https://berthub.eu/articles/posts/reverse-engineering-source...
I am not a scientist but my understanding is that the immune system targets only what it is trained against. It doesn't target anything which is non-self without being trained beforehand.
IMO if it was otherwise there would be no need for vaccine.
So I think that it is quite feasible to make a cell to produce a new protein, for example Zolgensma [0] works that way: It creates a new gene (hence a new protein) to replace a deficient one.
Lymphocytes, or T cells & B cells, are immune cells which have receptors on their surface that recognize a specific, tiny chunk of protein. For any given lymphocyte, the type of receptor on its surface is fixed for its life and that receptor is able to recognize exactly one distinct protein chunk.
Through some very cool mechanisms, very early on in life we all develop a massive number of lymphocytes, each of which recognize different protein chunks. I've not read any research which quantifies the scope here, but it's not unreasonable for the sake of a thought exercise to the assume that at one point we all have a lymphocyte receptor repertoire that's capable of recognizing every conceivable natural protein. Through what I consider one of my top 10 most jaw-dropping biological mechanisms [0], we cull the population of lymphocytes that recognize self, leaving us with a cell population capable of recognizing and responding to every non-self protein conceivable. No training is needed here, each of us at this exact moment has several T cell and B cell populations ready to recognize proteins produced by the next novel pandemic causing virus, whatever future protein mRNA vaccines might make, and whatever proteins these future mRNA therapies might produce.
What you refer to as training is probably more like immunological memory, which allows for a ramping up of the immune response on a quicker timeline. We give vaccines where it's generally not safe (or survivable) to wait out an effective immune response, because the disease causes so much havoc in the meantime. This doesn't really apply for the introduction of novel, useful proteins.
Your point about Onasemnogene abeparvovec is a very interesting one. I'm truly only guessing here, but people with SMA almost universally produce functional protein via SMN2, but very little of it. It's not enough to serve its function, but is perhaps enough for effective self-tolerance. I'm also not entirely sure about the timeline of self-tolerance development, it's possible Onasemnogene abeparvovec is given young enough to allow effective tolerance development!
Kinda. Programming but more like a browser JS.
You're not changing anything (your DNA) and whatever you're making is temporary and limited (especially because you can't have anything too complex or long as mRNA without that breaking up).
Anyway, AFAIK the mRNA breakthrough wasn't with the mRNA itself, but rather the delivery vehicle to avoid said immune response. I don't think your example of cystic fibrosis, and generally chronic illnesses, are prime targets for mRNA based therapy, as the genetic defect isn't causally targeted by mRNA tech. You would force all cells (with non-discriminatory vehicles) to produce the missing protein repeatedly. Much better target for gene therapy.
I don't think people see the future of mRNA in substituting into the delicate machinery of the cell continuously. I think the possible therapy target are of the type raw, simple one-hit wonders. Very much alike traditional medication approaches, with the twist of excellent drug delivery avoiding extracellular pharmacokinetics, e.g. vitamin C has very different, even antagonistic roles in the intra- vs extracellular space, and of course liver-fistpass and immunological clearance. mRNA tech is not gene therapy (although, there is of course effective overlap).
https://www.businessinsider.com/new-hiv-vaccine-could-be-mod...
So, that's great for vaccine production. Not so great for other diseases. Cystic fibrosis, for example, the body fails to make one specific protein. We (probably) can't just program it to produce that protein, because it would also train the immune system to target that protein. The best non-vaccine targets are probably cancer, where you want to get the immune system revved up against a tumor.