Researchers new to the epigenetic field will find CUT&Tag a significantly easier alternative to ChIP assays. CUT&Tag has tremendously benefited the epigenetic studies on rare and primary cell populations, generating high-quality data from very few cells. This protocol describes performing H3K4me1 CUT&Tag assays on mouse myoblasts isolated from mouse hindlimb muscles.
This protocol paper aims to provide the new researchers with the full details of using Cleavage Under Targets and Tagmentation (CUT&Tag) to profile the genomic locations of chromatin binding factors, histone marks, and histone variants. CUT&Tag protocols function very well with mouse myoblasts and freshly isolated muscle stem cells (MuSCs). They can easily be applied to many other cell types as long as the cells can be immobilized by Concanavalin-A beads. Compared to CUT&Tag, chromatin immunoprecipitation (ChIP) assays are time-consuming experiments. ChIP assays require the pre-treatment of chromatin before the chromatic material can be used for immunoprecipitation. In cross-linking ChIP (X-ChIP), pre-treatment of chromatin involves cross-linking and sonication to fragment the chromatin. In the case of native ChIP (N-ChIP), the fragmented chromatins are normally achieved by Micrococcal nuclease (MNase) digestion. Both sonication and MNase digestion introduce some bias to the ChIP experiments. CUT&Tag assays can be finished within fewer steps and require much fewer cells compared to ChIPs but provide more unbiased information on transcription factors or histone marks at various genomic locations. CUT&Tag can function with as few as 5,000 cells. Due to its higher sensitivity and lower background signal than ChIPs, researchers can expect to obtain reliable peak data from merely several millions of reads after sequencing.
CUT&Tag assay was invented to compensate for some overt flaws of ChIPs1. The two major disadvantages of ChIPs are 1) the bias introduced when fragmenting chromatin and 2) the incompetence to work with low cell numbers. X-ChIP assays rely on either sonication or MNase digestion to get chromatin fragments, whereas N-ChIP mostly uses MNase digestion to get nucleosomes. Sonication shows a bias towards relaxed chromatin locations such as promoter regions2, and apparently, MNase digestion also works more efficiently on relaxed chromatin fibers. Moreover, some reported that MNase digestion also shows a DNA sequence-dependent bias3. Therefore, at the input preparation step of ChIP assays, it is impossible to get chromatin fragments from all kinds of genomic locations in a perfectly random manner. Moreover, ChIP assays normally generate higher background signals compared to CUT&Tag and require over 10 folds more reads than CUT&Tag to accentuate where the peaks are1,4,5. This explains why ChIP experiments have to start with tremendously more cells than CUT&Tag. This is not a problem when studying cell lines as they can be repetitively passaged to achieve a very high cell number. However, the ChIP assay is definitely not a strong epigenetic tool to study rare or precious primary cell populations, although primary cells obviously hold more practical and medical implications.
While the long and complicated ChIP procedure discourages some researchers from learning or using this technique, people are more comfortable with easier assays such as immunocytochemistry (ICC) or immunofluorescence (IF). A CUT&Tag assay essentially resembles the process of ICC and IF experiments but only takes place in a test tube. CUT&Tag does not need fragmented chromatin to start with, and instead, the genome must be intact for antibody binding1. On the first day of a ChIP experiment, researchers normally spend up to 4 h preparing the fragmented chromatins from nuclei with sonication or MNase digestion before the short chromatin pieces can be mixed with antibody-beads4,5. In remarkable contrast, the first-day workload of a CUT&Tag procedure is to just immobilize the cells to Concanavalin-A beads and then to add the primary antibody onto the cell-beads. This only requires ~40 min1.
It is worth mentioning that Cleavage Under Targets and Release Using Nuclease (CUT&RUN) is an important alternative to CUT&Tag. CUT&RUN was established based on a similar working mechanism as CUT&Tag. In CUT&Tag, the antibodies guide the pA/G-Tn5 transposase to all the locations where the enzyme will each cut out a piece of chromatin and meanwhile tag it with library-making adaptors, while in CUT&RUN, the role of pA/G-Tn5 is played by the pA/G-MNase which only performs the cutting part of the job6. Therefore, compared to CUT&Tag, CUT&RUN requires an additional step which is to glue the library-making adaptors onto the pA/G-MNase-fragmented DNA pieces7,8. Due to the high similarities between CUT&Tag and CUT&RUN, researchers familiar with CUT&Tag will easily adapt themselves to performing CUT&RUN proficiently. However, it should be noted that some minor differences exist between CUT&Tag and CUT&RUN. CUT&RUN protocols normally use the physical concentration of salt (~150 mM) in the washing steps, while in CUT&Tag, the 300-Dig wash buffer is high in salt. Therefore, CUT&Tag is good at controlling the background when profiling histone marks/variants or transcription factors, as these proteins directly and strongly bind DNA1. CUT&Tag may run into problems when profiling chromatin-associated factors that do not directly bind DNA and show weak affinity to the chromatin. High salt washing steps in CUT&Tag may strip off chromatin-associated factors and cause no signals in the final output. Although there are successful cases where CUT&Tag can be used to profile some non-histone/non-transcription factor proteins9, we still recommend CUT&RUN over CUT&Tag to profile weakly-bound chromatin-associated proteins.
After mammals reach adulthood, their skeletal muscle tissues still contain muscle stem cells. During muscle injury, these stem cells can be activated and undergo cell number expansion and differentiation to regenerate damaged muscle fibers10. These stem cells are known as Muscle stem/satellite cells (MuSCs). After MuSCs are isolated from animals or once activated by muscle injury, they start proliferating and become myoblasts.
To obtain MuSCs from mouse skeletal muscle digest, MuSC surface markers such as Vcam1 (Cd106), Cd34, and α7-integrin (Itga7) are often used individually or in combinations to enrich MuSCs during fluorescence-activated cell sorting (FACS)11. It has been shown that Cd31–/Cd45–/Sca1–/Vcam1+ is probably the best marker combination to get >95% pure MuSCs12. FACS can isolate pure MuSCs right after fresh muscles are digested. However, if the experimental design does not require pure MuSCs right at their isolation, pre-plating is more cost-effective than FACS to obtain >90% pure myoblasts (MuSC progeny).
MuSCs freshly isolated from mice do not proliferate efficiently in Ham's F10 media supplemented with fetal bovine serum (FBS). To better expand the cell number of MuSCs and get sufficient myoblasts, bovine growth serum (BGS) should be used instead of FBS. However, if BGS is not available, Ham's F10 full media (containing about 20% FBS) can be mixed with an equal volume of T-cell conditional media to dramatically promote myoblast expansion13. Therefore, this protocol will also describe the preparation of T-cell media conditioned MuSC media.
Most importantly, this protocol provides a complete example of performing H3K4me1 CUT&Tag assays on mouse myoblasts isolated from mouse hindlimb muscles. Please note that this protocol also applies to other cell types and histone marks and histone variants, and readers only need to optimize the cell numbers or antibody amounts for their cases based on the enrichment of the specific histone marks or variants they study.
In order to be used in CUT&Tag or CUT&RUN, the Tn5 or MNase needs to be fused with protein A or protein G to make pA-Tn5, pG-Tn5, pA-MNase, or pG-MNase. Apparently, both protein A and protein G can be fused onto these enzymes at the same time to generate pA/G-Tn5 or pA/G-MNase. Protein A and protein G show differential affinities to IgGs from different species. Therefore, fusing protein A and protein G altogether onto the enzyme can overcome this problem and make the enzyme compatible with antibodies from multiple species.
The methods presented in this manuscript are all approved by the Institutional Animal Care and Use Committee of Guangzhou Laboratory. Mice used to generate this manuscript's representative results were housed and maintained in accordance with the guidelines of the Institutional Animal Care and Use Committee of Guangzhou Laboratory.
1. Myoblast isolation from mouse hindlimb muscles (Example of using 1 mouse)
2. Preparation of MuSC/Myoblast growth media (Example of preparing from 5 mice)
3. CUT&Tag with myoblasts (Example of 3 CUT&Tag reactions)
Before binding cells to Concanavalin-A beads, check the cell suspension under the microscope. Accordingly, after incubating the cells with Concanavalin-A beads, put the sample tubes on the magnetic rack, and the supernatant should be again observed using a microscope. This is to assess how efficiently the cells have been captured by Concanavalin-A beads. Wash buffer containing 7 x 105 cells/mL should look like Figure 1A under the microscope. In contrast, Figure 1B shows the supernatant after Concanavalin-A bead binding, where the beads have taken away all the cells. There is no point in proceeding into the following steps of CUT&Tag if the Concanavalin-A beads did not even capture cells efficiently. Wash buffer should not contain any serum content or EDTA/EGTA, as they can abolish the binding of cell membrane to Concanavalin-A beads. Bovine serum albumin, however, does not interfere with Concanavalin-A binding to cells.
After making the CUT&Tag libraries by PCR, a small aliquot of the library (3-4 µL) can be electrophoresed on a 1.2-1.5% agarose gel to test whether the library is good. Figure 2 shows the example of successful H3K4me1 CUT&Tag libraries we made. In a histone mark or histone variant CUT&Tag, the majority of cleaved units are apparently mono-nucleosomes, although two nucleosomes can also be cut out together. Three nucleosomes cut and tagged altogether would be very rare. Therefore, a nucleosome array containing mono-nucleosomes, di-nucleosomes, and very few tri-nucleosomes should be observed in the electrophoresis of histone mark CUT&Tag libraries. Nucleosomes contain 147 bp of DNA, but with the sequencing adaptors added by Tn5 transposase and library PCR, the total size of a CUT&Tagged mono-nucleosome should come by 300 bp, as shown in Figure 2.
Before spending money on high-throughput sequencing of the CUT&Tag libraries, another quality control method is to perform qPCR with locus-specific primers. This is equivalent to performing ChIP-qPCR. Moreover, the locus-specific primers in CUT&Tag qPCR are designed based on the same principles as ChIP-qPCR primers. Figure 3 shows the qPCR amplification of different locations on the Myod1 gene locus. Myod1 encodes the master transcription factor, MyoD, which determines the skeletal muscle lineage during myogenesis15. As shown in Figure 3, H3K4me1 is highly enriched at the core enhancer (CER) and distal regulatory region (DRR) of Myod1 locus. CER and DRR are very well-established enhancer regions for Myod116. This is consistent with the well-documented phenotype that H3K4me1 specifically marks commissioned enhancers17. Moreover, Figure 3 shows that the H3K4me1 level is very low at the proximal regulatory region (PRR) because the nucleosomes at the proximal region of active promoters are normally modified with H3K4me318 instead of H3K4me1. All the primer pairs for these assessed loci are previously published19,20, and the sequences can be found in Table 4. Normally, after the quality controls by library DNA electrophoresis and qPCR on specific loci, the researchers will be confident to proceed with sequencing.
Figure 1: Myoblasts resuspended in Wash buffer. Microscopic views taken (A) before and (B) after Concanavalin-A bead binding. Panel B shows there are almost no cells left in the suspension after Concanavalin-A bead incubation, indicating cells have been efficiently captured by the beads. The small particles in panel B are likely to be the excessive Concanavalin-A beads. Please click here to view a larger version of this figure.
Figure 2: Electrophoresis of CUT&Tag libraries. Three independent H3K4me1 CUT&Tag assays have been performed, and each was made into a sequencing library using Illumina primers and PCR. Each H3K4me1 CUT&Tag used 70,000 mouse myoblasts. Libraries mostly contain tagged mono-nucleosomes. Please click here to view a larger version of this figure.
Figure 3: qPCR amplification of different locations on Myod1 gene locus. qPCR shows that in mouse myoblasts, the H3K4me1 mark is highly enriched at the enhancer regions of the Myod1 gene, which aligns with published conclusions. CER and DRR are published enhancers of Myod1 which are located upstream of Myod1 promoter. Please note that the PRR does not have H3K4me1 accumulation, and this is also consistent with the fact that active promoters are mainly marked with H3K4me3. The enrichments were calculated by normalizing to the enrichment of H3K4me1 on "gene desert" of sample "Repeat 1". Please click here to view a larger version of this figure.
Name | Sequences (5’-3’): |
Oligo-Rev | phos-CTGTCTCTTATACACATCT |
Oligo-A | TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG |
Oligo-B | GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAG |
Table 1: Oligo sequences for making Adaptor-A and Adaptor-B to be mounted onto pA/G-Tn5 transposase.
Fraction | Volume |
Purified genomic DNA after CUT&Tag reaction | 21 µL |
N5XX primer (10 µM) | 2 µL |
N7XX primer(10 µM) | 2 µL |
NEBNext High-Fidelity 2× PCR Master Mix | 25 µL |
Total 50 µL |
Table 2: PCR setup for CUT&Tag library preparation using Illumina indexing primers. If using a commercial CUT&Tag kit that contains library PCR reagents such as DNA polymerase mix, follow the specific manual of that kit to set up the library PCR.
Step | Temperature | Time | Cycle number |
1 | 72 oC | 3 min | 1 |
2 | 95 oC | 3 min | 1 |
3 | 98 oC | 10 s | 9 to 14 |
4 | 60 oC | 5 s, then go to step 3 | |
5 | 72 oC | 1 min | 1 |
6 | 4 oC | hold | N/A |
Table 3: PCR program for CUT&Tag library preparation using Illumina indexing primers.
Location | Primer sequences (5’-3’) |
CER | For: GGG CAT TTA TGG GTC TTC CT |
Rev: CTC ATG CCT GGT GTT TAG GG | |
-15,000 | For: TGC CCA GAG CCT AGA ATC AT |
Rev: TCA TGC ATC CTT GCT GGA TA | |
-7,000 | For: GGC ATG GGA GGT TTA TAG CA |
Rev: ATG CCA CTA TGC AAT CCA CA | |
DRR | For: TCA GGA CCA GGA CCA TGT CT |
Rev: CTG GAC CTG TGG CCT CTT AC | |
PRR | For: GAG TAG ACA CTG GAG AGG CTT GG |
Rev: GAA AGC AGT CGT GTC CTG GG | |
CDS1 | For: CAT CTG ACA CTG GAG TCG CTT TG |
Rev: CAA GCA ACA CTC CTT GTC ATC AC | |
CDS2 | For: GTG AGC CTT GCA CAC CTA AGC C |
Rev: GTT GCA CTA CAC AGC ATG CCT G | |
IgH enhancer | For: ACC CTG GGA AGA CCA TAC TTA ATC T |
Rev: CCA TCC ACA CTC GTG CCT TA | |
Gene desert | For: TCC TCC CCA TCT GTG TCA TC |
Rev: GGA TCC ATC ACC ATC AAT AAC C |
Table 4: Locus-specific primers for testing H3K4me1 enrichment at different locations on the Myod1 gene.
The specific cell number required in a certain CUT&Tag reaction completely relies on the enrichment of the histone marks/variants or chromatin-binding proteins that are to be tested. Normally for very enriched histone marks such as H3K4me1, H3K4me3, and H3K27ac etc., 25,000-50,000 myoblasts are quite sufficient for one CUT&Tag reaction. However, some rare chromatin-binding proteins might require up to 250,000 cells. The cell number used in CUT&Tag assays is critical, which, if not handled well normally causes the failure of the assay. Using an excessive number of cells in a CUT&Tag reaction is detrimental, especially when testing very enriched histone marks. Starting with too many cells means insufficient antibodies allocated for each cell. CUT&Tag not only works with very low cell numbers, it can actually be applied at single cell level1.
Above all, using a reliable and validated antibody with high affinity and specificity is always the key to a successful CUT&Tag assay. Same as ChIP, Immunoprecipitation (IP), and IF assays, CUT&Tag also requires the type of antibodies that can recognize the naturally folded antigens1. In contrast, the antibodies that only recognize relaxed peptides in Western blot should be strictly avoided in the CUT&Tag.
If optimizing the cell number or using an immunofluorescence-validated or well-cited antibody still cannot solve the poor signal problem in a CUT&Tag assay, then look into whether the protein of interest is a chromatin-associated protein that weakly binds or does not even bind to DNA directly. In this case, as mentioned before, switch to using CUT&RUN instead of CUT&Tag. Alternatively, slight fixation, such as 5 min in 0.5% paraformaldehyde (followed by 10 min quenching in 0.1 M glycine), might be a method worth being tested. However, as we have never practiced the fixation in any CUT&Tag assays, nor have we even run into the need to fix the cells before performing CUT&Tag, no further comments can be made by us regarding cross-linking.
We understand that some versions of CUT&Tag protocols might claim that both RT (2 h) and 4 °C (overnight) for the primary antibody incubation step can generate decent results21. We however strongly recommend using 4 °C (overnight) as opposed to RT for 2 h. On the other hand, extending the incubation time longer than one night is also not recommended.
Regarding sensitivity, background controlling, and degree of complexity for researchers to handle, CUT&Tag is superior to ChIP assays. If some drawbacks have to be put for CUT&Tag compared to ChIP, then the only one is that CUT&Tag protocols do not include an internal normalization method such as using "Input" in a ChIP assay1. Therefore, when comparing peak heights between samples using CUT&Tag, researchers must equalize the cell number among samples to start with. Also, it is critical to include technical replicates.
This protocol presents the full details of performing CUT&Tag in mouse myoblasts by providing a successful example of H3K4me1 CUT&Tag in this cell type. The method described here can be directly applied to freshly isolated MuSCs without any modifications. Moreover, CUT&Tag experiments on histone marks, histone variants, or transcription factors in other cell types can also be carried out following this protocol. For myotubes or myofibers, the nuclei can be isolated first, and CUT&Tag can then be carried out on the myonuclei. When doing this, one can treat the nuclei the same way as cells are treated in this CUT&Tag protocol. Myotubes and myofibers are multi-nucleated cells, and we have not tested whether myotubes or myofibers can be directly bound onto Concanavalin-A beads, although this might be feasible.
The authors have nothing to disclose.
This work was supported by the Strategic Priority Research Program of the Chinese Academy of Science (XDA16020400 to PH); the National Natural Science Foundation of China (32170804 to PH).
bFGF | R&D Systems | 233-FB-025 | |
Collagen | Corning | 354236 | |
Collagenase II | Worthington | LS004177 | |
Concanavalin-A | Sigma-Aldrich | C5275 | |
Concanavalin-A beads | Bangs Laboratories | BP531 | |
Digitonin | Sigma-Aldrich | 300410 | |
Dispase II | Thermo Fisher Scientific | 17105041 | |
Fetal bovine serum | Hyclone | SH30396.03 | |
H3K4me1 antibody | abcam | ab8895 | |
Ham's F10 media | Thermo Fisher Scientific | 11550043 | |
Hyperactive Universal CUT&Tag Assay Kit for Illumina | Vazyme | TD903 | This kit has been tested by us to function well |
Magnetic rack for 1.5 mL EP tubes | Qualityard | QYM06 | |
Magnetic rack for 8-PCR tube stripes | Anosun Magnetic | CLJ16/21-021 | |
NEBNext High-Fidelity 2x PCR Master Mix | NEB | M0541L | For library-making PCR reaction |
pA-Tn5 | Vazyme | S603-01 | Needs to be mounted with adaptors before use |
Protease inhibitor cocktail | Sigma-Aldrich | 5056489001 | |
Proteinase K | Beyotime | ST535-100mg | |
RPMI-1640 media | Thermo Fisher Scientific | C11875500BT | |
Secondary antibody (Guinea Pig anti-rabbit IgG) | Antibodies-online | ABIN101961 | |
Spermidine | Sigma-Aldrich | S2626 | |
TruePrep Index Kit V2 for Illumina | Vazyme | TD202 | This kit provide Illumina N7XX and N5XX primers |
VAHTS DNA Clean Beads | Vazyme | N411 | Can substitute Ampure XP beads. Can purify CUT&Tag libraries and select DNA fragments by size |