Why don't mRNA levels measure protein synthesis rates?

Question

[Qroid]

I naively think of mRNA as part of the signal to produce a protein. For my naive thinking I have a naive factory analogy, where the mRNA is an order for that factory to make its product (a protein). The order is more transient than the product, i.e. the mRNA degrades relatively quickly and the protein goes on to persist for some time. In this model the amount of orders is literally the rate of production, and measuring the amount of orders in the factory is a measure of how fast the factory is producing. So by analogy I've thought of mRNA levels as correlating with protein synthesis.

What's wrong with my simple model? Why don't mRNA levels measure protein synthesis rates? Or can you measure protein synthesis with bulk rnaseq, but it's just super noisy and so not very good?

I know I'm vastly oversimplifying. Things like synthesis efficiency and degradation rates of mRNA and protein are all super variable. I accept that relationship between mRNA levels and synthesis rate should be noisy. But still, shouldn't a measure of the signal to produce a protein be a measure of that protein's synthesis rate? I brought this up off hand in a meeting and got strange looks.

Answer 1

[LChart]

Methods which profile translating RNA (e.g., by precipitating ribosomes -- Ribo-seq, ribo-lace, riboPLATE-seq) are measuring the instantaneous protein synthesis rate; just as PRO-seq/GRO-seq measure the instantaneous RNA synthesis rate. Papers profiling both total transcriptome and translatome (for instance https://www.nature.com/articles/s41598-019-47424-w) show correlations of ~0.85.

In other words, your model is pretty correct, and mRNA levels generally measure synthesis rates.

However once you start comparing mRNA to protein abundance, the correlation drops fairly substantially.

Edit: I should clarify that here "rate" is in terms of molecular species and not time: "per transcript" not "per millisecond"

Answer 2

[Istvan Albert]

mRNA does correlate with protein levels - the whole point of RNA-seq is precisely that.

If mRNA were not useful for that purpose, we would not perform RNA-seq experiments in the first place.

The question is more about how tightly are the two coupled. Evidently, if there is no mRNA, then none of the corresponding proteins could be produced, so that too, is valuable evidence.

Thus, in the end, the issue is more about how efficiently are abundant mRNAs translated to proteins.

That in itself is a separate field of research.

For some genes the correlation is very strong, for others much less so.

[Istvan Albert]

as for a summary:

the biggest disconnect between naive model and real biology is that it is ignoring how both mRNA and proteins can degrade at vastly different rates, and that several other biological phenomena can alter the rates of translation

[Qroid]

Thank you. I know RNA-seq correlates with protein amount, e.g. bulk rna-seq and mass spec have a median correlation of R~0.4. The question was about correlating mRNA with protein synthesis rate.

[Istvan Albert]

For what it is worth, in my opinion, in the naive wold view abundance and synthesis rates are identical concepts.

The naive view operates on a 1 to 1 correspondence: every transcript leads to a protein, and everything happens nearly instantaneously. This would be a simplified, naive model.

When we distinguish between the two concepts, there is no "naive thinking" anymore. The fact you want to account for different rates, makes the view non-naive :-=)

[Qroid]

Always nice to be labeled "non-naive"! :D

I know there's some mixing of rate and amount, i.e. a faster rate means more protein and a lower rate means less protein. My confusion is that rnaseq seems intuitively closer to a measure of the rate than the amount - you're measuring a transient signal about translation occurring. So, while it seems like an obvious rate signal, I don't see the rnaseq-rate connection discussed anywhere.

To me this explains some of the disconnect between rna-seq and mass spec, i.e. former is rate and the latter is amount. But the strange looks I got in the meeting and the fact I don't see it discussed like this in literature or elsewhere makes me think I'm confused. I'd love to find a paper or some data or something, so I can push back on those strange looks!