Fig. 5: Matr3 loss alters chromatin contacts to the nuclear structure.

a Reduced expression level of Mbd1 was revealed by RNA-seq from two independent experiments and Western blot. β-actin was used for a Western blotting control. b Chromatin occupancy of CTCF and Rad21 in parental and Matr3 KO cells near Mbd1 locus is shown. c CAPTURE experiment⁴³ was performed using sgRNAs directed to a putative cis-regulatory element near Mbd1 which is occupied by CTCF and cohesin, and an adjacent upstream, as a negative control (NC) (Figs. 5b and S4b). d Purification of biotinylated dCas9 and Western blot indicated that Matr3 interacted with the putative element (Mbd1) but not with the upstream control (NC). Relative enrichment of Matr3 protein at the putative element to the upstream control was quantified from three independent experiments. e Radial location of the Mbd1 gene was observed by 3D DNA FISH. For quantification, the nuclear radius was divided into five shells and the number of signals in each shell was counted. The graph is the result of 2–3 independent experiments and the change was significant under a two-sided t test (p = 0.006; Cohen’s d = 0.32). f Snapshot of Hi-C data at the Mbd1 locus showing compartments in parental and Matr3 KO cells and the difference in insulation score (parental in green and Matr3 KO in magenta). g, h Deletion of CTCF/cohesin binding region (shaded box in Fig. 5b) by CRISPR/Cas9 reduced Mbd1 expression. RNA (g) and protein (h) levels were examined by RT-PCR (n = 3) and Western blotting, and compared to Mbd1 KO cells. β-actin was used for a Western blot control. Data were the result of 2–3 independent experiments and error bars represent mean ±  s.d. Source data are provided as a Source Data file.