From cells to chromatin: capturing snapshots of genome organization with 5C technology

Abstract

In eukaryotes, genome organization can be observed on many levels and at different scales. This organization is important not only to reduce chromosome length but also for the proper execution of various biological processes. High-resolution mapping of spatial chromatin structure was made possible by the development of the chromosome conformation capture (3C) technique. 3C uses chemical cross-linking followed by proximity-based ligation of fragmented DNA to capture frequently interacting chromatin segments in cell populations. Several 3C-related methods capable of higher chromosome conformation mapping throughput were reported afterwards. These techniques include the 3C-carbon copy (5C) approach, which offers the advantage of being highly quantitative and reproducible. We provide here an updated reference protocol for the production of 5C libraries analyzed by next-generation sequencing or onto microarrays. A procedure used to verify that 3C library templates bear the high quality required to produce superior 5C libraries is also described. We believe that this detailed protocol will help guide researchers in probing spatial genome organization and its role in various biological processes.

Fig. 1. Mapping chromosome conformation at high resolution with 5C technology

(a) Schematic representation of the main experimental steps involved in 5C library production. 3C libraries are first prepared by capturing chromatin contacts in vivo with formaldehyde cross-linking, digestion with a restriction enzyme, inter-molecular ligation of cross-linked fragments with T4 DNA ligase, and reverse cross-linking. 5C libraries are next derived by annealing and ligating 5C primers onto predicted 3C junctions by ligation-mediated amplification (LMA; see text for details). (b) Flowchart of the 5C protocol steps described in this article. 3C library production has been described elsewhere and is presented in grey (see text for references).

Fig. 2. Example of a cellular 3C library titration

(a) Acceptable result of a cellular 3C library titration resolved on agarose gel and stained with ethidium bromide. This example shows the result of a 3C library analyzed by twofold serial dilution and amplified by endpoint PCR with GD34 and GD35 primers (Table 1). This primer pair recognizes neighboring fragments in a gene desert region (ENCODE region ENr313). Endpoint PCRs were resolved on a 1.5 % agarose gel containing 0.5 µg/ml ethidium bromide. (b) Gel quantification of the titration shown in (a). The volume of cellular 3C library stock is indicated on the x-axis. The intensity of PCR products stained with ethidium bromide is indicated on the y-axis. Ideal volume range is highlighted in orange.

Fig. 3. Examples of 3C library content profiling

(a) Profiling 3C libraries using random collisions. This example shows a “fixed point” analysis of two 3C libraries in a gene desert region. 3C libraries were from a differentiation time-course of NT2/D1 cells with retinoic acid as described previously [13]. “Fixed point” analysis was performed by combining the GD51 primer pair-wise with oligos GD52 to GD58 recognizing downstream BglII restriction fragments (Table 1). Interaction frequencies measured in the ENr313 gene desert region derive from random collisions, which decrease with increasing distance as this region is devoid of looping contacts. Error bars show the standard deviation (SD). (b) Profiling 3C libraries using known long-range interactions in a given locus. This example shows a “fixed point” analysis of the human HoxC cluster 5’ end in undifferentiated NT2/D1 cells as described previously [13]. Interaction frequencies between a region containing the HoxC9 gene promoter and downstream cluster genes confirm the presence of a looping contact with HoxC11 in this cell line. Error bars show the standard error of the mean (SEM). The x-axis indicates the linear genomic distance between interacting chromatin fragments. Interaction frequencies are indicated on the y-axis. Linear representations of the corresponding genomic regions and predicted BglII restriction patterns are shown above each graph.

Fig. 4. 3C to 5C library conversion

5C libraries are produced by annealing 5C primers at predicted 3C junctions in a multiplex setting followed by specific ligation of annealed primers with an NAD-dependent DNA ligase. The universal tails of 5C primers are illustrated as black and grey lines and are used to amplify libraries in a single PCR step. Universal tail and corresponding one-step PCR primer sequences depend on whether microarray hybridization or high-throughput sequencing is selected for analysis. Microarray forward and reverse tail sequences are T7: 5’-TAATACGACTCACTATAGCC-3’ and T3c: 5’-TCCCTTTAGTGAGGGTTAATA-3’ respectively. Reverse universal PCR primers used for microarray analysis must be fluorescently labeled at their 5’ ends as highlighted by the green star. High-throughput sequencing forward and reverse tail sequences are S1: 5’-CCTCTCTATGGGCAGTCGGTGAT-3’ and S2c: 5’- AGAGAATGAGGAACCCGGGGCAG-3’ respectively.

Fig. 5. Types of 5C experimental design

(a) Alternating scheme. Alternating forward and reverse 5C primers represent each consecutive fragment. (b) Anchored scheme. A reverse 5C primer represents a given fragment while forward 5C primers represent the rest. (c) Mixed scheme. This experimental design features both fixed anchored and alternating schemes.

Fig. 6. Controlled 5C library production

This example shows the results for the generation of 5C libraries from two cellular states and one control 3C library. “No ligase” and “ No template” controls for at least one library should be included in the initial 5C library production to verify the absence of contamination and the specificity of 5C product formation. 5C libraries were amplified with T7 and Cy3-labeled T3c primers and resolved on a 2.5 % agarose gel containing 0.5 µg/ml ethidium bromide. The “No 5C primer” control is not shown here.