Protein G magnetic beads and ChIP-seq

[jdub]

I am trying to make ChIP work, and my no-antibody control group (chromatin-protein + beads) enriched as much as my positive control (anti-RNA Pol II). Post-IP I washed each ChIP reaction in three different wash buffers a total of six washes. Is this a finesse step? I pipetted beads up and down to mix but not too terribly vigorously.

I was also told that blocking of the beads would not be necessary but maybe it is. Has anyone else found that helpful?

[pbluescript]

You can't do a no antibody control with Protein G beads. They will bind a lot of non-specific stuff unfortunately. You'd be better off using an antibody that is specific for something not found in your samples.

[jdub]

I use IgG as a negative control as well as a positive control antibody and enrich for a region where it should not amplify by PCR. I see lots of background with both and the no antibody control group tells me that could be a big reason why. So how can I avoid or reduce this, because it appears it will happen in every group regardless of antibody used.

[pbluescript]

You can do the standard stuff to try to reduce background like increasing salt concentration or changing the pH of the buffers or increasing the number and volume of the washes. You can also try less beads or switch to protein A beads. I actually had such a bad problem with background with protein G that I switched to the epoxy beads that Invitrogen sells. That fixed my problem.

[Inti]

I usually block the magnetic A or G beads only with BSA for 2 hours and bind the antibody first rather than put the AB in the sample and then the magnetic beads and the ChIPs looks very clean

[NGene]

If you do not want to switch to A protein beads, you do not want to use IgG-negative control and you do not want to block the beads, you might want to play with salt concentration and volume/repetitions of the washes.

I had such a bad issue too but I manage to get rid of the background by playing a bit with the above-mentioned parameters. Now I can have 10-fold differences between sample and No-Ab.

I hope it helps.