Protocol Online logo
Top : New Forum Archives (2009-): : Protein and Proteomics

Co-immunoprecipitation - PI3K (Oct/29/2009 )

Pages: 1 2 Next

Dear All,
I'm quite new to biochemistry, and I'm trying to co-immunoprecipitate p85alpha with my protein of interest.
I know from literature that they interact in neurons, but in this article they coimmunoprecipitated the endogenous protein with p85 and the other way round.
Since I'm studying mutants I can't immunoprecipitate the endogenous protein, but I have to transfect.
As you know neurons are quite hard to transfect, so I'm trying to do it with COS-7 cells.
I'm using a protocol suggested by Echelon for the immunopreciitation of active PI3K because I want to perform an ELISA assay with the immunoprecipitated protein.
Untill now, no results.
Is it possible that these two proteins interact only in neurons and not in COS-7?
My protein is tagged with mCherry.
It has been widely demonstrated that in various situations the tagged protein behaves like the untagged protein, moreover tha tag should be on the opposite terminus to the site of interaction, so I don't think that the tag can prevent interaction...but I'm getting nervous...
What do you think?
Thank you very much!!

-Starlight83-

Ohhhh this brings back a painful memory of grad school!!

First question: Can you do the straight IP?

What are your IP conditions? What buffer are you using? Are you vortexing your samples at any step? If so, don't vortex at any step of the Co-IP. What lysis buffer are you using? Have you tried a different buffer?

Are you IP'ing p85 or a your protein of interest?

-miBunny-

miBunny on Oct 30 2009, 03:11 AM said:

Ohhhh this brings back a painful memory of grad school!!

First question: Can you do the straight IP?

What are your IP conditions? What buffer are you using? Are you vortexing your samples at any step? If so, don't vortex at any step of the Co-IP. What lysis buffer are you using? Have you tried a different buffer?

Are you IP'ing p85 or a your protein of interest?


Ok
The IP is ok, the problem is the CO-IP
I tried 2 buffers:
buffer1: 20 mM Hepes pH 7.4, 100 mM NaCl, 1% Tryton
buffer2(the one suggested to maintain PI3K activity): 20 mM Tris-HCl pH 7.4, 137 mM NaCl, 1 mM CaCl2, 1mM MgCl2 and inhibitors
The IP works in the two buffers, but not the Co-IP
I vortex just at the last step, when I want to elute the complex in the loading buffer.
I'm IPing my protein of interest and I want to Co-IP p85

...

-Starlight83-

How long do you do your IP step? Over night or a few hours? If you can cut down the time that may help.

Instead of washing your beads with buffer, try washing with PBS (I needed to do this trick to get a really miserable CO-IP working). I am not sure why this worked, the proteins were together in lysis buffer during the interaction but I would lose the interaction if I washed the beads with lysis buffer. When a friend suggested this I thought he was nuts-until it worked.

Is there any stimulation step required to bring the two proteins together?

Have you tried IP'ing with p85 and probing with your protein of interest?

What lysis buffers were used in the neuronal study that showed interaction? Are you using the same ones?

-miBunny-

I put the antibody with the lisate o.n. and then I put the beads for 1 h.
Do you think that if I shorten the time of Ab incubation would be better??I would say the opposite..
Do you suggest to preincubate the antibody with the beas intstead of putting the antibody in the lisate and then the beads?
The two proteins interact without any stimulation, it is the activity of the immunoprecipitated PI3K that changes upon stimunation.
I didn't try the opposite IP because the Ab that I have against p85 is very expensive and works very poorly in IP. I should use a huge amount of Ab.
the lysis buffer that they use was specific for synaptosome, so I don't know if it could be good for total cell lisate.
Synaptosome are synapses isolated from brain through different steps if centrifugation.
I'm working on COS7 in culture.
I'm thinking to switch to neurons...
I want to work on young neurons (before they form synapses) in culture, so the condition is quite different from the condition of the article.
I must work in vitro because of the transfection of mutants...
:)

-Starlight83-

Starlight83 on Nov 4 2009, 05:50 AM said:

I put the antibody with the lisate o.n. and then I put the beads for 1 h.
Do you think that if I shorten the time of Ab incubation would be better??I would say the opposite..
Do you suggest to preincubate the antibody with the beas intstead of putting the antibody in the lisate and then the beads?
The two proteins interact without any stimulation, it is the activity of the immunoprecipitated PI3K that changes upon stimunation.
I didn't try the opposite IP because the Ab that I have against p85 is very expensive and works very poorly in IP. I should use a huge amount of Ab.
the lysis buffer that they use was specific for synaptosome, so I don't know if it could be good for total cell lisate.
Synaptosome are synapses isolated from brain through different steps if centrifugation.
I'm working on COS7 in culture.
I'm thinking to switch to neurons...
I want to work on young neurons (before they form synapses) in culture, so the condition is quite different from the condition of the article.
I must work in vitro because of the transfection of mutants...
:)



No one has pointed this out yet, but it is very possible that these proteins may not interact in COS-7 if they need intermediate proteins as well. In the original paper, did the authors just do co-IP of the endogenous proteins? Is there any evidence in a purified situation (GST pulldown, etc.) that these proteins interact directly? If the data is only from the coIP, this does not mean that the interaction is direct. Other proteins may be required for interaction, and these may be non-expressed in COS-7.
If your proteins of interest do directly interact, then the problem is surely with your IP protocol or with your IP antibodies, since the interaction of the transfected proteins should be far easier to visualize than endogenous. Are you using the same antibodies for IP as the other authors? Some antibodies will not work for IP due to masking of the epitopes due to folding or protein interactions. Do you have a positive control to ensure that your IP is working?

-Dr Teeth-

The longer the incubation, the more time for the complex to fall apart. That complex occurs in the cell and is "hopefully" maintained during the incubation (it is probably not occurring during the incubation).

I don't think pre-incubating the beads with antibody first would make much of a difference.

You may want to give thier conditions a try. I think Dr. Teeth has a good point about COS cells and the possibility that the interaction may not occur in COS cells. At this point the Co-IP is not working (although you know you are getting a good IP). Try replicating the assay that they are using in the same type of cells (same reagents, same antibodies, lysis conditons, etc).

Once you get this to work, you can begin to modify the assay to suit your needs in your cells.

-miBunny-

No one has pointed this out yet, but it is very possible that these proteins may not interact in COS-7 if they need intermediate proteins as well. In the original paper, did the authors just do co-IP of the endogenous proteins? Is there any evidence in a purified situation (GST pulldown, etc.) that these proteins interact directly? If the data is only from the coIP, this does not mean that the interaction is direct. Other proteins may be required for interaction, and these may be non-expressed in COS-7.
If your proteins of interest do directly interact, then the problem is surely with your IP protocol or with your IP antibodies, since the interaction of the transfected proteins should be far easier to visualize than endogenous. Are you using the same antibodies for IP as the other authors? Some antibodies will not work for IP due to masking of the epitopes due to folding or protein interactions. Do you have a positive control to ensure that your IP is working?


The proteins should interact direclty because there is another work in wich they perform GST pull-down with purified domains (the SH3 binding domain of my protein and the SH3 domain of PI3K)
My antibody is against the tag, so I suppose it doesn't mask the epitope...
I have another concern: my protein is a SV protein, since in COS7 there are no SVs my protein is cytosolic...Is it possible that the mislocalization cause a misfolding and prevent the association?
I don't have a positive control...with positive control you mean the IP with an Ab against PI3K?

-Starlight83-

miBunny on Nov 5 2009, 03:50 AM said:

The longer the incubation, the more time for the complex to fall apart. That complex occurs in the cell and is "hopefully" maintained during the incubation (it is probably not occurring during the incubation).

I don't think pre-incubating the beads with antibody first would make much of a difference.

You may want to give thier conditions a try. I think Dr. Teeth has a good point about COS cells and the possibility that the interaction may not occur in COS cells. At this point the Co-IP is not working (although you know you are getting a good IP). Try replicating the assay that they are using in the same type of cells (same reagents, same antibodies, lysis conditons, etc).

Once you get this to work, you can begin to modify the assay to suit your needs in your cells.


Ok I will shorten the incubation...
I'm not able to prepare synaptosome, but I'll find somebody that teach me how to do it...

-Starlight83-

I will wish you luck.......

When I was a grad student, I spent 6 months working out receptor/kinase and receptor/PI-3K co_IPs. The psychic scars still remain!

-miBunny-
Pages: 1 2 Next