CRISPR-Cas9 mutagenesis of regions upstream or downstream of AtCLV3 reveal that neither region alone is essential to its function.

(A) Representation of the AtCLV3 gene (right) and the 3.8 kb region up to the next upstream gene (left). The gRNA array spanning the 1.5 kb sequence proximal to the TSS of AtCLV3 is shown as purple arrowheads, and the gRNA array targeting the entire 3.8 kb region between the TSS of AtCLV3 and the next gene upstream is shown as pink arrowheads. A heatmap representation of the 11 CRISPR-Cas9 engineered 5’ AtCLV3^Reg5’ alleles including WT and Atclv3 null mutant. (B) A silique with three locules from the 5’ allele AtCLV3^Reg5’-7. A top-down view is shown in the inset. Scale bar, 1 mm. Inset scale bar, 500 um. (C) Representation of the AtCLV3 gene (left), and the 1.6 kb region downstream. gRNAs are represented by purple arrowheads. Heatmap representation of the three engineered 3’ AtCLV3^Reg3’ alleles. (A), (C) The alleles have been encoded, such that perturbations to the region are represented as the degree of sequence modification relative to WT within 20 bp windows. Inversions are shown in red in the encoding. The location of validated TFBSs (black and brown arrows), TFBSs predicted by FIMO (red arrowheads), and CNSs identified by Conservatory (blue arrowheads) and mVISTA (green bars) are shown. Locule number quantifications are represented by stacked bar plots and box plots. Box plots show the 25^th, 50^th (median) and 75^th percentiles, with outliers as black points. Number of fruits sampled (n) is shown to the left, and mean and standard deviation (sd) are shown to the right. Two-sided Dunnett’s compare with control tests were performed to compare engineered alleles to WT, and the p-values (p) are included to the right. ns, not significant.