CO-IP problem: non specific prot binding to G beads - (Feb/02/2009 )

Hello!
I'm trying to Co-IP two proteins, but the interacting protein is binding to the agarose G or A beads - so i cant have a negative control.
My procedure is the following: Transfect Myc tagged proten X in 293T cells, lyse, preclear, Incubate 2hrs with Myc antibody (9E10 from santa cruz), add beads, incubate for another hour, then wash and add Laemli buffer to beads, boil vortex load on gel. Detect with polyclonal against prot Y (endogenous).
I tryed to change the buffer, increasing salt up to 1M, preclear the cell lysate twice, but nothing seems to eliminate binding of Y to beads.
Right now my plan is to try cross-linking to preserve the complex (if there is any) and to try KCl- acetate buffer.
this however won't eliminate non-specific binding, so what would be your suggestions: any specific detergent, buffer, procedure?

Any help is appreciated, thanks in advance!
Dimilletronc

have you tried 0.1% tween-20? nonidet np-40? triton x-100?

one of them may help (i would try the tween first).

mdfenko on Feb 2 2009, 04:22 PM said:

Dimilletronc on Feb 2 2009, 08:14 PM said:

What concentration of detergent is in your IP buffer?
I use 450 mM NaCl and 1% detergent to lyse my cells. After which I dilute
in IP buffer containing 0.5% Triton.

ALSO, make the beads FRESH every time. Every time. I also make my IP buffer FRESH every time.
Beads go "off" quickly. Buffers go off more slowly, but off they go. Fresh beads makes a huge difference.
I know this will take you an extra hour but it's worth it.

mikew on Feb 7 2009, 06:55 PM said:

Well, if you make everything fresh, preclear for a reasonable amt. of time with 50 microliters beads the only reason I think you should be getting background is because you overexpressed your protein.
Also make sure your extracts are very clear and don't contain any lipids (smokey stuff in extract).
But please do not try the crosslinking and release under low pH. You will waste your time.
I have tried this several times and it simply doesn't work. I also asked several other people who have tried this. They didn't get it to work. I went to various institues looking for people who were successful in doing this. None were successful. One who claimed to be successful was lying. In fact, if someone claims they got it to work demand to see the result with proper controls showing the crosslinking worked and that the elution worked.
The crosslinking release under low pH to reduce background seems to be an urban legend.

mikew on Feb 9 2009, 01:57 PM said:

Hi,

The preclear during lysis may be the problem.

I get my lysate then dilute in IP buffer. Then preclear with beads for 2 hours.
The conditions during lysis are not necessarily the same as after adding IP buffer (cell membranes, NaCl concentration etc).

Also, in your 1st post you said you transfect a myc-tagged protein into your cells. This not endogenous protein. It's possible (likely) that the overexpressed my-tagged protein binds to the beads. If this interacts with your endogenous protein of interest it will bring the interacting protein to the beads indirectly.

Did you do a control Western for anti-myc? Is the transfected myc-tagged protein also on the beads?

Although the method of adding the antibody to the lysate and then adding beads is popular, I really don't like it. If you are using the 9E10 myc antibody I highly recommend the agarose conjugated available from santa cruz. I published a large Co-IP using this antibody and found the prebound to be much more reliable and it was easy to determine how much of the control IgG (also available prebound to agarose beads) I needed to use. I recommend you use about 10ul of the slurry, wash at least two times in PBS and then block in freshly made and filtered 5%BSA for at least one hour but I usually went overnight. Just set up the beads the day you transfect your myc construct. Wash the beads at least twice and add your lysate. Incubate at least four hours but again, I usually went overnight. Wash beads well three times with lysis buffer and boil the beads in sample buffer.

My first trick is I use an insulin needle to remove the washes. This way I can remove all the wash, thereby getting the most efficient wash and the insulin needle is too small to suck up beads in the vacuum aspirator. I never have to worry about loosing a pellet. Another trick is to either raise the salt or detergent in the wash to eliminate non-specific binding but it may actually help to wash with a low salt buffer. Some non-specific interactions are mediated by salt-bridges that won't be broken in high salt. You may have to do much optimization to find what works best for your situation. Finally, make sure you aren't centrifuging your beads too fast. I only spin my beads at 2000rpm for a few minutes to pellet them in between washes. Faster spins can alter or destroy the bead's binding capabilities allowing more non-specific binding.

How many ug of total protein lysate are you using with how many ug antibody? Are you getting good expression of the myc-protein? How does the myc IP look? Hopefully you are checking the IP even though the co-IP isn't working yet. Do you have any positive control you could check the IP for (ie: a known interacting protein. When I IP a cyclin protein, I always check for the Cdk before looking for my (hopefully) interacting protein).