We haven't tested the max no. of cells (eg. 4M cells per reaction, like in the ChIP assay). However, it's safe to use 250K cells per reaction, without scaling up beads, antibodies, and buffers.

We have tested dozens of cultured cell lines. We so far haven't noticed any difference of the assay efficiency among different cell lines.

Adherent cells need to be detached from the plate using Trypsin (Acctuse also does the job but the cell pellet is not as sticky so cells are easier to lose during the cell wash steps). Suspension cells are easier, just count and spin the cells. We didn't notice any difference in the assay efficiency between adherent and suspension cells.

We haven't done CUT&RUN with cells manipulated by shRNA yet. However, shRNA expressing cells are very stable (it's supposed to have been screened by antibiotics for days/weeks). CUT&RUN starts with native cells, it takes about 5-6 hrs from cell collection to DNA digestion. I don't think shRNA will lose its function within 5-6 hours. If you do have the concern, you may try fixing the cells (although I don't think it's necessary for your case).

Our kits our compatible with the protocol in this paper which describes a method to pre-treat plant tissues and get the nuclei ready.  After completing these steps, our kit can be used (no digitonin is required but it may help help permeabilize the cell membrane if the plant cells are not 100% homogenous.

Paper : https://link.springer.com/article/10.1007/s00497-018-00358-1  /  https://link.springer.com/article/10.1007/s00497-018-00358-1

In the paper referenced here, the Henikoff lab completed CUT&RUN on brain tissue using an automated protocol where the tissue was snap frozen and the it was thawed in CUT&RUN wash buffer to further break up clumps: https://epigeneticsandchromatin.biomedcentral.com/articles/10.1186/s13072-018-0243-8                         

We haven't tested our protocol with tissues yet. However, like the ChIP assay with tissue materials, the first thing that needs to be done before starting the experiment (no matter ChIP or CUT&RUN) is to prepare single cell suspension from the tissue sample, because digitonin may not work well on tissue chunks to permeabilize cellular membrane. No matter what procedues are used to isloate cells form a specific tissue material, this process may take time. Therefore, it might not be a bad idea to fix the tissue in advance to avoid significant changes of the protein-DNA binidng profile during the cell suspension preparation.

We haven't tested CUT&RUN on tissue sample or at a single cell level yet. However, for single cell/ultra-low input CUT&RUN on tissue samples, the below paper may provide a reference. https://doi.org/10.1016/j.cell.2019.03.014

It is possible to make all the buffers and purchase the additional necessary reagents on your own.  However, in talking to many homebrew customers, we have found that most do not want to make their own pAG-Mnase or the Spike-in DNA, which is why we offer those reagents as a separate bundle, #40366.  We also know that digitonin activity changes from lot-to-lot so homebrew customers might not want to validate every new lot of digitonin, which is why we will be releasing this as a stand-alone product, where we do the necessary validation work to ensure the activity remains consistent from lot-to-lot.  This product is scheduled to release in July along withe Concanavalin A Beads and Activation Buffer, Protease Inhibitor Cocktail (200X) #7012, RNAse A 7013, and Proteinase K #10012. 

No. We have tested pAG-Mnase amount titration and time course for digestion. It turns out that both conditions are very flexible. It means no significant changes to the results are seen when the amount of enzyme or digestion time changes in a relatively wide range. Therefore we don't suggest optimizing the protocol by working on adjusting pAG-MNase activities. I'd rather check if the antibody is CUT&RUN validated because Ab quality matters a lot to the success of the assay.

Isolating the nuclei first is unnecessary. The ConA beads bind to glycoproteins in membranes to attach cells or nuclei, as far as I know, there are more glycoproteins on cellular membranes than on the nuclear membranes. Also because CUT&RUN works on native cells, we hope to avoid unnecessary operations on cells in case the molecular events in the nuclei change upon the nuclei isolation. The nuclei isolation is only required for special materials (such as yeast or plant cells which have cell walls). Digitonin doesn't work well on cell walls so the enzyme and antibody cannot enter the nuclei if cells are not broken up in this scenario. I suggest fixing the cells first if nuclei isolation is performed. Otherwise, the protein-DNA interaction profile you see is not the real one at the native cell status. 

No, serum starvation is only needed when this treatment can induce your target protein binding with genomic DNA. 

MNase bias regarding chromatin status in not really a concern. Our SimpleChIP kits utilize MNase to digest/fragment the chromatin during the chromatin prep.  We have never seen any bias for the MNase toward euchromatin vs. heterochromatin in our ChIP or ChIP-seq experiments. Compared to MNase, the Tn5 used in CUT&Tag has more bias to open euchromatin regions, although it probably shows less bias when tethered to a target protein bound antibody. Similar in CUT&RUN, MNase access to DNA is limited by the antibody localization on DNA. We also emphasize that the CUT&RUN digestion should be performed at 0C (not even 4C), so MNase activity is very much limited compared to that at 37C (where MNase prefers to digest naked DNA with a high activity). 

We have done intensive valiation of our CUT&RUN protocol for proteins that are weakly bound to chromatin (cofactors), including Rpb1, SMARCC1, Brg1, RING1B, Bmi1, SUZ12, and EZH2, etc. Follow our protocol and choose a CUT&RUN validated antibody, you will get a good CUT&RUN result on cofactors.

See above. Although we don't belive cell fixation is needed for cofactors, you may try if you really want to. However, CUT&RUN protocol might need to be optimized for the fixed cells as our protocol is only valiated for living cells.

The question is incomplete. That said, CUT&RUN assay works for drug treated cells. Spike-in DNA can also be used to normalize the artifact that is induced by PTM removal inhibitors.

Based on the fact that CUT&RUN-NGS generates similar target protein binding pattern across genome as the ChIP-seq does, we believe Mnase is digesting the antibody binding loci but not random hypersensitivity sites. 

   

Normalization for CUT&RUN happens when the digestion is done. Yeast genomic DNA fragemnts are spiked in the CUT&RUN sample before DNA purification. In ChIP, both spike-in chromatin and spike-in antibody are added to the IP reacation. To compare them, ChIP spkie-in normalization occurs earlier in the experiment than CUT&RUN spike-in normalziation does, but the ChIP one is also more complicated than CUT&RUN one.

We suggest following our protocol #86652 for usage of spike-in DNA in qPCR and NGS. Please note that qPCR and NGS needs very different amount of spike-in DNA for detection. If both qPCR and NGS are required, please prepare two samples in parellel to satisfy different requirements of spike-in DNA amount.

We don’t recommend size selection during library preparation, no matter for CUT&RUN or ChIP-seq. Based our own experience and many customers' feedback, size selection significantly decreases both diversity and yield of the library. If you really concerned with large DNA fragments (>1kb),  decrease the extension time when amplifying the adaptor-ligated DNA sample during library prep. We see that 10-15s is enough for amplifying CUT&RUN library DNA, and no large DNA fragments (>1kb) have a chance to be amplified within such a short extension time.

We recommend the input sample as negative control for NGS, although both Input sample and IgG sample are acceptable for publication purpose. Input sample saves more reagents than the IgG antibody control does and it has a higher diverstiy. Customers' feedbacks show that IgG may produce the high background issue (at some gene loci of interest).

We use our SimpleChIP ChIP-seq DNA Library Prept Kit for Illumina (56795) and SimpleChIP ChIP-seq Multiplex Oligos for Illumina (29580, 47538)

Use shorter primers 60-80 bp instead of 100bp, but we have gotten it to work with primers that are 100bps

Yes. Our CUT&RUN-qPCR result is pretty compatible with the ChIP-qPCR result. However, to catch as much signal as possible, we suggest designing primer set with smaller amplicon size (ideally 60-80bp, ~100bp is also fine). The reason is that CUT&RUN DNA is smaller than ChIP DNA (size of TF/cofactor CUT&RUN DNA peaks around 80bp while size of histone CUT&RUN DNA peaks around 150bp).

We provide a positive and negative control antibody in the kit, H3K4me3 #9751 and Rabbit IgG Isotype Control #66362, both of which can be purchased as stand-alones.  The kit also provides both Human and Mouse positive Control primers, #7014, and #7015 which are also sold separately. The Sample Normalization Primers are only sold as part of the kit, and the amount of spike-in DNA that is provided in the bundle is only sufficient for NGS.  Henikoff indicates that he felt the DNA from CUT&RUN would be too small for qPCR, but we have successfully performed qPCR on all three protein types (histones, TFs, cofactors).  The qPCR is performed the exact same way as DNA from a ChIP experiment, with the only difference being that we recommend using primers that are 60-80 bps in length since CUT&RUN DNA is smaller than ChIP DNA.  That being said we have successfully performed qPCR with CUT&RUN DNA using primers that are 100bps in length.  

Use a negative gene locus as the control. In detail, design a primer set against the gene locus which is known not to be bound by the target protein. 

What is the best way to choose an antibody? Especially if there are none which have previously been used for ChIP-seq/CUT&RUN? Is it important for antibody to work for IP? (3)

You can try a ChIP validated antibody, but we have found that ChIP validation is not a guarantee that it will work for CUT&RUN.  Henikoff has indicated that IF might be a better indicator, but we have not validated this idea internally.

We haven't. But we have heard positive resutls from customers (eg. flag).

See Slide on CUT&RUN/Tag Comparison

Réactifs : 01 34 60 60 24 - tech@ozyme.fr

Instrumentation : 01 30 85 92 88 - instrum@ozyme.fr