Two questions: Magnetic Bead Aggregation and Chelex in ChIP assay - (Jan/04/2011 )

I have been using Millipore's magnetic beads during my ChIP and have been noticing that in some of my samples, after an O/N incubation, the beads will aggregate and float around in clumps rather than as a fine mixture. I can disperse the bead aggregates when I perform my first wash however I am not sure how this aggregation may affect my IPs. I was also curious if anyone has also had this issue. Any comments or insight would be useful.

My second question refers to an encounter with Chelex I just had........it seems that when I added the 10%Chelex mixture to my IPed material that, in one of the samples, twice the amount of resin was deposited....even though I pippetted 100ul. This particular sample also happened to be my 1% Input. Can anyone provide insight into the problems this may cause during my qPCR or data analysis?

Thanks for your time everyone,

Hi Chabhara,

I have noticed a similar problem when using magnetic beads from invitrogen (dynabeads protein G) interestingly the problem has only ever occured when using chromatin samples which have previously been sonicated and stored at -80.

When I have noticed this effect the IPs have generally failed - no enrichment about background with the negative IgG typically giving very high %IP values. Fresh chromatin put into IPs at exactly the same time work fine, so in my hands this is definitely a problem with frozen and thawed samples!

Not sure what to do about this problem but again any advice would be appreciated as effectively this is completely preventing me from using stored chromatin samples.

Jon

chabraha on Tue Jan 4 19:12:51 2011 said:

The easiest answer to your chelex issue is that you need to resuspend the chelex before pipetting for each sample. The only way I could see differing amounts of resin effecting your results is that, the more resin you add the less water you and the smaller your final recovered volume. While this would technically effect your final DNA concentration, if it only happened in one sample, it wouldn't effect your results more than 10%.

I use Dynalbeads Protein A from invitrogen for my ChIPs, and I have noticed that the beads will sometimes clump after the IP step, but it is always clumping for the same antibody. e.g. my FLAG antibody will never clump, but my K9Ac histone IPs always have clumps. I think it may just be depending on the antibody you're using. You might try a different chip dilution buffer to see if it will help reduce aggregation? In general I haven't noticed this to affect my results, at least the histone ChIPs. Also, I am always using frozen/rethawed chromatin.

jonny_boy on Wed Jan 5 11:47:50 2011 said:

Jonny boy,

Sorry to hear your having the same problem, I do get this aggregation with my thawed samples sometimes....and then sometimes not. I get similar %Input values for my ChIPed samples however my IgG values shoot though the roof by well over 10 fold. I don't really take into account my IgG %Input in my calculations to determine enrichment at different regions so as it is really low (like around .03-.05% Input) but it does concern me when my IgG is well within 10 fold of my 1%Input. I wonder if after diluting the thawed sample in ChIP dilution buffer and then re-spinning down samples may reduce the presence of aggregates that may have formed during the freeze-thaw. Or if maybe adding a little glycerol or non-ionic detergent could prevent or at least help with this problem. Let me know if you figure anything out.

angelawu,

Similar thing going on here. My H3K4me3 never clumps but my H3K27me3 and IgG will clump. Sometimes from a frozen sample and sometimes from a fresh sample. I hope that my O/N incubations are not too much for the beads. I have noticed that unlike the H3K27 and IgG antibodies the H3K4 antibody is stored in a glycerol containing buffer. Maybe that bit of glycerol helps keep things smooth in the IP? The only effects I have really seen is a drasctic increase in my IgG control IP which I guess I would expect to happen when aggregates are present. I talked to a post-doc in the lab next to me and he said it may have to do with salt concentration differences between samples. Maybe these differences occur when a sample is being sonicated or manipulated in a specific manner.......or when cell numbers are different between samples. Do you use traditional ChIP dilution buffer or something different?

KPDE,

Thanks for the reply....Yeah I try to invert the suspension several times between pippetting it in samples but sometimes I think I stick my the pipette tip towards the bottom where there tends to be more chelex. THe way I analyze my samples I'm hoping that there wont be too much of a difference. Instead of using %Input or fold increase over IgG as my final value I do qPCR for a region that does not contain my mark of interest and do the delta-delta Ct method to determine a fold increase at my target region over my negative region. So Im hoping the math works it all out. Any ideas on my magnetic bead issue? Do you see a problem with adding a small amount of glycerol or additional non-ionic detergent to my dilution buffer to try and eliminate aggregates?

chabraha on Thu Jan 6 00:08:38 2011 said:

Hi chabraha,

I don't think the storage buffer of the antibody would make a difference, at least not in my case, because I conjugate the antibodies to the beads and do some washes, before adding it to the chromatin for IP, so the storage buffer is already gone by the IP step. In my case it's definitely not the salt concentration between samples either, because the aggregation of beads occurs in different IPs from the same sample for me, where the salt concentrations should be the same. I use the same amount of chromatin for each IP, so the cell numbers should also be similar. The only thing that might make a difference is probably the buffer that I am IPing in. I am using a RIPA buffer as my IP buffer (so Tris-HCl, NaCl, EDTA, Triton-X, EGTA, SDS and Na-deoxycholate). Most of these seem common to the ChIP dilution buffer from the millipore kit... Perhaps you could play around with different quantities of detergent in your IP buffer to see if it resolves the aggregation issue?

About using IgG, I have stopped using it in favour of FLAG antibody instead, just because I've found IgG to be very sticky, and frequently will have high IP%, whereas using FLAG for the negative control IP has been much more consistent and low background in my hands. Also, lot to lot variation is very large for IgG antibody, since it's polyclonal (they probably just bleed a rabbit...) Maybe you can try using a specific negative control like FLAG or His-tag, assuming you are not pulling down a FLAG/His-tagged protein.

angelwau,

good advice on the anti-Flag as opposed to an IgG. I use the ChIP dilution buffer that is similar to the Millipore kit. After the next round of ChIPs I have to do I will try adding some glycerol and see how it goes. I will let you know the results.