Editing the 19 kDa alpha-zein gene family generates non-opaque2-based quality protein maize

Abstract

Maize grain is deficient in lysine. While the opaque2 mutation increases grain lysine, o2 is a transcription factor that regulates a wide network of genes beyond zeins, which leads to pleiotropic and often negative effects. Additionally, the drastic reduction in 19 kDa and 22 kDa alpha-zeins causes a floury kernel, unsuitable for agricultural use. Quality protein maize (QPM) overcame the undesirable kernel texture through the introgression of modifying alleles. However, QPM still lacks a functional o2 transcription factor, which has a penalty on non-lysine amino acids due to the o2 mutation. CRISPR/cas9 gives researchers the ability to directly target genes of interest. In this paper, gene editing was used to specifically target the 19 kDa alpha zein gene family. This allows for proteome rebalancing to occur without an o2 mutation and without a total alpha-zein knockout. The results showed that editing some, but not all, of the 19 kDa zeins resulted in up to 30% more lysine. An edited line displayed an increase of 30% over the wild type. While not quite the 55% lysine increase displayed by QPM, the line had little collateral impact on other amino acid levels compared to QPM. Additionally, the edited line containing a partially reduced 19 kDa showed an advantage in kernel texture that had a complete 19 kDa knockout. These results serve as proof of concept that editing the 19 kDa alpha-zein family alone can enhance lysine while retaining vitreous endosperm and a functional O2 transcription factor.

Keywords: CRISPR/cas9; QPM; lysine; opaque2; seed storage proteins; zeins.

Figure 1

Six gRNA sequences were used to edit multiple sites in the 19 kDa gene bodies. Target sites were selected that could effectively edit multiple locations, such that large deletions would be made that facilitate the use of PCR primers for genotyping and breeding. Sites noted here indicate a seed sequence match, or the 12 bp adjacent to the PAM.

Figure 2

PCR markers were used for identifying edited alleles. Separate markers were used to identify edits in each gene family. Both reactions are amplifying multiple individual gene copies. HiII and W22 wild‐type inbreds display a single band, whereas 19z‐10 and 19z‐11 display smaller and multiple bands due to family members being edited.

Figure 3

F2:F3 kernels demonstrated homozygosity and effective knockout of the 19 kDa alpha zein. SDS‐PAGE of individual kernels was used to identify which F2 plants were homozygous. Additionally, it was critical that the 19 kDa alpha zeins were affected.

Figure 4

Long‐read sequencing showed large deletions of the 19 kDa alpha‐zeins. The blue and orange highlights denote the region deleted from the respective lines. Each deletion is annotated with the number of base pairs.

Figure 5

(a) the chromosomal locations of the editing targets allowed for the discovery of segregating lines. (b) Kernel phenotypes of edited lines, wild type and QPM. Partially edited lines had comparable vitreousness to wild type.(c) images of whole ears from F3 lines. Images are not to scale.

Figure 6

TEM images of 20DAP endosperm. a and b show wild type W22. c and d show a total 19 kDa knockout. 19z‐10. Arrows denote malformed protein bodies. e and f show partial edit 19z‐10‐a1A2b, which has a wild family. g and h show 19z‐11‐a1a2B, which has a wild type B family. a, c, e, and g are at 10000x magnification. b, d, f, and hare at 25000x magnification.

Figure 7

Lysine levels in flour from normal maize, edited, and QPM. *p <0.05.

Figure 8

Amino acid levels in flour from normal, edited, and QPM. (a) Protein‐bound amino acids (b) free amino acids.