how to isolate non-crosslinked protein? - (Jan/02/2007 )

Hi,

I'm trying to see if under certain conditions a TF is more likely to be bound to DNA. In order to do so, I figure on transfecting cell line (HEK-293), growing in different conditions, and cross-linking with formaldehyde (using the ChIP protocol). I then want to SDS-PAGE + western blot free protein, and see if there is less unbound factor in treated cells. Does anyone have any tips on Do's or don'ts? Also, how can I get rid of the DNA+protein, and stay with only unbound protein?
Thanks, Benny.

---

um, it's a little confusing, what you are trying to accomplish... I see that you want to measure bound vs unbound protein, but a western the way you describe may not be the easiest choice?

can you use either EMSA or IHC? one of these techniques may be easier for you

---

I'll try to explain (reading my post over I see it really isn't quite clear) - I think a certain factor increases my TF's affinity for chromatin. EMSA won't help me, since it's not chromatin. What I'm thinking is that if the affinity for chromatin increases then less TF will be free in the cytosol. Thus, after crosslinking, there will be less free TF in the lysate - something I can assay with a western, assuming I can get rid of the DNA+bound protein. Any clearer?

---

Then I would suggest you to consider using an affinity column, or magnetic bead made of your target sequence.

---

yeah, but you can set it up like an EMSA; an EMSA is essentially a western, no?

when you are making your nuclear extraction, save the cytosolic fraction as well as well as your nuclear fraction and assay them both separately for your protein...the column idea is great, but wont' allow you to differentiate between bound and unbound in the nucleus. I personally think EMSA would work pretty well, just don't necessarily expect everything on your western to be the right MW if you choose to use the nuclear fraction as well; the bound form will be larger...but the idea is the same. I guess I don't really see the purpose of your croslinking; I would just do a cytosolic extraction and run a western with that fraction. what will crosslinking do to help you??? if you assay both nuclear and cytosolic fractions, your controls are essentially already built into the experiment and you can express your data as a ratio if you use densitometry to analyze the bands at the end...but I'm guessing you were planning on that anyway?

anyways, IHC would be valuable too but may require a bit of sophisticated data analysis software to get hard numbers, if the change caused by your treatment is small and needs more than naked-eye definition. stain your cells with labelled anti-TF. also use a nuclear stain. measure what ends up in the nucleus and what ends up in the cytoplasm

I dunno, maybe I'm being totally unhelpful, just an idea or two?

---

ihc is immunohistochemistry.

western blot is not very quantitative.

why do you have to separate them before the western? if you don't remove the crosslinked complex then the western should show two bands: the complex and the free protein (that's assuming that the crosslinking conditions don't create a lot of random complexes). if you can solublize the proteins after formaldehyde treatment then this should allow you to see at least an approximate ratio of bound to unbound protein.

(i started writing this post before aimikins' and genehunter-1's response and was distracted for some time before completing it, so if there is duplication or unnecessary statements, i apologize)

---

Thanks all for the input. mdfenko - that's what I'm aiming for. I just figured getting rid of the DNA/protein complexes would give cleaner results. Western will have to do for now as far as quantitation goes.

---