[echelon]

Why don't they use agrobacterium transfection, gene guns, or molecular techniques to directly transfer the resistance genes over? Do we not know what the genes themselves are?

I'm not criticizing their technique! Nor do I think I know better. This is amazing and valuable work, and I'd love to know more.

[space_fountain]

There is actually a parallel technique nearing the finish line that does sort of this. The gene they moved though is from wheat not another tree species, but still seems to give great blight resistance. Currently the barrier to it is entirely regularity. I believe it would be one of the first plants approved to just be released into the wild without something to prevent it reproducing naturally. The normal anti genetic engineering types are very opposed, but it seems likely it will get approved in the next few years.

https://www.nytimes.com/2020/04/30/magazine/american-chestnu...

[canjobear]

What authority can enforce approval or disapproval to release something into the wild? Suppose I own land next to a national park. Then can’t I plant whatever I want on that land, trusting that it will spread into the park?

[robwwilliams]

These are not GMO plants—just selectively bred plants. This is similar to almost all domestic animals and most plants.

[julienchastang]

There are two parallel efforts going on at present for restoring the American chestnut, a backcross effort and a GMO-based approach. This sub-thread is concerned with the GMO strategy.

[JshWright]

A buddy of mine has been involved in this project for years at SUNY ESF, and it's very exciting how successful it has been. There are dozens of test sites around central New York that have healthy (albeit very young) populations of American Chestnuts.

[notsapiensatall]

I've been wondering the same things for the past decade. It seems like it's very hard to get your gene into the right place so that it will be expressed at the right time, with enough consistency to make a stable cell line.

Also, it seems rare to find just one gene that encodes a property or compound. Usually you need several genes to form a pathway from a common precursor chemical to your target result. That pathway will be leaky, and it may take resources away from other important chemicals that the plant's lifecycle relies on.

It sounds expensive and time consuming - plants take a long time to compile, and you can forget about reproducible builds. Plus, once you have your GMO, you'll be expected to take care not to release it into the wild, and in today's world, a decent chunk of people will oppose your product because of their personal beliefs.

[culi]

Good question. I'd imagine something like blight resistance would probably not be a simple case of a few specific genes we can just switch over.

It could also just be technology. This breeding program has been around for 50 years or so and CRISPR only really took off around 2011. In addition, the original CRISPR-Ca9 caused a lot of unintended changes in DNA and it's really critical that the new trees are morphologically equivalent to the original trees in order for the birds, insects, and animals that depended on them to resume their interactions. Newer technologies like CRISPR-Nickase are much more precise but also very recently developed

[JshWright]

It actually is pretty "simple" in this case (the tree needs to produce a single chemical, oxalate oxidase). There is a program that has done this very successfully, and it just working through the regulatory process before wider distribution can happen. They have thousands of trees in (controlled) forest plots, including three generations of offspring.

[empyrrhicist]

That's the genetic engineering approach, but the Chinese ones have a different, more complex, less well understood, and potentially complementary resistance mechanism.

[JshWright]

Right, the genetic engineering approach is what this reply thread is about…

[culi]

That's very interesting. The video I linked showed that blight resistant trees had a very specific response to grow extra tissue around the infection site to stop it from spreading further so I assumed it was more of a multi-pronged defense

[empyrrhicist]

Yes - the Asian ones that evolved with the fungus have a more complex strategy, making a purely breeding-based program challenging. Separately, some clever genetic engineering allows the trees to break down oxalic acid, which prevents the fungus from being able to efficiently attack the plant cells.

Two separate things.

[pfdietz]

These two approaches could be combined, and probably will be naturally if both varieties are introduced to the wild.

[JshWright]

Indeed, part of the work being done by the team working on the genetic engineering approach is breeding Darling 58 with the Chinese backcrossed version.

[exmadscientist]

The things the gene gun can do turn out to be not the most useful things one could hope for from it. (It's a surprisingly brute-force device, though I guess that was all in the name if you actually took it at face value!) I don't see a ton of future for it as a generalist technology, though what things it does well it does very well.

CRISPR is... awful in different ways. (It seems to have a lot of trouble controlling how much material it transfers.) It's a laboratory curio or one-off stunt now, but I fully expect it will improve.

[tsol]

Honestly there's probably a benefit to just avoiding more controversial tech like GMOs. There will be that much less resistance to re-introducing them back into the wild

[ch4s3]

IIRC the breeding program predates those tools.

[at_a_remove]

If I recall, someone has jammed a wheat gene into a few of them in just that manner.

It's a tricky business, because chestnut had very particular properties, as wood, that made it quite valuable, and any gene interference, well ... who knows how it will change the resultant wood?

[systems_glitch]

Low-tech, non flash solution that delivers.

[robwwilliams]

They probably have no firm idea about the key genes involved in resistance. There could be half a dozen.

[adenner]

There are methods like GWAS that could help nail that down.

[mistrial9]

Do we not know ....