I am trying to look for a physical interaction between proteins "X" and "Y" using Co-IP. X is GFP tagged that I express in MDCK cells which normally don't express it. After transfecting the MDCKs, my expression levels of X are terrific. Whenever I IP a cell lysate with the transfected X on just my beads (protein A or G) I always see a band when I develop up it on a western blot indicating to me that it is non-specifically binding to the beads. I have tried pre-clearing my lysate, blocking my beads, and changing my detergent conditions which hasn't made any difference. Is it possible that I just have WAY WAY too much of protein X from my MDCKs and that it is just completely saturating the beads? Should I dilute my lysate before I add it to my beads?
Hi Charles, a couple of things spring to mind. If I understood correctly, you have bound your overexpressed X to your beads (by anti-GFP antibody?), mixed it with your lysate and then analysed by western blot? What size is the band as its normal to get antibody bands on a western blot from any IP - these will run at 55kD and 26kD and correspond to the heavy and light chain from your antibody.
As to whether you have saturated the beads, you should be able to find out the binding capacity of your beads and calculate accordingly. To check, just run the flowthrough after you have carried out the binding to make sure there is no unbound left.
One final thing, if your protein is too concentrated it may aggregate anyway, and these aggregates would spin down when you brought down your beads and might be responsible for the WB bands. You could get round this by using magnetic beads which don't need spinning.
Good luck.
Claire
Hi Charles,
If you are talking about non specific bound bands after developing by WB, I would suggest that you might like to take into account whether the protein X binds with any other proteins?
One more thing I would like to know whether you are seeing this non specific band above the desired loci. Also, are you taking GFP's molecular wight into account? For example if the non-specific band is above the desired location, and if your Ab does not consider GFP MW, then the GFP tagging is adding up it's MW, reducing mobility.
Hope this helps.
I wouldn't worry about the expression of X being too high. Saturating the beads with your protein X is what you want to do anyway. What's important is that you need enough concentration of protein Y since you may only have a small proportion of X-Y interacting in the cells. It's not clear what you are actually concerned about here though- are you worried that protein X is binding to the beads non-specifically i.e without antibody? It doesn't really matter how X is associated with the beads, as long as protein Y doesn't associate with your beads unless protein X is also present. Run your IP alongside your input and include all the controls to show that Y only IPs if X is there too. There will always be non-specific bands there too, e.g antibody and protein A/G may show on the gel. As long as your proteins are a different size from these you can ignore the non-specific bands.
Saturation can happen on several levels. I'm not sure what you are doing, because there are different ways in which co-IPs are done and people have different strategies. If proteins X and Y interact with each other in MDCK cells, you assume that X-GFP still interacts with Y in those cells, then when you make a lysate, you hope that X-GFP continues to interact with Y. So you incubate your centrifuged lysate with anti-GFP antibodies for 1 hour on ice, you spin briefly to remove unspecific protein aggregates, then you mix the supernatant with protein-A sepharose (or magnetic beads) and put it on a slow rotator in the cold room for 2 hours sustained capture by protein A, then you separate the immobilised protein A beads, wash them, separate them again etc.. and then you boil the pellet in sample buffer and run on a gel. How do you detect Y? Have you labelled your cells with 35S methionine and do you expect a radioactive band at a specific molecular weight in addition to the X-GFP band? Or do you have an antibody that specifically detects Y?
If this all makes sense and I'm on the right track, then here is what I think could happen with respect to saturation:
1) There is a limited amount of Y in your cells, so if you overload with X-GFP, and immunoprecipitate only a portion of the overexpressed X-GFP, then you loose a lot of Y interacting with those X-GFPs you failed to precipitate (either due to limiting anti-GFP antibodies or limiting spaces on proteinA beads).
2) In fact you can also overload with primary antibody, imagine you add so much anti-GFP serum that all the X-GFP (and associated Y) is bound to antibodies, but only 10% of those antibodies can be captured by protein A beads, then that will effectively dilute your signal on the proteinA beads by a factor 10. proteinA must be in excess of antibody. Antibody must be in excess of antigen (X-GFP). And X-GFP must be present in a reasonable level relative to protein Y that you expect to interact in vivo.
Ideally, a reviewer will say that you should do reciprocal IP, repeat everything, but switch Y and X around (make Y-GFP and test if it co-precipitates X)
This is my best guess to answer a question on saturation, but I agree with Amanda that your description doesn't specifically state what you do and what your concern is. I still hope the answer helps though.
- Badges
- Science method