DNA synthesis in E. coli is 20x faster than RNA synthesis at 1000nt/s vs 50nt/s. (Mirkin'05)

I find that perplexing since DNA polymerization has better proofreading than the RNA variety, which requires extra time as you need to backtrack, excise, and replace. Also, in bacteria there's almost no splicing, so that doesn't slow you down. And the 1000nt/s is a single replication fork. We are not talking about several replisomes together clocking 1knt per second.

Is there an explanation of this or, in the absence of evidence, hypotheses?

  • 1
    $\begingroup$ (shamelessly unreferenced) hypothesis: DNA synthesis needs to be faster, as you need to replicate huge amounts of DNA relative to a single RNA transcript. I believe that there's only a single origin in prokaryotes (compared to many in eukaryotes), so if there's a selective advantage towards growing/dividing/replicating faster, it makes sense that there's a difference in speed $\endgroup$
    – Luigi
    Commented Mar 22, 2016 at 12:12
  • $\begingroup$ It makes sense that replication should be fast, but that in itself doesn't mean replication should be faster than transcription. $\endgroup$
    – David
    Commented Mar 22, 2016 at 13:41
  • 1
    $\begingroup$ Good question. As Luigi was suggesting, I would guess it has to do with the amount of nucleic acid that must be synthesized. RNAs are tiny compared to DNA, so there is little selection pressure for speed. Also, multiple RNA polymerases can initiate simultaneously, effectively increasing the synthesis rate. $\endgroup$
    – Roland
    Commented May 22, 2016 at 5:25
  • $\begingroup$ To throw in another hypothesis: note that mutations in DNA tend to occur when RNA synthesis and DNA replication meet heads-on (antisense) ( nature.com/nature/journal/v535/n7610/full/nature18316.html ) Ultimately this would suggests that - if the speed of DNA synthesis was similar to the speed of RNA synthesis - there would be more head-on encounters and mutations $\endgroup$
    – tsttst
    Commented Jul 20, 2016 at 23:01

2 Answers 2


I wouldn't call them hypotheses, but the question is intriguing, and, as it seems to be ignored in the literature, I'll make a couple of suggestions.

  1. Perhaps it has something to do with recognition of termination signals in relation to selective pressure for speed in the two processes.

Perhaps if elongation of the RNA chain were any faster it wouldn't be able to respond to the rho-independent termination signals which are thought to be specific stem–loops in the DNA that cause the RNA polymerase to pause and fall off the DNA. The speeding RNA polymerase might disrupt the loops instead. The termination system for DNA replication may have evolved to be much more robust so as to be able to bring faster elongation to a stop. (Note: my suggestion of different ‘robustness’ in termination is not based on any data, it's pure speculation.)

If this were true, it would beg the question of why transcription hasn't evolved a more efficient termination system in concert with the a evolution of a smarter RNA polymerase. Perhaps the answer here is that there is no selective pressure for faster transcription — the rate is clearly sufficient to supply the needs of the cell — whereas the rate of DNA replication determines the time between cell divisions and thus the rate of growth, which is subject to selective pressure.

  1. Perhaps it has something to do with error frequency in relation to selective pressure for speed in the two processes.

First I'll make it clear that I think the proofreading argument in the question is a bit of a red herring. If the rate replication is 20x as fast as transcription, the occasional backtracking will not have much effect on the overall difference. However errors and proofreading could be involved in a different explanation.

RNA transcription does not have proof-reading. The length of RNA transcripts is such that the cell can tolerate the errors that occur. However if the speed transcription were increased it is reasonable to expect an increase in error frequency, which might be detrimental. The current rate of transcription may be a trade-off between efficiency and accuracy. Of course one can imagine a proof-reading system for transcription evolving if it were necessary, but, as discussed above, if the rate of transcription is adequate there would seem to be no selective pressure to produce.

  • 2
    $\begingroup$ Processivity may be an additional factor. I am not sure about the processivity of RNAP, though. $\endgroup$
    Commented Jun 20, 2016 at 17:06
  • $\begingroup$ I want to add an observation. Prokaryotes differ from us in that translation often begins before transcription ends. Prokaryotic translation links 17-21 aa/s, which means 51-63 nt/s are read in mRNA (Wiki). This is within the range of 40-80 nt/s of prokaryotic transcription (BioNumbers). Maybe there is a reason why they match. $\endgroup$
    – user38945
    Commented Apr 9, 2019 at 17:15
  • $\begingroup$ Sorry. I just read in (book.bionumbers.org/what-is-faster-transcription-or-translation) that single-molecule microscopy has shown that at least in E. coli, most translation is not coupled with transcription. $\endgroup$
    – user38945
    Commented Apr 9, 2019 at 17:21
  • 1
    $\begingroup$ @Jagoe — No need to apologize. The source you cite was not known to me and is very pertinent to this kind of question. Like many others, I had thought transcription and translation were coupled in prokaryotes. Selective use of images perhaps. (But I wish they wouldn't keep referring to The Central Dogma all the time.) $\endgroup$
    – David
    Commented Apr 9, 2019 at 20:00

This is because of the fact that DNA polymerase does the fast process that is 1000 nu/sec as compared to the RNA polymerase that 1000-2000nu/min because this process is done without discrimination and also DNA pol has exonuclease activity

  • $\begingroup$ Welcome to Bio. Could you expand your comment into a full answer by adding references and additional background to it? You might wish to go through the help center on answering questions. There it will become apparent that the SE network is not the place to answer posts with one-liners. -1 $\endgroup$
    – AliceD
    Commented Oct 4, 2017 at 7:59

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .