Fig. 2: Architecture of Cbp1:Cren7 chromatinization.

a Aggregate plots of Cbp1 and Cren7 ChIP-exo occupancy over 208 repeats from CRISPR arrays A and B. Signal for the plus and minus strand (relative to CRISPR array orientation) is plotted above and below the x-axis, respectively. The aggregate mean signal was calculated from the geometric mean of two biological replicates scaled to reads per million. b Cren7 recruitment to Cbp1:CRISPR DNA complexes depends on the flanking DNA downstream of the CRISPR repeat. EMSAs with templates bearing a CRISPR A repeat with 20 bp upstream and 20, 2 or 0 bp downstream flanking sequence and 12.5 nM Cbp1, 12.5 to 25 nM Cren7. The free DNA of the shorter templates appeared as double-band, possibly due to some denaturation at the incubation temperature of 75 °C. A representative gel from three replicate experiments is shown. c Orientation of Cbp1 binding to CRISPR repeats. Binding of Cbp1 and HTH-deletion mutants ΔHTH1 and ΔHTH3 to mutated CRISPR repeat sequences was assessed by EMSAs and the fraction of DNA bound by Cbp1 was plotted (mean of three biological replicates and standard deviation are shown). Double mutations were introduced into the CRISPR repeat sequence based on regions showing stronger sequence bias in the motif identified for non-canonical Cbp1 binding sites (Fig. 1b). The A13C/T14G double mutant was included as control for a region of CRISPR repeats that shows no sequence bias in non-canonical binding sites. To compensate for the overall lower affinity of the ΔHTH1 and ΔHTH3 variants, their concentration was raised to 160 and 80 nM, respectively. The effect of CRISPR repeat mutations on WT Cbp1 binding (black asterisks) and whether these effects are altered with HTH deletions in Cbp1 (red asterisks) was tested in likelihood ratio tests of nested beta regression models with Bonferroni multiple testing correction (see Methods section for details). *** and * denote p adj <0.001 and 0.01, respectively. d Cren7 recruitment to Cbp1:CRISPR repeat complexes depends on the HTH3 domain of Cbp1. EMSA assay testing Cren7 recruitment (12.5 to 25 nM) to CRISPR DNA-bound Cbp1 (12.5 nM), ΔHTH1 (50 nM) and ΔHTH3 (25 nM). Representative gels from three replicate experiments are shown. e DNase foot-printing assays corroborate Cbp1-dependent Cren7 deposition at the downstream spacer. Foot-printing assays were carried out with 12.5 or 25 nM Cbp1 and 25 or 50 nM Cren7. A G + A sequencing ladder is shown on the left as reference. Cbp1 and Cbp1:Cren7-induced protection against DNase cleaveage is indicated with open circles and hypersensitivity is indicated with full circles. A representative gel from three replicate experiments is shown.