Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Oct 20.
Published in final edited form as: Anal Chem. 2020 Oct 1;92(20):13661–13666. doi: 10.1021/acs.analchem.0c02550

Multiplexed and ultralow-input ChIP-seq enabled by tagmentation-based indexing and facile microfluidics

Chengyu Deng 1, Travis W Murphy 1, Qiang Zhang 1, Lynette B Naler 1, Alice Xu 2, Chang Lu 1,*
PMCID: PMC7578044  NIHMSID: NIHMS1631497  PMID: 32957776

Abstract

Epigenome constitutes an important layer that regulates gene expression and dynamics during development and diseases. Extensive efforts have been made to develop epigenome profiling =methods using a low number of cells and with high throughput. Chromatin immunoprecipitation (ChIP) is the most important approach for profiling genome-wide epigenetic changes such as histone modifications. In this report, we demonstrate microfluidic ChIPmentation (mu-CM), a microfluidic technology that enables profiling cell samples that individually do not generate enough ChIP DNA for sequencing library preparation. We used a simple microfluidic device to allow 8 samples to be processed simultaneously. The samples were indexed differently using a tagmentation-based approach (ChIPmentation) and then merged for library preparation. Histone modification profile for each individual sample was obtained by demultiplexing the sequencing reads based on the indexes. Our technology allowed profiling 20 cells and is well suited for cell-type-specific studies using low-abundance tissues.

Graphical Abstract

graphic file with name nihms-1631497-f0004.jpg

Introduction

Molecular biology within cells is not only affected by DNA sequences (genomics) but also by epigenomic regulations. Epigenomic regulatory mechanisms include DNA methylation, histone modification, higher-order chromatin organizations and regulations by non-coding RNAs1. These mechanisms do not directly change DNA sequence but strongly affect gene expression and cell state. Histone modifications, or post-translational modifications of histones, play pivotal roles in activating and inhibiting transcription2. Histone modifications define the “open/closed” state of chromatin and provide binding domains for transcription factors3. Deregulation of histone modifications has been associated with various diseases, including cancer4. Hence specific histone marks involved in disease development have become new targets for drug development and these studies have yielded promising clinical results5.

Chromatin immunoprecipitation followed by sequencing (ChIP-seq) is the gold standard technique to profile genome-wide histone modifications6. Major limitations of ChIP-seq include the large number of cells required (~106 cells) and time-consuming and complex process (~3 days). Numerous efforts have been made in the past years to reduce sample size to hundreds to tens of cells using ChIP or ChIP-free methods (iChIP7, MOWChIP-seq8–10, μChIP-seq11, STAR ChIP-seq12,13, CUT&RUN14–18, TCL19, SurfaceChIP-seq20, LIFE-ChIP-seq21). More recently, there have been several reports on single-cell methods that reveal epigenetic heterogeneity with 1,000 to 17,000 unique reads per cell (Drop-ChIP22, scChIC-seq23, scCUT&Tag24, uliCUT&RUN25, scChIP-seq26, ACT-seq27, sc-itChIP-seq28, coBATCH29). Although single-cell technologies are extremely informative for deconvoluting cellular heterogeneity, their data quality is generally compromised due to the lossy single-cell barcoding process. For example, a low-input technology like SurfaceChIP-seq produced 3 million unique reads on H3K4me3 using 30 cells20. This is 100 times more unique reads than pooling 30 single-cell data sets when the single-cell technology yields only 1000 unique reads per cell22. The low limit of the existing low-input technologies is often determined by the amount of ChIP DNA required for library preparation. For example, in our previous SurfaceChIP-seq work, the sequencing library preparation kit required a minimum of 10 pg DNA and our ChIP step collected about 10-15% of the genomic DNA. This means that we needed to start with at least 30 cells (containing roughly 5-6 pg genomic DNA per cells) in order to prepare the sequencing library. In this work, we aim to further lower the number of cells required for each sample by conducting multiplexing and combining multiple samples, taking advantage of tagmentation-based indexing and a simple microfluidic system.

In this work, we use a simple microfluidic device to facilitate simultaneous processing of multiple cell samples. The microfluidic ChIP process is combined with a tagmentation-based step (i.e. ChIPmentation30) to index each sample with unique sequencing adaptors. The produced ChIP DNA from various samples is subsequently pooled and used to produce one sequencing library. The sequencing data can be demultiplexed based on the indexes to produce ChIP-seq data sets on each sample. Because the ChIP DNA for library preparation was pooled from multiple samples, our approach allows each sample to contain less cells than the amount dictated by the library preparation requirement. Our method, referred to as microfluidic ChIPmentation or mu-CM, permits processing 8 assays in one run with as few as 20 cells per assay and the entire protocol can be finished in 7 h. Our technology provides new opportunities for the study of histone modifications in low-abundance samples in the context of precision medicine.

Experimental Section

Fabrication of the microfluidic device

Microfluidic device was fabricated by soft lithography using poly(dimethylsiloxane) (PDMS) (RTV615, Momentive). Design was drawn up using LayoutEditor (juspertor GmbH) and laser-plotted on photomasks (10,000 dpi). Features in the photomask were molded onto a silicon wafer (University Wafers) by photolithography. SU-8 2025 photoresist (MicroChem) was spun at 500 rmp for 10 s and then at 1500 rmp for 20 s to achieve a thickness of 60 μm. PDMS was produced by mixing 40 g of A and 4 g of B and poured onto the mold in a petri dish. PDMS was degassed using a vacuum pump at 60 mTorr for 1 h, and baked at 80 °C for 1 h to cure. Cured PDMS was peeled off from the mold and inlet/outlet holes were punched. The PDMS was finally bonded to a clean glass slide after air plasma treatment (PDC-32G, Harrick Plasma) of both surfaces, and baked at 80 °C for 1 h to achieve strong bonding.

Setup of the microfluidic device

The common outlet of the microfluidic chamber was connected with an infusion/withdrawal syringe pump (OEM Pump Half Case Module, Chemyx Inc) via perfluoroalkoxy alkane (PFA) tubing (1622L, ID: 0.02 in. and OD: 0.0625 in., IDEX Health & Science). A LabVIEW (LabVIEW 2019, National Instruments) program was used to change the flow rate and control the on/off of the syringe pump. Small-volume (< 3 μl) solutions, including chromatin and tn5, were directly added into inlet reservoirs. Large-volume solutions, including bead suspension, tagmentation buffer and washing buffer, were loaded into pipette tips attached to inlet reservoirs (as shown in Fig. 1c). Solutions were slowly drawn into all the chambers simultaneously by withdrawal of the syringe pump at 1.5 μl/min for small-volume solutions and 10 μl/min for large-volume solutions at the outlet. A cold pack and a hot plate (TC-124, Warner) were utilized to maintain the temperature of 4°C and 37°C for certain steps, respectively.

Figure 1 |. Overview of microfluidic ChIPmentation device and operation.

Figure 1 |

Open in a new tab

a. Schematic of the microfluidic device. The device contains 8 reaction chambers. Each chamber has an individual inlet, which is connected to a common outlet through a series of splits. Each chamber has supporting pillars to prevent collapse. b. Microscopic image of the microfluidic device. The stitched image is created by OLYMPUS CellSens. c. The device operation. Two major steps are shown: (i) Packing of magnetic IP beads. Pipette tips containing magnetic bead suspensions are inserted into the inlet reservoirs. A long magnet is placed under the device to retain the beads. Magnetic beads are drawn into the chambers under magnetic force and form a fluidized bead bed; (ii) Solutions are loaded into the chambers by withdrawal using a syringe pump at the outlet. Solutions slowly flow through the bead bed. d. Flow chart of the steps and corresponding molecular biology.

Cell culture

GM12878 was obtained from Coriell Institute for Medical Research. Cells were cultured in RPMI 1640 media (30-2001, ATCC) supplemented with 15% FBS (16000044, Life Technologies Corporation), 10,000 U/mL Penicillin-Streptomycin (15140122, Life Technologies Corporation) in an incubator maintaining 5% CO2 and 37 °C. Mycoplasma contamination was tested every 6 months using Universal Mycoplasma Detection Kit (30-1012K, ATCC).

Cell lysis and MNase digestion

Cells (20-1000) were suspended in 1 μl PBS (14190136, Life Technologies Corporation) and mixed with 1 μl lysis buffer [4% Triton X-100, 100 mM Tris (pH 7.5), 100 mM NaCl, and 30 mM MgCl2] containing freshly added 1% PIC (P8340-5ML, Sigma-Aldrich) and 1 mM PMSF (P7626-1G, Sigma-Aldrich). After 10-min incubation at room temperature, 0.2 μl of 500 mM CaCl2 and 0.2 μl of 100U/μl MNase (88216, Life Technologies Corporation) were added followed by another 10 min incubation. The reaction was stopped by adding 0.222 μl of 0.5M EDTA and incubated on ice for 10 min. Chromatin solution was stored on ice before ChIP assay.

Preparation of magnetic beads

372 μg Dynabeads Protein A (10001D, Life Technologies Corporation) (contained in 12.4 μl of bead suspension) were washed twice with 200 μl IP buffer (20 mM Tris-HCl, pH 8.0, 140 mM NaCl, 1 mM EDTA, 0.5 mM EGTA, 0.1% (w/v) sodium deoxycholate, 0.1% SDS, 1% (v/v) Triton X-100, 0.1% (v/v) Tween 20). Beads were suspended in 600 μl of IP buffer and mixed with 2 μg of antibody (H3K4me3: ab8580, Abcam; H3K4me1: 39498, Active Motif) at 4 °C on a rotator mixer at 24 r.p.m. for 1 h. After coating, beads were washed twice with 200 μl IP buffer and suspended in 120 μl IP buffer. These functionalized IP beads were used in one run using the 8-unit device.

Assembly of indexed transposome complexes

We followed a previous protocol for assembly of indexed transposome complexes31. Briefly, to prepare the T5/T7 transposon stock solution, 200 μl single-strand T5 or T7 transposon (IDT, standard desalting; Supplementary Table S1) at a concentration of 100 μM was annealed with 200 μl of 100 μM pMENTS, a 5’-phosphorylated 19-bp mosaic end complementary oligonucleotide, in the annealing buffer (10 mM Tris-HCl, 1 mM EDTA, 25 mM NaCl, pH 8.0). 95 °C for 5 min followed by a slow ramp to 25 °C at −0.1°C/sec was used for the annealing. T5 and T7 transposons have sequences that are complementary to P5 and P7 oligos in the Illumina sequencing platform, respectively, while T7 transposon contains an 8-bp index that is unique for each of 8 units/assays conducted in our mu-CM platform (Supplementary Fig. S2 and Table S1). 2 μl of EZ-Tn5 (TNP92110, Lucigen) was diluted with 8 μl of dilution buffer (50% glycerol, 50 mM Tris, pH 7.5, 100 mM NaCl, 0.1 mM EDTA, 1 mM DTT, 0.1% NP-40) to 600 nM. 5 μl of the diluted tn5 was mixed with 5 μl of 600 nM annealed T5 or T7 transposon (created from the stock solutions by dilution) and incubated at 37 °C for 1 h. Assembled T5 and T7 transposomes (10 μl each) were mixed finally to form 20 μl dimers and were stored at −20 °C till use. 8 μl was used in each experiment using the 8-unit device (with 1 μl added to each unit).

Microfluidic ChIPmentation

The microfluidic chambers were primed with IP buffer to prevent adsorption of protein on PDMS. A long permanent magnet (BZ082, 3” × 1/2” × 1/8”, K&J Magnetics) was attached to the glass side of the device using double-sided tape and antibody-coated IP beads were then loaded into the chambers under magnetic force and formed a fluidized bead bed. 2.5 μl of Mnase-digested chromatin was directly added into the inlet reservoirs. The chromatin solution flowed through the bead beds in the chambers by withdrawal of the syringe pump at a flow rate of 1.5 μl/min at the outlet. The ChIP step took around 40 min till the reservoirs were essentially dry. Washing buffer (20 mM Tris-HCl, pH 8.0, 150 mM NaCl, 2 mM EDTA, 0.1% SDS, 1% (v/v) Triton X-100, 0.1% (v/v) Tween 20) was flowed through the bead beds at a withdrawal flow rate of 10 μl/min for 5 min.

After washing, 10 mM Tris and Tagmentation buffer (10mM Tris, pH 8.0, 5 mM MgCl2, 35% (v/v) dimethylformamide) were then sequentially loaded for conditioning. 1 μl fully assembled indexed tn5 was added directly to the inlet reservoir and flowed through beads bed at an outlet withdrawal rate of 1.5 μl/min for 20 min. The device was put on a hot plate (TC-124, Warner) to reach 37 °C. RIPA LS buffer (10 mM Tris, pH 8.0, 140 mM NaCl, 1mM EDTA, 0.1% SDS, 0.1% DOC, 1% Triton x-100) was loaded at 10 μl/min outlet withdrawal rate for 5 min to stop tagmentation and the device was placed on a cold pack. Finally, the beads were flowed out of the 8 chambers under a flow rate of 100 μl/min and collected into one Eppendorf tube.

Purification of ChIP DNA

Beads were suspended in a mixture of 198 μl Elution buffer (10mM Tris, pH 8.0, 5 mM EDTA, 300 mM NaCl, 0.4% SDS) and 2 μl of 20 mg/mL Proteinase K (P2308-100MG, Sigma-Aldrich), and incubated at 65 °C for 1 h. DNA was purified by phenol extraction and ethanol precipitation, and eluted in IDTE buffer (10 mM Tris, pH 8.0, 0.1 mM EDTA).

PCR Amplification

20 μl ChIP DNA was mixed with 1.5 μl of P5 and P7 PCR primers (each at 25 μM) (IDT, standard desalting; Supplementary Table S1), 25 μl activated polymerase (KK2601, Kapa Biosystems) (Pre-heated at 98 °C for 30 sec), and 2.5 μl Evagreen (31000-T, Biotium). The qPCR was performed with the following program: 72 °C for 5 min; 98 °C for 30 s; n cycles of 98 °C for 10 s, 63 °C for 30 s; and 72 °C for 30 s; and a final extension at 72 °C for 1 min30. The amplification was stopped once the increase in the relative fluorescence units (RFUs) reached ~500. The library was purified using 0.85× SPRIbeads (B23317, Beckman Coulter) and eluted in 7 μl IDTE buffer.

Data analysis

Sequencing reads were trimmed by Trim Galore and aligned to hg19 genome using Bowtie232 with default setting. Duplicates, low-quality mapped reads, and reads in blacklist regions were removed before calling peaks with MACS233 using the following parameters: --nomodel --extsize 100 -q 0.0001. Pearson correlation was calculated using deepTools34 with default parameters.

Results and Discussion

We designed a very simple multiplexed microfluidic device made of PDMS/glass slide that has no valving system. Our device consists of 8 bell-shaped reaction chambers (each with a volume of ~ 180 nl) with each having an individual inlet and connected to a common outlet through a series of splits (Fig. 1). The operation only required an infusion/withdrawal syringe pump for liquid manipulation at the common outlet and avoided complex pneumatic control and valving system. The operation associated with each individual unit involved several steps: (i) Cells were lysed and chromatin was fragmentized with MNase in a tube with each tube containing 2.5 μl at the end of the step; (ii) A pipette tip containing 15 μl of antibody-coated magnetic beads was inserted into the inlet reservoir. The beads were pulled into the microfluidic chamber under the magnetic force generated by a long permanent magnet adhered to the glass slide side of the device using double-sided tape (Fig. 1c i). After the beads formed a fluidized bed in the chamber21, the pipette tip was removed from the inlet and the bulk of the solution was pipetted out with roughly 2 μl solution left in the reservoir; (iii) 2.5 μl chromatin solution was added into each inlet reservoir. The chromatin solution was pulled through the bead bed by a syringe pump connected to the common outlet that withdrew at a flow rate of 1.5 μl/min for 20-40 min until the reservoirs were mostly empty. In this step, ChIP occurred during which the chromatin fragments that contained targeted histone modification bound to its antibody on the IP bead surface. The small volume of the chamber and the small diffusion length within the bead bed facilitate efficient immunoprecipitation of target chromatin fragments; (iv) Washing was then conducted by plugging a pipette tip containing 150 μl washing buffer into the inlet reservoir and withdrawing solutions from the common outlet at a flow rate of 10 μl/min for 5 min (Fig. 1c ii). The washing removed nonspecific binding on the bead surface. (v) Next, the pipette tip containing the leftover washing buffer was then replaced by one containing 150 μl tagmentation buffer and the withdrawing was performed again (10 μl/min for 5 min). The pipette tip was then removed together with the remaining buffer inside. (vi) 1 μl of indexed tn5 transposase complexes (unique for each unit and sample) was loaded into the reservoir and then the entire device (including the thin magnet) was carefully placed on a hot plate that was set at 37 °C. The transposase complex solution was pulled through the bead bed by a syringe pump connected to the common outlet that withdrew at a flow rate of 1.5 μl/min until the reservoirs were mostly empty. The process took roughly 10-20 min (the tagmentation requires at least 10 min). During this step, Tn5 transposase complex cut and tagged the chromatin fragments on the bead surface with adapters that contained an index that was unique for an individual unit (Supplementary Table S1). After this step, the ChIPed chromatin fragments in each chamber/unit were tagged by unique indexes. (vii) After tagmentation, the magnet was removed and the magnetic beads were flushed out of the chambers and pooled together for DNA purification and PCR amplification to prepare one library from the pooled samples. The library was sequenced at a depth of 15 million total reads with single-ended 50 bp.

Using our method, we profiled two histone marks (H3K4me3 and H3K4me1) with various number of GM12878 cells (1000 cells per unit for H3K4me1, 100/40/20 cells per unit for H3K4me3) (Supplementary Table S2). Our mu-CM data show high correlation among units, with an average Pearson’s correlation of 0.90 for H3K4me1 with 1,000 cells per unit, 0.89, 0.67 for H3K4me3 with 100 and 40 cells per unit, respectively (Fig. 2a and Fig. 2b). The average Pearson’s correlation with ENCODE data sets were 0.75 and 0.70 for H3K4me3 with 100 and 40 cells, and 0.82 for H3K4me1 with 1000 cells, respectively (Fig. 2a and Fig. 2b). These correlations were generally lower than what we obtained using other low-input methods that did not involve ChIPmentation. For example, the Pearson’s correlation between MOWChIP-seq 100 cell data set and the ENCODE data was 0.82. All data showed high fraction of reads in peak (FRiP) (in the range of 8% to 28%) that is comparable to that of ENCODE data (Supplementary Fig. S1).

Figure 2 |. Mu-CM data using 1000, 100, and 40 GM12878 cells per unit.

Figure 2 |

Open in a new tab

a. Normalized H3K4me3 and H3K4me1 signals at house-keeping gene GAPDH using data generated with various sample sizes. ENCODE H3K4me3 data set (GSM733708) and H3K4me1 data set (GSM733772) are used for comparison. b. Pearson’s correlations calculated using normalized signal in promoter regions among our ChIP-seq data sets (1k H3K4me1, 100/40 cells H3K4me3) and ENCODE data. Promoter regions are defined as 2kb upstream and downstream from the transcription start sites.

Our indexing using ChIPmentation provided an approach to examine a very low number of cells. In our previous works, we needed at least 30-100 cells to start with in order to generate enough ChIP DNA to prepare a sequencing library8,20. With the tagmentation-based indexing, we produced enough ChIP DNA from 8 samples of 20 cells each that were indexed differently. By demultiplexing the sequencing data based on the indexes, we were able to produce ChIP-seq data using as few as 20 cells (Fig. 3). Among the 8 samples processed in one run in our device, only sample 8 presented fairly low quality (having an average correlation of 0.57 with the other 7 samples, Fig. 3b). The other 7 data sets were strongly correlated (average correlation ~ 0.75). For the 7 good data sets, each has an average of 3,886 peaks, and 500,000 unique reads. The decrease in the peak number and the correlation among data sets was likely due to the fact that 20-cell samples lost even larger fraction of their chromatin fragments to surface adsorption than samples containing higher cell numbers. It is also worth noting that there is a small probability for the assay to fail (1/8 in the case of Fig. 3a) with 20 cells per sample. Nevertheless, the results show that our mu-CM technology allows probing cell samples that do not provide enough ChIP DNA for library preparation by itself.

Figure 3 |. Mu-CM H3K4me3 data using 20 cells per unit.

Figure 3 |

Open in a new tab

a. Normalized H3K4me3 signals at house-keeping gene GAPDH. 8 replicates of 20 cell data were also merged together and labeled as ‘H3K4me3-merge’. Merged data was in excellent agreement with ENCODE data. b. Pearson correlations in promoter regions among the data sets and ENCODE data.

Conclusions

Here we demonstrate that mu-CM is an ultralow-input microfluidic platform for profiling histone modifications using as few as 20 cells. Our simple and multiplexed device requires minimal ancillary control system, permitting easy implementation in most biology labs. It greatly saves labor and can be performed by novices with little experience on microfluidics. Moreover, integrating ChIPmentation-based indexing enables pooling of ChIP DNA from multiple samples for library preparation and dramatically reduces the needed input cell number for each sample. Taken together, mu-CM is suitable for studies based on animal models or patients that generate low quantity of tissues.

Supplementary Material

Supporting information

Acknowledgements

This work was supported by US National Institutes of Health (NIH) grant R33 CA214176 (C.L.), P30 CA012197 (C.L.), and a seed grant from Virginia Tech Institute for Critical Technology and Applied Science (C.L.).

Footnotes

Supporting Information

Supplementary figures and tables: Average fraction of reads in peak (FRiP) calculated using featureCounts; Design of T5/T7 transposons and PCR primers; Transposon and primer sequences; Summary of mu-CM data.

Data Availability

The ChIP-seq data sets are deposited in the Gene Expression Omnibus (GEO) repository with the following accession number GSE152233; https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE152233

The authors declare no competing financial interest.

References

  • (1).Deng C; Naler LB; Lu C Microfluidic epigenomic mapping technologies for precision medicine. Lab Chip 2019, 19, 2630–2650. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (2).Lawrence M; Daujat S; Schneider R Lateral Thinking: How Histone Modifications Regulate Gene Expression. Trends Genet 2016, 32, 42–56. [DOI] [PubMed] [Google Scholar]
  • (3).Hyun K; Jeon J; Park K; Kim J Writing, erasing and reading histone lysine methylations. Exp Mol Med 2017, 49, e324. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (4).Feinberg AP The Key Role of Epigenetics in Human Disease Prevention and Mitigation. New Engl J Med 2018, 378, 1323–1334. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (5).Jones PA; Issa JPJ; Baylin S Targeting the cancer epigenome for therapy. Nat Rev Genet 2016, 17, 630–641. [DOI] [PubMed] [Google Scholar]
  • (6).Landt SG; Marinov GK; Kundaje A; Kheradpour P; Pauli F; Batzoglou S; Bernstein BE; Bickel P; Brown JB; Cayting P; Chen YW; DeSalvo G; Epstein C; Fisher-Aylor KI; Euskirchen G; Gerstein M; Gertz J; Hartemink AJ; Hoffman MM; Iyer VR, et al. ChIP-seq guidelines and practices of the ENCODE and modENCODE consortia. Genome Research 2012, 22, 1813–1831. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (7).Lara-Astiaso D; Weiner A; Lorenzo-Vivas E; Zaretsky I; Jaitin DA; David E; Keren-Shaul H; Mildner A; Winter D; Jung S; Friedman N; Amit I Chromatin state dynamics during blood formation. Science 2014, 345, 943–949. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (8).Cao ZN; Chen CY; He B; Tan K; Lu C A microfluidic device for epigenomic profiling using 100 cells. Nat Methods 2015, 12, 959–962. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (9).Zhu BH; Hsieh YP; Murphy TW; Zhang Q; Naler LB; Lu C MOWChIP-seq for low-input and multiplexed profiling of genome-wide histone modifications. Nat Protoc 2019, 14, 3366–3394. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (10).Cox MC; Deng CY; Naler LB; Lu C; Verbridge SS Effects of Culture Condition on Epigenomic Profiles of Brain Tumor Cells. Acs Biomater Sci Eng 2019, 5, 1544–1552. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (11).Dahl JA; Jung I; Aanes H; Greggains GD; Manaf A; Lerdrup M; Li GQ; Kuan S; Li B; Lee AY; Preissl S; Jermstad I; Haugen MH; Suganthan R; Bjoras M; Hansen K; Dalen KT; Fedorcsak P; Ren B; Klungland A Broad histone H3K4me3 domains in mouse oocytes modulate maternal-to-zygotic transition. Nature 2016, 537, 548–552. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (12).Zhang BJ; Zheng H; Huang B; Li WZ; Xiang YL; Peng X; Ming J; Wu XT; Zhang Y; Xu QH; Liu WQ; Kou XC; Zhao YH; He WT; Li C; Chen B; Li YY; Wang QJ; Ma J; Yin QZ, et al. Allelic reprogramming of the histone modification H3K4me3 in early mammalian development. Nature 2016, 537, 553–557. [DOI] [PubMed] [Google Scholar]
  • (13).Zheng H; Huang B; Zhang BJ; Xiang YL; Du ZH; Xu QH; Li YY; Wang QJ; Ma J; Peng X; Xu F; Xie W Resetting Epigenetic Memory by Reprogramming of Histone Modifications in Mammals. Mol Cell 2016, 63, 1066–1079. [DOI] [PubMed] [Google Scholar]
  • (14).Skene PJ; Henikoff S An efficient targeted nuclease strategy for high-resolution mapping of DNA binding sites. Elife 2017, 6, e21856. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (15).Skene PJ; Henikoff JG; Henikoff S Targeted in situ genome-wide profiling with high efficiency for low cell numbers. Nat Protoc 2018, 13, 1006–1019. [DOI] [PubMed] [Google Scholar]
  • (16).Janssens DH; Wu SJ; Sarthy JF; Meers MP; Myers CH; Olson JM; Ahmad K; Henikoff S Automated in situ chromatin profiling efficiently resolves cell types and gene regulatory programs. Epigenet Chromatin 2018, 11, 1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (17).Meers MP; Bryson TD; Henikoff JG; Henikoff S Improved CUT&RUN chromatin profiling tools. Elife 2019, 8, e46314. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (18).Schmid M; Durussel T; Laemmli UK ChIC and ChEC: genomic mapping of chromatin proteins. Mol Cell 2004, 16, 147–157. [DOI] [PubMed] [Google Scholar]
  • (19).Zarnegar MA; Reinitz F; Newman AM; Clarke MF Targeted chromatin ligation, a robust epigenetic profiling technique for small cell numbers. Nucleic Acids Res 2017, 45, e153–e153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (20).Ma S; Hsieh YP; Ma J; Lu C Low-input and multiplexed microfluidic assay reveals epigenomic variation across cerebellum and prefrontal cortex. Sci Adv 2018, 4, eaar8187. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (21).Murphy TW; Hsieh YP; Ma S; Zhu Y; Lu C Microfluidic Low-Input Fluidized-Bed Enabled ChIP-seq Device for Automated and Parallel Analysis of Histone Modifications. Anal Chem 2018, 90, 7666–7674. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (22).Rotem A; Ram O; Shoresh N; Sperling RA; Goren A; Weitz DA; Bernstein BE Single-cell ChIP-seq reveals cell subpopulations defined by chromatin state. Nat Biotechnol 2015, 33, 1165–1172. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (23).Ku WL; Nakamura K; Gao WW; Cui KR; Hu GQ; Tang QS; Ni B; Zhao KJ Single-cell chromatin immunocleavage sequencing (scChIC-seq) to profile histone modification. Nat Methods 2019, 16, 323–325. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (24).Kaya-Okur HS; Wu SJ; Codomo CA; Pledger ES; Bryson TD; Henikoff JG; Ahmad K; Henikoff S CUT&Tag for efficient epigenomic profiling of small samples and single cells. Nat Commun 2019, 10, 1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (25).Hainer SJ; Boskovic A; McCannell KN; Rando OJ; Fazzio TG Profiling of Pluripotency Factors in Single Cells and Early Embryos. Cell 2019, 177, 1319–1329. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (26).Grosselin K; Durand A; Marsolier J; Poitou A; Marangoni E; Nemati F; Dahmani A; Lameiras S; Reyal F; Frenoy O; Pousse Y; Reichen M; Woolfe A; Brenan C; Griffiths AD; Vallot C; Gerard A High-throughput single-cell ChIP-seq identifies heterogeneity of chromatin states in breast cancer. Nat Genet 2019, 51, 1060–1066. [DOI] [PubMed] [Google Scholar]
  • (27).Carter B; Ku WL; Kang JY; Hu GQ; Perrie J; Tang QS; Zhao KJ Mapping histone modifications in low cell number and single cells using antibody-guided chromatin tagmentation (ACT-seq). Nat Commun 2019, 10, 1–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (28).Ai SS; Xiong HQ; Li C; Luo YJ; Shi Q; Liu YX; Yu XH; Li C; He AB Profiling chromatin states using single-cell itChIP-seq. Nat Cell Biol 2019, 21, 1164–1172. [DOI] [PubMed] [Google Scholar]
  • (29).Wang QH; Xiong HQ; Ai SS; Yu XH; Liu YX; Zhang JJ; He AB CoBATCH for High-Throughput Single-Cell Epigenomic Profiling. Mol Cell 2019, 76, 206–216. [DOI] [PubMed] [Google Scholar]
  • (30).Schmidl C; Rendeiro AF; Sheffield NC; Bock C ChIPmentation: fast, robust, low-input ChIP-seq for histones and transcription factors. Nat Methods 2015, 12, 963–965. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (31).Amini S; Pushkarev D; Christiansen L; Kostem E; Royce T; Turk C; Pignatelli N; Adey A; Kitzman JO; Vijayan K; Ronaghi M; Shendure J; Gunderson KL; Steemers FJ Haplotype-resolved whole-genome sequencing by contiguity-preserving transposition and combinatorial indexing. Nat Genet 2014, 46, 1343–1349. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (32).Langmead B; Salzberg SL Fast gapped-read alignment with Bowtie 2. Nat Methods 2012, 9, 357–359. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (33).Zhang Y; Liu T; Meyer CA; Eeckhoute J; Johnson DS; Bernstein BE; Nussbaum C; Myers RM; Brown M; Li W; Liu XS Model-based Analysis of ChIP-Seq (MACS). Genome Biol 2008, 9, 1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (34).Ramirez F; Ryan DP; Gruning B; Bhardwaj V; Kilpert F; Richter AS; Heyne S; Dundar F; Manke T deepTools2: a next generation web server for deep-sequencing data analysis. Nucleic Acids Res 2016, 44, W160–W165. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting information
NIHMS1631497-supplement-Supporting_information.pdf (932.9KB, pdf)

RESOURCES