Hi everyone,
I have some trouble in my ChIP experiments. I want to study the recruitment of transcription factor and DNMTs on DNA. For this, I cross-link cell with 1% final of formaldehyde during 10 min. I stop reaction with glycin and I use the scrapping method to recover my cells and put the cells in a 1,5 ml tube. Next, for the lysis of the cells, I use a kit for ChIP (release by Ademtech, but I think, it's only available in France). After the final step of lysis, I sonicate my samples with the Bioruptor sonicator (two cylces of 10 min, 30 sec ON/30 sec OFF). I spin the samples 10 min at maximum speed. I save the supernatant and it's ready to use for my ChIP. After that, I use an automate which allow to do the different step of ChIP (Saturation of magnetics beads, antibody incubation, chromatin incubation, washing and elution step with proteinase K). For the antibody part, I use only ChIP-grade antibody (Abcam, active Motif or SantaCruz). For the negative control, I use GFP antibody or GST antibody or HA antibody (my cells are Glioblastoma cells line and they are not transfected with any construction). I use 3 µg of antibody for each ChIP. For the chromatin, I use 7 to 10 µg. After the elution/PK treatment step done, I decross-linked my samples at 65°C O/N in a shaking platform (300 rpm). Next, I purify my samples with "IP DNA purfication kit" of Active Motif and finally, I measure the DNA concentration of my IP with the Qubit HS sensitivity kit.
My problem is, the negative control (GFP, GST or HA) is always more DNA concentrated than in my IP antibody test samples. With the differents washing steps, no or very very few DNA/protein complex should be bind the negative control antibody and more should be bind with the IP antibody test.
I don't understand where is the problem, thank you for your help and sorry if I am not very clear with my explanations.
Do you know a positive control locus for your transcription factor of interest? Then you could design some qPCR primers and test for enrichment over background at specific regions. It could be that the negative control IPs are pulling down non-specific background DNA while your test samples are pulling down DNA from specific regions. It's fine to use the Qubit assay to quantitate your elutions, but you really need an assay that can better determine the specificity of your ChIP.
Also, what kind of beads are you using to pull down the antibody? Most people (including myself) find that Dynabeads give the least background.
I agree with Elisabeth, Dynabeads are good beads for IP. We tried the Ademtech one but we were not convinced by the efficiency as regard to Diagenode and Dynal ones. Don't juge the success of your experiment on the quantity of DNA IPed, your irrelevant antibody just have a high background rate, but it is still background. You can try IgG for negative control as well. Maybe you could try to increase the quantity of chromatine used to increase your signal. And finaly, you should use home made protocol as kits are developped for histone ChIP and usually give poor results with transcription factors.
Good luck
Thank you for your reply, i haven't tested the Dynabeads yet, but i will try. I have designed several qPCR primers for putative target sites for my transcription factor, but the result of my qPCR has a similar profile to the quantification with the Qubit of my DNA after IP. Indeed, for my negative control, CTs are less high than in my test samples. But for my DNMT1 ChIP (DNMT1 is more abundant in cells than a specific TF), the result is similar with a non-significant enrichement. But like you said, this is probably a problem with beads or with quantity of chromatin (even if 7 to 10 µg of chromatin is, for me, enough). I have just some difficulty to understand why the negative control antibody catch more DNA than with my transcription factor or DNMT antibody. The negative control antibody (GFP, GST or HA) recognize a specif sequence, like with my TF or DNMT antibody. So, the background should be pretty much similar. Maybe i believe overmuch in the efficiency of my antibody. Anyway, thanks a lot again for your reply.
Without knowing the specific brands of the antibodies you are using, might I suggest a couple other possibilities:
1) The control antibodies might have more basic framework regions. For instance, it is not unusual for human IgG1s to have isoelectric points around pI 8 or above. The basic antibody could stick non-specifically to the acidic backbone of the DNA. Try incubating the antibodies with synthetic or "naked" DNA that you know has no proteins bound to it and see if the antibodies still cause immunoprecipitation.
2) Your antibodies against the transcription factors may actually be contaminated with antigens that have hitchhiked along with the antibody during the purification process. This would be more of a problem if the antibodies were manufactured in eukaryotic cells given the evolutionary conservation of many eukaryotic nuclear proteins. If your manufactured lot of TF antibodies were already contaminated with TF from the cells they originated from, then there would be fewer antibodies available in the lot to actually bind the TF in your glioblastoma cells. That could artificially depress the signal compared to the negative controls. I assume your targeted TF binds a very specific sequence? If it does, you could try to check and see if your anti-TF or DNMT antibodies bind naked synthetic DNA consisting only of those sequences. If the antibodies do bind, it could mean you have antigen hitchhikers interfering with your experiment. You should try to wash the hitchhiker antigens off your antibody using a high salt wash (e.g. 2 M NaCl).
Good Luck.
- Badges
- Science method