Controls in ChIP experiments? - (May/09/2007 )

I want to start with ChIP experiments to find out, whether a certain epigenetic factor binds to my gene of interest. Now I wonder which controls i have to show, to proof my results. What control experiments do you always include in your experiments and why do you think they are important?

---

I'm ressurecting this thread becuase this has been a subject of debate recently in our lab. I've seen people use the following controls for the IP part of the ChIP protocol, some or all of them as negative controls:

1. Pre-innoculation serum (aka "IgG control) for each specific antibody
2. Sepharose beads without any conjugated antibody
3. Sepharose beads with conjugated protein A
4. Sepharose beads with conjufated protein G
5. Agarose beads without any conjugated antibody
6. Agarose beads with conjugated protein A
7. Agarose beads with conjugated protein G
8. Pre-blocked specific antibody with innoculating anti-sera peptide

Which do you use as negative controls? More than one?

I think this might be a point of confusion for people who are new to ChIP.

---

We use only one negative control per experiment. Our control is perform like other samples, only the antibody differ, to provide best measurement of the noise for the ChIP system: dynabeads with conjugated protein A and IgG antibody.

---

So, by per experiment do you mean per sonicated sample? I'm asking becuase for our experiments we often have multiple fixed pellets all taken from the same experimental sample (fixed cells are alloquoted into 1x10^7 fractions and then frozen as dry pellets). Thus, even for the sample experimental sample we might have 3 or 4 different sonications. Part of the discussion in our group is "do we run controls for each sonication?"

---

I would say one negative control for each ChIP experiment, not for each sonication.
If yours sonication parameters, yours cells and yours numbers of cells for each pellet are the same, I don’t think it would be a great difference between sonicated sample. Thus, for each tubes of ChIP the difference of noise would be quite the same. So one negative control (and one positive control=RNA pol.) would be sufficient. Moreover if you do one control for each sonication, the experiment would be a little bit “heavy” (it’s not an argument, but …)

---

Well, I agree with you completely. And we use the same positive control as well. Thanks!

---