Is my antibody ChIP grade? - (Mar/10/2010 )

What's the general consensus as to what constitutes a ChIP grade antibody? What experiments do I have to perform to prove that my antibody is ChIP grade? Thanks!

nigelcr on Mar 10 2010, 08:26 AM said:

A ChIP grade antibody is considered ChIP grade if it works in ChIP .

An experiment you can try is perform ChIP, reverse crosslinks but do not add proteinase K and run the eluted protein on Western to see if you pulled it down. It's not enough to just do a straight IP as the crosslinking can sometimes mask epitopes.

KPDE on Mar 11 2010, 04:42 PM said:

yes, make sure to use a polyclonal Ab. Polyclonal will help reduce the chance of masked epitopes during crosslinking

KPDE on Mar 11 2010, 02:42 PM said:

My thoughts exactly!

KPDE on Mar 11 2010, 02:42 PM said:

Thanks for the reply. I've done the experiment that you suggested and I do see my protein at the appropriate molecular weight using a polyclonal antibody on a western. However I do see a number of accessory bands, all of which are smaller than the protein of interest. I'm assuming these are one of 2 things: 1) a combination of degradation products and alternative isoforms; or 2) non-specific crap.

It gets a little more tricky when I look at actual ChIP data where I'm immunoprecipitating with my polyclonal antibody. Here's what I see for a gene that I expect to be regulated by my protein of interest:

Protein of interest ChIP % input = 0.3486%
PolII ChIP % input = 0.5147%
IgG ChIP % input = 0.0077%

Oh wow, my polyclonal antibody works for ChIP right? Well, I'm not so convinced. Here is what I see for the housekeeping gene PP1A:

Protein of interest ChIP % input = 0.6542%
PolII ChIP % input = 2.7785%
IgG ChIP % input = 0.031%

So I'm getting ~ 20 fold enrichment in the ChIP with my protein of interest compared to the IgG control in the housekeeping gene where my protein of interest has no business binding. I'm using a pretty generic ChIP protocol as far as I can determine. I am a novice at this and any constructive comments would greatly be appreciate. Thanks!

nigelcr on Mar 15 2010, 04:23 PM said:

This is probably a stupid question, but how sure are you that your protein of interest is not binding to the housekeeping gene (more specifically, what region are you looking at - intron, exon, 5' UTR etc...?). Genomw-wide we often see binding of specific transcription factors to regulatory regions of supposed housekeeping genes, often within the 5' UTR/promoter (and you do have a high PolII enrichment!). I am not too surprised about your results.

NemaToStella on Mar 16 2010, 10:50 PM said:

No, that definitely isn't a stupid question, and it is possible that my protein interacts with the PP1A promoter (the primers amplify the 5'-UTR and TSS of PP1A). I've shown in earlier experiments that my protein interacts with several general chromatin components, so it could interact with the promoter of this gene. However, I probably should have shared a little more data and details of my experimental setup, which may make my concerns a little clearer.

I initially did these experiments by transiently transfecting 293 cells with constructs encoding either HA-tagged protein of interest or lacZ-V5, and then running ChIPs in both cell cultures using either the protein-specific antibody or a supposedly ChIP-grade anti-HA antibody from Abcam (ab9110). These are the ChIP-qPCR data that I got for PP1A in both sets of cells:

293/HA-Gene of Interest:
Anti-HA ChIP % input = 0.8527%
Anti-protein of interest ChIP % input = 1.2289%
Anti-PolII ChIP % input = 4.9938%
IgG ChIP % input = 0.0085%


293/lacZ-V5:
Anti-HA ChIP % input = 0.4148%
Anti-protein of interest ChIP % input = 0.4307%
Anti-PolII ChIP % input = 7.6386%
IgG ChIP % input = 0.0058%

So, the ChIPs in the cells transfected with the HA-tagged protein look okay to me, with the protein-specific antibody behaving pretty much as I'd seen before. However, the concerning results are the ones for the cells transfected with lacZ-V5. As you can see, the ChIP-grade anti-HA antibody is ChIPing as much PP1A promoter DNA as the protein-specific antibody. Given that the HA-antibody shouldn't be detecting *anything* in this experiment, it leads me to conclude that either I'm doing something wrong (I'm currently trying other protocols and conditions) or the ChIP-grade anti-HA antibody isn't actually ChIP-grade.

I guess this brings me back to my original question: is there any uniform consensus as to what a 'ChIP-grade' antibody is? Clearly, the Abcam anti-HA antibody "works for ChIP" until I used it in a cell line where no HA was being expressed. Also (and probably more importantly for me ), if anyone has any ideas as to what is going on here and how I can get this experiment working I'd really appreciate it. Thanks in advance!!

For what it's worth I've had some luck using primers for LINE1 as a negative control. It seems to me the best way to determine if you have successfully performed a ChIP is to compare your %input from your gene of interest versus a negative control region (as suggested by KDPE in these forums). I've been using LINE1 and I've seen others use the 28srRNA gene. I've also had a little bit of success using an intergenic region ~40kbps upstream from my promoter of interest, although I have some difficulty getting clean qPCR data from such regions as there is very little binding in general...

MM

Mighty Mouse on Mar 17 2010, 06:08 PM said:

Thanks mighty mouse. If you have been doing ChIP with human cells would you be willing to share the sequence of the LINE1 primers?

Well I don't do ChIP with human cells, I do it with mouse brain. But here are the primer sequences, perhaps they'll be of some use to you:

LINE1-F 5’-AAA CGA GGA GTT GGT TCT TTG AG-3’
LINE1-R 5’-TTT GTC CCT GTG CCC TTT AGT GA-3’

MM