Are there examples of new ORFs in SARS-CoV-2 created by mutations?

The ORFs should not be present in the reference virus, but they should occur in a lineage occurring in the wild (at best, being part of the definition of the lineage). Of course those ORFs will somehow overlap with the well-known ORFs of the virus, potentially frame-shifted.

Bonus: Is there evidence that a protein is produced from the newly emerged ORF?


3 Answers 3


There is surely better but



  • An so on, in fact most "stop codons" in alternative frames seem to be mutated in some sequences. Of course most of those "new ORF" are probably not detectably translated. Introduction of new TRS has more chance to produce new subgenomic mRNA whence translation of new proteins.

  • If you only consider the consensus sequence of each pango lineage then you'll probably won't find much convincing event of this kind, except for the introduction of TRS in N which did happen in B.1.1, P.1, B.1.1.7,

    as well as a deletion causing the merging of ORF7b and ORF8 in B.1.1.318


The answer to this is both yes and no.

The reason I say yes and no is that SARS-CoV-2 exists as what is known as a quasispecies. Quasispecies are the population of virions that exist in an infection, with each separate virion containing a genome with its own mutations as a result of natural error rate of the RNA dependent RNA polymerase. When we sequence a virus and call the result its genome, what we are actually looking at is a consensus sequence, not the individual sequence of each virion that went into the sample from which the genome was derived.

Now, the problem with a quasispecies is that not all the virions produced in an infection are actually viable and can go on to produce more virions. In some cases there will be defective genomes, some more will have defective capsids or defective packaging of the genome and so on. These together make up a series of defective particles with a range of different terms depending on how they behave in an infectious process.

In many cases there are defective interfering particles, which miss parts of the genome or have genomes that are otherwise defective but are capable of infecting cells, but not capable of producing infectious particles themselves, and can compete with functional particles for infection (hence the interfering part).

In some of these described particles and within the sphere of the quasispecies will be virions that contain new ORFs. How those ORFs behave in an infection is very difficult to say as they are very hard to study. In most cases these new ORFs will cause a failure of some other (already) critical function and produce non-viable virions, but without examining any potential new ORF functionally in a lab we can't be certain that a potential new ORF does anything or indeed what the product might do.

For an estimate of the coding capacity of SARS-CoV-2, you can look at this study (1), published in Nature, which revealed that there probably are unknown ORFs within the virus. However, I suspect that these were there already and are not newly evolved as such, just newly identified.

  1. Finkel et al., Nature 589, pages 125–130 (2021)
  • $\begingroup$ I don't really see how it is related to the question. Please take a look at the ORF8 deletions in SARS, in particular the 29nt deletion present in the main epidemic strain but not in some earlier strains, causing a much shorter ORF8 protein, named ORF8a, and a new ORF8b ORF overlapping the end of ORF8a in a different frame. In the early strain ncbi.nlm.nih.gov/nuccore/AY390556.1 the long ORF8 is called ORF10-11 and in the epidemic strain ncbi.nlm.nih.gov/nuccore/NC_004718.3?report=genbank there is ORF8a and ORF8b $\endgroup$
    – reuns
    Commented Oct 21, 2021 at 15:51
  • $\begingroup$ @reuns - Why don't you post that as an answer yourself. The reason I posted this is because they genuinely do produce new putative ORFs all the time because of the quasispecies thing. However, a new ORF does not necessarily produce a functional protein or may alter the function (e.g. decrease expression) of a functional ORF, so you end up with non-functional or less functional particles that don't lead to anywhere in evolution terms. $\endgroup$
    – bob1
    Commented Oct 21, 2021 at 18:49
  • $\begingroup$ @reuns ORF8a/b was identified in SARS-CoV (the original one) so I don't think this is a new thing at all for CoV-2, which is what the question was about. $\endgroup$
    – bob1
    Commented Oct 21, 2021 at 20:31
  • $\begingroup$ I appreciate this answer although I hadn't really asked for it, because it broadens my view on the virus genome of SARS-CoV-2. I am still interested in more answers, describing situations for SARS-CoV-2 like the one exemplified by @reuns' comment for the original SARS virus. $\endgroup$ Commented Oct 22, 2021 at 12:35

Right now, there is a new preprint by Harriet Mears et al. on biorXiv with interesting new ORFs in SARS-CoV2 opened by some mutations.

  • The mutation N:R203K N:G204R found in pango lineage B.1.1 creates a transcription regulatory sequence (TRS-B) that triggers the expression of a proteine starting with N:M210 to the end of N. This proteine acts as an interferon antagonist. This ORF is named N:iORF3 by the authors.

Other mutations create new ORFs in nsp16 (coded by ORF1ab) and in the Spike gene.


You must log in to answer this question.

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