Lately, I've come across a number of papers where the control for AAV vectors (transgene or shRNA) was simply PBS. These were published in top journals - some of them preclinical research in prep for trials.
Is this now accepted practice? I would think empty AAV vector (better mutated transgene) or mutated shRNA AAV would be an optimal control?
@Simon N Waddington
PBS (or approriate excipient) seems to be the control which is acceptable as part of a preclinical data package i.e. acceptable to reviewers, regulators & industry. We've often agonised about other controls but their advantages over excipient alone. E.g. AAV is structurally stiffened by the enclosed DNA, and so full capsid will have different biological properties to empty capsid. & e.g. mutant transgene/shRNA might have unrelated & unexpected function.
Really? Regulators accept this? Wow
I would have thought the most appropriate would be a non-functional transgene (if this can be achieved with say some point mutations) or at a minimum a reporter gene? A scrambled (reshuffled sequence of the targeting oligo) control is absolute minimum standard in oligonucleotide therapeutics and needs to be used in in vivo experiments too.
Is nobody concerned about innate immune response to the vector backbone? How do you control for effects of the transgene versus vector?
At least in the shRNA field that's definitely an issue:
Article Bridge AJ, Pebernard S, Ducraux A, Nicoulaz AL, Iggo RInduct...
Could anyone please recommend any papers where they've looked at these issues? Shown that these aren't relevant? Or regulator guidelines with a definite acknowledgement that that is accepted?
With most gene therapy targets in the clinic and soon to enter the clinic, the preclinical experiments are demonstrating efficacy - efficiacy to express the missing gene, and then correct/prevent the pathology. And the effect is a one-off, and fairly brutal sledghammer biology. The only thing that would be expected from a scrambled control is a deleterious effect of unpredictable consequence.
For the clinical trial, one has to have in place release assays to ensure the gene therapy product expresses what it is supposed to express, and tox to ensure that the batch is, itself, not toxic.
I can easily imagine that for early, preclinical, probably even pre in vivo work (such as cell culture) a standard control virus might be considered smart (for example a reporter GFP construct or similar). This will confirm your construct works, and your vector alone does not cause adverse effects.
Of course, we already have a huge body of data that confirms this, so at this point "viral vector without any contents does nothing of note" isn't really even technically necessary to prove again.
Once you get into animal and human work, ethical and financial concerns become crucial: at this stage you've already shown the virally-conveyed transgene/construct works, while virus alone does not (cell culture).
What is the value of injecting mice with a pointless virus that you know WILL NOT work, just to show that the virus you have shown to be effective in culture is also effective in mice? The question you're asking is "does the in vitro effect translate to in vivo context?" not "does my negative control that does nothing also do nothing in an in vivo context?"
You are trying to confirm that your effect works in vivo, because ultimately that will be the scenario in human medicine.
When you get to actual clinical trials, this is even more important. Most viral vectors are one-shot deals: after treatment, you raise an immune response that prevents further viral therapy. It works first time or it does not work.
There is no way to ethically justify treating human patients with empty viral vector just to confirm it doesn't do anything: all you've done is exclude them from actual therapy options.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Plating