What is the RNA:Protein ratio for ONE gene over different RNA levels

Question

[sscholz8]

I've spent a long time trying to find the basic answer to this question:

What is the RNA:Protein ratio for ONE gene over different RNA levels?

In other words, if an inducible gene, such as lacZ expresses different amounts of RNA depending on the level on induction, how would the B-galactosidase protein concentration change with varying RNA concentration? Would the relationship be linear, quadratic, log?

I'm looking for experimental evidence. Thanks!!

Answer 1

[Devon Ryan]

There is no answer to your question, unfortunately. There's added regulation at the level of translation (not to mention protein turnover), so you will never find a single RNA:Protein ratio that's accurate. I suspect that this will still hold true for lacZ, since its turnover could easily vary by cell-type/experimental condition. It's likely that you could make an experiment-specific calibration curve, but that's likely the most you can hope for.

[sscholz8]

Thanks for your response! Yes, I agree that post-transcriptional and post-translational regulation may make this a question with a different answer on a gene-by-gene and condition basis. However, even for a single gene and condition in E. coli, is there ANY data for RNA and protein concentration relationship? The simplest case I can think of is just different induction levels on the lacZ, the amount of RNA and Protein resulting. I'd really love to see the protein concentration as a function of RNA concentration for just one gene/condition combination!

Answer 2

[mikhail.shugay]

If you assume the simplest model for mRNA translation, that is dp/dt = k * m - d * p, where k is the translation rate and d is protein degradation rate, the steady-state relation will be p = k * m / d. You can then look up those constants.

I suggest you have a look at this model: http://www.ebi.ac.uk/biomodels-main/BIOMD0000000065. Then you should let d/dt -> 0 (steady state) and resolve the equations (at a glance they'll be linear so no problem here) and get a direct relation of protein / mRNA

Update:

I was wrong about linearity of equations in the corresponding paper here. Those guys study feedback regulation, that often leads to non-linear equations, switch-like behavior and multiple steady states.

When bacterias are not growing, it has a single steady state. In this case k = 1.66 x 10^-2 1/min and d = 8.33 x 10^-4 1/min, so the ratio p / m = 1.99

When bacterias are growing rapidly, are 3 steady states in their model, see Appendix C Table 3 with mRNA and protein concentrations given in columns M and B respectively. One of those steady-states in unstable, i.e. minor fluctuations cause system to switch to one of two other steady states.

[sscholz8]

Thank you for the response! I agree with this model but, as you say, it is greatly simplified. The constants are not true constants - translation rate may vary as a function of RNA concentration - and this is what I am really trying to ascertain.

There is reason to believe that translation rate may vary depending on RNA concentration even beyond simple regulation of translation initiation. For example: In E. coli, ribosomes are localized to the poles of the cells. Therefore, a region of DNA in the nucleoid that is more highly transcribed may physically relocate to be proximal to the cell poles due to coupled transcription and translation. This would likely lead to a non-linear protein expression as a function of RNA concentration.

Of course, I haven't been able to find any data to support one view or the other so far!

[mikhail.shugay]

I've updated my reply, of course in this paper there is another layer of complexity added.. What you're saying, those spatial things, add even higher level of complexity, leading to system of partial differential equations (PDE). PDE models would lead to extremely beautiful and complex pictures, even the simplest of them.

However those models, even the simplest one, are in agreement with data in certain situations when constants are being fitted. And I think you can safely use those ratios if your know what your case is (e.g. dormant bacteria). This basically depends on what question you're trying to answer..

[sscholz8]

I appreciate your feedback! I'll take a closer look at the paper you've posted.