Regulation of DNA replication timing on human chromosome by a cell-type specific DNA binding protein SATB1

Abstract

Background: Replication timing of metazoan DNA during S-phase may be determined by many factors including chromosome structures, nuclear positioning, patterns of histone modifications, and transcriptional activity. It may be determined by Mb-domain structures, termed as "replication domains", and recent findings indicate that replication timing is under developmental and cell type-specific regulation.

Methodology/principal findings: We examined replication timing on the human 5q23/31 3.5-Mb segment in T cells and non-T cells. We used two independent methods to determine replication timing. One is quantification of nascent replicating DNA in cell cycle-fractionated stage-specific S phase populations. The other is FISH analyses of replication foci. Although the locations of early- and late-replicating domains were common between the two cell lines, the timing transition region (TTR) between early and late domains were offset by 200-kb. We show that Special AT-rich sequence Binding protein 1 (SATB1), specifically expressed in T-cells, binds to the early domain immediately adjacent to TTR and delays the replication timing of the TTR. Measurement of the chromosome copy number along the TTR during synchronized S phase suggests that the fork movement may be slowed down by SATB1.

Conclusions: Our results reveal a novel role of SATB1 in cell type-specific regulation of replication timing along the chromosome.

Figure 1. Replication timing of the human 5q23/31.

A. Experimental strategy for determination of replication timing. Asynchronously replicating cells were labeled with BrdU and sorted by FACS into six fractions (G1, S1–4, G2/M) on the basis of DNA content. Genomic DNA from cells in each fraction was extracted, and newly replicated DNA was immunoprecipitated with anti-BrdU antibody. Semi-quantitative PCR was carried out using the newly replicated DNA as template. Relative band intensity was quantified. The values in each fraction were normalized by the levels of BrdU-labeled mitochondrial DNA (mtDNA; replicated equally throughout the cell cycle) used as an internal control for the recovery of DNA in each sample. B. 40,000 cells (Jurkat and HL-60) sorted (upper) and collected on the basis of DNA content (G1, S1–4, G2/M) were stained with PI, and analyzed by FACS (lower). C. Validation of cell cycle fractionation. The known early (PGK1) or late (F9) replicating region is enriched in appropriate fractions in comparison with the level of mtDNA in HL-60. D. DNA replication timing on the human chromosome 5q23/31 region (3.5 Mb) containing the cytokine cluster region in Jurkat (T cell) and HL-60 (non T cell). The 2.2 Mb segment containing the cytokine cluster (130.3–132.5) replicates in G1 or early in the S-phase (S1 and S2), whereas the 0.9 Mb segment distal to the cluster and proximal to the centromere replicates late in the S phase (S3, S4 and G2). The mean locations of the timing transition region (TTR) are located at around 130.15–130.25 (yellow box) in Jurkat (TTR-J) and at around 129.95–130.05 (green box) in HL-60 (TTR-H), and are offset by 180 kb in the two cell types. The left boundary of early replicating region coincided with the transition of chromosomal synteny (see Fig. S2A). E. The results of replication timing assays with fractionated cells are shown for each location along the 3.5 Mb human chromosome. The locations of the 9 primers used are indicated along the 5q23/31 region shown to the right of the panels. Small red solid boxes show the peak timing fraction for each probe, and large red dotted boxes show the maximum timing transition segments for Jurkat (129.98–130.22) and HL-60 (129.52–130.16).

Figure 2. Analyses of replication timing by FISH.

A. Hybridization signals of replicating cells. SS, singlet-singlet: SD, singlet-doublet; DD, doublet-doublet. B. Replication timing analyzed by FISH at 5q23/31. Locations of DNA probes derived from 5q23/31used in this study are shown at the top. Human BAC clones were purchased from Invitrogen. A cosmid clone on the chromosome 12, cCl12–140, was kindly provided by Dr. Okumura (Nogami et al, 2000), and was used as a control for early replication. At least 200 BrdU-positive nuclei (S-phase) were counted for each probe. The signal patterns were classified into SS, SD, or DD. On the haploid segment of HL-60 (probes 2∼6), two signal patterns, singlet (S) and doublet (D), were observed. Replication timing of the human 5q23/31, estimated from the FISH analyses, is consistent with that of the cell cycle fractionation studies (see text for details). E, E/M M/L and L stand for early-, early/mid-, mid/late- or late-replicating, respectively.

Figure 3. Correlation between SATB1 expression and replication timing at TTR.

A. SATB1 expression in Jurkat, HL-60 and HeLaS3. Whole cell extracts, separated on 7.5% SDS-polyacrylamide gel, were blotted with anti-SATB1 antibody. Expression of SATB1 is high in T cell (Jurkat), and low or non-detectable in non-T cells (HL-60 and HeLaS3). B. Replication timing of the Probe 4 (see Fig. 2B and Table S1) in Jurkat, HL-60 and HeLaS3. This locus replicates in late-S in Jurkat (high SATB1) and in early-S in HL-60 (low SATB1) and HeLaS3 (very low SATB1).

Figure 4. Effect of SATB1 expression on replication timing at TTR: expression of SATB1 in HeLaS3 cells.

A. The procedure for obtaining HeLaS3 cells overexpressing SATB1. B. At 24 hr after transfection with pEF-mKO2 (upper panels) or pEF-mKO2-SATB1 (lower panels), cells were observed using a fluorescence microscope BioZero (KEYENCE). mKO2 fluorescence was observed mainly in the cytoplasm of HeLaS3 cells expressing mKO2 vector or in the nuclei of those expressing mKO2-SATB1. Red, mKO2 signal. C. Whole cell extracts were examined by western blotting using anti-SATB1 antibody. Lane 1, non-transfected HeLaS3, lanes 2 and 4, HeLaS3 transfected with pEF-mKO2; lanes 3 and 5, HeLaS3 transfected with pEF-mKO2-SATB1, lane 6, Jurkat. Lanes 2 and 3, 24 hr after transfection; lanes 4 and 5, 43 hr after transfection. D. Sorting of mKO2 positive cells by FACS Aria (BD Biosciences). Cells expressing mKO2-SATB1 were separated into viable (yellow) and dead (red) cells stained with PI, then mKO2 positive cells (green) among viable cells were further separated. E. Replication timing of HeLaS3 and HeLaS3 expressing mKO2-SATB1 at 43 hr after transfection. We analyzed replication timing of HeLaS3 transfected with mKO2-SATB1 at 24 and 43 hr after transfection. Only the data at 43 hr are shown for cells non-transfected (left panel) or transfected with mKO2-SATB1 plasmid (central panel). At least 200 BrdU-positive nuclei (S-phase) were counted for each probe. Replication timing in TTR (detected by Probe 4) changed from early (HeLaS3) to late (HeLaS3 expressing mKO2-SATB1) (indicated by the arrows). Expression of mKO2 did not affect the replication timing of the transition region (right panel).

Figure 5. Effect of SATB1 expression on replication timing at TTR: suppression of SATB1 expression in Jurkat cells.

A. The procedure for repression of SATB1 expression in Jurkat. B. Whole cell extracts were examined by western blotting using anti-SATB1 antibody. Lane 1, untransfected Jurkat; lane 2, Jurkat transfected with pRS vector; lane 3 and 4, Jurkat transfected with pRS-SATB1-shRNA1 and with pRS-SATB1-shRNA1+ pRS-SATB1-shRNA2, respectively. Cells were harvested at 72 hr after transfection. C. Replication timing was determined by FISH across the TTR. Only the data at the Probe 4 are shown. Replication timing in the transition region changed from late (Jurkat) to early (SATB1-depleted Jurkat) (indicated by the arrows). At least 200 BrdU-positive nuclei (S-phase) were counted.

Figure 6. Chromatin immunoprecipitation (ChIP) assays of SATB1 binding.

A. Locations of primers (a–e) used for ChIP assays. B. ChIP analyses were carried out by using anti-mouse SATB1 antibody (left panel) or purified mouse IgG1 control antibody (central panel) jn HeLaS3 cells stably expressing SATB1. Chromain-immunoprecipitated DNA was purified by MinElute (QIAGEN) and used for quantitative PCR. Error bars represent the mean and standard deviations based on four independent experiments. Relative ratio (SATB1/control) is shown as SATB1-specific binding (right panel).

Figure 7. Genome copy number analyses of TRR in synchronized cell population.

A. Locations of primers (a, b, c, d and f) used for copy number analyses. B. The procedure for quantification of copy number. C. Time course of DNA replication at various locations in and around TTR. Genome copy numbers of cells synchronously growing from double thymidine block were quantified at various loci at various timepoints. The level of DNA synthesis at 2 hr or 8 hr after release was set as 0% or 100% DNA synthesis, respectively.