Looking for endogenus expression by using custom antibody - (Dec/07/2009 )
Hi everybody!
I'm trying to detect one concrete protein by Western Blot in Mammalian cells (293t, Jurkat, SKNSH). With this respect, I've designed and produced an specific antibody against one epitope in this protein. To test it, I've used it with cells transfected with different expression vectors of my protein, althought these artificial constructions do not match my supposed complete endogenus protein. In every WB assay using transfected cells lysate, the result was OK. When I try to use the same antibody to detect endogenus expression in different cell lines, I allways detect one protein lighter (160kD) than the supposed endogenus (300kD). I assume it would be a splice variant or a caspase cleavage product. However, i've tested by RT-PCR that PMA or TNF, etc induce RNA expression, and that's the contrary I observe in WB: TNF and PMA reduces my endogenus protein expression. I've used MG132, PKC's inhibitors... and they failed to change the result. Otherwise, treatment with lamda phosphatase blocked protein degradation by PMA or TNF. I really don't have any idea what's happenig! Does anybody have any idea? Pleaaase!!!!
Gonzo on Dec 7 2009, 04:53 AM said:
I'm trying to detect one concrete protein by Western Blot in Mammalian cells (293t, Jurkat, SKNSH). With this respect, I've designed and produced an specific antibody against one epitope in this protein. To test it, I've used it with cells transfected with different expression vectors of my protein, althought these artificial constructions do not match my supposed complete endogenus protein. In every WB assay using transfected cells lysate, the result was OK. When I try to use the same antibody to detect endogenus expression in different cell lines, I allways detect one protein lighter (160kD) than the supposed endogenus (300kD). I assume it would be a splice variant or a caspase cleavage product. However, i've tested by RT-PCR that PMA or TNF, etc induce RNA expression, and that's the contrary I observe in WB: TNF and PMA reduces my endogenus protein expression. I've used MG132, PKC's inhibitors... and they failed to change the result. Otherwise, treatment with lamda phosphatase blocked protein degradation by PMA or TNF. I really don't have any idea what's happenig! Does anybody have any idea? Pleaaase!!!!
Being that the size is so different, how do you know that this is the protein your supposed to be seeing? Do you have controls on your blot, like a mutant that shouldn't express the protein at all because my first guess would be that this is merely a non-specific band. If you are worried that cleavage might be a problem, you could try an in-vitro experiment - make an in vitro translated protein that mirrors your protein of interest and then use it on a western to see if you get a band the correct size.
Hi Gonzo (q pasa killo),
how do you know that the protein you are detecting is indeed your target protein (160 is way to small for your protein, it could be anything!!)
In fact if the protein that you are detecting is not the native protein, but a degradation product or an inespecific protein you dont have to see an increase in the amount of protein when you use PMA.
How long do you transfer your gel? (sometimes for such a big protein you have to let the gel run a lot and then increase the transference time.For example when I want to see CBP or p300, instead of 1h per gel I run the transfer for 2.5 hours)
Is there anyway you can get the full lenght protein in a plasmid?
I dont get the lambda phosphatase result. What do you mean?? Do you incubate the lysates with lambda PPsa?
Hi, Laurequillo.
First of all, thank you for your very fast reply... I'll try to answer your questions in order, so:
1.- You are right. We have this possibility and the line i'm watching could be any protein. However, I've tried many modifications in WB method to make visible my endogenus protein: larger protein amount, different lysis buffer, sonication instead lysis using detergent... and the only result I obtained was this line. Studying my protein structure by bioinformatics, results in many sumolyation sites, together with caspase cleaving sites. So that's why I assume that the protein I'm wathching (160kD) could be a result of the original protein proccesing.
2.- By Rt-PCR PMA caused an increment of my protein mRNA. If PMA doesn't modify any postranslational proccessing, this RNA increment should correlate with an increment of my "proccesed 160kD line", isn't it?
3.- I did not try to transfer the gel for more than 1h. Maybe that could be a good option. I've asked some people for p300, and they didn't tell any special suggestion for this kind of proteins... Nowadays I'm running 6% acrylamide gels. Should I make them ligther?
4.- I'm trying to clonning my full protein in a plasmid, but that's becoming a hard work, because of the huge size of my insert. However, I can work with a mouse equivalent, and it results in many proccesed products.
5.- Lambda phosphatase: After my treatments (nothing and PMA) I proceed wih protein obtaining by lysis in an usual buffer. After that, I treat every sample with or not lambda phosphatase. The result suggests that treatment with lambda phosphatase makes more visible my 160kD protein. Is it possible that phosphorylation of my protein is blocking the recognition by my antibody?
Thank you very much. In the next days I'm gonna try to increment transfer time...
laurequillo on Dec 7 2009, 05:26 PM said:
how do you know that the protein you are detecting is indeed your target protein (160 is way to small for your protein, it could be anything!!)
In fact if the protein that you are detecting is not the native protein, but a degradation product or an inespecific protein you dont have to see an increase in the amount of protein when you use PMA.
How long do you transfer your gel? (sometimes for such a big protein you have to let the gel run a lot and then increase the transference time.For example when I want to see CBP or p300, instead of 1h per gel I run the transfer for 2.5 hours)
Is there anyway you can get the full lenght protein in a plasmid?
I dont get the lambda phosphatase result. What do you mean?? Do you incubate the lysates with lambda PPsa?
Hi, Smu.
Thanks for your reply. Actually I'm working now in the cloning of my entire protein in an expression vector. Once I've got it, it would be a very good idea to try one in vitro translation. However, since I don't have it yet, what kind of modifications should i do in my WB method to detect very large size proteins?
Thank you very much!!!!! This forum really works...
smu on Dec 7 2009, 04:30 PM said:
Gonzo on Dec 7 2009, 04:53 AM said:
I'm trying to detect one concrete protein by Western Blot in Mammalian cells (293t, Jurkat, SKNSH). With this respect, I've designed and produced an specific antibody against one epitope in this protein. To test it, I've used it with cells transfected with different expression vectors of my protein, althought these artificial constructions do not match my supposed complete endogenus protein. In every WB assay using transfected cells lysate, the result was OK. When I try to use the same antibody to detect endogenus expression in different cell lines, I allways detect one protein lighter (160kD) than the supposed endogenus (300kD). I assume it would be a splice variant or a caspase cleavage product. However, i've tested by RT-PCR that PMA or TNF, etc induce RNA expression, and that's the contrary I observe in WB: TNF and PMA reduces my endogenus protein expression. I've used MG132, PKC's inhibitors... and they failed to change the result. Otherwise, treatment with lamda phosphatase blocked protein degradation by PMA or TNF. I really don't have any idea what's happenig! Does anybody have any idea? Pleaaase!!!!
Being that the size is so different, how do you know that this is the protein your supposed to be seeing? Do you have controls on your blot, like a mutant that shouldn't express the protein at all because my first guess would be that this is merely a non-specific band. If you are worried that cleavage might be a problem, you could try an in-vitro experiment - make an in vitro translated protein that mirrors your protein of interest and then use it on a western to see if you get a band the correct size.
I've never done western blots on very large proteins and i don't work in mammals so many of the experiments that you've done so far go over my head. But unless you have some way of specifically showing that the 160 kDa protein is your protein of interest or some cleavage product thereof, then I'd be very skeptical. Some things you might want to consider trying: 1. do siRNA, miRNA or some other knock out method to obtain a line that should in theory knockdown expression of your protein. 2. Find growth conditions or treatments that result in down regulation of your protein - if you can see this at the RNA level by RT-PCR or a chip experiment then you would probably expect this to show up in your western as well. This would be a way to correlate the protein you see with some physiological treatment. 3. If you think this 160 kDa protein is very abundant then you might want to try to purify the protein and do peptide sequencing on it - I've never done this and I think it would be very tedious and hard work, but if your protein is very abundant and you can get it to IP with your antibody then it might be worth looking at. 4. You say that you've done some bioinformatics on the protein. Can you correlate the size of the protein you see with a particular cleavage site? If so, then maybe you should try using some truncated expression clones (since they would be smaller and easier to work with) and see if these produce cleavage products of the correct size that you would expect. This again would be correlative evidence.
That's all that I can think of for now and I need to leave soon. Hope that I've been of some help.