mRNA vs protein correlation issues, breaks central dogma - (Nov/16/2015 )

I had started with information gathered from a microarray which indicated that treatment with my compound increased expression of a certain gene (MAFB) about 3-fold.  When I verified with RT-PCR, I noticed I was consistently getting ddCt's of 2 (approximately 4-fold increase) in each individual experiment (I was having issues with getting consistent dCt's so I can do statistics, but that's another story), and my NTC did not .  However, Western blot for MAFB revealed that the protein level was not changing at all.  I have always noticed good qualitative correlation between mRNA and protein (that is, directions are the same, even if magnitudes aren't).  I also understand that there are some papers out there that note a poor correlation between mRNA and protein expression.  However, I did not expect this.  Has anyone else ever seen such poor correlation, and what explanations could there be for that?  I had previously found another protein whose mRNA was upregulated in RT-PCR to be upregulated in protein expression, so general translational issues don't seem to be a problem.

EDIT: OK, I think I found one possible solution.  Papers from PNAS (2006) and Nucleic Acids Research (2012) point to miR-130a as a known translational inhibitor of MAFB that does not immediately induce mRNA degradation through (as opposed to miRNAs that do by utilizing Argonaute).  The question then becomes whether miR-130a becomes upregulated in tandem with MAFB upon treatment with my compound.  Of course, my PI could then argue that it might not be a question worth considering, as it could take us in too many different directions.  What say you?  I'm beginning to think a microarray analysis that does not take miRNAs into consideration is sorely lacking in information.

I think it is good start to expect mRNA to correlate with protein, but not as a rule. 

I don't think it breaks any "dogma" (as much as the simple DNA->DNA->RNA->protein is not as simple and single-directional anyway and even Watson opposed calling it dogma in the first place), since it only suggests direction, not perfect correlation.

There are many different reasons for protein level or rather his activity not correlating with mRNA. I've seen mRNA degraded, expression silenced, not-silenced when it should be, regulated by translation, by miRNAs, (mis)regulation by folding, by inactivation, by (de)phosporylation... and not last by degradation of the protein, I've been working with a system where this was a primary way of regulation the transcription factor activity.

So, I just think there are so many solutions, and the fact that mostly we pretend that expression is all, doesn't mean it is always. Each system must be validated first by several means to asses the possible regulation that plays role (funny thing, these may even be different in different cells or conditions).

MAFB is a transcription factor, so is it possible to detect changes in its target genes? At least that should be functional output of a change in it's expression. Of course may not be, but you can see further if it's time to look for different regulatory mechanism.

Also, seems it is involved in differentiation of myeloid cells. Interesting thing I read about this was, that many lineages are defined by a "balance" of transcription factors securing commitment to a difference lineage, domination of one lead to suppresion of the other and vice versa  (this happens in myeloid x lymphoid and this gene of yours should be part of ), while balanced expression was a marker of an earlier developmental stage. Maybe this may also play some role, if this is such a balancing factor, the regulation may be more complex.

See for example:

http://www.nature.com/nri/journal/v7/n2/full/nri2024.html

http://www.ncbi.nlm.nih.gov/pubmed/11532392