Fig. 1

CENP-B interacts with Daxx and targets Daxx to centromeres for H3.3 loading. a FLAG IP from chromatin fraction of HEp2 cells expressing FLAG-HA-Daxx probed with CREST antibodies (left). Major band ~ 80 kDa was identified as CENP-B with specific antibodies (right). b Depletion of CENP-B prevents Daxx accumulation at centromeres. Representative images of HEp2 cells transfected with scrambled siRNA (left), CENP-B siRNA (middle) and CENP-C siRNA (right). Daxx (green) is associated with subpopulation of centromeres (CREST, red) in HEp2 cells (top row); treatment with MG132 (bottom row) increases number of PML bodies [36] and boosts association of Daxx with centromeres [36]. Depletion of CENP-B, but not CENP-C, reduced association of Daxx with centromeres in both tested conditions. PML nuclear bodies are marked with asterisks in the inserts. Quantitation of colocalization is presented in Additional file 3: Fig. S3. Graphs in the bottom panel show fluorescence intensity of Daxx (green) and CREST (red) signal along the region of interest (ROI). c Daxx association with α-satellite (CEN) and HS3 satellite (periCEN) were tested by ChIP assay. CENP-B depletion abolished Daxx association with CEN (CENP-B box positive α-satellite), while did not affect periCEN association (CENP-B box negative HS3). d ChIP analysis of H3.3 enrichment at CEN and periCEN in HEp2 cells expressing FLAG-HA-H3.3. Depletion of CENP-B reduced H3.3 enrichment at CEN, but not at periCEN. Depletion of Daxx and ATRX reduced association of H3.3 with both CEN and periCEN, confirming functionality of Daxx/ATRX complex as a histone chaperone at both genomic elements. Bars represent the mean ± SD (n = 3). Statistical analysis was performed by one-way ANOVA followed by Tukey’s multiple comparisons test. ****P < 0.0001; ***P < 0.001; **P < 0. 01; *P < 0.1; NS, nonsignificant