Assembly of CENP-A into centromeric chromatin requires a cooperative array of nucleosomal DNA contact sites

Abstract

We investigated the requirements for targeting the centromeric histone H3 homologue CENP-A for assembly at centromeres in human cells by transfection of epitope-tagged CENP-A derivatives into HeLa cells. Centromeric targeting is driven solely by the conserved histone fold domain of CENP-A. Using the crystal structure of histone H3 as a guide, a series of CENP-A/histone H3 chimeras was constructed to test the role of discrete structural elements of the histone fold domain. Three elements were identified that are necessary for efficient targeting to centromeres. Two correspond to contact sites between histone H3 and nucleosomal DNA. The third maps to a homotypic H3-H3 interaction site important for assembly of the (H3/H4)2 heterotetramer. Immunoprecipitation confirms that CENP-A self-associates in vivo. In addition, targeting requires that CENP-A expression is uncoupled from histone H3 synthesis during S phase. CENP-A mRNA accumulates later in the cell cycle than histone H3, peaking in G2. Isolation of the gene for human CENP-A revealed a regulatory motif in the promoter region that directs the late S/G2 expression of other cell cycle-dependent transcripts such as cdc2, cdc25C, and cyclin A. Our data suggest a mechanism for molecular recognition of centromeric DNA at the nucleosomal level mediated by a cooperative series of differentiated CENP-A-DNA contact sites arrayed across the surface of a CENP-A nucleosome and a distinctive assembly pathway occurring late in the cell cycle.

Figure 1

Mutations constructed for analysis of the CENP-A histone fold domain. (A) A diagram of the structural organization of the histone fold domain is shown at top over a sequence comparison between CENP-A and histone H3. Below is a table detailing the specific amino acid substitutions for each of the CENP-A mutants analyzed. Only the residues changed in CENP-A are shown. (B) Efficient expression of CENP-A mutants in HeLa cells. Plasmids containing mutant CENP-A sequences were transiently transfected into HeLa cells for analysis of protein expression by Western blotting with mAb 12CA5: (1) wildtype CA-HA1; (2) HN1; (3) HN2; (4) HH1; (5) HSA; (6) HSA5; (7) HSAΔ; (8) HSA3; (9) HH2; (10) HH2.1; (11) HH2.2; (12) HH2.3; (13) HSB; (14) HC. Molecular mass markers are shown (left) with sizes in kD noted.

Figure 2

Sequences required for centromere assembly are located in the central portion of the histone fold domain. Each panel shows a representative nucleus from a transfected cell visualized by confocal microscopy. CENP-A–HA1 derivatives were localized with mAb 12CA5 (green, left) and endogenous centromere proteins were detected with a human autoantiserum (red, right). The center of each panel shows a merge of the two immunofluorescence signals to evaluate antigen codistribution, where yellow indicates colocalization. WT, wild-type CENP-A; NΔ, amino-terminal deletion mutant CA-APA; hI, helix I mutant HH1; sA, strand A mutant HSA; hII, helix II mutant HH2; sB, strand B mutant HSB; C, carboxyl-terminal mutant HC. The lower right panel illustrates the relative positions of these structural elements in CENP-A. Note that CENP-A derivatives that fail to target often required increased exposure to collect an image of CENP-A–HA1 distribution, accomplished by increasing the slit width on the confocal microscope, resulting in an apparent increase in the total nuclear fluorescence intensity as compared with targeting derivatives. Nevertheless, for these nontargeting derivatives, no qualitative differences in distribution were observed between cells that expressed the transfected gene products at the limits of detection vs high level expressors.

Figure 3

Analysis of sequence components within the histone fold structures required for targeting. The strand A and helix II elements, as well as the N-helix of CENP-A, were further dissected to identify discrete sequences required for targeting. A–C show immunofluorescence results. Each panel shows a diagram of CENP-A with the test region highlighted in green over an alignment of the sequence of CENP-A (red) with the corresponding sequence of histone H3, with divergent residues highlighted in green. Typical immunofluorescence results are shown for individual mutants below the alignment. Images are as in Fig. 2, except they are oriented vertically with CENP-A mutants on top (green), endogenous centromeres at bottom (red), and a merge of both signals shown in the center of each group. Yellow indicates colocalization. (A) Strand A subregions. (B) Helix II subregions. (C) N-helix subregions. (D) A chart of the frequency of different localization patterns obtained for four CENP-A derivatives. Wide field images (161 μm × 215 μm) were collected using a ×40 objective. At least 80 transfected cells were examined for each construct and CENP-A–HA1 distribution was scored as described in the text. The frequency of cells in which the test protein was observed to be primarily localized (left group), detectably localized (center group), or unlocalized (right group) to the centromeres is plotted. A legend at the top left specifies each construct as it appears in Fig. 1. For control constructs WT and HC, there was a close correlation between total fluorescence intensity and the degree of nucleoplasmic staining, indicating that mislocalization of these proteins was due to overexpression.

Figure 4

Self-association in vivo indicates that CENP-A nucleosomes are homotypic. (A) Inducible expression of CENP-A– HA1 in stably transformed HeLa cells. HeLa tTA-CAHA cells were grown in the presence (U) or absence (I) of tetracycline for a period of 2 d. Total cell protein was then analyzed by Western blotting with mAb 12CA5 (left) and with a human anticentromere serum (right). The positions of CENP-A and the plasmidderived CENP-A–HA1 are noted (right). Induction results in accumulation of CENP-A–HA1. (B) HeLa tTA-CAHA cells were induced by removal of tetracycline as in A. A soluble chromatin fraction was prepared from isolated nuclei for immunoprecipitation as described in Materials and Methods. After immunoprecipitation with mAb 12CA5, fractions were analyzed by Western blotting with human anticentromere serum to reveal both endogenous and epitope-tagged CENP-A. Fractions are: nuclei, whole nuclei; chromatin, micrococcal nuclease solubilized nuclear extract; IP-sup, supernatant after immunoprecipitation; IP, proteins recovered by immunoprecipitation; -Ab-sup, supernatant from mock immunoprecipitation without mAb 12CA5; -Ab, proteins recovered by mock immunoprecipitation without mAb 12CA5. (C) Western blot analysis of an immunoprecipitation experiment after induction of CENP-A–HA1 to levels higher than endogenous CENP-A. Fractions are as in B.

Figure 5

Restriction of CENP-A expression to S phase abolishes centromeric targeting. (Top) Construction of plasmid pCA-TAG, which expresses CENP-A under the regulatory elements of a replication-dependent histone H3 gene, MH3.2-614. The construct was prepared by replacing the intronless histone H3 coding region with the CENP-A cDNA coding region, adding an HA-1 epitope (green) and maintaining both the promoter (arrow) and 3′ untranslated (loop) regulatory components of MH3.2-614. (Bottom) Results of immunofluorescence analysis of cells transiently transfected with pCA-TAG, with CENP-A–HA1 on the left (green), endogenous centromeres on the right (red), and a merge of the two signals in the center. Localization of CENP-A– HA1 at centromeres could not be detected in cells expressing the protein even at the lower limits of detection.

Figure 6

CENP-A expression occurs late in the cell cycle, after histone H3, and may be driven by a cell cycle–dependent promoter element. HeLa cells were synchronized at the G1/S boundary using a double thymidine block, and then released into S phase. Samples were collected at 2-h intervals for a period of 16 h, and total RNA was isolated. (A) CENP-A mRNA was detected using an RNase protection assay with a probe spanning the 5′ end of the cDNA, resulting in a protected band of the predicted size, 158 bp. A second, shorter protected band that paralleled the main band in abundance was observed, probably corresponding to an alternative transcription start site or promiscuous digestion of the 5′ end of the probe/transcript hybrid. Lanes are: M, markers; A, asynchronous culture; 0–16, time points in hours after release; P, undigested probe. (B) RNA was also analyzed by Northern blotting and probed with a histone H3 coding region probe. Samples correspond to the time course in A and are aligned under the appropriate lanes. (C) Signals from A and B were quantitated using a phosphorimager. The relative abundance of histone H3 (grey) and CENP-A (black) transcripts, using the lowest value for each transcript as the baseline, is plotted as a function of time. (D) Sequence of a segment of genomic DNA flanking the first exon of CENP-A, aligned with cell cycle–regulatory elements identified for three late S/G2-regulated genes. Regions of homology are boxed, and the distance of the last displayed nucleotide from the transcription start site, or the 5′ end of the cDNA for CENP-A, is shown at right.

Figure 7

Model of CENP-A targeting elements. (A) A model of a (CENP-A/H4)₂ heterotetramer. The sequences of CENP-A and histone H4 were modeled onto atomic coordinates of dTAF_II and projected as an alpha carbon backbone trace. Histone H4 (white) and CENP-A (green) are shown with targeting features highlighted: N-helix (peach), strand A (yellow), NH₂ terminus of helix II (tan), and COOH terminus of helix II (orange). The image in the top left is a “front” view perpendicular to the superhelical axis, while the top right shows a view down the superhelical axis. Below is a “bottom” view showing the dyad axis. DNA winds spool-like around the core starting from the left of the N-helix in the top right view, over the top to contact the strand A–N helix II segment of the “back” copy of CENP-A, across the dyad axis and up across the “front” copy strand A–N helix II segment, and over the top again to exit along the back N-helix. Compare with figures in Arents and Moudrianakis (1993). Images were generated using rasmol (Sayle and Milner-White, 1995). (B) Diagram of proposed DNA–CENP-A contacts in the nucleosome core particle. The 146-bp core particle associated DNA is shown. Histone H3 contact sites mapped by cross-linking (Mirzabekov et al., 1978) are shown as black bars, illustrating symmetric contacts made with each DNA strand by the two copies of histone H3 in the core particle. Approximate H3–DNA contact sites observed by x-ray crystallography (Richmond et al., 1993) are shown as grey bars, and the dyad axis is indicated by an orange line at the center. The structural segments of CENP-A involved in these contacts are denoted by colored bars as in A. (Arrowheads) Location of structural distortions in nucleosomal DNA noted by Richmond et al. (1984) and reviewed by Wolfe (1995), and thought to play a role in establishing the translational position of DNA on the histone octamer. Note the congruence of CENP-A targeting elements with DNA contact sites and their symmetry around the dyad axis.