Mitochondria have their own genome; the parent is talking about editing that genome which has several fairly unique challenges above normal chromosomal DNA editing.

That being said, there were recently some breakthroughs in this area: https://www.nature.com/articles/d41586-020-02094-x

Unfortunately we don’t have anything beyond editing yet :(

We’ve known TALENs can edit for a while, but that breakthrough made it a lot more efficient.

I guess I don't understand the distinction you're making between "editing" and "genetically transforming"?

You can make a base pair edit, but you can’t, for example, add in GFP

Add new DNA into the human mitochondria. If you could do that, you unlock a massive amount of science around how mitochondria function, kinda similar to “what I cannot create I do not understand.

From an engineer’s perspective, mitochondria are kinda the equivalent to virtual machines running on normal computers, as little “virtual cells” within cells. Their genomes are so stripped down that you can do some wild things with them.

> If you could do that, you unlock a massive amount of science around how mitochondria function

Note that 99% of the proteins contained in mitochondria are [coded] in the nucleus, only 1% is [coded] in the mitochondrial chromosome. It's extremely small, only containing 37 genes of which most are related to translation (but of course translation requires far more, e.g. elongation factors).

https://www.nature.com/articles/nrm2959

I guess that was a bit of a hyperbole. You can learn more about the mitochondria genetically engineering the nucleus, but the main problem is that there still is science to be done on the mitochondrial genome that is "locked" down by not being able to transform it, especially when it comes to stuff like how translation works within those mitochondria.

I get what you're saying, but I think you have a typo in there. Most of the proteins in the mitochondria are coded by nuclear DNA, but the proteins themselves end up in the mitochondria.

Thanks! Edited.

The number of genes encoded in the mitochondrial chromosome, 37, reminds me of the 29 states in Von Neumann’s universal constructor [1].

[1] https://en.wikipedia.org/wiki/Von_Neumann_universal_construc...

My answer to this question was ‘A John Scalzi Novel.’

But IIRC weren’t mitochondria subsumed by another cell for ATP output forever ago in our evolutionary development?

yep, and you can even make synthetic endosymbionts! (pretty new science too!)

Can we do this with nuclear DNA?

You can. if you want a new piece of DNA stably inserted into the nuclear genome, you have a couple of options:

https://en.wikipedia.org/wiki/Lentiviral_vector_in_gene_ther...

https://en.wikipedia.org/wiki/CRISPR

These next two do the same thing as CRISPR, only they are older technologies that take you longer to generate the probes in the lab to target your spot in the genome. The real advantage of CRISPR isn't what it does, that was already out there, it's that it's cheap and fast (comparatively) and an undergrad can do the whole experiment.

https://en.wikipedia.org/wiki/Zinc_finger_nuclease

https://en.wikipedia.org/wiki/Transcription_activator-like_e...

The main human genome is more than 6 orders of magnitude larger than the mitochondrial genome, if you measure by counting base pairs.

Edit the principal genes of the mitochondria (for example, increase the membrane's oxygen and CO₂ permeability) to make it extract more energy per unit time.

I don't think we know how to do that yet. But there's a big tradeoff mitochondria have to make in terms of either being more efficient at turning sugar into ATP or being more resistant to damage. Since most people in the first world have an easy time getting enough calories to survive it would make sense to lean into the resistance to damage side of things.