The S glycoprotein signal peptide in SARS-CoV-2 and the mRNA vaccine is encoded by:

           3   3   3   3   3   3   3   3   3   3   3   3   3   3   3   3
               !   !   !   !   ! ! ! !     !   !   !   !   !            

What is the motivation behind the CCA -> CCU modification between the virus and the vaccine?

  • 1
    $\begingroup$ I would argue that the "answer" you quoted is exactly the sort unresearched speculation this site is intended to avoid. ——— Codons "code" for amino acids not peptides. Peptides are composed of multiple amino acids — these terms are not synonyms! In addition, codon usage tables for humans are readily available and don't support the idea that use of CCU is strongly favored over CCA (1.75% vs 1.69%). $\endgroup$ – tyersome Dec 29 '20 at 0:07
  • 1
    $\begingroup$ @tyersome thank you for the feedback, great to know! (This is why such an answer should be posted as a separate answer and not within the question, so that people could vote on it.) $\endgroup$ – Franck Dernoncourt Dec 29 '20 at 0:10
  • 2
    $\begingroup$ Please re-read the tour and the Help. A question is a question. An answer is an answer. If the poster of a question thinks a particular answer is correct he can accept it. He may be right or wrong but he has made his judgment clear. Posting an answer as an appendix to a question completely negates the SE Model, which is why I have deleted it. $\endgroup$ – David Dec 30 '20 at 20:42
  • 2
    $\begingroup$ @David Thanks for removing it from the question. Franck originally posted it as an answer, but as it was comprised entirely of a quote from one source, with no original content, it was outside the referencing guidelines here: biology.stackexchange.com/help/referencing therefore I deleted it. Better to write your own answer, Franck, using your own words, referring to other content by reference rather than duplicating. $\endgroup$ – Bryan Krause Dec 30 '20 at 23:07

The article mentioned says:

"Why would you do that? As noted above, our immune system takes a very dim view of ‘exogenous’ RNA, RNA code coming from outside the cell. To evade detection, the ‘U’ in the RNA was already replaced by a Ψ.

However, it turns out that RNA with a higher amount of Gs and Cs is also converted more efficiently into proteins,

And this has been achieved in the vaccine RNA by replacing many characters with Gs and Cs wherever this was possible."

So codon optimisation - as Canadianer suggests.

RNA structural reasons I expect the sequence was also put through a folding analyser (for the mRNA that is) and that may be why not all the codons were optimised to C and G, and why there is an A -> U change as well.

If someone wants to use an online tool to show the folding differences, please feel free to add the graphics output to this answer.

It is also possible the designers didn't want the final sequence to exactly match the output of popular optimisation software, to reduce legal arguments about whether the software was used.

Differentiating between vaccine and viral sequences. Even though the vaccine uses a different base instead of uridine, when researchers check the sequences of mRNA in the future the base differences will make it easy to tell the source.


In case the mRNA keeps being found in someone's cells long term.

Testing for contamination in the production facility or research lab.

Testing efficacy in cell cultures exposed to virus.

Adding messages such as patent info.

  • 1
    $\begingroup$ Some of your speculations sounds quite plausible — would you please edit in some supporting references? Does psuedouridine really help the exogenous RNA evade detection? Also note that the question is about CCA->CCU so increased GC isn't relevant, but if you want to include it please put in supporting references. $\endgroup$ – tyersome Dec 29 '20 at 0:16
  • $\begingroup$ "reduce legal arguments about whether the software was used" why is that an issue? $\endgroup$ – Fizz Jan 3 at 14:59
  • $\begingroup$ A software company making optimisation software might make trouble, even if their software wasn't used. A bit like making sure your returned homework is very different from what you expect your classmates homework to be. $\endgroup$ – Polypipe Wrangler Jan 4 at 4:35

The other answer is a bit convoluted IMHO.

First note that as the article cited notes, that's not actually CCA -> CCU but CCA -> CCΨ; they only used ASCII for convenience. Furthermore, these three all code for proline as the article also notes.

Second, Ψ stands for pseudouridine and it's the golden discovery of Kariko and colleagues (which made mRNA vaccines [more] practical) that you need to use as many pseudo-Us as possible to ensure that the exogenous mRNA doesn't get destroyed prematurely and/or triggers downregulation of transcription.

Third, that's not even pseudo-U that's actually used in the Pfizer vaccine but N1-methyl-pseudouridine (variously abbreviated as N1mΨ or m1Ψ), which is even more "potent" in these terms than even pseduo-U. (According to a book on RNA therapeutics, m1Ψ was actually discovered and is patented by Moderna circa 2012-2013, before any academic publications.) Actually a 2017 publication found that independently of reduced immunogenicity, m1Ψ seems to be more active "by increasing ribosome pausing and density on the mRNA":

We show that in addition to turning off the immune/eIF2α phosphorylation-dependent inhibition of translation, the incorporated N1mΨ nucleotides dramatically alter the dynamics of the translation process by increasing ribosome pausing and density on the mRNA. Our results indicate that the increased ribosome loading of modified mRNAs renders them more permissive for initiation by favoring either ribosome recycling on the same mRNA or de novo ribosome recruitment.

And more research on the latter aspect (2019)

Global incorporation of different modified nucleotides often (but not always) markedly changed mRNA expression. This effect was seen across numerous synonymous coding variants of multiple proteins, in several different cell lines, and in vivo. m1Ψ generally gave higher expression than U or mo5U for the same sequence. Biophysical studies revealed that m1Ψ and mo5U have dramatically different and opposite effects compared to U (stabilizing and destabilizing, respectively) on overall mRNA folding, nearest-neighbor base-pairing thermodynamics, and secondary structure pattern as mapped by SHAPE.

There could be additional reasons, but these are probably marginal.

Now, canadianer asks an astute question below (which was not asked by the OP), whether it's (still) preferable to do UUU -> UUC (which normally it is) if m1Ψ is going to be globally substituted for U anyway. I don't actually know the answer to that; Pfizer or Moderna might have researched it, but I haven't found any publication on the trade-off.

The actual reason for the latter (especially since UUC is about 3 times as inefficient as UUU) might be that real goal is to slow down the protein folding there. The WHO doc does say that the protein is folded in a particular way "to ensure the S glycoprotein remains in an antigenically optimal pre-fusion conformation". In general, there's a trade-off between the speed of translation and accuracy of folding. (ΨUU is actually twice as efficient as UUU already, even with naturally occurring pseudouridine. In fact UUU is one of the most common places where one finds a naturally occurring pseudouridine.)

  • $\begingroup$ Then why change UUU->UUC, GUU->GUG, etc? $\endgroup$ – canadianer Jan 3 at 17:34
  • $\begingroup$ @canadianer: Frankly that one is even more mysterious because "The proofreading ratio of the UUC programmed ribosomes is one-third that of their UUU programmed relatives. From these results ribosomes programmed with UUC are apparently more prone to errors than those programmed with UUU" pnas.org/content/pnas/86/18/6888.full.pdf $\endgroup$ – Fizz Jan 3 at 18:09
  • 1
    $\begingroup$ I raised the question because all of the substitutions made in the coding sequence for the signal peptide are towards the most frequently used codons and tend to result in less methyl-pseudouridine in the final product, which seems to contradict your answer with respect to this specific sequence. I don’t know how they selected codons for the entire mRNA, but I wonder if for this sequence their selection was simply based on codon usage rather than trying to incorporate methyl-pseudouridine. Regardless, I found your answer quite informative. $\endgroup$ – canadianer Jan 3 at 18:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.