Large and stable genome edits at the sorghum alpha kafirin locus result in changes in chromatin accessibility and globally increased expression of genes encoding lysine enrichment

**Background:** Sorghum grain is deficient in the essential amino acid lysine due to the dominance of alpha-kafirin seed storage proteins, which lack lysine. Reducing alpha-kafirin accumulation triggers proteome rebalancing, increasing non-kafirin proteins and lysine content, but the underlying mechanisms are poorly understood. This study characterizes a gene-edited sorghum line with deletions at the alpha-kafirin locus to investigate transcriptional and chromatin changes during kernel development.

**Methods:** The edited line 19Q4-13 was derived from CRISPR/Cas9-edited T3 sorghum (Tx430 background) outcrossed to wild-type Tx430, then self-pollinated to obtain transgene-free homozygous edited progeny. PacBio long-read sequencing characterized structural variants at the alpha-kafirin locus. Developing kernels at 20 days after pollination were harvested from six edited and six wild-type plants. RNA-seq (12 libraries, average 76.4 million read pairs per library, 91.8% alignment rate) and ATAC-seq (12 libraries, average 21.5 million read pairs, 86.7% alignment after filtering) were performed. Differential expression and differential accessibility were analyzed using edgeR with TMM normalization and glmLRT at FDR < 0.1. Amino acid composition was measured from mature kernels (nine biological replicates per line).

**Key Results:** PacBio sequencing revealed that 8 of 16 alpha-kafirin gene copies carried edits, including a 1.35 kb deletion (k1C4), an inversion (k1C10), and complex structural rearrangements involving duplication and translocation of an inverted segment. The edited line showed a 43% increase in protein-bound lysine (from 14.7 to 21 nmol/mg flour, p < 0.05) and a 161% increase in free lysine (from 0.205 to 0.536 nmol/mg flour, p < 0.05). ATAC-seq identified 32,957 total accessible chromatin regions (21,843 in edited, 21,320 in wild-type), with 163 differentially accessible regions (DACRs). RNA-seq identified 184 differentially expressed genes (DEGs; 109 upregulated in edited, 32 downregulated, with 11 of the downregulated being edited alpha-kafirins). Among DEGs, the mean number of predicted lysine residues was 15.03 in upregulated genes versus 7.94 in downregulated genes (p = 0.003). Cross-referencing with maize studies identified 19 DEGs implicated in prior work, including a subtilisin protease (SbiRTX430.02G245400) found in three studies. Motif enrichment analysis showed enrichment of the ZmOPAQUE11 (O11) binding motif in accessible chromatin of edited lines, and 22 DEGs had syntenic orthologues implicated in O11 regulation, including five high-probability O11 targets.

**Clinical Implications:** While this is a plant science study without direct clinical application, the findings have implications for human nutrition and livestock feed. Enhancing lysine content in sorghum—a staple grain in developing countries and a gluten-free food source—could improve protein quality for populations dependent on sorghum-based diets. The identification of conserved transcriptional responses between sorghum and maize prolamin mutants may guide breeding or editing strategies to produce high-lysine grains without the negative agronomic traits (e.g., soft kernels) associated with previous mutants. The study also highlights challenges in editing tandemly duplicated gene families, informing future CRISPR applications in crops.