protein turn-over

[kli8808]

Hi there,

 

I am in a genetics lab and do not have much experience in biochemistry and please excuse my naive question.

I am planning to look at if my protein A is stabilized by another protein B in cell culture system. So I am going to do a pulse-chase experiment to compare the turn-over rate of protein A in the presence and absence of protein B. I will transfect cells with either A or A+B, radioactively label them, replace the radioactive media and them harvest the cell at different time points after media replacement. Cell lysates will be immunoprecipitated with anti-A antibody. Then the half-life of protein A will be monitored by SDS-PAGE. So my question is: if my hypothesis is true (A is stabilized when B is bound to it), shall I always see more A in the presence of B compared to A alone? Since B will be radioactively labeled and co-immunoprecipitated with anti-A antibody. Thank you very much and look forward to everyone's reply.

[CharonY]

Well, the protein complex can break up if you run a SDS-page (as it is denaturing). Assuming a highly specific antibody you should be able to see it though (i.e. a higher MW band for the complex, a lower for A alone). If you got A alone it (i.e. the complex breaks up during sample treatment) you should be in the clear. If you got semi-stable complexes it may add variance to your expression data.

Personally I would tend to go with a MS-based quantification (if only for convenience and higher sample throughput). But not every lab may have access to it...

[kli]

Well, the protein complex can break up if you run a SDS-page (as it is denaturing). Assuming a highly specific antibody you should be able to see it though (i.e. a higher MW band for the complex, a lower for A alone). If you got A alone it (i.e. the complex breaks up during sample treatment) you should be in the clear. If you got semi-stable complexes it may add variance to your expression data.

Personally I would tend to go with a MS-based quantification (if only for convenience and higher sample throughput). But not every lab may have access to it...

Hi CharonY, Thanks for your reply. That is my concern too. If protein complexes break up in a denaturing gel, it should be fine since I will only see a single band that indicates A. However, if I see a high-MW band that indicates AB complex, it's gonna affect my quantification of A's decay since some A are existing in the form of AB complex. So my question is there a way to break up complex on purpose right before running the gel? In that case, I should be able to tell between B stablizing A and B binding A. Thank you

[CharonY]

Assuming that it is a typical chaperone-like configuration I suspect that most bonds should easily break up during standard sample treatment (i.e. boiling before loading). If covalent bonds (i.e. sulphide bonds) are involved somehow it may help to add a reducing agent (such as DTT). However, this is rather rare.

You could run a few controls (which do not necessarily have to be labelled) with only A and A+B to check whether you see any mas shifts. You probably need that control anyway. If you get double bands you can introduce harsher conditions and see whether it vanishes.

[kli]

Assuming that it is a typical chaperone-like configuration I suspect that most bonds should easily break up during standard sample treatment (i.e. boiling before loading). If covalent bonds (i.e. sulphide bonds) are involved somehow it may help to add a reducing agent (such as DTT). However, this is rather rare.

You could run a few controls (which do not necessarily have to be labelled) with only A and A+B to check whether you see any mas shifts. You probably need that control anyway. If you get double bands you can introduce harsher conditions and see whether it vanishes.

Thank you very much for you reply. One last quick question, what if my B could also activate A transcriptionally. (I guess I could definitely test that by a RT-PCR or western.) But if that is the case, I would see more A protein in A+B than that in A alone right after the "pulse" phase. Can I get around that issue by looking at the turn-over rate/ half-life of A? Thank you.

[CharonY]

If the mode of interaction is not clear you may open up a giant can of worms. For example, another possibility to affect protein stability is post-translational modification. For example, by preventing or enabling ubiquitination degradation by proteasomes are initiated.

The half-life of a protein is the sum of these and many other factors and it would be very tricky to elucidate the mode of interaction based on this property.

 

But the first thing I would do is to figure out whether B has any domains that indicate any of these roles (such as DNA-binding).

[kli]

If the mode of interaction is not clear you may open up a giant can of worms. For example, another possibility to affect protein stability is post-translational modification. For example, by preventing or enabling ubiquitination degradation by proteasomes are initiated.

The half-life of a protein is the sum of these and many other factors and it would be very tricky to elucidate the mode of interaction based on this property.

 

But the first thing I would do is to figure out whether B has any domains that indicate any of these roles (such as DNA-binding).

I know B does activate A transcriptionally in vivo and AB do form heterdimer both in vivo and in vitro. My concern is that if A has an B-dependent enhancer within its coding sequence and it works with the basal promoter (hsp70 promoter, for example) in the vector, it might be difficult to interpret my data. I just want to make sure I can conclude B somehow stabilizes A at protein level (not activates A transcriptionally) as long as I see the half-life of A is extended.

[CharonY]

By activating transcriptionally I presume you mean activates transcription? If the starting protein amount (after the pulse) is very different, it may skew the results in unexpected ways.

Generally pulse-chase is sensitive to quite a number of factors (with time being a critical one) as many steps such as starvation prior to pulse or use of inhibitors can alter the physiological response of a cell (which in turn affects half life e.g. by activation the proteasomal degradation pathways). If it has dual functions you would probably really have to look at the result before one could figure out what to do.

 

For example, if A is produced in larger amounts than the degradation system can cope with, the half-time shifts to a higher value even without actual stabilization. On the other hand it is possible that stabilization by B is crucial but it is not similarly induced then a smaller fraction of A is going to be stabilized and the calculated half-life would appear lower (just to give some crude ideas off the top of my head).

In the end it would probably depend whether you actually can see a change in half-life and then you may have to alter experimental design, maybe using inhibitors to eliminate other sources of stabilization effects.