Dear Colleague(s)
this is really an extension of my previous question relating to Western blotting on Behalf of the same colleague
How valid would you consider a Western blot that does indeed demonstrate a band of the correct molecular weight but also a second band about 15KD smaller ? After all, in my experience of Western blots, this might be a bona fide truncated version of the actual 'specific' band; a steady state endogenous in vivo degraded product linked to metabolism of the real protein; or an exogenous degraded product associated with lysate instability, i.e. technical artefact
In the past I have either switched antibodies until I have found one that yields the correct band (although generally I have done this when I have not even obtained the correct band, but rather multiple non specific bands) or pre incubated my antibody with synthetic peptide (before probing) to show that one or more bands are indeed specific
However, this is not possible for this particular protein. Is it OK to accept relative changes in the specific band (in different sample lysates) in spite of a second smaller (unidentified band) or would you need to include a positive lysate (used as the reference on data sheet) and/or a negative lysate where for example the protein is not expressed and neither of the 2 bands can be detected (confirming that both are indeed linked to the 'real' protein) to confer validity on the blot
I know from experience this is a very common problem with Western blotting so all suggestions welcome
It would be better to use a negative lysate ( Samples from a knock out cells/animal). At the same time, If working with cell lines or primary cells, use siRNA to knockdown (KD) your target mRNA. If that particular protein has splice variants, both the method can reveal the splice variants or eliminate the nonspecific bands
The answer depends on the type of antibody. For a monoclonal antibody, you would observe a decrease of the signal if you use a higher dilution of the antibody, if it's just an unspecific binder with smaller size.
In case it is a truncated version or shorter splice variant of your protein, the epitope on both proteins would still be the same thus there should be no change in the signals for both bands.
Of course you can use a different antibody binding to a domain that is not present in the shorter protein then you won't see it any more, but you could also miss valuable information if that band happens to be a biologically relevant splice variant and not just an artifact.
Checking by siRNA knock-down is of course a nice validation method which you should consider if you plan to perform more experiments with that antibody. Remember that data (like knock-down with other cell lines) from the manufacturer is not necessarily helpful if your cells behave different or have a different expression profile for this protein.
The second band could be different posttranslational modification, alternative splicing or degradation. Whether you have to include it depends on the scientific question. Btw, in that case I would not call the band "non-specific".
Alternatively, it could be cross-reactivity to a different protein, in that case you can either ignore it or use a more specific antibody.
At any rate, you need to characterise this second band. Ideally, if the band is related to the protein of interest, you'd want to know whether it is physiologically active (in which case it would make sense to include it in calculations).
Another aspect: WB is at best a semi-quantitative method. ELISA or dot-blots would give more reliable results, but can't be used until you have determined specificity.
And finally: all antibodies used should be validated (doi:10.2144/000113382).
Dear Engelbert. That is a great answer
I am aware that the second band could be spevific; albeit truncated either for physiological reasons or an artifact. How to prove without changing antibodies. Some useful responses
- Badges
- Science method