mRNA versus protein expression. Why inverse correlation? - (Sep/28/2009 )

Hi all!
I test gene expression by qPCR followed by Western to test protein expression. I get inverse correlation with several proteins (down-regulation on mRNA level and up-regulation on protein level and opposite: up-regulation on mRNA lvel versus down-regulation on protein level). Is it as it should be or is something wrong with what I do? Thank you for answering!

katenkak on Sep 28 2009, 07:11 AM said:

Whether or not these data are accurate may depend on the genes or your protocol. There are numerous situations that could be occurring here.

1) you are not doing the qPCR properly
Are these validated primer sets, or did you design them yourself?
Are you targeting all splice variants of your genes of interest?
You may have non-specific amplification, primer dimers, or may be detecting splice variants not seen in the Western (or qPCR)
What is the amplification efficiency of your primer sets (control and reference)?
Approximately equal amplification efficiency between both the target and reference gene primer sets is necessary for accurate
determination of relative expression.
What is your reference gene(s) and are you sure the expression is stable?
Reference gene expression must be stable across treatments. Using multiple reference genes enhances accuracy.
How do the melt curves look?
You should have a single peak or comparable amplitude to other primer sets.
How are you calculating your relative expression?
You should use the Pfaffl equation to take into account the efficiency rather than the delta delta Ct method, which is less
accurate.
What are the Ct values?
Are the Ct values for these "strange" genes reasonable or are they too high (above 35, amplification of noise) or too low
(below 15, too much template)?
2) you are not doing the Western properly
How are you assessing changes in expression?
Are you calculating the changes in intensity using imaging software or eyeballing it? If you are using software, are you sure
that you are using it properly, keeping the size of the measurement box consistent between all readings and normalizing
against background intensity AND a loading control?
Are you normalizing the expression against B-Actin expression or other protein loading controls?
Are you sure you are looking at the correct band?

3) OR your data are correct
mRNA levels and protein levels certainly do not always correlate.
Downregulation of mRNA concurrent with upregulation of protein expression may occur when a protein half-life is increased due to stabilization--components involved with the protein's normal turnover may be disrupted or the protein may be stabilized through protein-protein interactions. Upregulation of mRNA concurrent with downregulation of protein expression is actually common among proteins that are downregulated by ubiquitination following a signaling event. These proteins are targeted for degradation via the 26S proteasome while their mRNA levels remain the same, or, in some cases, increase.
Basically, all of this boils down to: how sure are you of the accuracy of your experimentation, and, thus, your data?

katenkak on Sep 28 2009, 07:11 PM said:

It is also probable that you are on the right track. Initially, the protein level may be low, thus the cell is actively transcribing the corresponding gene in order to produce the needed amount of protein. As your protein of interest accumulates, it may act as a repressor of transcription, signaling that there is no need to synthesize the protein at a high rate since it is already present in certain amounts. Then lastly, when the required protein level is achieved, there would be no need to transcribe the gene and transcription can be totally shut down.

Thank you!

It is also probable that you are on the right track. Initially, the protein level may be low, thus the cell is actively transcribing the corresponding gene in order to produce the needed amount of protein. As your protein of interest accumulates, it may act as a repressor of transcription, signaling that there is no need to synthesize the protein at a high rate since it is already present in certain amounts. Then lastly, when the required protein level is achieved, there would be no need to transcribe the gene and transcription can be totally shut down.