ChIP backgroud issues (posted by Zebrafish) - (Feb/01/2009 )

Currently, I am conducting ChIP-PCR, but I am also suffering from some pratical difficulties. I find you are an old hand in this field and I registered in this forum to ask some help from you. Any reply will be highly appreciated!

1. When I get positive bands in blank control (No antibody added) and negative control (non-specific rabbit IgG was added), how to judge this dues to ChIP or PCR?

I know we can optimise PCR conditions to make these bands disappeared, but there must be a scale. If some false results come from ChIP, it should not be imputed to PCR. How to judge it and how to control it?

2. According to the Upstate ChIP kit (Cat# 17-610), 5μg non-specific IgG should be added as negative control while 1-10 μg specific antibody is added into another tube. I know the added amount of this antibody depends on it’s affinity to antigen, and it needs to be empirically determined, hence it varies. However, I think the amount of non-specific IgG and this specific antibody should be equal. Or, how non-specific IgG can serve as a control?

On the other hand, in the datasheet of Upstate antibodies, anti-H3K9me3 (Cat# 17-625) and anti-H3K27me3 (Cat# 17-622), there is no a concentration. Users are told to add ~4 μL per chromatin immunoprecipitation. If the added amount should be equal to that of non-specific IgG, if you are doing this assay, how do you deal with it?

(P.S. I have consulted the concentration according to the Lot# respectively, and Upstate said these are sera, and concentration can not be measured.)

3. Under the condition that I do not know an antibody concentration, I can use an efficient dilution of it. However, using the same antibody, anti-H3K9me3, from histone immunoblotting result, I concluded that 1:1500 volume to volume dilution is good enough to probe the antigen; in ChIP, if I add only 1μl of antibody, the volume to volume dilution is about 1:500. Do you think it is too high? In another word, should I adjust the amount of antibody in ChIP basing on the immunoblotting dilution?

4. It was reported that H3K4me3 could ‘turn on’ gene expression, and H3K9me3 could ‘turn off ’ genes. We used GAPDH as a negative control gene for H3K9me3. But I found that H3K9me3 is positive for GAPDH in some tumor cell lines (if there is no false result in ChIP). Do you have any information on H3K9 methylation status in cancer cells? And if you have ever tried H3K27me3 and H3K9ac?

Thank you!

Best

Zebrafish

Hi pcrman, thank you!
I really need help from all of you.

I find a paper on H3K9me3 in human cancer cells. It is:
Differentially expressed genes are marked by histone 3 lysine 9 trimethylation in human cancer cells
Oncogene (2008) 27, 2412–2421