When quantifying Western blots many people assume it’s ok to simply image your protein of interest, strip, re-probe for their loading control of choice and then use this to normalise, without considering the linear range of detection. However, there are several pitfalls associated with this, and more and more journals are now asking for a proper defined normalisation protocol, including evidence of linear detection, to be included with all manuscripts.
The linear range of detection is the region in which band intensity is proportional the amount of target protein on the membrane. As the amount of protein increases so should the band intensity. In the image below you can see where the issue lies, the linear range is marked with the blue box, outside of this region this proportional relationship is lost, especially at the extreme ends where the signal will be either completely over or under-saturated.
Things get more complex still when loading controls are factored in as Western blot quantification is only truly accurate if both the target protein, and loading control both have similar linear ranges. This is shown in the two graphs below; in the first where the linear ranges of detection overlap then quantification of proteins within this overlap zone will be fine. However, in the latter where they don’t overlap it is easy to see how changes in protein concentration can impact the band-intensity for one protein, but not the other. In this instance quantification is next to useless.
Determining the linear range is relatively easy, and can be achieved by taking a sample and performing a serial dilution. If the ranges overlap then determining the amount of sample to load is also similarly easy. However, if the ranges don’t overlap, as is common when assessing low-abundance proteins in association with a highly expressed housekeeping gene, then it becomes more difficult to accurately quantify. Furthermore, if your protein of interest is strongly up or down-regulated by your treatment then this can further interfere with your attempts to fit within the linear range.
What to do if your linear ranges don’t overlap
However, there are several options available. An often-overlooked step is change loading control. If you know your protein is expressed at relatively low levels then a loading control such as beta-actin will likely not be suitable. In these instances, using an organelle specific protein may prove more beneficial such as Lamin B1 or HSP60 as they should be less abundantly expressed, although care should be taken to confirm that expression of these is not altered by your treatments.
A second option is to transition from loading controls towards total protein quantification as a means of controlling for loading. By assessing the whole range of proteins over, or under-saturation becomes less of an issue generating a broader dynamic range. These stains include both in-gel and membrane stains and can be imaged in a variety of different ways. Furthermore, as discussed previously this technique is a great way to check both the quality of you protein separation and also transfer efficiency.
Finally, it is possible to expand the dynamic range during imaging. Film has a notoriously poor dynamic range and issues with inaccurate exposure timings can further complicate matters. Gel and membrane imaging systems, such as our cSeries. Imaging systems such as the c600 are able to image across a much wider dynamic range and therefore linear range of detection.