I want to detect interacting partners of a bacterial membrane protein and have a list of potential protein partners from a co-IP experiment. What is the best way to validate these results? Reverse co-IP may not be possible for all interacting partners and I want to be certain that it's worth chasing this list before going ahead with bacterial two hybrid.
It is not an issue of repetition, it is a matter of proper controls. When you have the proper controls, you can verify interactions by Western Blot to show that the putative proteins that are supposedly picked up in MS are really interacting (I don't know if proteins were just identified by PMF or also by MS/MS). A proper control for a bacterial membrane protein would be a similar type membrane protein to do the pull down with, to rule out artifactual interactions due to insufficient solubilization of membranes (this especially if the identified protein is also membrane-spanning or tightly associated with the membrane), or aspecific binding to hydrophobic patches of membrane proteins per sé. In addition, interactions can be validated by overexpression of both constructs with tags for identification and reciprocal coIP, including such proper controls.
Have you contemplated use of reversible cross-linking agents?
As far as replication is concerned, 3 independent demonstration of same results are commonly used as a measure of repeatability but that alone doesn't tell you if your interpretation of results are justified -
Thanks for your suggestions.
Tausif - Yes, I have used DSP and crosslinked the sample before the co-IP. When I treat it with DTT, the bands do disappear. I carry out crosslinking with the whole cell lysate and then go on with co-IP.
Dennis - The proteins were identified with LC/MS/MS. I include NP-40 in the buffer which hopefully doesn't disrupt interactions. My experimental setup includes:
1. wild-type strain with native level expression of the membrane proteins
2. a strain over-expressing the protein
3. a strain lacking the protein
The internal controls include:
1. Rabbit serum (Haven't been able to get my hands on mouse IgG since I use a Mouse monoclonal ab for my protein)
2. Bead control (No antibody added)
I could do the co-IP with another membrane protein, that sounds like a good control experiment to carry out.
- Badges
- Science topic
- Similar topics
- Biological Science
- Bioscience
- Biotechnology