Dephosphorylation and IP - (Jul/31/2008 )
Hey,
I performed an Co-IP and I see that the second protein comes down as a double band. I assume that it might be due to (partial) phosphorylation of the protein. What I wanna do are the following 2 things:
1) prepare lysates without phosphatase inhibitors (PPI) and add phosphatase to the cleared lysate. Then do the IP as before and see whether I still get the protein down.
2) Do IP in the presence of PPI and then incubate the beads with phosphatase. Load the sample with the phosphatase (in case this disrupts the interaction) and load the beads and see if the higher band at least partially did collapse.
Here are my questions:
1) What phosphatase do you recommend? CIP or lambda phosphatase?
2) What is the ususal incubation time? 30 min to 1h at 37˚C?
3) My lysis buffer is 20 mM Tris/HCl pH7.5, 150 mM KCl, 5 mM MgCl2, 1% Triton X-100. Will that work with the phosphatases? Do I need to add MnCl2 to the lambda?
4) Does anybody think it is a stupid experiment or setup? Suggestions are welcome.
Thanks.
I've had the same thing happen and I investigated the same issue with my IPs but you are making it much more difficult than needed. Do your IP again, the exact same way you did previously except do this one twice as big. Just increase everything by at least two and keep everything the same. In the final wash of your beads, make sure they are fully resuspended and evenly divide the sample into two. Boil one right away but then treat the other with the phosphatase. Remove the wash buffer and use the buffer that the phosphatase comes with. I used Lamda but the enzyme needed may depend on the protein you are investigating. You may want to try Lamda, CIP and PP1. Phosphatases are incubated at 30degree and anywhere from 30mins to an hour is plenty of incubation time. I don't recommend you wash the beads before boiling so try to keep the phosphatase reaction at a minimum volume. If the dephosphorylation disrupts the interaction and you wash the beads you will wash the protein away as well (good idea for future experiment if you do see a shift). Just be sure you treat your controls the same way. My IgG control suddenly had a band at the same size when treated with the phosphatase (sigh). Oh by the way, you may not necessarily see the band shift down. Some phosphorylation events can make a protein migrate faster so you will loose the bottom band with dephosphorylation.
my protein also comes as double band after IP, one is around 60 KDa and the other is 25 KDa.they told me it is the light chain of the antibody. don't know why.
ahhhhhh curtis!!
ahhhhhh curtis!!
what did I say? if you know anything then let me know.
This is a common problem in Co-IPs and you need to be careful to avoid it. When you do the Co-IP you use an antibody (which is protein) to pull your protein of interest from the lysate. You then boil the beads which removes all protein (including the antibody you used to IP) from the bead. When you load the sample onto the gel you are also loading the IP antibody. An antibody is made of two types of proteins, heavy chain and light chain. These proteins run around 50kDa and 25kDa but there can be some sight variation is size between different antibodies.
There are two potential problems because of these proteins now being on your membrane. First, if you used a rabbit antibody for the IP and you use a primary antibody produced in rabbit and then an anti-rabbit secondary antibody in the western, it recognizes and binds ALL rabbit proteins, including your primary antibody for the western and the rabbit proteins which made up the antibody you used to IP. You now get a hugely robust signal at 50 and 25kDa which more times than not will interfer with you getting good blots of your protein of interest. Also, this holds true for all species, not just rabbit. If you IP with a mouse antibody, you generally don't want to western blot with a mouse antibody. However, there are times when you have no choice because the only available antibodies are of the same species. I only know of one potential way around this but it doesn't always work (at least not for me). You can try to use protein A-HRP as your secondary. Protein A (supposedly) only binds antibodies in their native conformation. The IP antibody proteins on your membrane have been denatured in the SDS-PAGE and shouldn't bind the protein A-HRP. However, I've even had this light up my chains.
Which is the second problem, generally the IP antibody proteins are a huge amount of protein relative to the proteins from the IP itself. You should ponceau your membrane after transfer and you usually will be able to see the bands of antibody protein at 25 and 50kDa. Sometimes when a protein is just in such huge amounts, it non-specificially binds to other proteins, like your antibodies. I've seen this with GST and BSA. The best thing you can do here is use as little antibody in your IP as you can and this may take quite a bit of optimization.
The thing is, if you can visualize and get clean images of your proteins of interest without signal coming from the heavy or light chain interfering, it's not a big problem but you just got really lucky. More times than not the signal from the chains will interfer and it seems like I'm always looking for proteins that are around the same size. There's only one other solution that I know of and that is chemically cross-linking your IP antibody to the bead before doing the IP. It becomes covalently bonded to the bead and doesn't come off when you boil the beads. Hence no IP antibody in your western and you can use whatever antibodies you want in the western. There are kits and protocols but I've never gotten them to work well but then again I've never spent a whole lot of time on it. Some people swear by them.
Well~ I sincerly hope this made sense and helped you understand the two bands you are seeing in your IP. Now you know that when you want to do an IP, you have to think ahead to the western and figure out how to make the two compatable. p.s. This is one reason why people use tags and put tags on proteins that they express. There are many commercially available antibodies to tags and you can always find at least a mouse and a rabbit.
What??? You clearly know the problem and yet you won't help? You won't take a few minutes out of your schedule, even though you obviously aren't that busy, to explain? What? Did you always know everything? Didn't someone, at some point in your early scientific career have to explain this to you? Why are you here if not to help and be helped? Why do you think it is so beneath you to stop and help someone who is clearly new to science and/or lab work? Again, why are you on this forum? You need to give to get. I sincerly hope you do not work with students and have no desire to ever do so.
Thanks for all the explanation rkay447
Just a thought...but you could crosslink your pull-down antibody with DSS (disuccinimidyl suberate) to your bead. I have found this to work well. The reduction in Heavy and Light chain is significant on the western blot. The protocol is described in "Use of Immunomatrix methods to improve protien-protein interaction detection" by Qoronflech et al., 2002 in the Journal of Biomedicine and Biotechnology. Essentially, the Fc region covalently bonded to the Bead with the DSS.
I am sorry for my previous remark. i know its unhealthy here. i apologize again. but i feel its up to to each individual to learn and understand science no matter how your mentor instructs you. i did not mean to hurt curtis feelings, but simply enjoyed his newbie funny ignorance.
of course i did not follow it up this post until now but i will avoid such statements hereafter.
cheers.
p.s i love this forum