Fig. 2

On- and off-target hypermethylation of tumour suppressor genes in primary cells. a–d EPIC array at targeted promoter regions of a HIC1, b PTEN, c RASSF1 and d CDKN2A. gRNA (grey arrows point towards PAM sequence) in relation to CGI (in green) and TSS of genes, percentage methylation from dCas9 3A3L∆ targeted cells (pink line) compared to 3A3L targeted cells (black line), 10 days post-transfection of primary myoepithelial cells from donor 1 and after magnetic sorting on day 2 (mean ± SEM; n = 3; excluding error bar if smaller than plotted point). e Scatter plot of EPIC data comparing dCas9 3A3L∆ to dCas9 3A3L targeted cells 10 days post-transfection. Significantly hypermethylated probes from HIC1, PTEN, RASSF1 and CDKN2A genes are highlighted on the plot ( ≥ 20% increase; n = 3; Bumphunter; p < 0.01; Benjamini-Hochberg (BH) correction). f Overlap of the 946 EPIC array probes significantly hypermethylated in dCas9 3A3L targeted cells vs. 3A3L∆ (inner ring) compared to the percentage overlap of all probes in the EPIC array (outer ring) with CGIs, CGI shelves, CGI shores or CpG intergenic (top doughnut) and exons, introns or intergenic regions (bottom doughnut). Significant differences were calculated using regioner in R. g, h Gene expression of g p16 and h other target genes in primary myoepithelial cells from donor 1, 10 days after transfection with dCas9 3A3L∆ (purple) or dCas9 3A3L (orange). qPCR data is normalised to ACTB expression and shown as fold change compared to average gene expression in early passage cells (dotted line; mean ± SEM, n = 3; T-tests. *p < 0.01 dCas9 3A3L compared to 3A3L∆; ^$ p < 0.01 dCas9 3A3L or 3A3L∆ compared to early passage cells). RNA-seq data is shown as fold change of target gene transcripts from 3A3L or 3A3L∆ compared to average RPKM from early passage cells (dotted line; mean ± SEM, n = 3). Significance is from DESeq2 analysis with BH correction: *p < 0.001 3A3L compared to 3A3LΔ targeted cells; ^$ p < 0.001 3A3L or 3A3L∆ compared to early passage cells