3D genomic architecture reveals that neocentromeres associate with heterochromatin regions

Abstract

The centromere is an important genomic locus for chromosomal segregation. Although the centromere is specified by sequence-independent epigenetic mechanisms in most organisms, it is usually composed of highly repetitive sequences, which associate with heterochromatin. We have previously generated various chicken DT40 cell lines containing differently positioned neocentromeres, which do not contain repetitive sequences and do not associate with heterochromatin. In this study, we performed systematic 4C analysis using three cell lines containing differently positioned neocentromeres to identify neocentromere-associated regions at the 3D level. This analysis reveals that these neocentromeres commonly associate with specific heterochromatin-rich regions, which were distantly located from neocentromeres. In addition, we demonstrate that centromeric chromatin adopts a compact structure, and centromere clustering also occurs in vertebrate interphase nuclei. Interestingly, the occurrence of centromere-heterochromatin associations depend on CENP-H, but not CENP-C. Our analyses provide an insight into understanding the 3D architecture of the genome, including the centromeres.

Graphical abstract

Figure 1.

Applying the 4C-seq analysis to DT40 cells. (A) An experimental scheme for 4C-seq analysis. The cells were fixed with 1% PFA, and the genome was initially digested with a six-base cutter restriction enzyme (RE; first RE digestion) and ligated (first ligation). The sample was decross-linked, and the genome was digested for a second time with a second four-base cutter restriction enzyme (second RE digestion) and ligated (second ligation). A primer set was prepared for a viewpoint (the region of interest where the interaction will be studied), and reverse PCR was performed (4C PCR). The PCR products were sequenced by a high-throughput sequencer and mapped to the reference genome. (B) Calculation of the SOR. The multiple reads mapped per fragment (≥1) were treated as a single positive read, and the rate of the number of positive restriction sites to the number of all the possible restriction sites in a region was calculated. For instance, if there were four or three restriction sites in a 150-kb region and two positive sites were in these regions, the SORs would be 0.5 (2/4) or 0.66 (2/3), respectively. (C) Representation of the cell lines used in this study. The native centromere (red) on the Z chromosome is located at the 42.6-Mb region in WT DT40 cells. The native centromere was removed, and the surviving cells were isolated and cloned. Each clone forms a neocentromere (blue) at a different position. The neocentromeres were formed at 3.8-, 35-, and 55-Mb regions in #BM23, #1320, and #1304 cell lines, respectively.

Figure 2.

4C profiles determined using different viewpoints in different cells containing various neocentromeres. (A) 4C profiles determined using #BM23 cells (second horizontal panel) and #1320 cells (third panel) with the 3.8-Mb region as a viewpoint (VP: Z3.8M-H-L). Top: ChIP-seq profile with CENP-A around the 3.8-Mb region in #BM23 cells. The two profiles were merged (fourth panel). Subtraction data: SOR values in #BM23 cells minus SOR values in #1320 cells are shown (bottom). Arrowheads indicate the 8- and 26-Mb regions. The position of the neocentromere in each cell line is indicated as broken lines. (B) 4C profiles determined using #BM23 cells (second horizontal panel) and #1320 cells (third panel) with the 3.8-Mb region as a viewpoint (VP: Z3.8M-H-R). Arrowheads indicate the 8- and 26-Mb regions. Although the same region was used for determining the 4C profiles in A, a different set of primers was used here (Table S1). (C) 4C profiles determined using #1320 cells (second horizontal panel) and #BM23 cells (third panel) with the 35-Mb region as a viewpoint (VP: Z35M-H-L). Top: ChIP-seq profile with CENP-A around the 35-Mb region in #1320 cells. The two profiles were merged (fourth panel). Subtraction data: SOR values in #1320 cells minus SOR values in BM23 cells are shown (bottom). Arrowheads indicate the 8- and 26-Mb regions. The position of the neocentromere in each cell line is indicated as broken lines. (D) 4C profiles determined using #1320 cells (second horizontal panel) and #BM23 cells (third panel) with the 35-Mb region as a viewpoint (VP: Z35M-H-R). Arrowheads indicate the 8- and 26-Mb regions. Although the same region was used for determining the 4C profiles in C, a different set of primers was used (Table S1).

Figure 3.

The native centromere on the Z chromosome associates with the 8- and 26-Mb regions. (A) 4C profiles determined using WT DT40 cells with the 42.6-Mb region (centromere region) as a viewpoint (top). Arrowheads indicate the 8- and 26-Mb regions. (B) ChIP-seq data on the Z chromosome with an antibody against H3K9me3 in DT40 cells. Arrowheads indicate the 8- and 26-Mb regions on the Z chromosome. Positions of the viewpoints (VPs) for Figs. 3 C, 4, and S3 experiments are shown. (C) 4C profiles determined using WT DT40 cells with the 8-Mb region (VP: Z8M-H-L) and 26-Mb region (VP: Z26M-H-L) as viewpoints. Arrowheads indicate the native centromere on the Z chromosome. (D) Analyses with two-color FISH in WT DT40 cells. Green-labeled BAC clone #279C6, comprising the 42.6-Mb centromere region, and red-labeled BAC clone #116B8, comprising the 26-Mb region, were used as probes (left). As a negative control, red-labeled BAC clone #279G5 comprising the 62-Mb region was used (right). Signals are shown by arrows. DNA was stained with DAPI (blue). Bar, 10 µm. The graph shows the distance between the two probes for each cell line.

Figure 4.

The neocentromeres are associated with heterochromatin-rich regions. (A) 4C profiles determined using #BM23 cells (top horizontal panel) and #1320 cells (second panel) with the 8-Mb region as a viewpoint (VP: Z8M-H-L). Detailed information pertaining to the VP is also represented in Fig. 3. The two profiles were merged (third panel). Subtraction data: SOR values of #BM23 cells minus SOR values of #1320 cells are shown (bottom). Arrows indicate the 3.8- and 35-Mb regions. The position of the neocentromere is indicated as broken lines. (B) 4C profiles determined using #BM23 cells (top horizontal panel) and #1320 cells (second panel) with the 26-Mb region as a viewpoint (VP: Z26M-H-L). Detailed information pertaining to the VP is also presented in Fig. 3. The two profiles were merged (third panel). Subtraction data: SOR values in BM23 cells minus SOR values in #1320 cells are shown (bottom). Arrows indicate the 3.8- and 35-Mb regions. The position of the neocentromere is indicated as broken lines. (C) Analyses with two-color FISH in #1320 and #BM23 cells. Green-labeled BAC clone #206E12, comprising the 35-Mb region (centromere in #1320 cells), and red-labeled BAC clone #261B8, comprising the 8-Mb region, were used as probes. Signals are shown by arrows. As a negative control, BAC clone #279G5 comprising the 62-Mb region was used. DNA was stained with DAPI (blue). Bar, 10 µm. The graph shows the distance between the two probes for each cell line.

Figure 5.

The neocentromere forms a compact structure. (A) High-resolution 4C profile (red bars) within the centromeric region determined using #BM23 and #1320 cell lines. Two viewpoints (3.8M-H-L and 3.8M-H-R) were selected for the 3.8-Mb region. The CENP-A ChIP-seq profiles (blue) in #BM23 and #1320 cell lines around the 3.8-Mb region are also shown. The values on the vertical axis represent the number of sequence reads. HindIII sites around this region are also shown. Arrows show the viewpoint positions. (B) High-resolution 4C profile (red bars) within the centromere region determined using #BM23 and #1320 cell lines. Two viewpoints (35M-H-L and 35M-H-R) were used for the 35-Mb region. The CENP-A ChIP-seq profiles (blue) around the 35-Mb region in the #BM23 cells and #1320 cells are also shown. Arrows show the viewpoint positions. The values on the vertical axis represent the number of sequence reads. HindIII sites around this region are also shown.

Figure 6.

Neocentromeres associate with the centromeres of other chromosomes with nonrepetitive centromeres. (A) The number of sequence reads and SOR values containing the centromere sequence on chromosome 5 determined using six different viewpoints (A–F) in #BM23 and #1320 cell lines. The position of each viewpoint is shown. (B) The numbers of sequence reads and SOR values containing the centromere sequence on chromosome 27 determined using six different viewpoints (A–F) in #BM23 and #1320 cell lines. The position of each viewpoint is shown.

Figure 7.

Neocentromeres associate with the centromeres of other chromosomes with repetitive centromeres. (A) The number of sequence reads near repetitive centromere sequences on chromosome 1 determined using six different viewpoints (A–F) in #BM23 and #1320 cell lines. The position of each viewpoint is shown. (B) The number of sequence reads near repetitive centromere sequences on chromosome 2 determined using six different viewpoints (A–F) in #BM23 and #1320 cell lines. The position of each viewpoint is shown. (C) Measurements of numbers of centromere foci at various cell cycle stages using CENP-T as a centromere marker. Right image is a representative immune fluorescence image of DT40 cells with anti–CENP-T (red). Cell cycle stages are determined by EdU labeling (left; green). DNA was stained with DAPI (blue). Bar, 10 µm. The graph shows numbers of centromere foci at various cell cycle stages.

Figure 8.

CENP-H but not CENP-C contributes to neocentromere–heterochromatin interactions. (A) Analyses with FISH using CENP-H– or CENP-C–conditional knockout #1320 cell lines. CENP-H and CENP-C are degraded upon adding auxin. Auxin was added at time 0, and green-labeled BAC clone #206E12 comprising the 35-Mb region (centromere in #1320 cells) as well as the red-labeled BAC clone #261B8 comprising of the 8-Mb region were used as probes. Signals are shown by arrows. DNA was stained with DAPI (blue). Bar, 10 µm. The graph on the right shows the distance between the two probes in each cell line at the indicated times after the addition of auxin. (B) 4C profiles determined with CENP-C–conditional knockout #1320 cell lines with the 35-Mb region as a viewpoint in the absence (top: CENP-C ON) or presence (middle: 2 h after auxin; CENP-C OFF) of auxin. The two profiles were merged (bottom). Arrowheads indicate the 8- and 26-Mb regions. The neocentromere was located at the 35-Mb region in these cells. (C) 4C profiles determined using #1320-based CENP-H–conditional knockout cell lines with the 35-Mb region as a viewpoint in the absence (top: CENP-H ON) or presence (middle: 2 h after auxin, CENP-H OFF) of auxin. CENP-H ON cells were treated with nocodazole for 2 h. The two profiles were merged (bottom). Arrows indicate the 8- and 26-Mb regions. The neocentromere was located at the 35-Mb region in these cells.

Figure 9.

A proposed model of 3D architecture of the genome including the centromere in DT40 cells. The centromere is located at the 42.6-Mb region, and the H3K9me3 rich regions are located at 8- and 26-Mb regions on the chicken Z chromosome at the 1D level. Although the centromeric region is estimated to lie within 30–40 kb (∼200 nucleosomes) according to analyses with ChIP-seq using anti–CENP-A, there exist ∼30 CENP-A–containing nucleosomes, indicating that the CENP-A–containing nucleosomes are scattered in the centromeric region. Although distances between the centromere and heterochromatin regions are long at the 1D level, the centromere physically associates with regions of heterochromatin in the interphase nuclei. The nucleosomes containing CENP-A might indulge in self-interactions to form a compact structure. The formation of the 3D genomic architecture, including the centromere, depends on the CENP-H–associated complex but not on CENP-C.