Protocol Online logo
Top : New Forum Archives (2009-): : SDS-PAGE and Western Blotting

Protein is present, but not detected by antibody - (Mar/03/2014 )

In short, we have produced a fusion protein in a transgenic plant system (let's call it, pepA-pep B). The protein sequences are originally of bacterial origin, but codon optimized for expression in plants. In immuno-blots, anti-pepA Ab successfully binds to the fusion protein -- but the anti-pepB Ab does not. However, we know that the full-length protein is present because anti-pepA Ab detects a ~50 kDa band, which is our expected size.

To determine whether any deleterious mutations had occurred, we submitted our protein for identification by mass spectrometry, with the aim of detecting the two regions (pepA and pepB). Although it is present as a low-abundance protein, two different companies independently identified the pepB sequence, and we successfully attained sequence coverage spanning the entire length of the protein. This indicates that both peptides are present.
We are thinking that perhaps the failure of the anti-pepB Ab is due to the fact that the antigen used to originally produce it (by a commercial vendor) was sourced from bacterial cells. Production of pepB in plants may have caused changes in the protein’s post-translational modification pattern, therefore preventing plant pepB from being recognized by antibodies against bacterial pepB.
We also immunized mice with the plant-expressed fusion protein. The results were the same: responses were detected using anti-pepA Ab, but nothing was observed using anti-pepB Ab.
Is the cause likely to be antigen-antibody incompatibility? If so, are there any suggestions for how we can overcome this problem? We've been struggling with this issue for a long, long time now, and some kind of resolution would be nice!



As you said its a fusion protein, is it possible that the fusion is causing a different folding of pepB? Antibody has been generated with only pepB but may be pepA is creating some hindrance to the epitope to which your antibody should bind. It is just an idea, I may be wrong. 


Can you generate antibody against the fusion protein only? If yes, you can show your antibody binds to pepA and pepB alone and also to fusion protein but it does not bind to another control fusion protein, something like that.


Do you have any evidence that anti-pepB works in a western with either bacteria or plant-produced pepB?  If not, it is possible that the antibody has a conformational epitope that is abolished with the protein is linearized in the SDS-PAGE.  I don't think that the folding or sterics should pose much of a problem in SDS-PAGE due to the denaturing nature of the system.


Are there any other antibodies to pepB you could try?


"We also immunized mice with the plant-expressed fusion protein. The results were the same: responses were detected using anti-pepA Ab, but nothing was observed using anti-pepB Ab." - what do you mean?  How do you know that the mouse is not creating any anti-pebB antibodies?


Thanks for the replies. With regards to protein folding, we run SDS-PAGE under denaturing conditions, and boil the samples prior to loading. Apart from the gel, the plant protein extraction buffer itself comprises many of the same components as standard gel loading buffers (4% SDS, 8 M urea, 50 mM Tris-HCl, 5% BME, 20% glycerol). Recently we also tried out an unmasking buffer (2% SDS, 62.5 mM Tris-HCl, 100 mM BME) directly on the blot, but we didn't observe a band even for the bacterial pepB antigen that we use as a positive control. So, that protocol will require a fair bit of optimization.


On that note: we always load the bacterial antigen alongside our transgenic samples, and yes, anti-pepB detects bacterial-produced pepB just fine.


Sorry I wasn't clear about the animal immunizations. Turns out I was a little confused myself, since the ELISAs were conducted by a former colleague almost a year ago. It turns out that the plates were coated with only a linear epitope of pepB, not the whole pepB antigen itself, and is therefore inconclusive. (That said, ELISA plates coated with bacterial pepA antigen react very nicely to the sera from immunized mice -- so that part of the fusion protein seems to be working alright.)


We're now thinking of cloning the exact same plant codon-optimized sequence into a bacterial expression vector; hopefully codon preference won't be a deterrent. If the bacterial pepA-pepB fusion protein is recognized by anti-pepB Ab, we'll know that the issue is related to possible PTM mechanisms in the plant system. Maybe then we can use the bacterial fusion protein as an immunogen to raise our own antibodies in rabbits. If the bacterial version isn't recognized either, however... we're back to square one.


As far as antibodies are concerned, we've only tried a polyclonal from Abcam thus far. Other commercial vendors do supply it, but in general the prices are a bit too steep for us to play trial and error. The full-length pepB protein has never been expressed in plants before, so trying to figure out what works and what doesn't is proving to be a challenge.


At this stage it no longer matters whether plant pepB is immunogenic or not -- I'll just be happy with a positive Western Blot to corroborate our mass spec findings and prove that pepB is, in fact, being expressed.


Hi - I would recommend double checking how the Abcam antibody was created. If a non-denatured full length bacterially expressed protein was employed as the immunogen, the primary epitope that comprises the bulk of the polyclonal antibody you are using may be discontinuous and not linear. In other words, both neuron and Missle might be on to something and you may be looking at having an issue detecting the denatured protein with this particular antibody.


To address this possibility, you should be able to employ a technique that does not require denaturing conditions for a "western" - either a dotblot or an ELISA should work, you just have to isolate a cell lysate under non-denaturing conditions. For plant material and getting around the cell wall problem, liquid nitrogen immersion followed by mechanical shearing and reconstitution in a non-denaturing buffer and centrifugation to remove bulk solids could work (yeast researchers do this routinely - there are many protocols for this available - typically called yeast lysis by liquid nitrogen grinding). Then go ahead and spot or dot your lysate onto your membrane with appropriate controls, dry it and proceed with a standard immunoblotting technique. I would leave out any detergents, BME or urea.


If that doesn't work, then a plant specific post translational modification(s) might be blocking your epitope - I think it's unlikely that the fold of pepB is the source of the problem, since you are using denaturing SDS-PAGE and immunoblots for detection. Do you have any information from your mass spec results to demonstrate this? If you can identify certain likely PTM culprits, you may be able to unmask them with certain enzyme treatments - again, if the only goal is to corroborate your mass spec data, then this might allow you to detect a signal.


Have you considered using a tag for pepB, such as HA? These are small and usually innocuous and you could subsequently assess function compared to your initial fusion protein (untagged) to ensure genetic complementation. There are many anti-HA antibodies available commercially ( for example). If you simply want to demonstrate/corroborate expression, this would fit the bill.


Hope this helps