Supplementary Figure 3: Ectopic deposition of CENP-A into open and active chromatin at G1 does not function as a seeding hotspot for neocentromere formation.

(a) Number of non-α-satellite CENP-A SICER binding sites called at G1 or G2 at different fold thresholds (above background). (b) Human DLD1 cells with auxin degradable CENP-A^AID and a doxycycline-inducible CENP-A^WT4, were synchronized at G1 using the CDK4/6 inhibitor PD-0332991 (also known as Palbociclib) or at mitosis using nocodazole, following addition of doxycycline. The experiment was repeated independently three times with similar results. (c) Read mapping data of CENP-A^TAP ChIP-seq at G1 (red) and G2 (blue), at the chromosomal location of a known patient derived neocentromere⁵ found in chromosome 13. The experiment was repeated twice independently with similar results. A third human neocentromere, identified in line MS4221, has been identified to lie within a 400 kb neocentromere at position 86.5 to 86.9 Mb on chromosome 8 in hg19^5,6 (corresponding to 85.78–85.88 Mb in hg38). However, a gap and segmental duplications that appear in this region precluded precise analysis of CENP-A mapping at this neocentromere. (d-g) Fold enrichment of CENP-A^TAP chromatin in randomly cycling cells or at G1 (d, e) and CENP-A^LAP chromatin (f, g) at G1 at different genomic locations. SICER peaks ≥ 5-fold supported between two replicates were analysed for their enrichment level at different genomic locations, compared to the level of enrichment at these sites by chance. (h, i) Number of CENP-A^TAP (h) and CENP-A^LAP (i) SICER peaks ≥ 5-fold that overlap with ‘HOT’ regions in the human genome in G1 and G2 synchronized cells. Data shown are from two biologically independent experiments. Source data for d-i can be found in Supplementary Table 4.