Chromatin Immunoprecipitation (ChIP).

Cells (2 × 10⁸) were collected, washed in PBS, and then fixed with 1% paraformaldehyde in PBS for 20 min at 37°C. The fixed cells were washed in PBS, and chromatin was solubilized in urea as described (35). Chromatin was sonicated in a volume of 2 ml to an average DNA length of 600–1,300 bp by using a Heat Systems Ultrasonics W220 at setting 6 with the small probe and 10–15 pulses of 15 s separated by cooling on ice. Crosslinked chromatin with a buoyant density of ≈1.4 gm/ml was purified by equilibrium centrifugation in CsCl as described (36).

Aliquots of chromatin containing about 30 μg of DNA were adjusted to volumes of 50 μl in RIPA buffer (1% Triton X-100/0.1% sodium deoxycholate/0.1% SDS/150 mM NaCl/5 mM EDTA/50 mM TrisHCl, pH 8.0/1 mM PMSF) and incubated with 1.5 μg of rabbit IgG; 1.5 μg of anti-EBNA1 mAb, EBNA1.OT1x (37), obtained from Jaap Middledorp (Free University Medical Center, Amsterdam, The Netherlands); 1.5 μg of rabbit antibody against human Mcm2, affinity-purified by using BM28 antigen (38); or 1 μl of rabbit antiserum specific to Orc2 (39), Orc3 (40), or Orc4 (39) and kept at 4°C for 4 h. Antibody complexes were recovered by adding 10 μl of 50% (vol/vol) protein A-Sepharose (Sigma), rocking for 1 h at 4°C, and pelleting. Pellets were washed five times with RIPA buffer adjusted to 750 mM NaCl, twice in LiCl buffer (0.25 M LiCl/0.5% Triton/1% SDS/1% sodium deoxycholate/1 mM EDTA/10 mM TrisHCl, pH 8.0), twice in TE (10 mM TrisHCl, pH 8.0/1 mM EDTA), and finally resuspended in 100 μl of TE. Recovered chromatin was de-crosslinked by adding SDS to 0.5% and proteinase K to 1 mg/ml and incubating at 37°C overnight followed by 8 h at 65°C. After extracting with 1:1 phenol/chloroform, DNA was recovered by ethanol precipitation in the presence of 20 μg of glycogen as carrier. PCR was performed by using 1/10th to 1/1,250th of the ChIP DNA in 50 μl by using 2.5 units of Taq polymerase (Roche Biochemicals) for 35 cycles of 1 min at 95°C, 1 min at 60°C, and 40 s at 72°C. The PCR products were separated by electrophoresis in 2% agarose gels (NuSieve 3:1, BioWhittaker), which were stained with SYBR Green or SYBR Gold (Molecular Probes) and digitally photographed.

The regions amplified and the sequences (B95–8 strain) of upper and lower strand primers (5′ to 3′) were as follows: DS or ΔDS (for Figs. ​Figs.11 and ​and2),2), 435 bp or 315 bp, EBV 8758–9192, ODJ3 gggggcgtcacctgaaaccttgtttt and ODJ11 gccctcttgagggtccgcttatcggt; DS (for Fig. ​Fig.3),3), 254 bp, EBV 8762–9115, OJY216 gcgtcacctgaaaccttgttttcg and OHX-8987 tgggagatatcgctgttccttaggaccct; E1 5′, 334 bp, EBV 107656–107989, OJY129 aatccccatccctaccgtcca and OJY130 ggccatttccaggtcctgtacc; E1 3′, 248 bp, EBV 109726–109973, OJY127 atcagggtgactgtgtgcagcttt and OJY128 caccaacagcacgcatgatg; Qp, 155 bp, EBV 62398–62552, OJY205 cgctttgcgaaaacgaaagtg and OJY206 gggtccccaaacatacaccg; and hph, 182 bp, 275–456 of GenBank K01193, OJY132 tctccgacctgatgcagctc and OJY131 ttccccaatgtcaagcacttcc. To obtain even, simultaneous amplification by PCR, the primer pairs had to be used at different concentrations, which required adjustment for different chromatin preparations or for use with different thermal cyclers. For the experiments shown, the concentrations were as follows. Fig. ​Fig.1:1: DS, 400 nM; E1 5′, 46 nM; E1 3′, 32 nM; hph, 36 nM. Fig. ​Fig.2:2: Raji, DS, 430 nM; E1 3′, 32 nM; Qp, 260 nM; P3-ΔDS-33:ΔDS, 120 nM; E1 3′, 41 nM; Qp, 600 nM. Fig. ​Fig.3:3: DS, 100 nM; E1 5′, 47 nM.

Open in a separate window

Figure 2

The association of ORC and MCM with oriP requires the replicator, DS. ChIP analysis of 30 μg of chromatin from Raji cells or BL30 cells carrying the EBV mutant P3-ΔDS-33. Preimmune sera (PI) corresponding to the antisera against Orc4 and Orc3 were used for B, lanes 3 and 6, respectively. The amounts of ChIP DNA used for PCR were 1/250th (lanes 1, 3, 5, and 7) and 1/1,250th (lanes 2, 4, and 6) (A) and 1/250th (lanes 2, 3, 5, 6, and 8) and 1/1,250th (lanes 1, 4, and 7) (B). The amounts of DNA tested from input chromatin were for Raji, 0.06, 0.3, and 1.5 ng (lanes 8–10 in A, lanes 9–11 in B) and for P3-ΔDS-33, 0.1, 0.5, and 2.4 ng (A) and 0.05, 0.25, and 1.25 ng (B).

Open in a separate window

Figure 3

Cell cycle dependence of the association of Mcm2 with DS of oriP. DG75 cells carrying p818 were separated by centrifugal elutriation into four populations enriched for cells in G₁, early S, late S, and G₂; 30 μg of chromatin from each was analyzed by ChIP with antibodies against Orc2, Mcm2, or EBNA1, or with nonimmune rabbit IgG, as indicated. Flow cytometric analysis of the cells for DNA content is shown (Left) in the conventional manner. For this experiment, a 254-bp region bordering DS was amplified to detect the presence of DS, rather than the 435 bp spanning DS as done for Figs. ​Figs.11 and ​and2.2. 1/50th (lanes 2, 4, and 6) or 1/250th (lanes 1, 3, and 5) of the recovered DNA were tested. For lanes 8 and 9, 2.4 ng and 12 ng of DNA from input chromatin were tested.

The Association of ORC and MCM with oriP Requires the Replicator DS.

A 120-bp deletion removing most of DS from the EBV chromosome was shown to eliminate detectable origin activity at oriP. During latent infection, the circular chromosome of such ΔDS EBV mutants is replicated exclusively by forks that initiate away from oriP (34). ChIP was performed on crosslinked chromatin from BL30 cells carrying the EBV mutant P3-ΔDS-33 and, for comparison, on chromatin from the EBV-positive cell line Raji, in which oriP is functional (42). For this experiment, the recovery of DS DNA was compared with two regions of the EBV chromosome that are far from oriP, the region 3′ to the EBNA1 gene and the promoter, Qp. EBNA1 binds to two regulatory sites at Qp, so this region was also precipitated by the EBNA1 antibody, although less efficiently than was DS (Fig. ​(Fig.22A, lanes 6 and 7). The ΔDS deletion removed all four EBNA1 binding sites at DS, but chromatin containing the DS deletion was still precipitated by the EBNA1 antibody, with a lower efficiency, because the amplified region is only 1 kb away from FR, to which EBNA1 also binds; for this experiment, the average DNA length after sonication was close to 1 kb.

As shown in Fig. ​Fig.22A, antibody against Orc2 specifically precipitated DS chromatin from Raji cells but did not precipitate the corresponding region of chromatin of the ΔDS mutant. The antibody against Mcm2 caused an enrichment of DS over the two control regions in chromatin from Raji. In contrast, with ΔDS-33 the Mcm2 antibody did not cause any enrichment of the DS region preferentially compared with the two control regions (Fig. ​(Fig.22A, lanes 4 and 5). Qp was recovered as efficiently by the EBNA1 antibody in the case of P3-ΔDS-33 as it was with Raji, so it is unlikely that the failure of the antibodies against Orc2 and Mcm2 to selectively precipitate chromatin containing the DS deletion was because of any fault with the chromatin preparation. The results concerning Orc2 were corroborated by a similar experiment by using antibodies against two other ORC subunits, Orc3 and Orc4. As shown in Fig. ​Fig.22B, both antibodies selectively precipitated DS chromatin prepared from Raji cells but not from cells carrying the ΔDS mutant. Again, the recovery of Qp with the EBNA1-specific antibody validated the ChIP assays. From these experiments, we can conclude that ORC and MCM require DS, the oriP replicator, to associate specifically with this region of the EBV chromosome.

How Might ORC Associate with the Replicator of oriP?

The key issue is where ORC associates at or near DS. Functionally DS is at most 119 bp long and includes two pairs of EBNA1 binding sites (8), with each pair spanning 39 bp that have been shown by in vivo footprinting to be occupied by EBNA1 throughout the cell cycle (ref. 48; H.X. and J.Y., unpublished data). The remaining DS DNA that are not bound by EBNA1, 15 bp between the pairs of EBNA1 sites and at most 13 bp at each end, are too short to accommodate ORC (30), and because these regions include sites that are hypersensitive to cleavage by DNase I, they are unlikely to be bound continuously by proteins in vivo (48). The in vivo footprinting data did not reveal signs of ORC binding immediately flanking DS. If ORC binds to DNA outside of DS (to nonreplicator DNA), then DS must in some manner facilitate ORC binding. Alternatively, if ORC is present at DS, then it must be bound indirectly. This is of interest in either case because it is not known how sites of ORC binding are determined on human chromosomes (33).

In light of this report, considerable circumstantial evidence suggests that EBNA1 is responsible for recruiting ORC to DS or its vicinity. Not only is EBNA1 the only protein that appears to bind to DS directly (48), the EBNA1 binding sites at DS are the only sequences that are essential for replicator activity (8). Replicator function requires the presence of two EBNA1 binding sites that are spaced exactly 21 bp apart center to center (J. Bashaw and J.Y., unpublished data), implying that a precise structure must be formed by two adjacent EBNA1 dimers. EBNA1 represses transcription by binding to Qp (49) and fosters episome segregation by binding to FR of oriP (10–13). However, EBNA1 does not support replication at these sites (refs. 6 and 8 and unpublished data) or appear to recruit ORC (Fig. ​(Fig.22 and unpublished data), which might be because EBNA1 dimers bind at these sites with spacings other than 21 bp. EBNA1 cannot activate replication from oriP in rodent cells, implying a failure of EBNA1 to interact with a necessary host factor. This host factor could be either ORC or something that facilitates ORC binding.

We thank Jaap Middledorp for providing the mAb, EBNA1.OT1x, and David Kowalski and Joel Huberman for insightful discussions and for their comments on the manuscript. This work was supported by National Institutes of Health Grants CA14322 (to J.Y.), CA60499 (to A.D.), and CA16056 (to the Roswell Park Cancer Institute Biopolymer and Flow Cytometry Facilities).