My problem is I got lots of signal in each sample, include IgG antibody pull down. I am using mouse brain samples so not sure if you have enough cells for pull down. Did you increase crosslink time to see any changes? Did you run gel shift before ChIP?

Vladimir, thanks for your advise. But I use different antibodies for experimental proteins, and for a (+)control-antibody, and nowhere I have a signal. Of course, may be proteins of interest don't present at this region... (+)control antibodies are, ChIP grade, as I said, and they are good in immunofluorescense.

Dear Xiuli,

I used 3-10 million cells, and 10 million cells - is a good amount of cells, as described in several protocols.

My crosslink time varied between 5-11 minutes, and no difference was observed.

Hi Evgeniy,

you might have crosslinked too long (the epitope is not anymore accessible), or too short (you lose the proteins during the chromatin prep). I would suggest doing some time course using only the positive control and one negative control to find the conditions where the ChIP works.  You can also vary the crosslinking temperature (e.g. room temperature versus 4°C) or the formaldehyde concentration. I usually had good results with TFs in mammalian cells using 1% and 30-45 minutes crosslinking at 4°C.

In addition, your IP-buffer or the one you use for washing might be too stringent. Also here, you could test other buffers (with less salt or detergents) using just positive and negative controls.

Finally, your input does not look that great. How diluted is that? I would expect such a band if you dilute the input 1:1000 or even higher.

Vladimir, thanks, I'll try, however, I think, that some signal (may be not very big) must be present with this antibodies. So, as for me, it seems to be a some simple step, when I loose the sample or a signal.

Hello Achim,

Thanks a lot for your detailed answer! I'll try 4°C for crosslinking, but what timepoints can I choose in this step except 30min?

As for buffers - I used 2 protocols with different standart buffers, although, may be my antibodies are too sensetive to some detergents.

I took 10% of chromatin for input, and I took 10% of input sample per PCR. So, it is 1% of DNA at PAGE for inputs.

Hi Evgeniy,

as you crosslinked 5-11 minutes so far (at room temperature, I guess), it could be, that this was too short. You could try 20 minutes at room temperature and 30, 45 and 60 minutes at 4°C. Depends, how much samples you want to process.

I usually use RIPA buffer (10 mM Tris-HCl (pH 8), 1 mM EDTA, 0.5 mM EGTA, 1% Triton X-100, 0.1% Na-Deoxycholate, 0.1% SDS, 140 mM NaCl plus proteinase inhibitors) for IP and washes. To get it less stringent, one can leave out the SDS, or reduce the Triton or the NaCl (100 mM).

 If this is 1/100 of the input, I would expect a fat band after ca. 30 cycles of PCR.

Dear Achim,

Ok, thanks, I'll try this timepoints. I work with mouse embrionic stem cells. There are some successful articles, where cells were cross-linked for 10-15 minutes. So, I think, that some signal I should receive.

Last time I used this buffers:

Lysis buff.1

0.25%Triton X-100, 10mM EDTA, 10mM Tris-HCl (8.0), PMSF+PIC

Lysis buff.2

200mM NaCl, 10mM EDTA, 10mM Tris-HCl (8.0), PMSF+PIC

After this two buffers I receive nuclei, and use IP buffer for sonication and IP:

200mM NaCl, 0.5% Triton X-100, 0.5% NP40, 0.05%DOC, 20mM Tris-HCl (8.0), PMSF+PIC

So, they are not very stringent - they don't contain SDS, but concentration of NaCl is 200mM

I didn't said that 1% for input is a 1% of DNA that I use per one ChIP with one antibody.

By the way, may be it will be more gently for proteins and chromatin, if I'll choose restriction digest for specific regions near my region of interest against sonication?

Hi Evgeniy,

I do not think that sonication is the problem. It is very unlikely, that you destroy your epitope like that. Or, of course it depends how much you sonicate. If you sonicate too long and the sample heats up, you decrosslink the chromatin, and you loose indeed your proteins. You should always have ice cold water in the Bioruptor. We routinely do 3 or 4 cycles (depends on the cell type) of 5 minutes (30´´ on, 30´´ pause) in 200 microl of buffer and replace the water after each cycle with fresh ice cold one.

Your buffers look ok, at least not too stringent. By the way, another interpretation of your result could be, that the buffers are not stringent enough, as you always get an amplification product (also with the non-specific antibodies). Maybe with RIPA the bands in the negative control would disappear?

Hello Achim,

I use standart sonication according to Sigma ChIP protocol -  3 cycles for 10 minutes with High power (30s ON/ 30s OFF) with replace ice after each cycle in 100ul of sample. So, may be problem is at this step in time of cycle? But I can't say that temperature of water was rather hot (appx. 35 degrees in some cases).

I don't think that this bands are a real signal - there were 34 cycles of PCR, so it seems to be a background (besides, they are in case of negative control primers).

Hi Evgeniy,

I do not think that the sonification was the problem. Usually it takes hours to achieve efficient decrosslinking just by heating.

In any case, there are several parameters you can adjust in your ChIP: time of crosslinking, crosslinking temperature, the buffers you use and maybe also the sonification conditions. Sometimes it also helps to add glass beads to the sonification, to reduce cycles and run time (they will help chopping the chromatin). You just need to play around a bit, to find the conditions that your antibodies will work.

Good luck!

May be problem associated with a small amount of antibodies? Everytime I used only 1ug of abs.

That should be enough. But of course, one can use more. It depends on the quality of the antibodies. And on the volume of chromatine you use for IP. I usually do 300 microl. IPs and use the antibody (1 microg/microl) 1:100, i.a. 3 microl. If the antibody is very good, also less. In rare cases, I use more. Also here, you need to try different conditions, to get the optimum yield.

Hello Achim,

Thanks for Your reply.

First time I used 15ug of chromatin per ChIP (appx. 10mln of cells) in volume of 400-600ul + 1ug of abs.

Second time I used appx. 5ug of chromatin per ChIP (3 mln of cells) in volume of 900ul + 1ug of abs.

So, I think it's very possible that there are too little amount of abs...

Dear collegues, tell me, please, what ratio of cell number and antibody amount do you use in Your ChIPs?

For example, last time in positive control, I used 3mln of cells with 1ug of abs in volume 1ml.

May be my problem related with this ratio or with this big volume?

Hi Evgeniy,

that sounds indeed like you used too little antibody in a too big volume. We use usually 2-3 million of mouse ESCs per IP, in 300 microl. with ca. 3 microg of antibody (or even more for some antibodies), as written above.

Hello Achim!

Thanks a lot for Your advice. It really seems that there were too little amount of abs in big volume... I want to do dot-blot with chromatin before IP and after IP with different amount of abs. 

Could You tell me, please, what buffer and volume do You use for Lysis and sonication? Is it FA-buffer like in abcam protocol? http://www.abcam.com/ps/pdf/protocols/x_chip_protocol.pdf

And one more question - what kind of TF did You check for Your IP with 2-3 million of mouse ESCs? Is it rare or abundant TFs?

Thanks in advance!

Hi Evgeniy,

we do a two step lysis, first lysing the cells, then the nuclei. The nuclear lysis buffer is also the sonication buffer. The buffers are: Cell lysis buffer (5 mM PIPES (pH 8), 85 mM KCl, 0.5 % NP40, proteinase inhibitors) and Nuclear lysis buffer (50 mM Tris-HCl (pH 8), 10 mM EDTA, 0.8 % SDS, protease inhibitors). After sonication the chromatin is diluted to obtain RIPA buffer. The protocol can be found here: http://cshprotocols.cshlp.org/content/2006/4/pdb.prot4560.abstract

Click full text of the new Carey version on top. The volumes depend on the number of cells you start with. Using a bioruptor, one should take up the final nuclear pellet in not more than 1 ml (when starting with 50-100 million cells). If you start with less cells, you can also use 250 microl here.

But my students now like to use one of the Kits available, e.g. the Chip-IT kit by Active Motif. To get better IPs though, we often change buffer for the IP back to RIPA.

We do and did ChIPs for CTCF, Polycomb proteins, PolII, Estrogen receptor alpha, Gata factors, DNMTs and of course various histone modifications. These are all more abundant.

Hello Achim,

Thank You so much for the protocol.

I have some questions about it:

1) Do You use a Dounce homogenizer when incubate cells with cell and nuclei lysis buffers?

2) There is an 11 step with chromatin measuring. If I understood correctly - it means concentration of DNA after reverse cross-linking?

3) Also, if I understood correctly - You dilute a nuclei pellet of 50-100mln of cells in 1ml of Nuclei Lysis Buffer. Than You take a small aliquot of this after sonication, do reverse cross-linking, measure a concentration of DNA. Than You take some volume with 100ug of chromatin and dilute it to 300ul with RIPA buffer?

Thanks in advance!

Hi Evgeniy!

1) No, no Dounce homogeniser is needed. The cells are anyway fixed, thus the solution will not get viscous. Just resuspend the pellets in the respective buffers and leave them on ice.

2) This can also be measured without de-crosslinking, which will give you a rough estimate of the chromatin concentration. This can also be useful to compare different batches of chromatin, when you know that sonication works and you did set all conditions, in order to use the same amount of chromatin. Nevertheless, it is more accurate to do this with a de-crosslinked aliquot, which is more time consuming, of course.

3) Yes, this is basically what we do. Or, after sonication, the sample is diluted (to get the SDS concentration down). We found it important for efficient fragmentation that SDS is between 0.8-1% during sonication. But this is of course too high for doing IPs. Therefore a dilution step is necessary (or a buffer exchange using e.g. Amicon filters). I did chromatin also with much less cells, down to 1 million. You just need to downscale the volumes. You can also just work with cell numbers (without measuring ODs - if you can measure the cell number before fixation; with attached cells that are fixed in the plate this is difficult, of course) and use a certain amount of your final diluted chromatin for IP that corresponds to a certain cell number (e.g. 1 million). The volume should not be too big for the IP (not more than 500 microl). Thus it is good to have the cells for sonication in a small volume (less than 1 ml), as you nee to dilute 8-fold.

Hello Achim,

Thanks a lot again for detailed answer.

I'd like to clarify - 1) What volume, appx. do You take from 1ml of nuclei lysate after sonication to check a sonication quality?

2) You said that use PIPES (8.0) for cell lysis buffer. Is it critical to use exactly PIPES? Also it's said that effective buffering range of this buffer is 6.1-7.5.

3) You said earlier that use RIPA buffer for washes. Do You use just RIPA, or something else for washing, like in abcam protocol with low salt and high salt washing buffers? 

RIPA buffer

50 mM Tris-HCl pH8

150 mM NaCl

2 mM EDTA pH8

1% NP-40

0.5% Sodium Deoxycholate

0.1% SDS

Protease Inhibitors (add fresh each time)

Wash buffer

0.1% SDS

1% Triton X-100

2 mM EDTA pH8

150 mM NaCl

20 mM Tris-HCl pH8

Final wash buffer

0.1% SDS

1% Triton X-100

2 mM EDTA pH8

500 mM NaCl

20 mM Tris-HCl pH8

Also in Your protocol it's said that washes are with high-salt buffer and TE.

4) And the last one - Do You use 1%SDS and 100mM NaHCO3 for elution?

Also now I have just a "Next Gen Seq Sepharose" from Sigma. May be somebody worked with this Sepharose, and know work specifics with this kind of Sepharose?

Hi Evegniy

1) 50 microl should be enough.

2) I have no idea how essential PIPES is. This comes from the ChIP-paper we adapted the protocol from (http://mcb.asm.org/content/19/12/8393.long). It works and I never changed that.

3) I always used RIPA for the washes. This protocol was adapted from our older protocol. I did not notice that they use other buffers. But of course you can also try different washing conditions using different salt concentrations, if you still have background problems. In one of our first protocols, there was also a final washing step with a LiCl-buffer (0.25M LiCl, 0.5% NP 40, 0.5% Na-Deoxycholate, 1mM Na-EDTA, 10mM Tris-HCl (pH 8)). The final wash with TE is done to remove salt and detergents for the decrosslinking with heat and proteinase K.

4) I never use a special elution buffer. After the wash with TE, the beads are resuspended in a small volume of TE and then crosslinks are reversed (with heat or heat plus proteinase K). The I directly did chloroform-phenol extraction and precipitated the DNA. In these days people mostly use spin columns for that.

Dear Achim,

So, do You wash beads after IP with RIPA 3-4 times and after this do You wash once with TE?

Hi Evgeniy,

here is the protocol of the IP part as I do it:

1. Use ca. 100 μg of chromatin for one immunoprecipitation. For abundant proteins like histones also less chromatin can be used (30-50 μg). Dilute the appropriate amount of chromatin in 300-500 μl of RIPA buffer.

2. Add 25 μl of Protein A/G agarose beads using a cut pipette tip (wide aperture). Incubate for 1-2 hours at 4 °C for pre-clearing and centrifuge in a microfuge at 12000-14000xg for 10 minutes at 4°C.

3. Transfer the supernatant to a new tube and add the appropriate amount of antibody, usually 1 μg (or more) of affinity purified antibody (dilutions of 1:100 - 1:500). Use the same amount of pre-cleared chromatin in the controls, without the addition of antibody (for mock- and input-control), pre-immune serum or an appropriate non-specific antibody. Incubate the samples from 2-3 hours to overnight at 4°C on a rotator or rocker.

4. Centrifuge the samples in a benchtop centrifuge at 12000-14000xg for 10 minutes at 4°C. Transfer the IPs to new tubes.

5. Add 25 μl of the 50% Protein A/G agarose beads and incubate for a further 2-4 hours.

6. Pellet the beads by centrifugation (500xg for 2 minutes) in a benchtop centrifuge. Transfer the supernatant of the no-antibody control to a new tube and leave on ice. This material can serve as total input control if you are short of manterial. Discard the other super-natants.

7. Wash the beads five times with 600 μl of RIPA buffer, once with 600 μl LiCl-buffer (optional) and once with 600 μl TE (pH 8.0), collecting the beads between washes with brief centrifugations. Finally, resuspend the beads in 100 μl of TE.

8. Incubate samples overnight at 65°C.

9. The next day, adjust samples to 0.5% SDS and 0.5 mg/ml Proteinase K and incubate for a further 3 hours at 50°C.

10. Add one volume of phenol-chloroform-isoamylalcohol, vortex for 2 minutes and centrifuge at 12000-14000xg for 8 minutes. Back-extract the phenol phase by adding an equal volume of TE (pH 8.0) and vortex. Combine the aque-ous phases and add one volume of chloroform-isoamylalcohol, vortex for 2 minutes and centrifuge at 12000-14000xg for 8 minutes.

11. Precipitate the DNA by adding GlycoBlue to 100 μg/ml as carrier, 1/10 volume of 3M sodium acetate pH 5.2 and 2.5 volumes of 100% ethanol. Incubate at -20°C for two hours to overnight.

12. Collect the DNA by centrifugation at 12000-14000xg for 15 minutes at 4°C, and wash the pellet in 800 μl of 70% ethanol. Repeat centrifugation and discard the supernatant.

13. Allow the pellet to air dry for 5 minutes. Redissolve the precipitated DNA in ca. 50 μl of TE and store at 4°C (to avoid DNA precipitation, do not freeze).

By the way, step 4 (the centrifugation step before adding the beads) can be very important, in particular, if you always get an amplification (also in the negative controls). Sometimes DNA can precipitate from the chromatin during ON incubations and you will also spin this down during the washes. If you spin before adding the beads and transfer the super to a new tube, you get rid of this!

Dear Achim,

Thank You very very much!

I'll write here my result.

Dear collegues,

Now I have one big problem - antibodies don't bind with proteins in ChIP.

I used 3 different ChIP-grade abs for ChIP, so I don't think that problem is related with abs. My IP-buffer: 50mM Tris-HCl, 200mM NaCl, 1% NP-40 or 1% Triton, 0,25% DOC, protease inhibitors. I dilute chromatin probe 8-fold with this buffer, do a ChIP, wash sepharose twice with IP-buffer, elute abs and proteins and check supernatant and elution on western blot for proteins of interest, and I have all signal in supernatant.

By the way, my sepharose bind antibodies very well, becouse most  signal of abs present in elution. So problem is related exactly in antigen-antibody binding, and I think it's becouse of IP-buffer composition.

May be somebody knows the answer?

Thanks everyone in advance!