Increasingly, more articles are being retracted because of duplicate or otherwise dubious Western blots. Consequently, journals are asking for original Western blots and authentication of antibodies. Many labs rely upon data from companies. However, many antibodies continue to circulate that simply do not work as advertised. Thus, the scientific community must be vigilant and organized to maintain high quality Western blots. But how best to do this? Many labs, including my own, no longer use film. Instead, we scan blots directly and save them as digital files. Further, many labs, including my own, will physically cut blots to probe with different antibodies. I for one will only allow this once the antibody is rigorously vetted for authenticity using overexpression in cells, knockdown in cells or knockout in mice, preabsorption of antibody with peptides (when available) prior to probing, as well as cell/tissue positive/negative controls. In the case of authenticating an antibody, we use the entire blot to serve as the "authentication blot" in case a journal or reader wishes to know how we know the antibody works, as advertised. Once such a blot is filed away, we cut blots so as to streamline experiments in the most parsimonious way without compromising results. The problem with this and thus the question before the community is: will journals accept this approach or must EVERY experiment be saved as a full length (uncut) blot for purposes of submission to a journal which require original blots?
In our model, we would publish the original authentication experiment in a full length blot but then subsequent experiments would be cut down blots showing, for example, the protein of interest and then the attending loading control.
I welcome all feedback on this important issue.
An important issue - thank you for posting that.
Each step you described is right and I at least try to do the same in my lab.
However, the real problem is the amount of time and money one has to spend on the validation experiments.
I try to purchase antibodies that are certified as validated with KO cell lines and try to run a positive control if possible. However, sometimes there is a problem with post translational modification or an unusual splicing that produce a signal but of a "wrong' molecular weight.
I think that these issues require setting up a free- accessible database in which the antibody, cell type, experimental conditions etc. would be listed. Such database should be open to everyone and a possibility of adding new data should be created.
The question is - how.
Any suggestions?
Thank you again for tackling this issue, stay safe!
Hi Agnieszka
Agreed! The field I am in (smooth muscle biology) was duped into thinking a certain protein (Myocardin) of predicted MW = 100 kDa indeed migrated at that size. However, as detailed in one of my research posts, this protein migrates at 150 kDa (likely due to some PTMs). I am still battling with authors and vendors over antibodies to this single protein! Frustrating..
A super idea you suggest and we have such a database in my lab so that new people are aware of what is good and what to avoid. Wouldn't it be great if there was a mother-of-all antibody database? Just as you describe with notes as to validations, use in ChIP, IHC etc, dilutions.
I think we all need to confess that sometimes, maybe more often than we would like to admit, the antibody staining or the western result is not what we think it is!
Joseph - I feel for you...
I will try to approach the ResearchGate people for advice. Perhaps they would be willing to host such a platform; who knows...?
As for the myocardin: there are ELISAs, that detect it: https://www.biocompare.com/pfu/110627/soids/137913/ELISA_Kit/Myocardin. And with ELISA, there are no 100 vs. 150kDa issues.
Or use gene silencing to prove that your 150 kDa is what it is.
Or boil your sample really well and long in the running buffer before putting on ice and then on the SDS-PAGE.
Cheers,
Agnieszka
- Badges
- Science topic
- Similar topics
- Social Science
- Media
- Journalism