Any particular gene may or may not expressed similarly in any stress condition. Maybe if tubulin is good for brown algae in salinity stress, its not necessary equally good for pH stress. Should we go for absolute quantification? The normalization can make a change in result and complicate things. Why do we think the normalization is necessary? We normalize the total RNA before cDNA preparation, is that not enough to normalize the expression?
If you can make sure that all your samples contain cDNA from (more or less) the same amount of cells, then: yes, no further normalization is required. However, reverse transcription is often a problem. Although identical amounts of totalRNA may be introduced, the yields can be quite different.One would need to purify the cDNA and quantitate it again. My gutt feeling says that this approach would be more robust.
Many samples just do not allow a precise quantitation of RNA and/or cDNA, for instance laser-microdissected samples. Here, the amounts can easily vary be orders of magnitude, and the quantification of some internal reference gene is the only way to correct for this.
Bustin has a nice paper about this topic:
http://www.gene-quantification.de/bustin-2002.pdf
Interestingly, the reference genes must of course not be altered in their expression by the experimental conditions, and this must be checked - but to check this, one has to have many samples (controls and treated) with a *known* amount of cDNA. This is impossible for small samples and might be not achievable for rare samples. But if it is possible, then the question remains why the reference genes should the subsequently used for normalization, since the proper normalization of the cDNA is possible.
- Badges
- Science method