Negative Control for ChIP realtime PCR in Mouse - (Feb/05/2009 )
KPDE on Feb 7 2009, 09:56 PM said:
jiro_killua on Feb 5 2009, 02:39 PM said:
cellcounter on Feb 5 2009, 11:14 AM said:
jiro_killua on Feb 5 2009, 10:55 AM said:
So I need a pair of primers that amplifies a region in the MOUSE genome that has no H3K4 trimethylation
Couldn't see any in PubMed....
Any suggestions will be welcomed.
Thanks
H3K4 trimethylation is a mark of any actively transcribed gene that has polII binding.
So, any gene that is silent in the cells/tissue of your interest will be a good negative control.
I would run away from all ubiquitously expressed genes (beta-actin, gapdh, laminb1 etc). Look for something like neurone-specific gene if you are working on erythroid cells and so on so forth.
So what non-expressing gene would be good if I'm working on liver, kidney and lung?
Shouldn't be any rhodopsin expression in those tissues. I think that would make a great negative control. Maybe b-globin as well but you have to look out with that one. Some primers may amplify several other globins as well which can screw up your quantitation. You just have to check your candidate primers with virtual PCR first.
Rhodopsin sounds good. Nah to b-globin. All tissues have RBCs in their blood vessels and many of them could be nucleated depending upon the age or treatment of the mice. So avoid using as negative control genes expressed in blood cells , that includes WBC and platelets. Same goes with genes expressed in blood vessel wall (endothelial) too.
TanyHark on Feb 8 2009, 09:10 AM said:
KPDE on Feb 7 2009, 09:56 PM said:
jiro_killua on Feb 5 2009, 02:39 PM said:
cellcounter on Feb 5 2009, 11:14 AM said:
jiro_killua on Feb 5 2009, 10:55 AM said:
So I need a pair of primers that amplifies a region in the MOUSE genome that has no H3K4 trimethylation
Couldn't see any in PubMed....
Any suggestions will be welcomed.
Thanks
H3K4 trimethylation is a mark of any actively transcribed gene that has polII binding.
So, any gene that is silent in the cells/tissue of your interest will be a good negative control.
I would run away from all ubiquitously expressed genes (beta-actin, gapdh, laminb1 etc). Look for something like neurone-specific gene if you are working on erythroid cells and so on so forth.
So what non-expressing gene would be good if I'm working on liver, kidney and lung?
Shouldn't be any rhodopsin expression in those tissues. I think that would make a great negative control. Maybe b-globin as well but you have to look out with that one. Some primers may amplify several other globins as well which can screw up your quantitation. You just have to check your candidate primers with virtual PCR first.
Rhodopsin sounds good. Nah to b-globin. All tissues have RBCs in their blood vessels and many of them could be nucleated depending upon the age or treatment of the mice. So avoid using as negative control genes expressed in blood cells , that includes WBC and platelets. Same goes with genes expressed in blood vessel wall (endothelial) too.
If I picked a non-expressing gene, and see the % Input is really low, but there's still a difference between control and treatment, what does that mean?
like 0.0001% vs 0.00003%
If I then normalize the data using this "looks-like-noise", the difference between my real samples might be gone...
What's the interpretation?
I am having a similiar problem. I am using 2 cell types, fibroblasts and a B cell tumor cell line. I am doing ChIP for H3, H3K27me3 and H3K4me3. The validation works perfectly in all of my fibroblast samples, but completely fails in all of my B cell samples! For validiation for H3 I am using RPL30, for H3K27me3, MyoD1, and for H3K4me3, Hoxc11. Why does it work great for fibroblasts and not B cells?
Also, I have read that GapDH is considered a positive control for H3K4me3. I used it as my negative for enrichment and it works really nicely. In fact, when I am using my fibroblast cells, if I compare enrichment between GapDH and Hoxc11, I get on average around a 200 fold difference! (but essentially NO difference in B cells....errrrr)
Anyone out there have an issue validating in multiple cell lines?