

  1. Phenol:Chloroform:Isoamyl Alcohol extraction twice → spin at 19800xg for 10min.

  2. Ethanol precipitation: 0.1x volume of sodium acetate and 2.5X volume of 100% ethanol → -80°C for > 1hr → spin at 19800xg for 15min at 4°C → wash pellet by 75% ethanol → spin at 19800xg for 5 min at 4°C → Air dry pellet for 10min.

  3. Resuspend pellet in 50uL of TE buffer (+ 1X RNase solution) and incubate for 30min at 37°C.

  4. ZymoClean to purify DNA.

  5. Run DNA samples on 3% Nusieve agarose DNA gel.

  6. Size selection of the band between 250 – 350 bp.

  7. ZymoGel purification and dissolve final product in 17 µL of elution buffer.

  8. Quantify the input DNA by Qubit.

IX. Library construction by “with-bead” method

2) A-tailing

3) Adapter ligation

4) Streptavidin beads purification

  1. Wash 2.5 µL of beads per sample (100mL culture) by 1X TBW twice.

  2. Resuspend the washed beads in 150 µL 2X BW.

  3. Mix with 150 µL of adapter-ligated DNA sample.

  4. Rotate for 15 min at room temperature.

  5. Wash by 500 µL of 1X TBW twice.

  6. Rinse by 200 µL of MBuffer#2.

  7. Resuspend in 15 – 20 µL of EB buffer.

5) On-beads PCR

X. Deep sequencing by Illumina PE-50

Solutions and Enzymes

  • YPD: yeast extract/peptone/dextrose

  • 37% Formaldehyde (Sigma Aldrich # 252549)

  • DSG (disuccinimidyl glutarate) (ThermoFisher #20593)

  • EGS (ethylene glycol bis(succinimidyl succinate)) (ThermoFisher #21565)

  • 2.5M Glycine (Sigma Aldrich #G7126)

  • Buffer Z: 1M sorbitol, 50mM Tris pH 7.4

  • 14.3M 2-Mercaptoethanol (Sigma Aldrich # M6250)

  • Zymolyase solution: 10mg/ml in Buffer Z; lasts up to 2 weeks at 4°C (Sunrise Science #N0766555)

  • MBuffer#1: 50mM NaCl, 10mM Tris-HCl pH 7.5, 5mM MgCl2, 1mM CaCl2, and freshly complete with 0.5mM spermidine, 1mM B-ME, and NP-40 (percentage is determined by titrating the ratio of pellet / sup).

  • Micrococcal Nuclease (Worthington Biochem): resuspended from lyophilized powder at 20 U/μl in Tris pH 7.4. Aliquot into tubes upon first use and freeze at –80°C.

  • 0.5M EDTA (Life technology # AM9261)

  • MBuffer#2 (NEBuffer#2): 50mM NaCl, 10mM Tris-HCl pH 7.5, 10mM MgCl2

  • Shrimp Alkaline Phosphatase (rSAP) (New England Biolabs # M0371)

  • T4 DNA Polymerase (New England Biolabs # M0203)

  • T4 Polynucleotide Kinase (New England Biolabs # M0201)

  • Biotin-14-dCTP (Life Technologies # 19518018)

  • Biotin-14-dATP (Life Technologies # 19524016)

  • T4 DNA Ligase (New England Biolabs # M0202)

  • Exonuclease III (E. coli) (New England Biolabs # M0206)

  • 20X Proteinase K solution: TE with 20 mg/ml proteinase K and 50% glycerol. Store in -20°C.

  • Elution buffer (EB): 10 mM Tris-HCl pH 7.5

  • TE buffer: 10 mM Tris-HCl pH 8.0, 1 mM EDTA

  • End-It DNA End-Repair Kit (EpiCentre BioTechnologies # ER81050)

  • Exo-Minus Klenow DNA Polymerase (EpiCentre BioTechnologies # KL111)

  • Fast Link DNA Ligation Kit (EpiCentre BioTechnologies # lk6201)

  • Dynabeads® MyOne Streptavidin C1 (Life Technologies # 65001)

  • KAPA HiFi HotStart ReadyMix (KAPA Biosystems # KK2601)

ACKNOWLEDGEMENTS

We thank L. Mirny and members of the Rando lab for insightful discussions. Work was supported in part by NIH grant GM079205 to OJR. GF and AG were supported by R01 GM114190 and U54 DK107980 of the 4D Nucleome Program. T-HSH is an HHMI international student research fellow.

REFERENCES

  1. 1.↵
    Friedman, N. & Rando, O.J. Epigenomics and the structure of the living genome. Genome research 25, 1482–1490 (2015).
    OpenUrlAbstract/FREE Full Text
  2. 2.↵
    Dekker, J., Marti-Renom, M.A. & Mirny, L.A. Exploring the three-dimensional organization of genomes: interpreting chromatin interaction data. Nature reviews. Genetics 14, 390–403 (2013).
    OpenUrlCrossRefPubMed
  3. 3.
    Dekker, J. & Misteli, T. Long-Range Chromatin Interactions. Cold Spring Harbor perspectives in biology 7, a019356 (2015).
    OpenUrlAbstract/FREE Full Text
  4. 4.↵
    Horn, P.J. & Peterson, C.L. Molecular biology. Chromatin higher order folding-wrapping up transcription. Science (New York, N.Y 297, 1824–1827 (2002).
    OpenUrlAbstract/FREE Full Text
  5. 5.↵
    Luger, K., Mader, A.W., Richmond, R.K., Sargent, D.F. & Richmond, T.J. Crystal structure of the nucleosome core particle at 2.8 A resolution. Nature 389, 251–260 (1997).
    OpenUrlCrossRefPubMedWeb of Science
  6. 6.↵
    Kornberg, R.D. & Lorch, Y. Twenty-five years of the nucleosome, fundamental particle of the eukaryote chromosome. Cell 98, 285–294 (1999).
    OpenUrlCrossRefPubMedWeb of Science
  7. 7.↵
    Hughes, A.L. & Rando, O.J. Mechanisms Underlying Nucleosome Positioning in vivo. Annual review of biophysics (2014).
  8. 8.↵
    Tremethick, D.J. Higher-order structures of chromatin: the elusive 30 nm fiber. Cell 128, 651–654 (2007).
    OpenUrlCrossRefPubMedWeb of Science
  9. 9.
    Song, F. et al. Cryo-EM study of the chromatin fiber reveals a double helix twisted by tetranucleosomal units. Science (New York, N.Y 344, 376–380 (2014).
    OpenUrlAbstract/FREE Full Text
  10. 10.
    Dorigo, B. et al. Nucleosome arrays reveal the two-start organization of the chromatin fiber. Science (New York, N.Y 306, 1571–1573 (2004).
    OpenUrlAbstract/FREE Full Text
  11. 11.
    Ghirlando, R. & Felsenfeld, G. Hydrodynamic studies on defined heterochromatin fragments support a 30-nm fiber having six nucleosomes per turn. Journal of molecular biology 376, 1417–1425 (2008).
    OpenUrlCrossRefPubMedWeb of Science
  12. 12.
    Routh, A., Sandin, S. & Rhodes, D. Nucleosome repeat length and linker histone stoichiometry determine chromatin fiber structure. Proceedings of the National Academy of Sciences of the United States of America 105, 8872–8877 (2008).
    OpenUrlAbstract/FREE Full Text
  13. 13.
    Collepardo-Guevara, R. & Schlick, T. Chromatin fiber polymorphism triggered by variations of DNA linker lengths. Proceedings of the National Academy of Sciences of the United States of America 111, 8061–8066 (2014).
    OpenUrlAbstract/FREE Full Text
  14. 14.
    Grigoryev, S.A., Arya, G., Correll, S., Woodcock, C.L. & Schlick, T. Evidence for heteromorphic chromatin fibers from analysis of nucleosome interactions. Proceedings of the National Academy of Sciences of the United States of America 106, 13317–13322 (2009).
    OpenUrlAbstract/FREE Full Text
  15. 15.↵
    Cheng, T.M. et al. A simple biophysical model emulates budding yeast chromosome condensation. eLife 4, e05565 (2015).
  16. 16.↵
    Dekker, J., Rippe, K., Dekker, M. & Kleckner, N. Capturing chromosome conformation. Science (New York, N.Y 295, 1 306–1311 (2002).
    OpenUrl
  17. 17.↵
    Hsieh, T.H. et al. Mapping Nucleosome Resolution Chromosome Folding in Yeast by Micro-C. Cell 162, 108–119 (2015).
    OpenUrlCrossRefPubMed
  18. 18.
    Li, G. et al. Extensive promoter-centered chromatin interactions provide a topological basis for transcription regulation. Cell 148, 84–98 (2012).
    OpenUrlCrossRefPubMedWeb of Science
  19. 19.↵
    Lieberman-Aiden, E. et al. Comprehensive mapping of long-range interactions reveals folding principles of the human genome. Science (New York, N.Y 326, 289–293 (2009).
    OpenUrlAbstract/FREE Full Text
  20. 20.
    Sexton, T. et al. Three-dimensional folding and functional organization principles of the Drosophila genome. Cell 148, 458–472 (2012).
    OpenUrlCrossRefPubMedWeb of Science
  21. 21.
    Dixon, J.R. et al. Topological domains in mammalian genomes identified by analysis of chromatin interactions. Nature 485, 376–380 (2012).
    OpenUrlCrossRefPubMedWeb of Science
  22. 22.
    Nora, E.P. et al. Spatial partitioning of the regulatory landscape of the X-inactivation centre. Nature 485, 381–385 (2012).
    OpenUrlCrossRefPubMedWeb of Science
  23. 23.
    Le, T.B., Imakaev, M.V., Mirny, L.A. & Laub, M.T. High-resolution mapping of the spatial organization of a bacterial chromosome. Science (New York, N.Y 342, 731–734 (2013).
    OpenUrlAbstract/FREE Full Text
  24. 24.↵
    Mizuguchi, T. et al. Cohesin-dependent globules and heterochromatin shape 3D genome architecture in S. pombe. Nature (2014).
  25. 25.
    Wang, C. et al. Genome-wide analysis of local chromatin packing in Arabidopsis thaliana. Genome research 25, 246–256 (2015).
    OpenUrlAbstract/FREE Full Text
  26. 26.
    Eagen, K.P., Hartl, T.A. & Kornberg, R.D. Stable Chromosome Condensation Revealed by Chromosome Conformation Capture. Cell 163, 934–946 (2015).
    OpenUrlCrossRefPubMed
  27. 27.↵
    Marbouty, M. et al. Metagenomic chromosome conformation capture (meta3C) unveils the diversity of chromosome organization in microorganisms. eLife 3, e03318 (2014).
    OpenUrlCrossRefPubMed
  28. 28.↵
    Rao, S.S. et al. A 3D map of the human genome at kilobase resolution reveals principles of chromatin looping. Cell 159, 1665–1680 (2014).
    OpenUrlCrossRefPubMedWeb of Science
  29. 29.↵
    Lajoie, B.R., Dekker, J. & Kaplan, N. The Hitchhiker’s guide to Hi-C analysis: practical guidelines. Methods 72, 65–75 (2015).
    OpenUrlCrossRefPubMed
  30. 30.↵
    Duan, Z. et al. A three-dimensional model of the yeast genome. Nature 465, 363–367 (2010).
    OpenUrlCrossRefPubMedWeb of Science
  31. 31.↵
    Ansari, A. & Hampsey, M. A role for the CPF 3’-end processing machinery in RNAP II-dependent gene looping. Genes & development 19, 2969–2978 (2005).
    OpenUrlAbstract/FREE Full Text
  32. 32.↵
    O’Sullivan, J.M. et al. Gene loops juxtapose promoters and terminators in yeast. Nature genetics 36, 1014–1018 (2004).
    OpenUrlCrossRefPubMedWeb of Science
  33. 33.↵
    Scherrer, R., Louden, L. & Gerhardt, P. Porosity of the yeast cell wall and membrane. J Bacteriol 118, 534–540 (1974).
    OpenUrlAbstract/FREE Full Text
  34. 34.↵
    Henikoff, S., Henikoff, J.G., Sakai, A., Loeb, G.B. & Ahmad, K. Genome-wide profiling of salt fractions maps physical properties of chromatin. Genome research 19, 460–469 (2009).
    OpenUrlAbstract/FREE Full Text
  35. 35.↵
    Benedetti, F., Dorier, J., Burnier, Y. & Stasiak, A. Models that include supercoiling of topological domains reproduce several known features of interphase chromosomes. Nucleic acids research 42, 2848–2855 (2014).
    OpenUrlCrossRefPubMedWeb of Science
  36. 36.↵
    Doyle, B., Fudenberg, G., Imakaev, M. & Mirny, L.A. Chromatin loops as allosteric modulators of enhancer-promoter interactions. PLoS computational biology 10, e1003867 (2014).
    OpenUrl
  37. 37.↵
    Lee, K., Hsiung, C.C., Huang, P., Raj, A. & Blobel, G.A. Dynamic enhancer-gene body contacts during transcription elongation. Genes & development 29, 1992–1997 (2015).
    OpenUrlAbstract/FREE Full Text
  38. 38.↵
    Nagano, T. et al. Comparison of Hi-C results using in-solution versus in-nucleus ligation. Genome biology 16, 175 (2015).
  39. 39.↵
    Gavrilov, A.A. et al. Disclosure of a structural milieu for the proximity ligation reveals the elusive nature of an active chromatin hub. Nucleic acids research 41, 3563–3575 (2013).
    OpenUrlCrossRefPubMedWeb of Science
  40. 40.
    Imakaev, M.V., Fudenberg, G. & Mirny, L.A. Modeling chromosomes: Beyond pretty pictures. FEBS letters 589, 3031–3036 (2015).
    OpenUrlCrossRefPubMed
  41. 41.↵
    Denker, A. & de Laat, W. The second decade of 3C technologies: detailed insights into nuclear organization. Genes & development 30, 1357–1382 (2016).
    OpenUrlAbstract/FREE Full Text
  42. 42.↵
    Tsankov, A.M., Thompson, D.A., Socha, A., Regev, A. & Rando, O.J. The role of nucleosome positioning in the evolution of gene regulation. PLoS biology 8, e1000414 (2010).
    OpenUrlCrossRefPubMed
  43. 43.↵
    Yuan, G.C. et al. Genome-scale identification of nucleosome positions in S. cerevisiae. Science (New York, N.Y 309, 626–630 (2005).
    OpenUrlAbstract/FREE Full Text
  44. 44.↵
    Imakaev, M. et al. Iterative correction of Hi-C data reveals hallmarks of chromosome organization. Nature methods 9, 999–1003 (2012).
    OpenUrlCrossRefPubMed
  45. 45.↵
    Naumova, N. et al. Organization of the mitotic chromosome. Science (New York, N.Y 342, 948–953 (2013).
    OpenUrlAbstract/FREE Full Text
  46. 46.↵
    Sofueva, S. et al. Cohesin-mediated interactions organize chromosomal domain architecture. The EMBO journal 32, 3119–3129 (2013).
    OpenUrlAbstract/FREE Full Text
  47. 47.↵
    Enervald, E., Lindgren, E., Katou, Y., Shirahige, K. & Strom, L. Importance of Poleta for damage-induced cohesion reveals differential regulation of cohesion establishment at the break site and genome-wide. PLoS genetics 9, e1003158 (2013).
    OpenUrl
  48. 48.↵
    Hu, B. et al. Biological chromodynamics: a general method for measuring protein occupancy across the genome by calibrating ChIP-seq. Nucleic acids research 43, e132 (2015).
    OpenUrlCrossRefPubMed
Back to top
Posted August 24, 2016.
Download PDF

Supplementary Material

Email
Micro-C XL: assaying chromosome conformation at length scales from the nucleosome to the entire genome
Tsung-Han S. Hsieh, Geoffrey Fudenberg, Anton Goloborodko, Oliver J. Rando
bioRxiv 071357; doi: https://doi.org/10.1101/071357
Reddit logo Twitter logo Facebook logo LinkedIn logo Mendeley logo
Citation Tools

Subject Area