I’ve detected a consistent discrepancy in the apparent expression of Beta Actin and Beta Tubulin in the livers of Snell dwarf mice and controls. This discrepancy exists even after ensuring that the tissue lysis is carried out at the same Lysis buffer: Tissue Weight ratio for the dwarf and control mice (adjusting for the smaller size of dwarf livers), and then normalizing loading using a BCA assay (lysis efficiency appears equal in the two genotypes).
There is no consistent opinion on this issue among my colleagues. I want to see if ResearchGate has an opinion.
Some of my colleagues think that this apparent loading discrepancy is likely due to differences in the organ architecture caused by the smaller size of the dwarf mice, which affects the apparent expression of such controls when normalizing to tissue weight or soluble protein (BCA). They believe that this issue can be reasonably remedied by manually adjusting the loading to equalize the Beta Actin between the genotypes (e.g. load more protein for the dwarfs).
I and others are concerned that such a manipulation would itself introduce a possible bias to the data, especially since these genes are intended to be used as loading controls.
What is a good way to solve this? I am looking for other loading controls as I type this (e.g. GAPD).
Data from an earlier experiment corroborates such a loading discrepancy at the mRNA and protein level; at-least for Beta Actin.
The plot thickens... GAPD shows a clear difference by genotype, but in the opposite direction of Beta Actin and Beta Tubulin...
Thanks for the insight, I'll try taking a look at the paper soon. I have often used BCA quantification as a proxy for protein loading, as Ponceau S blot staining proved difficult to reliably quantify using ImageJ (I suspect it saturates). I'm aware that others in the comparative biology field have sometimes used either Ponceau S or India Ink in cases where a cross-species loading control is unavailable.
Ultimately, however, my specific problem is behind me, as I no longer with with these particular model systems. That said, I don't believe my former field ever reached an adequate conclusion as to what was the most appropriate means of normalization.
Hello you two,
as Christian might remember from our last talk at the Proteom Forum in Potsdam, we might have an elegant and valid solution for this problem. The issue regarding HKP is not a new one. Developing this normalization method researchers have been aware of the fact, that a HKP is only an insufficient crutch to the original idea of the Western blot, namely to normalize with the total protein. As the total protein could not be detected together with the target at this time crutches like HKP and staining techniques have been necessary. A lot of researchers somehow forgot about this fact over the time.
With the possibility to multiplex in different fluorescence channels, this limitation is now not longer present. Therefore we developed our Smart Protein Layers using fluorescent labeling to detect the total protein and the target at the same time. Additionally we also included a sample standard to monitor all kinds of variables that can effect the signal intensity of the total protein and the target. So we normalize the TP first and then use this to normalize the target.
Have a look on our website to find out more:
https://www.dyeagnostics.com/site/en/products/spl/
I hope we can help,
Kristin
P.S.
Dear Christian, can you send me your paper. I would love to read it.
An interesting technology. I've always been concerned that whole-lane detection might be non-linear, as it's difficult to perform a background subtraction when the region of interest is essentially the entire blot rather than a discrete band. That said, I've never tested it exhaustively.
Unfortunately, the academic field suffers from considerable inertia, so I suspect that the development of a new standard for protein normalization will be met with lukewarm interest. Unfortunately.
- Badges
- Science method