|Home | About | Journals | Submit | Contact Us | Français|
Identifying and characterizing highly accessible chromatin regions assists in determining the location of genomic regulatory elements and understanding transcriptional regulation. In this chapter we describe an approach to map accessible chromatin features in plants using the Assay for Transposase Accessible Chromatin, combined with high throughput sequencing (ATAC-seq), which was originally developed for cultured animal cells. This technique utilizes a hyperactive Tn5 transposase to cause DNA cleavage and simultaneous insertion of sequencing adapters into open chromatin regions of the input nuclei. The application of ATAC-seq to plant tissue has been challenging due to the difficulty of isolating nuclei sufficiently free of interfering organellar DNA. Here we present two different approaches to purify plant nuclei for ATAC-seq: the INTACT method (Isolation of Nuclei TAgged in Specific Cell Types) to isolate nuclei from individual cell types of the plant, and tissue lysis followed by sucrose sedimentation to isolate sufficiently pure total nuclei. We provide detailed instructions for transposase treatment of nuclei isolated using either approach, as well as subsequent preparation of ATAC-seq libraries. Sequencing-ready ATAC-seq libraries can be prepared from plant tissue in as little as one day. The procedures described here are optimized for Arabidopsis thaliana but can also be applied to other plant species.
Plants are sessile organisms that must precisely regulate their transcription in response to their environment, as well as for proper development, growth, and homeostasis. Transcription is associated with regions of relatively open chromatin, in which cis-regulatory elements such as enhancers and promoters can recruit transcription factors and RNA polymerase II to transcribe DNA . Binding of transcription factors to DNA generally results in the depletion of nucleosomes, rendering these regions hypersensitive to nucleases. Characterizing such regulatory regions throughout the genome has therefore relied on methods that combine enzymatic digestion of nuclear DNA and high-throughput sequencing, such as microccocal nuclease sequencing (MNase-seq, see Chapter 10) and DNase I Hypersensitivity sequencing (DNase-seq) [2, 3]. Alternatively, regulatory regions can be inferred by Chromatin Immunoprecipitation sequencing (ChIP-seq, see Chapter 5) where antibodies are used to pull down transcription factors or histone marks associated with active transcription .
An improved method for identifying accessible regions of chromatin and transcription factor binding is the Assay for Transposase-Accessible Chromatin with high-throughput sequencing (ATAC-seq) [5,6]. This method uses a hyperactive Tn5 transposase to integrate preloaded sequencing adapters into regions of open chromatin (Fig. 1A). ATAC-seq is a fast protocol with simple library amplification steps and requires very small amounts of starting material, making it a vast improvement over alternative methods. However, a drawback of this protocol is that the hyperactive Tn5 transposase also targets sources of extranuclear genetic material, including the genomes of mitochondria and chloroplasts. This decreases the proportion of reads that map to the nuclear genome, reducing the amount of information that can be used to identify regulatory regions of open chromatin. Such extranuclear reads must be discarded at the start of the data analysis process, diminishing the efficiency of the assay both in terms of cost and in effective use of materials. To gain the maximum efficiency of this powerful procedure, input material free from extranuclear genetic material, such as purified nuclei, is the ideal input for ATAC-seq
In this chapter, we describe the use of two different methods to isolate either total nuclei from tissues or nuclei from specific cell types of Arabidopsis thaliana (Fig. 1B). To isolate total nuclei from plant tissue we use extraction buffers with a non-ionic detergent to lyse organelles, followed by sucrose sedimentation to further purify the nuclei . This method of nuclei isolation can be done in any lab on most plant tissues. However, these partially purified nuclei still contain some organellar DNA in addition to nuclear DNA, which reduces the efficiency of Tn5 transposition to nuclear DNA and results in fewer sequencing reads that map to nuclear DNA. In addition, we describe the Isolation of Nuclei TAgged in specific Cell Types (INTACT) method to isolate nuclei from tissue or from specific cell types . This system uses two transgenes for nuclear targeting for affinity purification: 1) the Nuclear Tagging Fusion (NTF) construct, which encodes a fusion of WPP nuclear envelope-targeting domain, a Green Fluorescent Protein (GFP), and the Biotin Ligase Recognition Peptide (BLRP); and 2) an E. coli biotin ligase (BirA), which biotinylates the BLRP tag. The BirA is expressed from a constitutive promoter while the NTF is expressed either from a constitutive or cell type-specific promoter. The specificity of the NTF promoter determines which cell types will have biotinylated nuclei and can then be isolated by affinity purification with streptavidin-coated magnetic beads . A key advantage of the INTACT approach is not only that the isolated nuclei have less organellar DNA contamination, but also that this method can be used to selectively isolate nuclei from specific cell types. While INTACT is a powerful technique, it does require that stable transgenic lines containing BirA and NTF cassettes for the cell type of interest are available, which are time-consuming to generate and can be limiting for many species. Even so, the protocol described here, particularly ATAC-seq using sucrose sedimentation-purified nuclei, can readily be adapted for chromatin profiling in any plant species.
Users should either begin at section 3.1 for affinity purification of nuclei using INTACT, or at section 3.2 for isolation of total nuclei. In either case, the purified nuclei are used for tagmentation by Tn5 transposase in step 3.3. All procedures are carried out at room temperature (25 °C) unless otherwise specified.
This work was supported by the National Science Foundation Grant no. 1238243. We thank Paja Sijacic and Shannon Torres for helping to optimize the protocol for nuclei isolation and for suggestions on the manuscript.
1This protocol is optimized for 3 g of root or 0.5 g of leaf tissue from Arabidopsis thaliana. Ground leaf tissue contains more debris, relative to roots, and therefore requires a lower amount of starting material to obtain highly purified nuclei. INTACT may also be performed on fresh tissue by chopping the tissue in NPB as opposed to grinding to a fine powder using liquid nitrogen. However, this approach does require the use of fresh tissue. The number of samples that can be run through INTACT purification simultaneously is mainly limited by the capacity of the DynaMag 15 magnetic rack used for nuclei capture. Up to four separate samples can be processed in parallel using one DynaMag 15 magnetic rack.
Using an INTACT line with nuclei labeled in the root epidermal non-hair cell type, approximately 200,000 purified nuclei can be obtained from 3 g of roots. Larger amounts of tissue can be used for purifying nuclei from less abundant cell types, and this generally only requires adjustments to the amount of streptavidin beads used and the volume of solution used for bead capture. See  for more details on variations in the INTACT procedure.
2After isolating the bead bound nuclei, keep the sample on ice while quantifying the nuclei from the aliquot in Subheading 3.1 Step 12. Do not freeze the isolated nuclei before doing tagmentation and library preparation. Freezing and thawing of isolated nuclei can disrupt protein-DNA interactions.
3After DAPI staining, nuclei purified by INTACT can be easily identified and counted using a hemocytometer. The ideal setup for visualizing nuclei is under a mix of dim white light and DAPI channel fluorescence. The dim white light allows for visualization of the hemocytometer grid and the beads, and the DAPI fluorescence allows for the visualization of nuclei. A sample image of isolated bead-bound nuclei is shown in Fig. 1C. A nucleus is identified as a circle that fluoresces in the DAPI channel and has several beads clustered around it. Minimal cellular debris or contaminating unbound nuclei should be observed in the final product. These contaminants may be further reduced by using fewer beads and by increasing the volumes of NPB and NPBt used during purification as described in Note 1.
We have successfully used as few as 20,000 to as many as 200,000 INTACT-purified nuclei in this procedure without altering any other parameters of the protocol presented here.
4This protocol is optimized for less than 1 g of root or 0.5 g of leaf tissue. Ground leaf tissue contains more debris relative to roots, and therefore requires a lower amount of starting material to obtain purified nuclei. As with the INTACT protocol, sucrose sedimentation of nuclei may also be performed on fresh tissue by chopping the tissue in NPB as opposed to grinding to a fine powder using liquid nitrogen. However, this approach does require the use of fresh tissue. We recommend starting with the minimum amount of tissue needed to obtain the required number of nuclei (e.g. 50,000 per ATAC-seq reaction).
5Proper separation of nuclei from other cellular debris requires the nuclei to pass through the sucrose cushion during centrifugation. The NEB3 resuspended nuclei should therefore be placed gently on top of NEB3 layer present in the tube. After centrifugation, the contaminating organelles and debris may be visible at the top of the tube. If leaf tissue was used, the top layer will become greener after centrifugation and the pellet will become noticeably less green than it was prior to centrifugation.
6After DAPI staining, nuclei purified by sucrose sedimentation can be identified and quantified using a hemocytometer. A mixture of DAPI-channel fluorescence and white light illumination allows the stained nuclei and the hemocytometer grid to be seen simultaneously. A sample image of isolated nuclei is shown in Fig. 1C. A nucleus is identified as a punctate circle with strong DAPI fluorescence. The nucleus is typically ~5 μm in size and can be easily identified at 200X and 400X magnifications. Cellular debris may be observed in the final preparation, but this generally does not affect the outcome of the ATAC-seq procedure. To reduce cellular debris contamination, starting tissue can be chopped with a razor blade (see Note 4) and/or additional NEB3 wash steps may also be done by repeating Subheading 3.2 steps 7–9 for a second sucrose cushion centrifugation.
7Ensure that all work surfaces, pipettes, and reagents needed for amplification and library preparation are free of DNA contamination. For library amplification, unique barcoded adapters are used for each sample if multiple libraries are to be sequenced in an individual flow cell lane. The sequences of all primers can be found in the supplementary material of .
8The number of PCR cycles needed to amplify ATAC libraries is determined by the PCR reaction in Subheading 3.4 step 5. We recommend using the minimum number of cycles necessary to obtain a sufficient molar amount of library for Illumina sequencing. This must be determined empirically and will also depend on the number of libraries to be pooled for sequencing.
9The ratio of Ampure XP PCR Purification beads to PCR volume determines the size of purified DNA fragments isolated. The 1.5 Ampure bead to PCR reaction ratio results in the isolation of DNA fragments shown in Fig. 2A. Using ratios that have higher proportions of beads may result in purification of sequencing adapters and PCR primers, which can negatively affect sequencing.
10A drying time of 5 minutes is generally sufficient to remove all traces of ethanol from the beads, but this time may vary based on humidity and room temperature. Georgia is very humid in the summer. Ensure that all ethanol has evaporated before moving on to the next step. Do not allow beads to dry to the extent that the pellet begins to crack.
11Libraries can generally be visualized by agarose gel electrophoresis followed by ethidium bromide staining. Sensitivity can be greatly increased by staining the gel with Sybr green stain or using an Agilent Bioanalyzer or equivalent instrument, if available.
The libraries that we have prepared using this method generally present as a DNA smear starting at ~180 bp and ranging to greater than 1 kb, with peak intensity between ~180 – 500 bp (See Figure 2A). The original publication on ATAC-seq  reported a nucleosome-like periodicity in the library size distribution, but we have not observed this phenomenon as assayed by either electrophoresis or estimation of fragment size distribution based on distance between paired-end sequencing reads, as shown in Fig. 2B. This lack of observed nucleosome fractions may be due to size selection of library fragments by Ampure XP beads and the low transposase to nuclei ratio described in this protocol.
12Paired-end sequencing is recommended in order to maximize the number of transposase integration events that can be observed in a given sample and to allow measurement of the length of the sequenced fragments (Fig. 2B).
To identify open chromatin regions in Arabidopsis, users should aim to obtain at least 10–20 million reads per library that map to the nuclear genome. For transcription factor footprinting the number of nuclear genome-mapping reads should be increased to at least 100 million per library.
When using sucrose sedimentation for nuclei purification, users should expect ~50% of reads to map to the nuclear genome, while the use of INTACT purification will increase this number to > 90%.
13Sequencing reads are checked for overall quality using FastQC (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/) or equivalent. The reads are aligned to the TAIR10 Arabidopsis thaliana genome (https://www.arabidopsis.org/download/index-auto.jsp?dir=%2Fdownload_files%2FGenes%2FTAIR10_genome_release) using Bowtie2 (http://bowtie-bio.sourceforge.net/bowtie2/index.shtml). The resulting SAM file is converted to a binary BAM file, which is sorted and indexed using Samtools (http://samtools.sourceforge.net/). The quality of the resulting BAM file, including fragment size distribution, is analyzed using Picard Tools (https://broadinstitute.github.io/picard/). Alignment data is visualized using the Integrated Genome Viewer (http://software.broadinstitute.org/software/igv/). For ease of visualization, BAM files were converted to BigWig files using DeepTools BamPECoverage tool (http://deeptools.readthedocs.io/en/latest/index.html). Downstream analyses of ATAC-seq data include calling peaks with HOMER (http://homer.salk.edu/homer/index.html), editing BED files with bedtools (http://bedtools.readthedocs.io/en/latest/) and identifying transcription factor footprints using pyDNase (http://pythonhosted.org/pyDNase/).