Functional cooperation between FACT and MCM is coordinated with cell cycle and differential complex formation

Abstract

Background: Functional cooperation between FACT and the MCM helicase complex constitutes an integral step during DNA replication initiation. However, mode of regulation that underlies the proper functional interaction of FACT and MCM is poorly understood.

Methods & results: Here we present evidence indicating that such interaction is coordinated with cell cycle progression and differential complex formation. We first demonstrate the existence of two distinct FACT-MCM subassemblies, FACT-MCM2/4/6/7 and FACT-MCM2/3/4/5. Both complexes possess DNA unwinding activity and are subject to cell cycle-dependent enzymatic regulation. Interestingly, analysis of functional attributes further suggests that they act at distinct, and possibly sequential, steps during origin establishment and replication initiation. Moreover, we show that the phosphorylation profile of the FACT-associated MCM4 undergoes a cell cycle-dependent change, which is directly correlated with the catalytic activity of the FACT-MCM helicase complexes. Finally, at the quaternary structure level, physical interaction between FACT and MCM complexes is generally dependent on persistent cell cycle and further stabilized upon S phase entry. Cessation of mitotic cycle destabilizes the complex formation and likely leads to compromised coordination and activities.

Conclusions: Together, our results correlate FACT-MCM functionally and temporally with S phase and DNA replication. They further demonstrate that enzymatic activities intrinsically important for DNA replication are tightly controlled at various levels, thereby ensuring proper progression of, as well as exit from, the cell cycle and ultimately euploid gene balance.

Figure 1

Subassemblies of MCMs form distinct, DNA unwinding-competent complexes with FACT. (A) Western blot analysis of HeLa whole cell extracts (left panel) and nuclear extracts (right panel), as well as the different immunocomplexes targeted by control (2B12), α hSpt16p (8D2), and α SSRP1 (10D1) mAbs. Immunoblotting was done using the indicated antibodies against pan-MCM or individual subunits. The amount of the Input is equivalent to 1/40 the IP. The identity of the protein band, marked by the asterisk, is unknown. The chart summarizes the constituents of the different FACT immunocomplexes. (B) Immunocomplexes were isolated as in (A) and subjected to DNA helicase assay. The reaction was conducted on a radiolabeled, 17-mer oligonucelotide annealed on the M13 single-stranded DNA ("HD", heteroduplex DNA). Upon protein removal, reaction mixtures were resolved by native gel. The locations of the annealed and displaced (ssDNA) substrates on the gel are shown. Displacement of the annealed substrate by heat denaturation is also shown (Boiled). (C) The DNA helicase activity of the FACT-MCM complex is ATP-dependent. The displacement of 17-mer oligonucleotide (SS) from the heteroduplex substrate (HD) by the 10D1-immunocomplex was assayed in the presence (lane 2) or absence (lane 3) of ATP, or in the presence of ATP-γ S (lane 4). (D) HeLa cell extracts were prepared from control (lanes 2 & 3), MCM4^RNAi (lane 4), or MCM3^RNAi (lane 5) cells. The displacement activity of the mock- (lane 2) or 10D1- (lanes 3-5) immunocomplex isolated from these extracts is shown.

Figure 2

Fractionation profile of distinct FACT-MCM complexes. HeLa nuclear extracts (Figure 1A) were subjected to gel filtration chromatography using Sephacryl S-400. Generated fractions (numbers as indicated) were immuno-probed with the specified antibodies. Selected fractions were further subjected to immunoprecipitation using 10D1 and 8D2, and subsequently probed with the indicated antibodies. Peaks 1, 2, and 3, as shown in the 10D1 IP and 8D2 IP panels, denote the approximate fractions in which copurification of MCM with FACT was observed.

Figure 3

Functional attributes of the two distinct FACT-MCM assemblies. (A) HeLa WCE was subjected to immunoprecipitation with a control (lane 2), 8D2 (lane 3), or 10D1 (lane 4) antibody. Lane 1 is the extract input for the IP (1/40). Existence of pre-RC components (ORC1 and Cdc6) and Cdc45 in these immunoprecipitates was detected by specific antibodies. (B) ChIP was performed as described in Methods. Sonicated chromatin fragments were prepared from cells at different stages: asynchronous (lanes 1, 4, 7 & 10), G₁/S (lanes 2, 5, 8 & 11), and G₂/M (lanes 3, 6, 9 & 12). Immunoprecipitation was done with either control (IgG, lanes 7-9), 8D2 (lanes 4-6, 10-12), or 10D1 (lanes 1-3) antibody. Products from final PCR analysis using primers specific to lamin B2 origin (lanes 1-9) or to a non-transcribed region (lanes 10-12) were resolved by 1.5% agarose gel. Precipitated products are shown in the upper panels, DNA input in the lower. (C) Degree of origin association of the two immunocomplexes was quantitatively determined by the normalized ratio of amplified origin sequence, with the ratio of the asynchronous immunoprecipitate represented as 1. The histograms summarize such calculation and compares the origin binding of the FACT immunocomplex (8D2, left; 10D1, right) at each cell cycle stage. Data are averaged ± standard deviations of three independent experiments. (D) ChIP was performed as in (B). Sonicated chromatin fragments were prepared from control, MCM3^RNAi, and MCM4^RNAi cells (HeLa). Immunoprecipitation was done with either 8D2 (top panel) or 10D1 (middle panel) antibody. Presence of origin fragment was detected by PCR. PCR products from input DNA are shown in the bottom.

Figure 4

FACT-associated MCM4 undergoes a cell cycle-dependent phosphorylation change. (A) HeLa cells were immunoprecipitated with mAb 8D2 (lanes 1 and 2) and 10D1 (lanes 3-6). Immunoprecipitates were further subjected to buffer (lanes 1, 3, and 5) or alkaline phosphatase treatment (lanes 2, 4, and 6). The presence of MCM proteins was detected by Western blot analysis using anti-pan MCM (lanes 1-4) and α MCM4 (lanes 5 and 6) antibodies. The positions of MCM 3/4/6 are distinguished by arrowheads. The smear of immunoreactive signlas for MCM4 is also highlighted. (B) FACT-associated immunocomplex was isolated by α SSRP1 mAb 10D1 from HeLa cells at different cell cycle stages: exponentially growing (lane 1), G₁/S phase (lane 2), S phase at 1 hr after double-thymidine block (lane 3), S phase at 3 hr (lane 4), and M phase (lane 5) and G₂ (lane 6). The presence of MCMs and FACT heterodimers was detected by the specified antibodies. The top four panels are immunoblots for the immunoprecipitates, and the bottom two panels are for the lysate inputs. The two anti-Mcm4 blots are from different exposure lengths of signal development (longer exposure on top).

Figure 5

Catalysis and formation of the FACT-MCM complex is regulated in a cell cycle-dependent manner. (A) FACT-associated immunocomplexes were isolated by mAbs 10D1 (left panel) and 8D2 (right panel) from HeLa cells at different cell cycle stages: exponentially growing (As), G₁/S phase, and M phase and G₂. The immunoprecipitate was subsequently subjected to DNA helicase assay as in Figure 1B. Displacement of the annealed substrate by heat denaturation is also shown ("B"). (B) Indirect immunofluorescence analysis was performed on HeLa cells to observe localization of endogenous MCM4 and SSRP1 using the respective antibodies. Nuclear DNA was counter-stained by DAPI. Cells in indicated phases of the cell cycle were examined: early G₁, G₁/S, mid- and late- S phase (cells at 2 and 4 hours after G₁/S release, respectively). Images were generated by laser scanning confocal microscopy.

Figure 6

The FACT-MCM interaction is reduced in response to growth inhibition. Immunoblot showing the levels of FACT, MCM4, and MCMs in K562 cell extracts (lanes 1 & 2) and in 10D1 immunoprecipitates (lanes 3 & 4). Before extraction, K562 cells were either uninduced (lanes 1 & 3) or induced to differentiate with 3-day treatment of 2 mM sodium butyrate (NaBu) (lanes 2 & 4). The number at the bottom illustrates the relative levels of MCM co-immunoprecipitated by the 10D1 antibody. The degree of co-IP ("relative MCM co-IP"), expressed as a ratio in band intensity of total MCM to FACT, was normalized to the control treatment group.

Figure 7

Hypothetical paradigms of the regulatory mechanisms underlying the functional cooperation between FACT and MCM complexes. (A) Putative models for the distinct roles of the two identified FACT-MCM complexes. As shown in Figure 3, the two assemblies may be recruited to the origin region in a temporal fashion, with FACT-MCM2/3/4/5 preceding FACT-MCM2/4/6/7. Both complexes are enzymatically active, and they may exert their activities, in two possible modes (I & II), at distinct and/or sequential stages during the early phase of DNA replication initiation (see Discussion). (B) The FACT-MCM complex constitutes an important activity that facilitates chromatin DNA unwinding during DNA replication [29]. Functional attributes of the FACT-MCM complexes are coordinated with the cell cycle. The temporal mode of such regulation is depicted in this diagram, in which the relative levels of the indicated functional attributes are represented. Origin association of the two FACT-MCM complexes reflects their distinct roles during origin establishment and replication initiation (Figure 3). DNA helicase activity and protein mitotic phosphorylation are inversely regulated to ensure proper integration and resetting of the complex activity within the cell cycle (Figures 4 & 5). Finally, the formation of the FACT-MCM complex is associated with cell proliferation and S phase (Figure 6). ND, characteristics of the indicated functional attributes in the quiescent state (or upon exit from cell cycle) have not been determined.