Multi-target genome editing reduces polyphenol oxidase activity in wheat (Triticum aestivum L.) grains

Introduction Polyphenol oxidases (PPO) are dual activity metalloenzymes that catalyse the production of quinones. In plants, PPO activity may contribute to biotic stress resistance and secondary metabolism but is undesirable for food producers because it causes the discolouration and changes in flavour profiles of products during post-harvest processing. In wheat (Triticum aestivum L.), PPO released from the aleurone layer of the grain during milling results in the discolouration of flour, dough, and end-use products, reducing their value. Loss-of-function mutations in the PPO1 and PPO2 paralogous genes on homoeologous group 2 chromosomes confer reduced PPO activity in the wheat grain. However, limited natural variation and the proximity of these genes complicates the selection of extremely low-PPO wheat varieties by recombination. The goal of the current study was to edit all copies of PPO1 and PPO2 to drive extreme reductions in PPO grain activity in elite wheat varieties. Results A CRISPR/Cas9 construct with one single guide RNA (sgRNA) targeting a conserved copper binding domain was used to edit all seven PPO1 and PPO2 genes in the spring wheat cultivar ‘Fielder’. Five of the seven edited T1 lines exhibited significant reductions in PPO activity, and T2 lines had PPO activity up to 86.7% lower than wild-type. The same construct was transformed into the elite winter wheat cultivars ‘Guardian’ and ‘Steamboat’, which have five PPO1 and PPO2 genes. In these varieties PPO activity was reduced by >90% in both T1 and T2 lines. In all three varieties, dough samples from edited lines exhibited reduced browning. Discussion This study demonstrates that multi-target editing at late stages of variety development could complement selection for beneficial alleles in crop breeding programs by inducing novel variation in loci inaccessible to recombination.

Table S4: Details of three sgRNAs designed to target PPO1 and PPO2 genes in wheat.
Table S5: Predicted off-target effects of the selected sgRNA used in this study.
Table S7: Editing events in seven PPO1 and PPO2 genes in selected T1 and T2 'Fielder' individuals.
Table S8: Editing events in seven PPO1 and PPO2 genes in selected T1 and T2 'Guardian' individuals.

Table S9:
Editing events in seven PPO1 and PPO2 genes in selected T1 and T2 'Steamboat' individuals.

Table S11: Mean brightness values of dough samples.
Table S12: Mismatches between the sgRNA described by Zhang et al. 2021 and the seven PPO1 and PPO2 genes in the 'Fielder' genome.
Table S13: Mismatches between the sgRNA used in the current study and orthologous PPO1 and PPO2 genes from closely related species.
Table S14: Primers used for PCR in the current study.
Table S3: Alleles of PPO1 and PPO2 genes in 'Fielder'.Wild-type alleles are predicted to encode full-length, functional PPO proteins with all conserved domains.
Table S4: Details of three sgRNAs designed to target PPO1 and PPO2 genes in wheat.Quality scores include: RS2 -Rule Set 2 quality score derived from the algorithm described by (Doench et al., 2016).Overall score -weighted average of RS2 and cutting frequency determination from the wheatCRISPR tool (Cram et al., 2019).CRISPRP score -quality score determined by the CRISPR-P 2.0 tool (Liu et al., 2017).The total number of T0 plants represents the number of independentlyderived plants confirmed by PCR to carry the transgene.A T0 plant was considered "edited" when induced edits in at least one target PPO1 or PPO2 gene were detected by Sanger sequencing.All results described in this manuscript used the first sgRNA in this table.Table S5: Predicted off-target effects of the selected sgRNA used in this study (CGTGGTGCGCGAAGAAGATG).Cutting Frequency Determination (CFD) represents the predicted cutting efficiency of an off-target sequence relative to the on-target sequence, ranging from 0 (no predicted activity), to 1.0 (full predicted activity).The first row represents the sgRNA selected sequence targeting all seven PPO1 and PPO2 genes (CFD = 1).The gene IDs of other genes targeted in the promoter are listed.Off-target effects were predicted using the wheatCRISPR tool (Cram et al., 2019).
Table S10: Alleles of PPO1 and PPO2 genes in 'Kronos'.PPO gene sequences were used as queries in BLAST searches against the Kronos Elv1.1 genome assembly.

PPO2-A1 TraesFLD2A01G514300
Wild-type Stop codon introduced at amino acid 7. Table S12: Mismatches between the sgRNA described by Zhang et al. (2021) and the seven PPO1 and PPO2 genes in the 'Fielder' genome.This sgRNA matches 100% only with PPO1-A1, PPO1-B2 and PPO1-D1.The protospacer sequence (5'-GAAGAAGACGCTGCTGTTCC-3') designed to the forward strand is between positions 1,291 and 1,310 bp downstream of the initiating ATG codon of PPO1-A1 based on the 'Fielder' gene model.Mismatches are highlighted in bold.PAM sequence is in red.

Gene name
Fielder gene ID sgRNA sequence Number of mismatches Table S13

Figure S1 :
Figure S1: Position of PPO1 and PPO2 genes on chromosome 2B in the wheat landrace 'Chinese

Figure S1 :
FigureS1: Position of PPO1 and PPO2 genes on chromosome 2B in the wheat landrace 'Chinese Spring' and the variety 'CDC Landmark'.Gene positions are drawn to scale and homologous genes linked by lines determined using the Triticeae Gene Tribe microhomology tool(Chen et al., 2020).

Table S1 :
Presence/absence variation in PPO1 and PPO2 genes in 17 common wheat varieties.

Table S2 :
(Liu et al., 2020)3)e Spring' and 'Fielder'.Gene names are presented according to the guidelines of the Wheat Initiative(Boden et al., 2023)and consistent with previously described allelic variation.Previously proposed gene names are listed(Liu et al., 2020), alongside gene IDs from Chinese Spring (IWGSC v1.2) and Fielder (v1) genome assemblies.Genes that are absent from the 'Chinese Spring' genome assembly are marked with a dash (-).The seven target PPO1 and PPO2 genes are separated from non-target PPO genes by a horizontal line.

Table S6 :
PPO activity in wild-type, T1 and T2 lines from the varieties 'Fielder', 'Guardian' and 'Steamboat'.Mean PPO activity, the number of biological replicates (n), standard deviation and twotailed Student's T-test result showing the difference with each corresponding wild-type line are shown.

Table S9 :
Editing events in five PPO1 and PPO2 genes in selected T1 and T2 'Steamboat' individuals.Where mutations are heterozygous, both allele types are described, separated by "/".

:
Mismatches between the sgRNA used in the current study and orthologous PPO1 and PPO2 genes from closely related species.Mismatches are highlighted in bold.PAM sequence is in red.* The mismatch in ScPPO1 alters the PAM sequence from CCC to CCT, which is still permissible for editing.

Table S14 :
Primers used for PCR in the current study.To amplify the PPO1-D1 and PPO2-D1 genes, different PCR assays were used in different genotypes due to sequence variation.