Proper western blotting - (Mar/16/2013 )

Why do journals accept western blots without positive, negative, and loading controls as well as requiring whole gels be shown? Almost no one validates their antibodies it seems, I really don't understand why journals accept gels that show "a representative of triplicate run gels". That just seems ripe for data manipulation to me. And many blots I see never even have positive or negative controls. What is the 100% conclusive way to run a western blot? And don't even get me started on densiometry and trying to quantitate proteins using it. Just a horrible technique many people don't seem to know how to do properly .

Many journals ask for the original gel scans to be supplied at the time of submission, as well as some requiring that the gels scans are included as supplementary material.

I agree with you regarding densitometry, it is pretty dodgy, though the signal from the newer fluorescent antibodies is much much more linear, and hence better for densitometry, than that of HRP/alkaline phosphatase.

How do you mean conclusive? it depends on the experiment - are you looking to see if there is a band (i.e. presence/absence)?, are you doing IP/co-IP? Dose response?

I guess "conclusive" was a bad choice of word. What I really was trying to ask I guess was, what is the 100% correct way to do and present a WB? My understanding is that EVERY blot should have a positive, negative, and loading control. Why don't I see them then that much in literature? Vendors don't seem to rigorously validate their antibodies, so it should be up to the authors to validate their target is being hit.

Also, is there a link you have describes these new "linear" antibodies? I'm curious. AFIK, if one is going to do densiometry readings, then a standard curve is needed on your gel. Do the new ABs no longer require this?

I guess for some proteins it's really difficult to obtain a positive control (or even a negative control)... I use AB's in other applications than WB though (flow cytometry, immunofluorescence etc.) and I really had and have trouble finding cells etc. that do express the proteins I work with for sure !

Tabaluga on Sun Mar 17 18:25:18 2013 said:

Do you have trouble because of the ABs? Or because of the protein? How do you make sure your ABs really work and really do bind to the protein of interest? Just take the vendor's word for it? What should one do to validate their Abs? SiRNA experiments if one can not purchase the protein outright to do a positive control on a gel?

Trouble because I can hardly get my hands on cell lines or tissues which are definitely positive for the proteins (not very common ones). Or else there are expensive cell lines available but we wouldn't think of buying them. I agree it would be better science to do so, though. And yes, if I see a plausible staining (right intracellular location, plausible quantitative values when comparing with literature etc.), then I take it for "the AB is really binding my protein of interest"
As for your question on what you can really do to validate your AB, I'm curious to hear what other people here think about it.

To validate an antibody is quite simple - all you need is the peptide it was raised against and a + and - control. Run + and - controls on a gel side by side in two wells each for both (i.e. run Ladder, +,-, Ladder, +,-), transfer, cut the membrane in 2, each half containing one of the replicates and block membranes. Pre-incubate an aliquot of the antibody with an excess of the peptide it was raised against, then simultaneously probe the membranes, one in pre-incubated antibody and one with antibody alone. The antibody alone membrane should show the specific band in the + lane and the same signal in the corresponding lane should be removed in the pre-incubated antibody membrane.