They are manufacturing errors more than they are mutations. Since the vaccine and virus use the same machinery to make the proteins, they both result in proteins with errors.
> the viral piece of DNA [of Adenovirus based vaccines] - deriving from an RNA virus - is not optimized to be transcribed inside of the nucleus. ... Thus, it could well be that the Spike open reading frame of SARS-CoV-2 is potentially disrupted by arbitrary splice events when transcribed inside the nucleus. Most, if not all, of these undesirable splice events would produce shorter protein variants, ... leading to soluble Spike protein variants.
So the error happens during DNA to RNA transcription in the nucleus. The Covid-19 virus is RNA based so how are the same splicing events occurring when it replicates itself in our cells?
Edit: There is further mention in the paper of predicting and then detecting potential splice sites in the "wildtype Spike" (which I think means the RNA of the spike protein of the actual virus) and they found several potential splice sites, but not as many as those predicted / found in the AZ and J&J vaccines. It's still unclear how those splice sites could be activated if the virus replicates entirely outside the nucleus but there is probably some information I'm missing here.
mRNA is the blueprint to a protein. You can get the mRNA into your body directly (super cool novel tech) or using a carrier virus that has the blueprint as DNA (eg. a modified chimpanzee cold in astrazeneca's covid vaccine). Here, the body first has to turn DNA into mRNA.
The researchers show that the latter approach gives rise to errors in prodcing proteins, errors that can cause severe side effects.
Note that this is a preprint awaiting actual publication in a journal... but surely itll pass review.
As far as I know the vector vaccines were tested on animals. Nothing was skipped.
The issue is more that this side effect is so rare that it didn’t register even during the quite large phase 3 trials. This is to be expected. If side effects are rare enough you just won’t find them during trials. That’s why it’s important to monitor the vaccine rollout for all vaccines.
Also, your statement is highly misleading at best.
No idea what the US did but the EMA in Europe didn’t relax any of its standards and often drug testing is mostly slow because it’s costly. Here cost wasn’t really an issue and things that are usually done sequentially were done in parallel. That saves time.
If you want to use the VAERS data, the side effects, include death, are not really that rare, in fact, these are the most dangerous vaccines ever produced based on that data.
If that data isn't good enough, you'll have to provide data that is.
TLDR: mRNA vaccines turn their payload into proteins in the cytosol.
Conversely, DNA vaccines (astrazeneca, JNJ) must first use the nucleus to turn their DNA into mRNA. This incurs so-called post-~translational~ (edit: post-transcriptional) modification to the genetic material prior to protein synthesis.
As it turns out, some of these unforseen modifications create soluble spike protein fragments that bind to epithelial cells. This in turn causes inflammatory, and coagulant, activity, which leads to the lethal side effects that are sometimes observed.