ChIP-Seq is the bread and butter of histone and transcription factor research. It has seen countless modifications to allow it to perform under all the variables biology can throw at it, but with that specialization has come limitations to generalization.
Now the methodological maestros in the lab of Christoph Bock, which brought us single-cell methylome sequencing, have created an all purpose ChIP-Seq method that can handle all the challenges thrown at it.
This new method, called ChIPmentation takes advantage of tn5 transposase library prep. This ChIP-Seq library prep technique has been used to create the assay for transposase-accessible chromatin using sequencing (ATAC-seq), which has allowed for the indirect study of open chromatin but has not yet been applied to directly study histone modifications and transcription factors.
Co-first author Christian Schmidl shares, “We first tried to tagment the purified ChIP-DNA, but we found it difficult to set up robust tagmentation reactions with the low and very variable amounts of fragmented DNA recovered by ChIP. The trick is not to purify the DNA from the ChIP assay for tagmentation-based library preparation, but to tagment the chromatin during the immunoprecipitation step while still bound to the antibody and magnetic beads.”
The advantages behind a transposase library prep for ChIP-Seq are:
- Fragmented and immunoprecipitated chromatin, instead of the standard purified (protein free) immunoprecipitated DNA, has the sequencing compatible adapters tagged on.
- There are no sequencing adapter dimers.
- A much lower input of DNA is required for library prep, thus enabling the study of rare cell types.
- It requires only a single downstream purification before library amplification.
- The process is quicker than ‘Greased Lightning’.
Since the tagmentation is done on the immunoprecipitated and bead bound chromatin, the chromatin proteins appear to prevent extensive tagmentation. The team then used ChIPmentation across a 25-fold difference in tranposase concentrations to assess performance by examining:
- Size distribution of ChIPmentation libraries.
- Size distribution inferred from paired-end sequencing reads.
- Read-mapping performance.
- Concordance between sequencing profiles and signal correlations.
Using these metrics ChIPmentation was applied to study five histone mods (H3K4me1, H3K4me3, H3K27ac, H3K27me3, and H3K36me3) and four transcription factors (CTCF, GATA1, PU.1 and REST). When compared to standard ChIP-Seq it was found that:
- The ChIPmentation profiles were concordant with ChIP-Seq and biological replicates.
- There was a large reduction in the number of cells required, with the histones marks (H3K4me3 &HK27me3) generally needing 10,000 cells and 100, 000 cells for transcription factors (CTCF and GATA1).
- Just like standard ChIP-seq the outcome is dependent on antibody quality.
- It doesn’t require optimization of the tagementation for the different antibody optimized protocols, making it easy to ‘tag on’ to existing protocols.
The data also correlated with ATAC-seq and DNase-seq profiles after bioinformatic correction for the inherit biases of transposase. Once they fixed that all up they were able to match up the transcription factor footprints and also made some new observations on GATA1 and PU.1. Finally, they found that the average distance between tagmentation events was ~10bp, suggesting that nucleosome occupancy may be able to be inferred.
Future Applications: Transcription Factor Footprints and High Res Nucleosome Maps
The group shares that “Transcription factor (TF) footprinting can be used to understand high-resolution interactions of TFs with DNA beyond just describing a binding event. In addition, techniques that probe accessible chromatin including DNase-seq and ATAC-seq can infer factor binding without actually immunoprecipitating the factor. They do so by assessing regions that are protected from reactions (DNase digestion or transposase insertion in DNase-seq and ATAC-seq, respectively), creating an impression of the TF presence — a “footprint”.”
“Since ChIPmentation also uses a tagmentation reaction and this is performed on chromatin with DNA-binding proteins still in it, we have seen patterns that resemble footprints from TFs on ChIPmentation data from TFs. Furthermore, we hypothesise that an additional utility of TF footprint detection in ChIPmentation data will be the ability to distinguish real TF binding from indirect binding, which cannot be detected solely by ChIP.
“In a recent publication by the Greenleaf lab at Stanford it is shown that ATAC-seq generates a highly structured pattern of DNA fragment lengths and positions around nucleosomes. The correlation of this pattern with the ATAC-seq signal, along with elegant bias correction and background modeling allowed them to detect “nucleosome-bound” and “nucleosome-free” regions, inferring nucleosome positions with high precision.”
“ChIPmentation data from histones is different from ATAC-seq in that the majority of signal is depleted of the “nucleosome-free” signal, due to the enrichment of the histone mark by the ChIP procedure. We are interested in pursuing methods to model the ChIPmentation signal around nucleosomes, integrating bias and background correction, procedures which will be also useful for the more accurate detection of TF footprints. If it works, we would be able to position nucleosomes with high precision in all regions that are precipitated with a histone mark of choice
Co-first author André Rendeiro shares that “We are working on computational methods to better understand the biological relevance of these high-resolution patterns obtained by ChIPmentation. The usefulness of the high resolution patterns is under investigation at the moment, and it is highly dependent on good background model for the inherent Tn5 sequence bias.”
The authors elaborate that “The combination of various data from several histone marks on the same cells could greatly expand the amount of positioned nucleosomes and allow to tag each with the modifications it contains, generating a more complete chromatin map. Additionally, a great advantage is that regions typically beyond the reach of ATAC-seq (repressed chromatin and gene bodies) would still be profiled. This information would again, come along “for free” with ChIPmentation in addition to a normal ChIP-seq readout.”
The Clinical Potential of ChIPmentation
Bock concludes that “The protocol opens the possibility to analyse patient samples on a clinical timescale. With the current protocol it would take around 3 days from blood draw to analysed data, and we are working at an even faster version of ChIPmentation at the moment. Hence, we reach the timeframe to use personalized epigenomics to support clinical decision making, *e.g* by profiling the regulatory landscapes of patient cells and how they react to a certain drug. An advancement from the clinical perspective is the fact that ChIPmentation has relatively low input requirements, and could therefore be used to profile rare cell populations or minimize invasiveness to get material for personalized epigenomes.”