Extended Data Fig. 8: Characterization of HeLa Sororin-AID and HeLa NIPBL-AID cells.

a, Western blot for Sororin and GAPDH of HeLa Sororin-AID cells synchronized to G2 and either treated with auxin (+) or H₂0 (-) as well as western blot for NIPBL and GAPDH of HeLa NIPBL-AID cells synchronized to G2 and either treated with auxin (+) or H₂O (-). This experiment was repeated two (Sororin-AID samples) and three (NIPBL-AID samples) more times with similar results. Uncropped images are displayed in Supplementary Fig. 1. b, Cell cycle analysis of HeLa Sororin-AID and HeLa NIPBL-AID cells synchronized to G2 as indicated in Extended Data Fig. 3d, treated with auxin. Panel shows a FACS plot of cells stained for pH3S10 to mark mitotic cells and propidium iodide to measure DNA content. Gates for different cell cycle stages are shown and the indicated numbers reflect percentage of cells that were measured. This experiment was repeated three more times with similar results. c, Contact probability of all contacts at different genomic distances of HeLa NIPBL-AID cells synchronized to G2 (n = 4 biologically independent experiments) that were treated with auxin and HeLa WT cells synchronized to G2 (n = 11 biologically independent experiments). d, Chromosome congression analysis of NIPBL- and Sororin-depleted cells. AID-tagging often reduces protein levels already before the addition of auxin^59,60, which for NIPBL might impair sister-chromatid cohesion establishment during S-phase. We therefore performed metaphase congression analysis by time-lapse microscopy of WT HeLa cells, HeLa Sororin-AID cells and HeLa NIPBL-AID cells stained with SiR-DNA. HeLa Sororin-AID cells were treated with auxin before the final S-phase and HeLa NIPBL-AID cells were treated with auxin after the final S-phase. Panel shows the cumulative frequency of cells congressing their chromosomes in metaphase after entering mitosis in a RO3306 wash-out. Pooled replicates are shown from n = 2 biologically independent experiments. e, Cis sister and trans sister contacts of n = 11 biologically independent, merged G2 wildtype samples, n = 4 biologically independent, merged G2 NIPBL-degraded samples and n = 3 biologically independent, merged Sororin-degraded samples at a representative region on chromosome 5 are displayed alongside the ___location of TAD boundaries (see Methods for details). The strong accumulation of trans sister contacts close to the diagonal in NIPBL-depleted cells indicates frequent contacts between sister chromatids and a tighter alignment. Owing to normalization of contacts to the marginal count per row, trans sister contacts appear less frequent at larger genomic distances. Bin size of matrix is 150 kb. f, HiCrep⁵⁷ analysis of all, cis sister and trans sister contacts of all replicates of HeLa NIPBL-AID (n = 4 biologically independent experiments) or Sororin-AID (n = 3 biologically independent experiments) cells treated with auxin. Bars show the mean of all comparisons.