Regulation of origin recognition complex conformation and ATPase activity: differential effects of single-stranded and double-stranded DNA binding

Abstract

The Saccharomyces cerevisiae origin recognition complex (ORC) is bound to origins of DNA replication throughout the cell cycle and directs the assembly of higher-order protein-DNA complexes during G(1). To examine the fate of ORC when origin DNA is unwound during replication initiation, we determined the effect of single-stranded DNA (ssDNA) on ORC. We show that ORC can bind ssDNA and that ORC bound to ssDNA is distinct from that bound to double-stranded origin DNA. ssDNA stimulated ORC ATPase activity, whereas double-stranded origin DNA inhibited the same activity. Electron microscopy studies revealed two alternative conformations of ORC: an extended conformation stabilized by origin DNA and a bent conformation stabilized by ssDNA. Therefore, ORC appears to exist in two distinct states with respect to its conformation and ATPase activity. Interestingly, the effect of ssDNA on these properties of ORC is correlated with ssDNA length. Since double-stranded origin DNA and ssDNA differentially stabilize these two forms of ORC, we propose that origin unwinding triggers a transition between these alternative states.

Fig. 1. ORC binds ssDNA in an ATP-independent manner. ORC EMSAs were performed using a radiolabeled 244 bp dsDNA containing a wild-type ARS1 origin (wt dsDNA, lanes 1–5), a labeled 295 nt ssDNA containing wild-type ARS1 (ssDNA, lanes 6–13) and a labeled 244 bp dsDNA containing ARS1 with a mutation in the ORC-binding site (mut dsDNA, lane 14). ATP (50 µM final) was added to reactions in lanes 2–5 and 10–14. Two-fold titrations of ORC resulted in 2.5 ng (lanes 2, 6 and 10), 5 ng (lanes 3, 7 and 11), 10 ng (lanes 4, 8 and 12) or 20 ng of ORC per reaction (lanes 1, 5, 9, 13 and 14). The binding reactions were electrophoresed on a native polyacrylamide gel to separate bound and unbound DNA.

Fig. 2. ORC–ssDNA binding and ORC–dsDNA binding are mutually exclusive. The ssDNA and dsDNA EMSA was repeated in the presence or absence of one of three unlabeled competitor DNAs: an ssDNA M13 circle (lanes 3, 7 and 10), wild-type dsDNA (lanes 4, 8 and 11) and mutant dsDNA (lanes 5 and 12). Competitor DNAs were present at a 10-fold molar excess over the labeled DNA, and ATP (50 µM) was included as indicated.

Fig. 3. ORC preferentially binds longer ssDNAs. (A) A 295 nt end-labeled ssDNA was incubated with or without ORC. Anti-ORC polyclonal sera and beads coupled to protein G were used to immunoprecipitate ORC and the associated ssDNA. ‘B’ and ‘F’ represent the bound and free DNA, respectively. (B) ORC was incubated with end-labeled ssDNA that was 113 or 295 nt long, and was precipitated as described. ORC binds these two (full-length) molecules with similar efficiencies under these experimental conditions. (C) ssDNAs of 113 and 295 nt were cleaved with S1 nuclease to generate random populations of ssDNA molecules of different lengths (input DNA, I). The resulting ssDNA was then incubated with ORC and immunoprecipitated as before. Input, bound and free DNAs were electrophoresed on a denaturing polyacrylamide gel. ssDNA lengths (in nucleotides) are shown on the right. (D) Schematic representation of the 113 and 295 nt ssDNAs (dashed lines) with respect to the pARS1/WT sequence (Marahrens and Stillman, 1992). The ARS1 ACS, B1, B2 and B3 elements are indicated by boxes labeled A, 1, 2 and 3, respectively. The 295 nt ssDNA encodes the top (A-rich) strand of ARS1, whereas the 113 nt ssDNA encodes the bottom strand. The 3′ end of each ssDNA is indicated by an arrowhead, and the radiolabeled 5′ end is indicated with an asterisk. The 90 nt position is indicated on each oligonucleotide with vertical arrows.

Fig. 4. ssDNA stimulates ORC ATPase activity and induces a conformational change in ORC. (A) Apparent association constants of ORC for ssDNA oligonucleotides of indicated lengths were determined using EMSA. The K_A was calculated by taking the inverse of the concentration of free ORC at half-maximal binding. Lengths in nucleotides are indicated below each oligonucleotide. (B) The rate of ATP hydrolysis by ORC was measured in the absence of DNA, in the presence of origin-containing dsDNA (the 244 bp dsDNA, see Materials and methods) or in the presence of oligonucleotides of various lengths. ATPase rates were normalized to the rate of hydrolysis seen in the absence of DNA. The averages and standard deviations for three experiments are shown. (C) Proportion of ORC molecules in the bent conformation as determined by electron microscopy (Figure 5). ORC was examined in the absence of DNA, in the presence of dsDNA or in the presence of ssDNA of various lengths. The total number of ORC molecules counted in at least two experiments is shown below each bar.

Fig. 5. Electron microscopy of ORC. Transmission electron microscopy of ORC was performed on samples negatively stained with uranyl acetate in the absence of DNA (A and C) or in the presence of the 96 nt DL15 ssDNA (B and D). Low magnification images show that, in the absence of DNA, ORC is primarily in a straight or extended conformation (A), whereas in the presence of ssDNA, many complexes adopt a bent or curved conformation (B). (C) High magnification images of straight ORC molecules in the absence of DNA. (D) High magnification images of bent ORC molecules in the presence of ssDNA. The scale bar represents 40 nm for (A) and (B) and 17 nm for (C) and (D).

Fig. 6. ssDNA length is correlated with the strength of ORC–ssDNA interactions. Oligonucleotides made up of multiple copies of a non-origin DNA (50% GC bases) were used to determine the apparent association constant with ORC (A) or the effect on the rate of ORC ATP hydrolysis (B) as described in the legend to Figure 4. The GC-45, GC-55 and GC-65 oligonucleotides were 45, 55 and 65 nt long, respectively. The longer oligonucleotides each include all of the sequences of the shorter oligonucleotides with 10 nt of additional ssDNA.