|Home | About | Journals | Submit | Contact Us | Français|
Poly(ADP-ribose) polymerase1 (PARP1) is a global regulator of different cellular mechanisms, ranging from DNA damage repair to control of gene expression. Since PARP1 protein and pADPr have been shown to persist in chromatin through cell cycle, they may both act as epigenetic markers. However, it is not known how many loci are occupied by PARP1 protein during mitosis genome-wide. To reveal the genome-wide PARP1 binding sites, we used the ChIP-seq approach, an emerging technique to study genome-wide PARP1 protein interaction with chromatin. Here, we describe how to perform ChIP-seq in the context of PARP1 binding sites identification in chromatin, using human embryonic kidney cell lines.
PARP1 is an abundant and ubiquitous nuclear enzyme that catalyzes the transfer of poly(ADP-ribose) (pADPr) moiety from NAD to either a protein acceptor or an existing pADPr chain (1). PARP1 regulates many cellular functions, including stress-induced apoptosis, DNA damage detection and repair (2–4), transcriptional regulation and chromatin remodeling (5–7), and the control of gene expression by the induction of chromatin loosening at targeted genetic loci (8). The distribution of PARP1 in chromatin is nonrandom and globally regulates transcription (9). The dynamic regulation of poly(ADP-ribose) polymerase 1 protein binding to chromatin is mediated by nucleosomal core histones (10). For example, PARP1 and histone H1 exhibit a reciprocal pattern of chromatin binding at many RNA polymerase II-transcribed promoters (11). Since PARP1 is involved in the regulation of so many cellular mechanisms, we were inspired to study its genome-wide locations in the human genome in interphase and mitotic cells. Results of this work reveal the true loci of PARP1 in mitotic chromatin, allowing us to further understand the molecular mechanisms of PARP1-dependent processes.
In order to identify PARP1 protein binding sites in the human genome, we applied chromatin immunoprecipitation followed by sequencing (ChIP–seq). In ChIP–seq experiments, the precipitated ChIP-DNA fragments of interest are sequenced directly. In comparison to microarray, ChIP–seq has higher resolution, generates fewer artifacts, and provides greater coverage and a larger dynamic range. ChIP-seq studies have been used to characterize transcription factor binding (12–14), genome-wide nucleosome positioning (15), and to determine epigenetic changes (16). ChIP-seq technology does not require very long sequencing reads. Large numbers of short reads (35 bp) are sufficient for mapping binding sites in most organisms. Therefore, Illumina/Solexa and ABI/SOLiD have been favored over Roche/454 because they both generate millions of very short reads (about 35 bases/read), whereas Roche/454 generates fewer reads, but longer length (200–300 bases/read). These three main sequencing technologies are utilized on the basis of their applications. As a control, input DNA, consisting of nonimmunoprecipitated, sonicated and cross-linked DNA, has great importance in ChIP-Seq studies, as ChIP DNA samples are normally scored against the input DNA for transcription factor binding site (TFBS) identification (17). Even after successfully extracting ChIP-seq raw data, determination of binding sites from the data remains a formidable challenge. Therefore, many research groups published different algorithms that allow determining binding sites (18 – 24).
ChIP-seq can be divided in to the following steps (Figure 1): 1) ChIP; 2) Library preparation (end repair; addition of an ‘A’ base to the 3′-end of DNA fragments; ligation of adapters to DNA fragments; amplification of adapter-modified DNA fragments and gel purification; pre-sequencing control assays (enrichment check using positive/negative control primers)); and 3) library sequencing (annotation, sequence of DNA and validation by quantitative PCR (qPCR)).
ChIP combines immunoprecipitation of chromatin fragments and qPCR of precipitated DNA to map the binding sites of protein-DNA interaction in vivo. ChIP procedures fall into two main categories: those that use native chromatin prepared by nuclease digestion (designated N-ChIP) and those that use chromatin in which DNA and proteins are cross-linked, either chemically or with UV light (designated X-ChIP). Each procedure has its own advantages and drawbacks. Here, we outline the methods currently in use in our laboratory to isolate and immunoprecipitate cross-linked chromatin from cultured cells and to isolate and analyze immunoprecipitated protein and DNA.
Human embryonic kidney 293 (HEK293) cells were cultured in two plates of 10 cm (for the Western blot analysis) and four plates of 15 cm for ChIP. At confluence of 70%, half of the plates (for Western blot and ChIP) were treated with 60ng/mL nocodazole for 18hrs to synchronize the cells in prometaphase stage. The morphology of treated and untreated cells was the same (Figure 2A, B).
Before moving forward to ChIP-seq, precipitated ChIP DNA was checked by qPCR (StepOnePlus™, Applied Biosystems) using known PARP1 binding genes (positive control) (see Note 10).
Here, we describe the library preparation protocol for Illumina sequencing platforms, an important step for ChIP-seq. This protocol is based on the Illumina Sample Preparation Kit for Genomic DNA with minor modifications. Currently, most ChIP-Seq studies have been performed on Illumina. During Illumina library generation, oligonucleotide adapters are introduced at the ends of the small ChIP DNA fragments. Protocols differ depending on the sequencing platform used; 454/Roche, Solexa/Illumina, SOLiD/ABI, and Helicos each use different strategies to create a library representing the population of short DNA fragments selected by ChIP.
End repair is performed using the “End-It DNA End-Repair Kit” (Epicentre Biotechnologies). The End-It™ DNA End-Repair Kit was used to convert DNA with damaged or incompatible 5′-protruding and/or 3′-protruding ends to 5′-phosphorylated, blunt-end DNA for next-generation DNA sequencing adapters (Figure 3C).
To verify that a library has maintained a specific enrichment of ChIP target sites, an input library derived from the 10% input sample serves as a control to normalize the qPCR data to, in turn, determine the relative enrichment of a given target.
Using 50–200 ng of input chromatin extract, construct an input library as described above. Using qPCR analysis, check the ChIP-seq library as well as input library (see Note 11).
Analyze the qPCR results by manually determining the cycle threshold for each reaction across the plate within the linear range of the amplification curve. Divide the relative DNA amount of the ChIP-seq library by the relative DNA amount of the input library for a given primer set. The resultant quotient corresponds to the enrichment value of a target in the library over the input library. The enrichment value for a target primer set should be at least 20-fold greater than the enrichment of a negative control primer set.
On average, 10–20 million uniquely mapped tags are sufficient to identify all binding sites for a site-specific binding transcriptional factor. Ideally, these reads should come from two independent ChIP samples, with the binding sites identified in each replicate having at least a 60% overlap. After the sequencing is performed, the short tags (approx. 25–50 nt) are mapped to the human genome (hg18, 2006). The tags that map uniquely to only one location in the genome are selected, and then the unique tags are extended to the average size of the library fragments (approx. 200 nt) and mapped into consecutive maps running the length of each chromosome. The collected data can be visualized using the UCSC browser (http://www.genome.ucsc.edu) or the Affymetrix Integrated Genome Browser (http://www.affymetrix.com/partners_programs/-programs/developer/tools/download_igb.affx) (Figure 4). Target sites can be identified using a variety of peak calling methods (12–14, 19).
The ChIP-seq experiment can be validated by qPCR using primers designed to amplify a mapped sequence to the reference genome of a transcription factor. ChIP samples of specific, nonspecific and input are used as template. These input samples should be diluted 1:10.
ChIP-seq is an unbiased technique for genome-wide mapping of transcription factor binding sites, nucleosome mapping, and determining histone modifications at a high resolution. For successful ChIP-seq, both antibody specificity and sonication size of chromatin play very important roles. Before experimentation begins, specificity of antibody should be checked by Western blotting. The size of DNA fragments after sonication should be in range of 150–250 bp. Analysis of the ChIP-seq data explores unknown new binding sites and the location of binding sites relative to nearby potential new genes that may affect the functions of transcription factors.
We thank D. Martin and K. Pechenkina for comments on the manuscript. We also thank Greg Donahue (University of Pennsylvania Medical School) and Yan Zhou (Fox Chase Cancer Center) for advice on ChIP-seq data analysis. The research was supported by grants from the National Institutes of Health (R01 DK082623) to A.V.T.
1The choice of an antibody is very important for successful ChIP-seq. Therefore, it is essential to keep detailed records on all antibodies that include their catalog and lot numbers. We checked antibody by Western blot before the ChIP experiment. The major band on the blot should be the correct size for the protein of interest. Antibodies that work in a test immunoprecipitation will most likely work in a ChIP assay (Figure 5B).
2We designed the 100–150 bp oligos for amplicon. If possible, design replicate oligos for positive and negative controls, as well as for the test.
3Micrococcal nuclease or sonication can be used to shear the chromatin. Both techniques work fine. The protocol discussed here is based on shearing by sonication.
4To sonicate the chromatin, we used BioRuptor on a high setting for 5 cycles of 10 min with 30 sec on/off. Sonication depends on cell number, time, solution volume, and extent of cross-linking. To optimize sonication time, we collected 20 μL aliquots of chromatin after each 10 min cycle and analyzed size of DNA fragment on 1.5% agarose gel (Figure 5C).
5The size of sonicated chromatin should be in the range of 150–250 bp.
6We did not use agarose beads blocked with salmon sperm DNA as this could cause sequencing of the salmon sperm DNA. This would result in low-quality ChIP-seq data. Here, we used agarose beads from Invitrogen.
7If possible, use nonspecific and specific antibody from the same batch.
8After proteinase K (Invitrogen) and RNAse (Qiagen) digestion, purify the DNA by Qiagen PCR purification kit. This step prevents the DNA loss and saves time.
9The precipitated ChIP DNA should be quantitated by Quant-iT™ PicoGreen assay (Invitrogen) before proceeding to library preparation.
10Before library preparation, oligo pairs should be checked by SYBR Green-based q-PCR. To do this, use input DNA. For the dissociation curve, there should be a single peak of desired amplicon.
11Input DNA should be processed with specific and nonspecific antibody for library preparation. Input library is critical for determining a baseline genome for identification of binding sites.