Hi-C2B: Optimised detection of chromosomal contacts within synchronised meiotic S. cerevisiae cells

Hi-C, a genome-wide chromosome conformation capture assay is a powerful tool used to study three-dimensional genome organisation by converting physical pairwise interactions into counts of pairwise interaction. To study the many temporally regulated facets of meiotic recombination in S. cerevisiae the Hi-C assay must be robust such that fine- and wide-scale comparisons between genetic datasets can be made. Here we describe an updated protocol for Hi-C (Hi-C2B) that generates reproducible libraries of interaction data with low noise and for a relatively low cost.


Introduction
Chromosome conformation capture methodologies based on "3C"-a "one-to-one" approach for studying contact frequencies between known genomic sequences-have advanced our understanding of chromosome and nuclear organisation in many organisms (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)14).Prior to the evolution of 3C, chromosome structure was studied via a combination of electron microscopy, immunofluorescence microscopy and the global, spatial patterning of proteins of interest determined by chromatin immunoprecipitation (ChIP).Now, chromosome conformation can also be analysed with respect to the underlying genomic sequence unveiling a wealth of information about the structural properties and spatial conformations of chromosomes within individual or populations of cells.
Early discoveries attributed to the 3C assay include identification of "flexible" chromosomes within S. cerevisiae (1); observations of enhancer-promoter interactions for gene regulation at the β-globin locus (2), furthered by studies that showed enhancer-promoter interactions require tissue-specific transcription factors (3)(4).Architectural chromatin loops have been uncovered as stable chromosome structures mediated, in mammals, by a partnership between ubiquitously expressed CTCF protein and the cohesin complex (5)(6)(7)(8)(9).On the back of such discoveries, and a greater accessibility and availability of high-throughput DNA sequencing technologies, development of many 3C-related techniques-including Hi-C-have transpired.Each 3C-based method customises the technique to provide independent, detailed documentation of the spatial organisation of chromosomes (10).
To capture chromosome conformation via 3C-related methodologies, the same initial approach is employed.Briefly, intact cells are isolated and subjected to treatment with formaldehyde in order to cross link physical DNA-protein and protein-protein interactions occurring in three-dimensional space (1).Cross-linked material is digested with a 6-or 4-bp cutting restriction enzyme, the latter increasing the spatial resolution of resulting data.Subsequent adaptations of the Hi-C method have also employed the use of MNase (Micro-C) to generate Hi-C interaction maps at nucleosome resolution (11).In all methods, fragmented DNA is subjected to in situ ligation prior to cross link reversal.At this stage, a 3C "template" is obtained which contains both linear and circular polymeric concatemers made from genomic segments reshuffled with respect to their spatial proximity at the time of fixation (10).Following generation of a 3C "template", 3C-based techniques differ in their detection and quantification of 3C ligation junctions (10).
Like classic 3C, Hi-C employs one of a number of restriction enzymes to digest cross-linked DNA to generate 5' overhangs that are subsequently filled with a biotin-labelled nucleotide (Fig. 1) (12).Bluntend fragments are then ligated to produce products where the ligation junction is marked with the biotin label, referred to as the 3C template (12).A Hi-C library is then created by mechanical shearing followed by enrichment steps before paired-end high-throughput sequencing, and mapping reads to a reference genome (12).The result: a two-dimensional matrix of the sequence-specific interactions occurring within the three-dimensional nuclear space (12).
Hi-C, the genome-wide adaptation of 3C, was introduced more than a decade ago (Hi-C1) (1).Hi-C2, described as an optimised method for the production of high-resolution chromosome conformation, includes a number of changes to the Hi-C1 protocol, particularly to the blunt-end ligation reaction (13): 1) removal of additional chromatin solubilisation by SDS prior to blunt-end ligation, with the aim to preserve nuclear structure thus reducing the occurrence of random ligation events; 2) reduced reaction length; 3) abolition of large-scale dilution during ligation which was included in the original Hi-C protocols to avoid random intermolecular ligation events.Hi-C1 in situ ligation reactions are diluted 15-fold whereas the Hi-C2 protocol documents just a 2-fold dilution; finally, 4) Hi-C2 includes a molecular crowding agent, polyethylene glycol (PEG) during ligation.
The methodological adaptations of Hi-C2 yielded a protocol with smaller processing and reaction volumes-thereby, dramatically reducing the overall cost of Hi-C library preparation (13).In addition, smaller processing volumes make handling of samples much easier such that more samples can be processed in parallel.Performing Hi-C1 restricted handling to a maximum of two samples per assay.
However, approximately eight samples can be processed in a single run of Hi-C2.Despite these benefits, we found that employing a Hi-C2-like protocol in S. cerevisiae generally yielded libraries with high levels of inter-chromosomal (trans) Hi-C interactions, indicative of random ligation.
Here we present a composite chromosome conformation capture assay that combines beneficial aspects of both the Hi-C1 and Hi-C2 protocols in order to generate high quality, reproducible S. cerevisiae Hi-C interaction maps with high yield, throughput and relatively low cost.For a summary of changes see Table 1.
6. Collect cells, again by centrifugation using a table top centrifuge (1 minute, 14,000 g, room temp).
Carefully remove the supernatant using a pipette and snap freeze the pellets on dry ice (see Note 6).
2. Wash cells in 1 mL ice-cold spheroplasting buffer by pipetting up and down (do not vortex, the cells are fragile at this stage).

Collect cells by centrifugation using a table top centrifuge (2 minutes, 2,500 g, room temp).
Carefully remove supernatant using a pipette (see Note 7).
5. Incubate for 10 minutes at 35˚C, inverting slowly after five minutes.11.Add 8 µL DpnII, mix thoroughly by inversion and incubate overnight in a thermomixer at 37˚C, 300 rpm.The final volume at this stage will be ~460 µL.

Fill-in, Blunt-end Ligate, Reverse Crosslinks
1. Incubate at 65˚C for 20 minutes in order to deactivate DpnII.(14) for libraries assayed by Hi-C1 ±PEG or Hi-C2 ±PEG (where Hi-C2 -PEG is equivalent to Hi-C2B, reported here).Addition of PEG-8000 to either protocol increases the rate of artifactual intermolecular ligation (detectable as increased trans contacts and higher matrix background (C), which is unwanted in Hi-C libraries).B) Table showing the percentage of reads that pass all Hi-C Pro filters as a fraction of the starting read value (all pairs) for Hi-C libraries assayed by Hi-C1 or Hi-C2 as described above.Increasing the rate of intermolecular ligation through the addition of PEG in either protocol generates an increase in the percentage of reads that pass all Hi-C Pro filters, but this increase in ligation efficiency is disproportionately made up of artifactual inter-molecular ligation products (detectable as increased trans contacts) at the expense of true intra-molecular proximity ligation events.C) Hi-C interaction maps from meiotic S. cerevisiae cells prepared by Hi-C1 or Hi-C2 as described above.Hi-C contact maps of chromosome 6, 11, and 7 are plotted at 10 Kb resolution.The black arrows indicate positions of the centromeres.

3 .
Quench formaldehyde by adding 1.3 mL of 2.5 M glycine (~0.35 M final) and incubate for a further 5 minutes at 30˚C with shaking (250 rpm) (see Note 4).

8 .
Solubilise chromatin by adding 3.8 µL 1% SDS and mixing gently by pipetting up and down to avoid making bubbles.Incubate at 65˚C for 5 minutes.

9 .
Cool mixture by placing it on ice for 5 minutes immediately after incubation with SDS.10.When the mixture has cooled, add 43 µL of 10% Triton X-100 and invert gently, 2-3 times to quench SDS.Incubate for 15 minutes at 37˚C, 300 rpm (see Note 9).

Fig 2 .Fig 3 .
Fig 2. Comparisons between Hi-C1 and Hi-C2B data output A) Fragment size distribution of short-and long-range intra-chromosomal (cis) contacts versus long-range inter-chromosomal (trans) contacts as reported by the Hi-C Pro pipeline (14) for libraries assayed by Hi-C1 or Hi-C2B methodologies.The table shows the percentage of reads that pass all Hi-C Pro filters as a fraction of the starting read value (all pairs).B) Hi-C interaction maps from meiotic S. cerevisiae cells prepared by Hi-C1 or Hi-C2B.Assessment of wide-scale patterns of Hi-C interaction highlight the similarity in output data from either methodology.Hi-C contact maps of chromosome 6, 11, and 7 are plotted at 5 Kb resolution.The black arrows indicate positions of the centromeres.C) Assessment of known chromosome conformations demonstrates similarities in the finescale Hi-C interaction patterns from libraries prepared by Hi-C1 or Hi-C2B.Arm compaction: axial compaction is indicated by the width of the main diagonal relative to the fixed-width black clamp.Chromosome misalignment: dashed pinked lines outline the interchromosomal cross-shape structures that manifest due to chromosome misalignment (present in meiotic recombination mutants for example).Inter-chromosome interactions: enhanced domains of inter-chromosomal contacts indicated by green arrowheads.Punctate interactions: focal grid-like patterns that arise from loops generated by meiotic cohesin (Rec8) (18); prominent examples are indicated by white arrowheads.Hi-C contact maps from meiotic S. cerevisiae cells are plotted at 2 Kb resolution.