In western blots, do loading controls have to be proteins that are in equal amounts in each lane?

Question

I have recently done a Western Blot and I am doing data analysis on my blots. I am studying the protein GSK3, which is a phosphoprotein.

I have labelled my membrane against active GSK3 and inactive GSK3. I have also labelled the membrane against total GSK3. I have used Ponceau S as a loading control. From my Ponceau S staining I have seen that there are variations in the amounts of protein loaded per lane. So there are variations in the amounts of total GSK3 per lane (the intensities of the bands are not all equal).

I am wondering if the band intensities of total GSK3 are not consistent among lanes, in this case would using total GSK3 be a good loading control to normalize my data on? I have read for the case of phosphoproteins, you can normalise to the total protein (using an antibody that recognises all forms of the protein of interest).

I have read that loading controls are endogenous proteins that are unaffected by experimental conditions and used as an indicator of sample concentration. But if the amounts to total GSK3 vary between the lanes, how can this be a good loading control? Similarly, why do many researchers use Ponceau S as a loading control, if the band intensities among lanes vary (if you don't load equal amounts of sample)?

In other words, for a protein to be a loading control, does it have to be in equal amounts for all your samples/lanes?

Any insights are appreciated.

Answer 1

I'm aware that this question is over 3 years old at my time of writing this answer.

While @Minnow has one answer - it's not one that applies in your situation. That answer applies if you are using a defined standard protein detected by an antibody - usually something like B-actin or nucleolin, where you can compare intensity of bands relative to that single standard in each lane. However, even in this case you generally aim to have the same amount of protein loaded in each lane as it simplifies analysis and makes a more convincing story if aiming for publication of your research (as you should be doing...). The convincing story is important as people struggle to determine changes in target protein expression visually if the standards are not the same, which means reviewers will have a hard time and your analysis will be suspect.

In the case of something like Ponceau S staining, which is a non-specific stain for proteins, you do generally need same amounts of proteins in each lane so that you can compare the expression for the same reasons as outlined above. If you are doing densitometry analysis and using this to determine relative expression of the proteins in each lane, then you should choose one well-defined band on the Ponceau S stain in each lane and compare to that, not compare all lanes to a single band in one lane. Also be aware that this technique does not allow for inter-blot comparison!

Be aware that densitometry is an inexact science, so saying that bands are 1.1 increased expression relative to a control is problematic statistically (how did you choose the bands to determine as your reference? what if band areas are different sizes? was the image saturated at any point? What file format are you using for analysis (use raw or tif, not JPG - and you should know the difference and why not the last one!)?, while bigger differences might not be. Convincing reviewers of differences needs visual confirmation in the blot as well as a graphical representation.

There are whole websites (e.g. Retraction Watch) devoted to problematic/fraudulent results in papers, many of which are western blot results. Don't become one of the papers on there for dodgy analysis!

Answer 2

Controls do not need to loaded in equal amounts per say, but they do need to be expressed relatively equally between lanes. They need equal expression, such that controls can be used to normalize values of the experimental proteins.

It is common to use housekeeping genes, such as beta actin, which are usually expressed equally between treatments to control for loading. Important to ensure treatments don't affect those proteins.

For examples and rationale, see:

https://www.abcam.com/primary-antibodies/loading-control-guide

https://www.bio-rad-antibodies.com/western-blot-loading-controls-antibodies.html

https://www.labome.com/method/Loading-Controls-for-Western-Blots.html