Generating broad-spectrum tolerance to ALS-inhibiting herbicides in rice by base editing.

Herbicide-tolerant rice varieties generated by genome editing are highly desirable for weed control. We have used a cytosine base editor to create a series of missense mutations in the P171 and/or G628 codons of the acetolactate synthase (ALS) gene to confer herbicide tolerance in rice. The four different missense mutations in the P171 codon, P171S, P171A, P171Y and P171F, exhibited different patterns of tolerance towards five representative herbicides from five chemical families of ALS inhibitors. For example, P171S and P171A had lower levels of tolerance than P171Y and P171F to bispyribac but not to the other herbicides. Interestingly, a novel triple mutant (P171F/G628E/G629S) had the highest tolerance to all five tested herbicides. Field trials showed that both P171F and P171F/G628E/G629S could potentially be used with nicosulfuron. Our work illustrates an effective way of using base editing to generate herbicide tolerance in elite rice varieties.

Generation of imidazolinone herbicide resistant trait in Arabidopsis.

Recently-emerged base editing technologies could create single base mutations at precise genomic positions without generation DNA double strand breaks. Herbicide resistant mutations have been successfully introduced to different plant species, including Arabidopsis, watermelon, wheat, potato and tomato via C to T (or G to A on the complementary strand) base editors (CBE) at the P197 position of endogenous acetolactate synthase (ALS) genes. Additionally, G to A conversion to another conserved amino acid S653 on ALS gene could confer tolerance to imidazolinone herbicides. However, no such mutation was successfully generated via CBE, likely due to the target C base is outside of the classic base editing window. Since CBE driven by egg cell (EC) specific promoter would re-edit the wild type alleles in egg cells and early embryos, we hypothesized the diversity of base editing outcomes could be largely increased at later generations to allow selection of desired herbicide resistant mutants. To test this hypothesis, we aimed to introduce C to T conversion to the complement strand of S653 codon at ALS gene, hosting a C at the 10th position within the 20-nt spacer sequence outside of the classic base editing window. While we did not detect base-edited T1 plants, efficient and diverse base edits emerged at later generations. Herbicide resistant mutants with different editing outcomes were recovered when T3 and T4 seeds were subject to herbicide selection. As expected, most herbicide resistant plants contained S653N mutation as a result of G10 to A10. Our results showed that CBE could create imidazolinone herbicide resistant trait in Arabidopsis and be potentially applied to crops to facilitate weed control.