Initiation of translation precisely at the AUG start codon may be more of a suggestion than a rule. With fairly recent discoveries regarding inexact transcription, this isn’t really surprising, but still interesting to see.
The Cell article can be found here:
https://www.cell.com/cell/fulltext/S0092-8674(19)30499-4?_returnURL=https%3A%2F%2Flinkinghub.elsevier.com%2Fretrieve%2Fpii%2FS0092867419304994%3Fshowall%3Dtrue
“Because of our study, we can now ask very important questions: what do all these new proteins do? Do they have important functions in our body or are they waste side-products of translation that can damage our cells?”
I go for option B, though “damage our cells” seems out of left field.
I have read parts of the paper, and the technology itself is really interesting. However, they use a cancer derived cell line, and I don’t think they can extrapolate from these experiments to normal cells in a functioning organisms. Cancer cells are known to have higher rates of dysfunctional transcription and translation, so that has to be weighed into the results. I would have liked to have seen them look at normal, non-immortalized cultured cells, or perhaps at something like zebrafish. I think they need a more relevant biological setting before they can start to make general statements about out-of-frame translation.
I’ve been weary of this kind of “far more complex than previously thought” hype often seen in popular press articles on science. They say that about almost anything. It invariably turns out nobody is able to point out who used to think it was far simpler than it is.
And I’d lean towards it being noisy, as opposed to functional translation too most of the time we see heterogenous translation.