We present a protocol for assay for transposase-accessible chromatin with high-throughput sequencing (ATAC-seq) specifically on adipocytes using nucleus sorting with adipose tissues isolated from transgenic reporter mice with nuclear fluorescence labeling.
Assay for transposase-accessible chromatin with high-throughput sequencing (ATAC-seq) is a robust technique that enables genome-wide chromatin accessibility profiling. This technique has been useful for understanding the regulatory mechanisms of gene expression in a range of biological processes. Although ATAC-seq has been modified for different types of samples, there have not been effective modifications of ATAC-seq methods for adipose tissues. Challenges with adipose tissues include the complex cellular heterogeneity, large lipid content, and high mitochondrial contamination. To overcome these problems, we have developed a protocol that allows adipocyte-specific ATAC-seq by employing fluorescence-activated nucleus sorting with adipose tissues from the transgenic reporter Nuclear tagging and Translating Ribosome Affinity Purification (NuTRAP) mouse. This protocol produces high-quality data with minimal wasted sequencing reads while reducing the amount of nucleus input and reagents. This paper provides detailed step-by-step instructions for the ATAC-seq method validated for the use of adipocyte nuclei isolated from mouse adipose tissues. This protocol will aid in the investigation of chromatin dynamics in adipocytes upon diverse biological stimulations, which will allow for novel biological insights.
Adipose tissue, which is specialized for storing excess energy in the form of lipid molecules, is a key organ for metabolic regulation. The strict control of adipocyte formation and maintenance is vital for adipose tissue function and whole-body energy homeostasis1. Many transcriptional regulators play a critical role in the control of adipocyte differentiation, plasticity, and function; some of these regulators are implicated in metabolic disorders in humans2,3. Recent advances in high-throughput sequencing techniques for gene expression and epigenomic analysis have further facilitated the discovery of the molecular regulators of adipocyte biology4. Molecular profiling studies using adipose tissues are challenging to conduct due to the heterogeneity of these tissues. Adipose tissue consists primarily of adipocytes, which are responsible for fat storage, but also contains various other cell types, such as fibroblasts, endothelial cells, and immune cells5. In addition, the cellular composition of adipose tissue is dramatically altered in response to pathophysiological changes such as temperature and nutritional status6. To overcome these problems, we previously developed a transgenic reporter mouse, named Nuclear Tagging and Translating Ribosome Affinity Purification (NuTRAP), which produces GFP-tagged ribosomes and mCherry-tagged biotinylated nuclei in a Cre recombinase-dependent manner7. The dual-labeling system enables one to perform cell type-specific transcriptomic and epigenomic analysis with tissues. Using NuTRAP mice crossed with adipocyte-specific adiponectin-Cre lines (Adipoq-NuTRAP), we previously characterized gene expression profiles and chromatin states from pure adipocyte populations in vivo and determined how they are altered during obesity7,8. Previously, NuTRAP mice crossed with brown and beige adipocyte-specific Ucp1-Cre lines (Ucp1-NuTRAP) allowed us to characterize the epigenomic remodeling of the rare thermogenic adipocyte population, beige adipocytes, in response to temperature changes9.
ATAC-seq is a widely used analytical method to assess genome-wide chromatin accessibility.The hyper-reactive Tn5 transposase used in ATAC-seq allows for the identification of open chromatin regions by tagging sequencing adapters in the chromatin-accessible region of nuclei10. ATAC-seq is a simple method, yet it provides robust results and is highly efficient even with low-input samples. It has, thus, become one of the most popular epigenomic profiling methods and has contributed to the understanding of the regulatory mechanisms of gene expression within diverse biological contexts. Since the original ATAC-seq protocol was created, various ATAC-seq-derived techniques have been further developed to modify and optimize the protocol for various types of samples. For example, Fast-ATAC is designed for analyzing blood cell samples11, Omni-ATAC is an optimized protocol for frozen tissue samples12, and MiniATAC-seq is effective for early-stage embryo analysis13. However, applying the ATAC-seq method to adipocytes, especially from tissue samples, is still challenging. In addition to the heterogeneity of adipose tissue, its high lipid content may interfere with efficient recombination reactions by Tn5 transposase even after nucleus isolation. Furthermore, the high mitochondrial content in adipocytes, particularly in brown and beige adipocytes, causes high mitochondrial DNA contamination and wasted sequencing reads. This paper describes a protocol for adipocyte-specific ATAC-seq using Adipoq-NuTRAP mice (Figure 1). By taking advantage of fluorescence-labeled nucleus sorting, this protocol allows the collection of pure populations of adipocyte nuclei away from other confounding cell types and the efficient removal of lipids, mitochondria, and tissue debris. Hence, this protocol can generate cell type-specific high-quality data and minimize waste from mitochondrial reads while using a reduced amount of input and reagents compared to the standard protocol.
Animal care and experimentation were performed according to procedures approved by the Institutional Animal Care and Use Committee of Indiana University School of Medicine.
1. Preparations before beginning the experiment
2. Nucleus isolation
3. Nucleus sorting
4. Tn5 Tagmentation (Table 1)
5. DNA purification
NOTE: The following procedure uses the PCR purification kit mentioned in the Table of Materials. Any other similar DNA purification methods can be used.
6. PCR amplification (Table 2)
NOTE: The primers used in this study are listed in Table 3.
7. Real-time quantitative PCR (qPCR) test (Table 4)
NOTE: This step aims to determine the additional cycles needed to amplify the DNA. It is optional but highly recommended, especially for new experiments.
8. Additional PCR amplification
9. Second DNA purification using the PCR purification kit
10. DNA fragment size selection
NOTE: Use solid phase reversible immobilization beads, as described below. Any other similar DNA purification beads can be used according to the manufacturer's instructions. The SPRI bead suspension should be completely resuspended and equilibrated at RT with rotation before use.
11. DNA quantification using a fluorometer
12. Library quality check by high-sensitivity electrophoresis systems
13. Quality check of the ATAC-seq library by using targeted qPCR
14. Sequencing
To analyze adipose tissue using this ATAC-seq protocol, we generated Adipoq-NuTRAP mice that were fed chow diets; we then isolated adipocyte nuclei from epididymal white adipose tissue (eWAT), inguinal white adipose tissue (iWAT), and brown adipose tissue (BAT) by using flow cytometry. The isolated nuclei were used for tagmentation, followed by DNA purification, PCR amplification, quality check steps, sequencing, and data analysis, as described above. The purpose of this representative experiment was to profile the chromatin accessibility of pure adipocyte populations isolated from different fat depots.
We observed that the mCherry-labeled and GFP-labeled adipocyte nuclei were clearly distinguishable from the nuclei of other cell types isolated from the adipose tissues of an Adipoq-NuTRAP mouse using flow cytometry (Figure 2A–C). There were differences in the adipocyte fractions depending on the type of adipose depot: ~50% in eWAT, ~30% in iWAT, and ~65% in BAT (Figure 2D)7. We collected 10,000 nuclei within 2-10 min per sample and used them for the ATAC-seq procedures. We ran a total of 10-13 PCR cycles (five cycles of the first PCR and five to eight cycles of the second PCR, Figure 3A). If the samples require more than a total of 15 PCR cycles, this may indicate low-quality samples (Figure 3B). The size distribution analysis of the ATAC-seq libraries demonstrated multiple peaks corresponding to the nucleosome-free region (NFR) and mono, di, and multi-nucleosomes (Figure 4A–C), with average sizes of ~500-800 bp. Poor-quality samples typically showed mostly NFR with no or few nucleosomal peaks (Figure 4D). The quality check tests by the qRT-PCR analysis showed 10-20-fold enrichments with the positive genomic elements near adipocyte marker genes such as Adipoq, Fabp4, Plin1, and Pnpla2, while no enrichment was shown with the negative control (Figure 5). We also observed enrichment with the thermogenic gene Ucp1 specifically in BAT but not in eWAT or iWAT (Figure 5).
After the sequencing and analysis, we identified ~55,000 peaks combined from three different adipose depots (eWAT, iWAT, and BAT). The mitochondrial reads were <2%, and the fraction under the peaks was ~18%-44% (Figure 6A, B). Poor-quality samples may have significantly higher mitochondria reads and/or a lower fraction of reads in the peak regions. Visual inspection of the ATAC-seq library tracks revealed multiple strong peaks with high signal-to-noise ratios near adipocyte marker genes, such as Adipoq, Plin1, and Fabp4, from all the adipose depots (Figure 7A–C). We also observed strong ATAC-seq peaks at the brown adipocyte maker Ucp1 locus in the BAT sample, but these peaks were not observed in the eWAT or iWAT samples (Figure 7D). All these data indicate successful ATAC-seq data generated from adipocyte nuclei.
Figure 1: A schematic flowchart of adipocyte-specific ATAC-seq using NuTRAP mice. The steps unique to this protocol are highlighted in the red shaded boxes. Please click here to view a larger version of this figure.
Figure 2: Representative FACS gating illustrating the isolation of mCherry/GFP-tagged adipocyte nuclei from the adipose tissues of an Adipoq-NuTRAP mouse. (A–C) Flow cytometry analysis of isolated nuclei from the eWAT, iWAT, and BAT of an Adipoq-NuTRAP mouse.(D) Quantitative analysis of nuclei from adipocytes and non-adipocytes in the eWAT, iWAT, and BAT of an Adipoq-NuTRAP mouse. Data are mean ± SEM (n = 3). These nuclei count data are from Roh et al.7. Abbreviations: FACS = fluorescence-activated cell sorting; GFP = green fluorescent protein; eWAT = epididymal white adipose tissue; iWAT = inguinal white adipose tissue; BAT = brown adipose tissue; NuTRAP = nuclear tagging and translating ribosome affinity purification; Adipoq-NuTRAP = NuTRAP mouse crossed with adipocyte-specific adiponectin-Cre line; FSC-A = forward scatter-peak area; FSC-H = forward scatter-peak height; SSC-A = side scatter-peak area; FITC-A = fluorescein isothiocyanate-peak area. Please click here to view a larger version of this figure.
Figure 3: Representative amplification curves from qRT-PCR. (A) Good-quality sample that needs seven additional amplification cycles. (B) Poor-quality sample that need ≥15 additional cycles. Abbreviation: qRT-PCR = quantitative reverse transcription PCR. Please click here to view a larger version of this figure.
Figure 4: Size distribution profiles of the ATAC-seq libraries. (A–C) Good-quality and (D) poor-quality libraries are shown as representative results. Abbreviations: ATAC-seq = assay for transposase-accessible chromatin with high-throughput sequencing; FU = fluorescence units; bp = base pairs; NFR = nucleosome-free region; eWAT = epididymal white adipose tissue; iWAT = inguinal white adipose tissue; BAT = brown adipose tissue. Please click here to view a larger version of this figure.
Figure 5: Quality control qPCR analysis of the ATAC-seq libraries. Enrichments for promoters and enhancers near the general adipocyte markers Adipoq, Fabp4, Plin1, and Pnpla2 or the brown adipocyte marker Ucp1. Abbreviations: ATAC-seq = assay for transposase-accessible chromatin with high-throughput sequencing; NC = negative control; eWAT = epididymal white adipose tissue; iWAT = inguinal white adipose tissue; BAT = brown adipose tissue. Data are mean ± SEM (n = 2). Please click here to view a larger version of this figure.
Figure 6: Mitochondria reads and fractions under the peaks of the ATAC-seq libraries. (A) Mitochondria reads (%) and (B) fractions under the peaks (%) from eWAT, iWAT, and BAT. Abbreviations: ATAC-seq = assay for transposase-accessible chromatin with high-throughput sequencing; eWAT = epididymal white adipose tissue; iWAT = inguinal white adipose tissue; BAT = brown adipose tissue. Please click here to view a larger version of this figure.
Figure 7: ATAC-seq signal tracks at the loci of representative genes. (A–C) General adipocyte marker genes. (B) Brown adipocyte-specific marker. Abbreviations: ATAC-seq = assay for transposase-accessible chromatin with high-throughput sequencing; eWAT = epididymal white adipose tissue; iWAT = inguinal white adipose tissue; BAT = brown adipose tissue. Please click here to view a larger version of this figure.
Table 1: Components of the Tn5 master mixture. Please click here to download this Table.
Table 2: Components of the PCR master mixture and initial PCR cycling conditions. Please click here to download this Table.
Table 3: List of the barcoded primers for PCR amplification. Please click here to download this Table.
Table 4: Components of the qPCR master mixture. Please click here to download this Table.
Table 5: qPCR cycling conditions. Please click here to download this Table.
Table 6: Second PCR cycling conditions for additional amplification. Please click here to download this Table.
Table 7: Primer sequences and targeted qPCR cycling conditions for the quality check of the ATAC-seq library. Please click here to download this Table.
Table 8: Analysis of the qPCR results for the quality check. Please click here to download this Table.
In this paper, we have presented an optimized ATAC-seq protocol to assess adipocyte-specific chromatin accessibility in vivo. This ATAC-seq protocol using the Adipoq-NuTRAP mouse successfully generated adipocyte-specific chromatin accessibility profiles. The most critical factor for successful and reproducible ATAC-seq experiments is nucleus quality. It is critical to immediately snap-freeze the dissected adipose tissues in liquid nitrogen and store them safely at −80 °C without thawing until use. It is also important to prevent adipocyte nucleus damage during nucleus isolation and sorting. The samples must be handled gently via low-speed centrifugation and minimal pipetting during the entire process of nuclei isolation.
Careful gating during nuclei sorting is crucial to selectively isolate the mCherry/GFP-tagged adipocyte nuclei (Figure 1). It is important to ensure a clear separation between the mCherry/GFP-positive and negative populations. Likewise, gentle handling and rapid sorting are necessary to preserve the nature of the chromatin in the nuclei. If the fluorescence is weak, sorting takes too long, and/or the adipocyte fractions are out of the anticipated ranges, this indicates that there may be issues with the samples or problems during the nuclei isolation procedures. This protocol can be modified depending on the experimental circumstances. First, the number of nuclei required for the ATAC-seq can go down to 1,000 or up to 100,000. If there are less than 5,000 nuclei, using a further reduced volume (0.6 µL) of the TDE I enzyme would prevent over-tagmentation. For numbers of nuclei equal to or higher than 50,000, the volume of the Tn5 master mixture can be doubled to 50 µL, which is the same volume as used in the original protocol10.
Second, other nucleus labeling systems, such as INTACT (isolation of nuclei tagged in specific cell types) mice, can be used instead of NuTRAP mice16. As INTACT mice express GFP-tagged SUN1 nuclear membrane protein, GFP-tagged nuclei can be isolated by sorting and used for ATAC-seq using the same methods described in this protocol. Any other nuclear labeling methods can be adopted as well. Lastly, in vitro-cultured adipocytes can also be used for ATAC-seq following this protocol. For this, cells grown on dishes/plates are scraped, collected using NPB, and then subjected to the same downstream procedures-except that sorting should be done only based on size, complexity, and Hoechst, not mCherry/GFP as described previously15.
This adipocyte-specific ATAC-seq protocol using the NuTRAP technique poses a limitation as it does not work without fluorescent labeling; thus, analyzing human samples will not work with this technique. However, it is still possible to perform ATAC-seq with adipocytes isolated by tissue digestion and flotation methods. It should be noted that working with human floater adipocytes has some limitations, such as contamination by other cell types when using low-speed centrifugation and the loss of large adipocytes when using higher-speed centrifugation. This protocol can be applied to any cell type for which specific Cre lines are available, allowing efficient cell type-specific chromatin profiling even with limited cell inputs17. In conclusion, this improved ATAC-seq protocol will be helpful to researchers wishing to evaluate chromatin accessibility from specific cell types within heterogeneous tissues ex vivo, in addition to the adipocytes we discussed here.
The authors have nothing to disclose.
This work was supported by the IUSM Showalter Research Trust Fund (to H.C.R.), an IUSM Center for Diabetes and Metabolic Diseases Pilot and Feasibility grant (to H.C.R.), the National Institute of Diabetes and Digestive and Kidney Diseases (R01DK129289 to H.C.R.), and the American Diabetes Association Junior Faculty Award (7-21-JDF-056 to H.C.R.).
Animals | |||
Adiponectin-Cre mouse | The Jackson Laboratory | 28020 | |
NuTRAP mouse | The Jackson Laboratory | 29899 | |
Reagents & Materials | |||
1.5 mL DNA-LoBind tubes | Eppendorf | 86-923 | |
100 µm cell strainer | Falcon | 352-360 | |
15 mL tubes | VWR | 525-1071 | |
2x TD buffer | Illumina | 15027866 | |
384-well PCR plate | Applied biosystem | 4483285 | |
40 µm cell strainer | Falcon | 352-340 | |
50 mL tubes | VWR | 525-1077 | |
AMPure XP reagent (SPRI beads) | Beckman Coulter | A63881 | |
Bioanalyzer High Sensitivity DNA kit | Agilent Technologies | 5067-4626 | |
Clear adhesive film | Applied biosystem | 4306311 | |
DNase/RNase-free distilled water | Invitrogen | 10977015 | |
Dounce tissue grinder | DWK Life Sciences | 357542 | |
DTT | Sigma | D9779 | |
DynaMag-96 side skirted magnet | Thermo Fishers | 12027 | |
FACS tubes | Falcon | 28719128 | |
HEPES | Boston BioProducts | BBH-75 | |
Hoechst 33342 | Invitrogen | 2134015 | |
KCl (2 M) | Boston BioProducts | MT-252 | |
Magnetic separation rack for PCR 8-tube strips | EpiCypher | 10-0008 | |
MgCl2 (1 M) | Boston BioProducts | MT-200 | |
MinElute PCR purification kit | Qiagen | 28004 | |
NEBNext High-Fidelity 2x PCR master mix | BioLabs | M0541S | |
NP40 | Thermo Fishers | 28324 | |
PCR 8-tube strip | USA scientific | 1402-4708 | |
Protease inhibitor cocktail (100x) | Thermo Fishers | 78439 | |
Qubit dsDNA HS assay kit | Invitrogen | Q32851 | |
Sucrose | Sigma | S0389-1KG | |
SYBR Green I (10,000x) | Invitrogen | S7563 | |
TDE I enzyme | Illumina | 15027865 | |
Instruments | |||
Flow cytometer | BD Biosciences | FACSAria Fusion | |
Qubit fluorometer | Invitrogen | Q33226 | |
Real-Time PCR system | Thermo Fishers | QuantStudio 5 |