I have this question that I havent been abIe to answer by Iooking at the Iiterature, everyone seems to assume something Im obviousIy missing, which is: when you are doing an assay of Reference genes , you gonna need to verify that the gene is getting same expression in both conditions, the IogicaI thinking wouId say that I have to take into account amount of bacteria (same number in both conditions) for do so, but many research I read takes same amount of RNA from both conditions into consideration, what happens if a condition (like Im sure is mine) is gonna reduce metaboIism and many genes are going to be negativeIy transcribed, Im pretty sure I wouIdnt yieId same ammount of RNA, so, it is ok to take in this case same ammount of bacteria, and then do the whoIe protocoI with different ammount of RNA? my thinking is biased to do so, if someone has faced a crossroad Iike this Iet me know, thanx to aII.
An ongoing argument in the field. The short answer is that there is no absolute standard.
It makes sense to keep the number of bacteria constant if you can, but the issue then becomes the RNA isolation method - no method can reliably purify all of the RNA available (and the amount of RNA recovered often varies widely), so a constant number of bacteria is not a suitable lone control - you'd have to repeat the assay many times to prove significance (which you might be in a position to do, but you'd need to be extremely careful with your RNA prep and even then I doubt reviewers would accept it).
And so someone somewhere along the line began to use reference genes, accompanied by the inherent claim that no change in genome or environment would have a significant effect on these genes. Like you, I don't buy that.
So I believe that the best control for most situations is to normalise against total RNA, though in your situation where you seem confident that global gene expression will be reduced under treatment conditions, this method will 'fail'. I guess microorganisms probably display a wider relative variation in global gene expression than most vertebrate tissues.
If you are interested in the expression of your gene as a proportion of the total RNA produced, you can still normalise against total RNA. But if you want to know in real terms (how many transcripts are produced in each treatment relative to the others) then here's a crazy idea which might just work: why don't you prepare a sort of investigative curve of expression to examine the consistency of a reference gene which you don't believe is specifically targeted by the treatment?
You could do this by mixing bacterial cultures of equal density serially (100-0, 75-25, 50-50, 25-75, 0-100), isolating RNA uniformly and normalising against total RNA. If the real-terms expression of the reference gene under investigation changes between treatments (ie. if global RNA production changes), it should be possible to determine the factor by which it changes and incorporate that into your assay for the gene of interest.
If the reference gene expression (real terms) doesn't change between treatments, you'd see a linear investigative curve with R-squared value close to 0 (100-0 expression value normalised against total RNA would equal 50-50 and 0-100 values).
However if the reference gene expression does change between treatments (real terms), you'd see a higher R-squared value.
You can then work out a correction factor to compensate for change in real term reference gene expression.
Then when you run the assay for the gene of interest and normalise against the reference gene using the calculated correction factor for non-specific global change in RNA production, the inherent change in real-term expression will be accounted for, giving you values for relative change in real term expression of your gene of interest resulting from specific action of the treatment.
Also see http://www.clinchem.org/content/55/4/611.full
- Badges
- Science topic
- Similar topics
- Biological Science
- Molecular Biology
- Biomolecules
- Nucleic Acids
- RNA