Mutation breeding: still a long shot

This is why we love the internets. Because when we say “one to keep an eye on,” what we really mean is, “This seems odd beyond belief”. And fortunately for us, people who really know their stuff are indeed keeping an eye on things. Matthew V Di Leo commented on yesterday’s post, and has given us permission to elevate his comment to a post, which we are doing because it is too long merely to feature.

Calling mutation breeding a “technical revolution,” particularly this implementation of it, appears beyond absurd to me.

I don’t have access (yet) to the full publication that this appears to come from, but from what I can tell they did the following.

1. Took a rice variety that was already well known to have a set of traits (pale green leaves and semi-dwarfism) that were associated with high yields
2. Mutated a huge number of individuals of another elite line that does not have these traits
3. Looked for new mutations that mimic the original trait
4. Identified the mutations associated with the phenotype with resequencing and created MAS markers to allow them to clear out the extra mutations

A couple points:

1. They didn’t discover agronomically valuable genes de novo–they simply identified them within a known variety. This could also be done by genetic mapping approaches or simply by crossing the plants through traditional breeding and selecting for the right phenotype. Their approach probably speeds up the process by a year or two at most, and might be much more expensive.
2. This only worked at all because they had a simple, highly heritable trait that was incredibly easy to measure. This is very rare. Something complex and subtle like yield or salinity resistance ((Which is what all the fuss is about. Ed.)) would be incredibly hard or impossible to do with their approach.
3. No matter how you move better genes into your plant, whether by traditional breeding, mutation approaches, or genetic engineering, the total variety development time is rarely much less than 10 years–especially due to the need for multi-year/environment trials.

Thanks again to Matt, who will surely post something more detailed on his own blog.