Low yield after IP - (Jul/19/2013 )

Hello maybe one of you can help me.
I am trying to perform ChIP experiments but I always have a very low yield in the IP sample. My starting material is around 7-8 µg (which I can also see )and in the IP I get something between 0.5-15ng. The strange thing is that the results are very inconsistent. I use Magnetic Dynabeads and a commercial available antibody against an HA-tag (the same batch for all IPs). What can be the reasons for the varying yield?
Thanks for your help

your yield is low. I have never run ChIP, but I often run IP and co-IP and I know that some antibodies are sensitive to certain type of chemicals. my anti-flag M2 antibody was sensitive to many types of detergents and it caused a lot of problems for me. I discovered this after 3-4 months of failed experiments. we use protein A/G however. Maybe if you explain your protocol we might be able to help better.

thankls for your reply,
my protocol is based on this http://tryps.rockefeller.edu/Protocols/ChIP-seq_ver4.pdf.
For the IP my buffer contains: 10mM Tris-Hcl (pH8), 350mM NaCl (is this to stringent?) , 1mM EDTA (pH8), 0.5mM EGTA, 0.1% Na-Deoxycholate, 0.5% N-lauroylsarcosine and 0.3% Triton X-100. I incubate with 50µl Protein G Beads (+ 10µg ANtibody) ON at 4°C.
The puzzling thing is that the IP sometimes works and sometimes not. I don't exactly know at what point (and how) I can do propper test and or controls.
Do you have any suggestions?

I wouldn't do IP with triton. It's a strong detergent, eventhough it's only 0.3%. I prefer 1% CHAPS or digitonin. What I'm suggesting is just based on my own terrible experience, maybe your problem lies somewhere else. Triton broke down my antibody bound to protein G and I didn't realize for a long time.

your link isn't working by the way.

do you add antibody to your sample directly? or you add antibody to protein G first? I prefer the second method.

Thanks for you reply. The whole thing is really frustrating.
I bind the antibody usually one day before the experiment to the beads (ON at 4°C). But I also tried to bind them just prior of the experiment (4h at RT). Both of these worked pretty nice for another antibody
I also use Triton-X in the same concentration for another antibody (what is the detergent doing at this step anyway?). There it works also good. But maybe the anti-HA antibody just doesn't like Triton-X. In order to find out whether the antibody get's dissociated from the beads I was thinking of doing a TCA precipitation with the non bound supernatant followed by WB against the antibody. Does it make sense?

p.s. here is the protocol (I hope the link is working now) http://tryps.rockefeller.edu/Protocols/ChIP-seq_ver4.pdf

- on page 2, you add 50 ul of beads to 100 ul of 10% triton x-100. the author of this protocol must have been joking.
- on page 2, you are supposed to wash with RIPA for 7 times. Just as I thought. this is a lot man, we washed with 0.5% CHAPS 3 times only.
- your elution steps look OK to me, everything that is happening is before that. I didn't block protein G. I noticed there is not much difference between blocked beads and non-blocked. Protein G is better than A for IgG3 antibodies, but for the rest doesn't really matter.

I suggest you to do your Chip base on this protocol blow:
http://labs.genetics.ucla.edu/fan/Protocols_ChIP.htm

If your antibody is sensitive, you can decrease amount of SDS in Lysis Buffer from 1% to 0.5%.

Low yeild if qPCR of Input is good.
1- Too much sonication or bad condition for sonication(Power, Input, Time, Temperature)
2- Sensitive Antibody.
3- High Ionic Strength
4- Week Fixation of DNA to Protein Complexe (Tethering )