Mapping cis- and trans- chromatin interaction networks using chromosome conformation capture (3C)

Abstract

Expression of genes can be controlled by regulatory elements that are located at large genomic distances from their target genes (in cis), or even on different chromosomes (in trans). Regulatory elements can act at large genomic distances by engaging in direct physical interactions with their target genes resulting in the formation of chromatin loops. Thus, genes and their regulatory elements come in close spatial proximity irrespective of their relative genomic positions. Analysis of interactions between genes and elements will reveal which elements regulate each gene, and will provide fundamental insights into the spatial organization of chromosomes in general. Long-range cis- and trans- interactions can be studied at high resolution using chromosome conformation capture (3C) technology. 3C employs formaldehyde crosslinking to trap physical interactions between loci located throughout the genome. Crosslinked cells are solubilized and chromatin is digested with a restriction enzyme. Chromatin is subsequently ligated under conditions that favor intramolecular ligation. After reversal of the crosslinks, the DNA is purified and interaction frequencies between specific chromosomal loci are determined by quantifying the amounts of corresponding ligation products using polymerase chain reaction (PCR). This chapter describes detailed protocols for 3C analysis of chromatin interactions in the yeast Saccharomyces cerevisiae and in mammalian cells.

Figure 1

Schematic representation of the chromosome conformation capture (3C) technology. Formaldehyde is used to treat either yeast or human cells resulting in covalent crosslinks between interacting DNA fragments via protein interactions (indicated by gray square and black circle). The crosslinked cells are then digested, followed by ligation under dilute DNA concentrations. The crosslinks are then reversed and the DNA is purified. The PCR products are detected (PCR primers indicated by arrows) and can then be quantified by gel electrophoresis.

Figure 2

Titration and gel quantification of yeast 3C and control templates. The left (control template) and right panels (3C template) represent a titration and quantification using two primer pair combinations (detecting interactions between loci separated by 10 kb and 80 kb respectively). The amount of PCR product is plotted versus the template DNA concentration in micrograms. For both primer pair combinations the control template yields similar amounts of PCR product while the 3C template yields significantly more PCR product for the primer pair that detect interactions between loci separated by 10 kb. The linear range for PCR amplification is found to the left of the gray line. The template DNA concentration chosen should be enough to yield sufficient PCR product to allow visualization and quantification on a gel, but should be within the linear range of PCR detection. The same template concentration must be used for all subsequent PCR reactions.

Figure 3

Hypothetical results for a 3C looping experiment. (A) Analysis of a looping interactions between two elements (a and b) under two conditions: The open circles indicate the expected pattern of interactions when a and b engage in a specific looping interaction. The black circles indicate the expected pattern of interactions when and b do not engage in a specific interactions. For both datasets a number of interactions were determined between element a and element b as well as a set of interactions between element a and loci located in between elements a and b and loci located downstream of b. In both cases, a interacts highly with sites very close to it and this interaction frequency decreases as the genomic distance between loci increases. However, in the case of the looping interaction, a local peak of high interaction frequencies is observed ~55 kb which is indicative of a looping interaction at the site of the sequence element b. As well as determining whether a looping interaction is present, information regarding the flexibility and the level of compaction can be inferred from the graphs (3,4). (B) A hypothetical model inferred from the datasets in (A).