Characterization of cell lines. (A) Structure of Mcm-fluorescent protein (Mcm-mEm) fusion proteins. Each Mcm-mEm contains a full-length Mcm2-7 cDNA (Kimura et al., 1996), followed by a flexible linker, the fluorescent protein, a second flexible linker, and an optimized peptide substrate (BLT) for the E. coli biotin ligase enzyme (BirA). (B) Homogenous expression of Mcm4-mEm. After transfection of the expression cassette shown in A, cells were selected in the presence of dox to establish cell lines in the absence of Mcm-mEm expression. Dox was then removed from aliquots of each cell line for 48 h to induce tagged protein. Shown is Mcm4-mEm fluorescence merged with a phase contrast image. Bar, 10 µm. (C) Cell cycle regulation of soluble and insoluble fractions of endogenous and tagged Mcm4-mEm. Cells were synchronized in mitosis by shake-off and detergent-extracted at the indicated times after replating. Aliquots of cells were plated into aphidicolin for 16 h to arrest cells at the G1/S phase border (G1/S Aph), and a portion of those cells were released into S phase for 6 h (G1/S+6). The soluble and insoluble fractions (Fig. S4 A) were subjected to immunoblotting with the indicated antibodies. Both tagged and endogenous Mcm4 are completely soluble and exhibit a molecular weight shift during mitosis, as expected (Pereverzeva et al., 2000; Okuno et al., 2001). During G1, both tagged and endogenous Mcm4 present as a doublet band when sufficient care is taken to inhibit phosphatases, as expected (Komamura-Kohno et al., 2006). Insoluble PCNA tracks S phase; note that aphidicolin arrest results in increased detergent extractability of PCNA. β-Tubulin and LaminB are used as loading controls for the soluble and insoluble fractions, respectively. Both tagged and endogenous Mcm4 are reduced in the insoluble fraction 6 h after release from aphidicolin, as expected (Okuno et al., 2001). (D) Autogenous regulation of Mcm4-mEm. Mcm4-mEm–expressing cells were grown in the indicated concentrations of dox, and whole cell extracts were subjected to immunoblotting with an anti-Mcm4 antibody. Anti–β-tubulin was used as a loading control. (E) Coprecipitation of Mcm4-mEm with endogenous Mcm subunits in chromatin from late G1 phase cells. Cells either expressing (+) or not expressing (−) Mcm4-mEm were synchronized in mitosis and collected 8 h after release into G1 phase, and Mcm4-mEm (indicated with a black arrow) was precipitated from the solubilized chromatin fragments (Chromatin-Bound; Fig. S4 A). Whole cell extracts (WCE) from the same cells are also shown. Mcm4-mEm–expressing cells were grown in 0.5 ng/ml dox for 24 h before harvesting to eliminate all cytoplasmic Mcm4-mEm expression. Note that endogenous Mcm4 did not pull down with the tagged Mcm4, which indicates that double hexamers containing both tagged and untagged subunits are rare under these conditions, but this does not imply that they do not exist. Their abundance may depend on the ratio of tagged and untagged proteins bound to chromatin (e.g., see Fig. S4 C). (F) RFP-PCNA displays cell cycle–dependent punctate patterns in Mcm-mEm cell lines. Shown are cells in very early (initial appearance of PCNA foci), early, mid, or late S phase displaying the characteristic spatial patterns of PCNA (red) at sites of ongoing DNA synthesis. Note that Mcm4-mEm (green) is distributed heterogeneously in very early S phase but more homogeneously as S phase proceeds. PCNA is distributed uniformly in both G1 and G2 phase (Fig. 2, B and C). Bars, 10 µm.

Cell cycle dependence of Mcm4 recovery. Mcm4 FRAP recoveries vary significantly throughout the cell cycle. The IF, calculated from the FRAP intensity 30–60 s after bleaching, is minimal after mitosis, but gradually increases to >60% as G1 phase progresses. Once S phase begins, the IF begins to drop, returning to 0% by G2 phase. Inset graphs show the time (x axis) in log scale. n, number of cells analyzed to calculate the IF and standard deviation of the mean. The gray bars indicate the initial appearance of PCNA foci.

Mcm4 is cumulatively loaded onto chromatin throughout G1 phase. (A) A cell line stably expressing Mcm4-mEm and H2B-mCherry fusion proteins was established, and FRAP experiments were conducted at various times after mitosis (operationally defined here as the time at which Mcm4 was observed to reenter the cell nucleus after nuclear membrane reformation). H2B, Mcm4, and overlay images are shown for five representative cells (i–v) at different time points after mitosis. Bars, 10 µm. (B) FRAP curves and IFs for H2B-mCherry and Mcm4-mEm are shown for the five representative cells in A. (C) Using the Mcm4-mEm/H2B-mCherry cell line, 71 cells underwent FRAP experiments between 0 and ∼500 min after mitosis, and the IF was calculated for both Mcm4 and H2B. Although it is difficult to track the entire length of G1 phase by this method, a linear slope fit (broken line) to all 71 data points would extrapolate to ∼70% Mcm4 loaded by 700 min, which is close to the length of G1 phase and is consistent with cells deemed to be in very late G1 phase by H2B colocalization (Fig. S5 B).

Other Mcm subunits also display a cell cycle dependence of the IF. Mcm2-, Mcm3-, Mcm4-, Mcm6-, or Mcm7-mEm cells were tracked out of mitosis, and FRAP experiments were conducted at various times in the cell cycle. The immobile Mcm fraction was calculated from each FRAP curve, and the data were binned according to the subunit and the number of minutes after mitosis. The number of cells analyzed for each cell cycle stage and the standard deviation of the mean are shown at the top of each bar (error bars).

Protocols for cell cycle–specific FRAP in the absence of drug synchrony. (A) A cell expressing Mcm4-mEm was imaged emerging from mitosis (top). FRAP experiments were conducted on both sister cells ∼5 h later (bottom). (B) A late G1 cell expressing Mcm4-mEm was imaged before a FRAP bleach (Pre-bleach), and at different time points after the bleach to measure recovery. PCNA foci appear between 30 and 60 min, which indicates entry into S phase. (C) A late S phase cell expressing Mcm4-mEm was imaged until all PCNA foci had cleared, followed by a FRAP experiment in the subsequent G2 phase. PCNA foci disappear between 70 and 135 min, which indicated entry into G2 phase. Note that for live cell images, exposure time is minimized to 1/10th of a second, so resolution is compromised. Bars, 10 µm.

Mcm4 recovery results from eviction during replication. (A) The IF was measured from a set of short FRAP experiments (each lasting <1 min) on a pool of cells at various stages of progression through S phase. The mean IF gradually decreased and was well fit by a line or single weak exponential decay (broken line), yielding a decay time of ∼13.5 h. The number of cells analyzed for each time interval is shown above the bar. (B) Long FRAP experiments (>150 min each) were performed on 10 cells in G1 phase that were all ∼100 min from entering S phase. The mean FRAP recovery after S phase entry was well fit with a line or single weak exponential (broken line), yielding a recovery time of ∼13.6 h. The consistency of this recovery time and the IF decay time suggests that Mcm4 undergoes little or no exchange throughout S phase. The gray bar indicates the appearance of PCNA foci. (C–F) Replication arrest prevents Mcm4 unloading. (C) Cells were cultured in roscovitine for 24 h, and a cell lacking PCNA foci but displaying heterogeneous Mcm4-mEm distribution (indicating G1 arrest) was subjected to FRAP and tracked for 403 min with no detectable IF recovery. (D) Cells were cultured in aphidicolin for 24 h, and a cell displaying PCNA foci characteristic of very early S phase as well as heterogeneous Mcm4-mEm distribution (Fig. 1 F) was subjected to FRAP and tracked for 296 min with no detectable IF recovery. (E) A cell cultured in aphidicolin for 24 h and displaying PCNA foci characteristic of early S phase and homogeneous Mcm distribution was subjected to FLIP/FRAP, as described in the text. The FLIP laser was directed in the broken yellow circle, and the FRAP bleach was directed along the solid yellow line. Similar results were obtained with roscovitine-arrested cells. Note that, although we observed a substantial reduction in the intensity of PCNA foci during aphidicolin arrest, as described previously (Görisch et al., 2008), characteristic focal patterns of PCNA were still discernable. Bars, 10 µm. (F) Quantification of the average bound fractions for groups of cells treated and analyzed as illustrated in C–E, and imaged for a minimum of 200 min. The number of cells analyzed and the standard deviation of the mean are shown above each bar (error bars). FRAP curves similar to those shown in Fig. 3 showed no recovery of the IF throughout the imaging period, so the bound fraction at the end of the entire imaging period is shown for simplicity. Note that for FLIP/FRAP of cells in early to mid S phase (ES/MS), the IF is actually considerably lower, but is measured after a significant fraction of the soluble molecules have been bleached.

Sites of DNA synthesis are sites of low chromatin and Mcm4 density. (A) Lack of colocalization of PCNA and Mcm proteins in living cells. Mcm4-mEm/RFP-PCNA–expressing cells in very early S phase were subjected to FLIP to reduce the fluorescence of the soluble pool of molecules as in Fig. 4. After FLIP, colocalization analysis of the immobile Mcm4-mEm and RFP-PCNA fractions was performed, and the correlation coefficient is indicated in the Post-FLIP overlay. (B and C) Similar FLIP colocalization analysis for RFP-PCNA/H2B-mCherry (B)- and Mcm4-mEm/H2B-mCherry (C)-expressing cells. The yellow circles represent where the FLIP laser (488 nm) was directed. (D) Strip FRAP was performed in cells stably expressing Mcm4-mEm/RFP-PCNA (top), Mcm4-mEm/H2B-mCherry (middle), and H2B-eGFP/RFP-PCNA (bottom) in early S and G2 phase. In each experiment, the IF was measured along the length of the bleach strip (third column), and this was correlated (via Pearson’s correlation coefficient, fourth column) with the prebleach fluorescence along the bleach strip (first two columns). Mean results are shown together with the standard error of the mean (error bars with number of cells for each displayed) in the bar plot below. Bars, 10 µm.