Monday, February 16, 2009

[Techniques] Analyzing the genome-wide chromatin landscape: ChIP-Seq


New technologies arise every few years that help us address important biological questions from a new angle. An example of this is the DNA microarray. First developed in the 1990s, this important tool has been essential for example, for profiling gene expression in specific cell types.
Nowadays, high-throughput DNA sequencing serves as an example of such important tools. In 2007, Johnson et al.1 described a technique they called “ChIP-Seq” (from CHromatin ImmunoPrecipitation and SEQuencing) to study the binding sites of a specific transcription factor across the entire human genome using the advance DNA sequencing technology from Solexa/Illumina.

Before explaining what “ChIP-Seq” is, let me tell you a little about the high-throughput technologies used before that to study the binding sites of proteins to DNA.
The most common method of locating these sites in vivo is known as chromatin immunoprecipitation (ChIP). In this technique, cells are treated with a reagent (typically formaldehyde) that crosslinks proteins to DNA. Every protein that is bound to DNA in the moment the reagent is applied will remain bound to DNA. Afterwards, the chromatin is isolated, sheared and incubated with an antibody directed to the protein of interest. This will precipitate the protein and the DNA bound by it (as they are crosslinked). After reverse-crosslinking, the precipitated DNA is analyzed. At first, this technique was restricted to studying if a particular protein was bound to a determined gene (to its promoter, protein coding region, etc), so the analysis of the precipitated DNA fragments was restricted to PCR (with the gene’s-specific primers) to see if the protein indeed allowed the precipitation of that sequence (which would result in an enrichment of the gene in the precipitated DNA vs the control, where no specific antibody is used).
About 6 years ago, a more ‘genomic’ approach was derived from this technique, resulting in the “ChIP-chip” method. The difference with the previously described technique is that after the incubation with the antibody, all the precipitated DNA fragments are used as probes on a DNA microarray (chip). This allows for a high-throughput analysis of the DNA binding sites of a specific protein (limited of course, by the sequences represented on the chip, See "Advantages" below) instead of studying single genes.

In a way to improve this technique, Johnson et al.1 replaced the “chip” by direct DNA sequencing.
What does this mean? After the precipitation and reverse-crosslinking steps, DNA fragments are sequenced (in that specific paper, using Solexa/ Illumina technology). After sequencing, the reads are mapped to the genome to determine their locations. In this way, the genome-wide DNA binding sites of a particular protein can be assessed [See figure]. The control simply omits the antibody (again, you will be looking for enrichment in your treated sample when compared to the control).
What are the advantages of ChIP-Seq vs ChIP-chip? First, you avoid all complications arising from array hybridization (probes with different optimal temperatures for binding to their complementary strands, probes that hybridize to more than one DNA sequence, interference of hybridization by DNA secondary structure). Second, you are no longer limited to what’s represented on a chip. Tiling arrays may limit such an advantage, though. Nevertheless, ChIP-Seq is cheaper (compared, for example, to whole-genome human tiling arrays)2.
Finally, you can apply ChIP-Seq regardless of whether a microarray chip has been developed for a particular species.
This technique, along with the fact that prices of high-throughput DNA sequencing are continuously becoming more accessible, will allow for the identification of the binding sites of not only transcription factors, but chromatin remodeling complexes, structural components, etc, across an entire genome.

[Image is from ref 2.]

1Johnson DS, Mortazavi A, Myers RM, Wold B (2007) Genome-wide mapping of in vivo protein-DNA interactions. Science 316, 1497.
2Fields S (2007) Molecular biology. Site-seeing by sequencing. Science. 316(5830):1441-2.