It's interesting that they did not just undo the inherited disease. I assumed that, especially with Sickle Cell where we have a good understanding of how it works, they would go into Chromosome 11 and put it back how it "should" be with CRISPR. But instead they apply a workaround, ensuring continued fetal haemoglobin production.
The article does not mention whether that's because putting Chromosome 11 back with CRISPR is harder, or whether for some reason that wouldn't fix the problem.
That's also what struck me as odd, but having maintained legacy systems for a long time I can see why it's sometimes preferable to use a proven if inelegant workaround than going in and try to fix the actual defect, especially if you have limited debugging options.
It actually follow a common pattern in evolution. Doing big changes are hard, difficult and risky. Tinkering and doing small changes over time works for most things and any legacy system, as long they are not directly harmful, can remain as a byproduct. There is even a word for it in biology called spandrel.
Analogies to IT and development aren’t that appropriate here tbh.
This isn’t a legacy system, if you want to keep the IT analogy this would be failing over to a different system or replacing an existing solution that doesn’t work with a competitive alternative rather than fixing all your bugs.
It is much easier to wreck things with Crisper (in this case the regulatory region that turns of fetal hemoglobin) that to really go in and alter one or a couple of specific base pairs. What they did is nice but the latter would be the holy grail.
If the holy grail is only a few years away, then yes. Normal CRISPR editing rely on non-homologous-end-joining (NHEJ) which is not an absolutely exact process. However, there are now base editors (look up prime editing) that edit a single position with very high specificity. Takes a few years from basic research to the clinics but they will be everywhere soon.
"Holy grail" not because of how it would work on this particular disease, but because it would allow basically limitless editing of any gene in any living human, allowing us to fix all hereditary genetic diseases, and do a whole lot of other things besides.
One reason may be that a general fix that produces good hemoglobin is much more efficient than attempting to fix the original problem. People can have a near-limitless number of mutations that lead to defective proteins, so creating fixes for all those "bugs" can be very expensive. A general fix that works around the "bug" is cheaper, consistant, and more scalable.
Another way to look at it: Would a per-patient fix be possible? Maybe. Is it worth debugging and fixing everyone's crappy DNA to do that? Probably not.
That’s true even if the mutation is the same it’s hard to actually fix it, while reactivating or over-expressing a gene that already exists and works is much easier.
For this specific case it’s also better because lesser coverage might still produce sufficient results since fetal hemoglobin out competes adult hemoglobin.
I think in general CRISPR would be patch over rather than a point fix.
The article does not mention whether that's because putting Chromosome 11 back with CRISPR is harder, or whether for some reason that wouldn't fix the problem.