Constitutive centromere-associated network contacts confer differential stability on CENP-A nucleosomes in vitro and in the cell

Abstract

Eukaryotic centromeres are defined by the presence of nucleosomes containing the histone H3 variant, centromere protein A (CENP-A). Once incorporated at centromeres, CENP-A nucleosomes are remarkably stable, exhibiting no detectable loss or exchange over many cell cycles. It is currently unclear whether this stability is an intrinsic property of CENP-A containing chromatin or whether it arises from proteins that specifically associate with CENP-A chromatin. Two proteins, CENP-C and CENP-N, are known to bind CENP-A human nucleosomes directly. Here we test the hypothesis that CENP-C or CENP-N stabilize CENP-A nucleosomes in vitro and in living cells. We show that CENP-N stabilizes CENP-A nucleosomes alone and additively with CENP-C in vitro. However, removal of CENP-C and CENP-N from cells, or mutating CENP-A so that it no longer interacts with CENP-C or CENP-N, had no effect on centromeric CENP-A stability in vivo. Thus, the stability of CENP-A nucleosomes in chromatin does not arise solely from its interactions with CENP-C or CENP-N.

FIGURE 1:

CENP-N protects CENP-A nucleosomes from micrococcal nuclease digestion. (A) Comparative MNase digestion analysis of H3.1- and CENP-A nucleosomes, reconstituted with a 166–base pair DNA fragment derived from the 601 sequence. After digestion (quenched with 50 mM EDTA), samples were analyzed in a gel shift assay prior to DNA extraction (left panel) and after DNA extraction (right panel). The length of the DNA fragments was determined according to the standard DNA ladder curve obtained from the Bioanalyzer (Agilent). (B) MNase digestion analysis of CENP-A nucleosome (with the same DNA sequence as in A) in the absence and presence of CENP-N^1-289. The reactions were performed and analyzed as in A. (C) MNase digestion analysis of CENP-A nucleosome (147–base pair DNA derived from α satellite DNA) in the absence and presence of CENP-N^1-289. (D) The DNA ends of the CENP-A nucleosome are important for the proper orientation of CENP-N^1-289 on the CENP-A nucleosome. CENP-A nucleosomes were reconstituted with either 147 or 127 base pairs of 601 nucleosome-positioning DNA. CENP-N^1-289 was mixed with CENP-A nucleosome at a 3:1 ratio. After a 5-min incubation at 37°C, samples were analyzed by 5% native PAGE.

FIGURE 2:

CENP-N increases the stability of CENP-A nucleosomes against dissociation in vitro. (A) CENP-N^1-289 increases the stability of CENP-A nucleosomes at increased ionic strength. CENP-A nucleosome (100 nM) was mixed with either buffer or 200 nM CENP-N^1-289 at the indicated NaCl concentration. Nucleosome stability was quantified by measuring the intensity of all bands, including shifted bands (right panel; normalized as described under Materials and Methods), n = 3. All assays were performed in a final buffer containing 20 mM Tris HCl, pH 7.5, the indicated NaCl concentration, 5% glycerol, 0.5 mM EDTA (no detergent). The ∼25% of nucleosomes that are not protected in the presence of CENP-N likely reflects the amount of unbound CENP-A nucleosome (right panel). (B) CENP-N^1-289 stabilizes CENP-A nucleosomes against the combined effects of dilution and heat treatment. Left panel: native PAGE. CENP-N^1-289 was mixed with CENP-A nucleosomes at a molar ratio of 3:1. As a control, *CENP-N^1-289 indicates CENP-N^1-289 that was denatured by heating at 55°C for 5 min before mixing with CENP-A nucleosomes. The same amount of sample was loaded immediately after treatment. All bands including shifted bands were quantified to determine the percentage of remaining nucleosome, and error bars are derived from three independent gels (n = 3). Intensity at 647 nm was measured. (C) CENP-A mono-nucleosomes on an EM grid are greatly stabilized in the presence of CENP-N^1-289. The CENP-A nucleosome sample (2.5 µM) was mixed with 7.5 µM CENP-N^1-289, and the control was adjusted with buffer. Red boxes indicate nucleosome-shaped particles. Yellow arrows show free DNA. Scale bar = 50 nm. The blue box highlights the area from the left micrograph. The intact particles were counted, and the numbers are listed in Table 1.

FIGURE 3:

The CENP-A nucleosome-binding domains of CENP-N and CENP-C have additive effects on stabilizing CENP-A nucleosomes in vitro. (A) MNase digestion analysis for complex AC and complex ANC. CENP-C^426-537 was mixed with CENP-A nucleosome at a 2:1 ratio to form the complex AC. CENP-N^1-289 was mixed with preformed AC complex at 2:1 ratio to form the ANC complex. DNA fragments after digestion were analyzed as in Figure 1A. (B) CENP-N^1-289 and CENP-C^426-537 showed additive effects in stabilizing the CENP-A nucleosome against dilution and heat. CENP-N^1-289 or CENP-C^426-537 was mixed with CENP-A nucleosome at molar ratios of 3:1. The buffer was adjusted to a final 20 mM Tris HCl, pH 7.5, 150 mM NaCl, 2% glycerol, 0.5 mM EDTA. Dilution and heat treatments were done as described in under Materials and Methods. Samples were kept at 4°C for 2 h before analysis by native PAGE. (C) Quantification of the percentage of stable nucleosomes under different nucleosome concentrations without heat (from the upper panel in B, n = 2. The raw signal (intensity of fluorescence signal at 647 nm) of nucleosome or complex after dilution (150 and 75 nM) was normalized to the signal from samples before dilution (at 300 mM NaCl). (D) The stabilizing effect of CENP-N and CENP-C on CENP-A nucleosome, as demonstrated by cryo-EM. Both CENP-N^1-289 and CENP-C^426-537 were mixed with CENP-A nucleosome at molar ratio 3:1 to form the ANC complex. Buffer was adjusted to the same condition for both samples. Concentration of both samples was 2.5 µM. Red boxes indicate nucleosome-shaped particles. Scale bar = 50 nm. Blue box highlights the area from the left micrograph. The intact particles were counted, and the numbers are listed in Table 1.

FIGURE 4:

CENP-C and CENP-N degradation has no significant effect on centromeric CENP-A maintenance. (A, B) Cell lines containing AID tagged CENP-N, CENP-C, or both were treated with IAA to degrade the indicated proteins. Blue bars represent cells not treated with IAA. Cells were maintained in IAA beginning just after mitosis (red bars) and harvested in an early S-phase thymidine (thym) arrest (A) or after mitosis (mustard bars) or in early S phase (green bars) and harvested in a G2-phase roscovitine (rosco) arrest (B). Centromeric CENP-A immunofluorescence signal was normalized to the no-IAA signal. (C) Degradation of CENP-N inhibits new CENP-A assembly but not preexisting CENP-A in chromatin. The CENP-N AID-sfGFP cell line containing a stably integrated SNAP-tagged CENP-A was either fluorescently labeled or quenched according to the schematic (left panel). Green bars represent the time of synthesis of the fluorescent population of SNAP-tagged CENP-A. IAA was added as in A. Centromeric TMR-Star intensity represents the fluorescent population of SNAP-tagged CENP-A. Data are presented as mean ± SEM for three independent replicates. *p < 0.05.

FIGURE 5:

CENP-C or CENP-N alone does not affect the salt-extractability of centromeric CENP-A. (A) Schematic of experimental workflow. Cells were treated with different concentrations of KCl to extract CENP-A from chromatin. Cells were then fixed and stained for CENP-A before imaging. (B) IAA alone has no effect on CENP-A extractability. CENP-A extraction with increasing salt concentration in the parental OsTIR1 expressing cell line with no AID tagged proteins. (C–E) Salt extractability curves for three different cell lines where CENP-C, CENP-N, or both were tagged with AID. Cells were left untreated or treated with IAA to degrade AID-CENP-N (C), AID-CENP-C (D), or both (E). Centromeric CENP-A immunofluorescence signal was normalized to the no-IAA, 150 mM KCl signal. Data are presented as mean ± SEM for three independent replicates.

FIGURE 6:

R80A G81A mutation does not affect CENP-A or CATD chimera maintenance at centromeres. (A) Schematic of constructs stably integrated into cells for this experiment. Mustard-colored bars represent CENP-A sequences. Blue-colored bars represent H3.1 sequences. (B) Schematic of labeling scheme used for C. Cells were seeded on coverslips at the same density 5 d before harvest. Cells were labeled with TMR-Star for 15 min on different days so that the fluorescent population of CENP-A is diluted to different degrees when cells are harvested at day 0. (C) Centromeric TMR-Star intensity was determined by microscopy. Signals across the different cell lines were normalized to the intensity at 0 d since labeling. Intensities were not background subtracted because of variable nuclear background. There are no significant differences between the four curves. Data are presented as mean ± SEM for three independent replicates.