Selective bypass of a lagging strand roadblock by the eukaryotic replicative DNA helicase

Abstract

The eukaryotic replicative DNA helicase, CMG, unwinds DNA by an unknown mechanism. In some models, CMG encircles and translocates along one strand of DNA while excluding the other strand. In others, CMG encircles and translocates along duplex DNA. To distinguish between these models, replisomes were confronted with strand-specific DNA roadblocks in Xenopus egg extracts. An ssDNA translocase should stall at an obstruction on the translocation strand but not the excluded strand, whereas a dsDNA translocase should stall at obstructions on either strand. We found that replisomes bypass large roadblocks on the lagging strand template much more readily than on the leading strand template. Our results indicate that CMG is a 3' to 5' ssDNA translocase, consistent with unwinding via "steric exclusion." Given that MCM2-7 encircles dsDNA in G1, the data imply that formation of CMG in S phase involves remodeling of MCM2-7 from a dsDNA to a ssDNA binding mode.

Figure 1. CMG causes leading strand stalling ~20 nucleotides from a DNA inter-strand crosslink

(A) Cartoon depicting the DNA sequence surrounding the ICL in pICL^Inter. Red line, inter-strand cross-link. Blue arrow, StuI cleavage site, which is used to map leftward leading strands in (C). The sequence of the longest leading strand detected at the 10 minutes time point (see C) is shown in red letters, and the product generated after the leading strand advances towards the ICL after 15 min (see C) is shown in green letters. Blue letters, sequence differences between pICL^Inter and pICL^Intra. (B) Same as (A), except for pICL^Intra. Red bracket, 1,2 intra-strand cross-link. The sequence of the longest leading strand seen in (C) is shown in red letters. (C) Mapping leading strands near DNA inter- and intra-strand cross-links. pICL^Inter (lanes 1-8) or pICL^Intra (lanes 9-16) was incubated sequentially in HSS and NPE containing [α–³²P]dATP. At the indicated times after NPE addition, replication intermediates were digested with StuI, and separated on a DNA sequencing gel alongside a sequencing ladder generated with primer M (see Figure S1A). The distance of sequencing products from the bold G in panel (A) and T in panel (B) is indicated on the right. (D) Kinetics of Mcm7 binding to an ICL. pICL^Inter was replicated as in (A) but lacking radioactivity, and samples were withdrawn for Mcm7 ChIP using ICL proximal (pink) and control (purple) primer pairs (see plasmid cartoon). The relative ChIP signal adjacent to the ICL (pink circles) and distal to the ICL (purple triangles) was plotted. In parallel reactions containing [α–³²P]dATP, replication intermediates were digested with AflIII, and separated on a DNA sequencing gel alongside a sequencing ladder generated with primer S (see Figure S1A). The leading strands stalled between −20 and −40 and at the −1 position (see Figure S1B) were quantified and plotted (blue diamonds and grey squares). Error bars represent the standard deviation of three experiments.

Figure 2. Replisomes converging on an ICL do no interfere with each other

(A) Locations of restriction sites, the nitrogen-mustard-like ICL, biotins, and sequencing primers on pICL^Lead/Lag. Figure S2A presents two alternative scenarios for how replication forks might interact at an ICL. (B) pICL^Lead/Lag was pre-incubated with buffer or streptavidin, as indicated, and replicated in egg extracts in the presence of [α–³²P]dATP. At the indicated times after NPE addition, replication intermediates were digested with Stu I and separated on a DNA sequencing gel alongside a sequencing ladder generated with primer M. The distance of products from the ICL is indicated on the left of the gel. White arrows on the DNA sequencing ladder indicate the location of biotins. Red bracket, leading strand arrest 24-50 nt from the ICL in the absence of SA. Orange bracket, leading strand arrest 70-80 nt from the ICL in the presence of SA (30-40 nt from the outermost biotin). Green arrow (−1 postion), leading strands that have advanced to the ICL. Black arrow (−41 position), leading strands that have advanced to the outermost biotin-SA complex. Figure S2B shows that DNA polymerase ε can advance to within one nt of a biotin-SA complex on the leading strand template. (C) pICL^Lead/Lag was pre-incubated with buffer or streptavidin, as indicated, and replicated in egg extracts in the presence of [α–³²P]dATP. At the indicated times after NPE addition, replication intermediates were digested with AflIII (see Figure 2A) and separated on a DNA sequencing gel alongside a sequencing ladder generated with primer S. Products of the rightward fork are shown. The leftward fork was efficiently arrested by the biotin-SA (see Figure S2C).

Figure 3. Biotin-SA complexes located on the leading strand template but not on the lagging strand template arrest the replisome

(A-F) The 3′ to 5′ ssDNA translocation (A, B, C) and dsDNA translocation models (D, E, F) for CMG make different predictions regarding how the leftward moving replisome (CMG, green; DNA polymerase, grey) will interact with SA molecules bound to pICL^Lead/Lag, pICL ^Lead, or pICL^Lag (see main text). On all plasmids, the rightward replisome (not shown) will be prevented from approaching the biotin-SA complexes by the ICL. The yellow line in (D, E, F) represents the “ploughshare” postulated to split the duplex as it emerges from the central channel. (G) pICL^Lead/Lag, pICL ^Lead, or pICL^Lag was pre-incubated with buffer or streptavidin, as indicated, and replicated in the presence of [α–³²P]dATP as in Figure 2B. SA was not displaced from pICL^Lag during replication (Figure S3).

Figure 4. Single-molecule analysis of replisome collision with leading and lagging strand-specific roadblocks

(A) Reaction scheme for the replication of immobilized λ DNA in a microfluidic flow cell using a single pair of diverging replisomes. SYTOX Orange and dig-dUTP detection of replicated DNA are indicated schematically. (B, C) λ DNA containing a QDot on the bottom strand (19 kb from the end) or top strand (15 kb from the end), as indicated, was immobilized within a microfluidic flow cell and replicated as depicted in (A). After protein removal, total DNA (SYTOX Orange), dig-dUTP (fluorescein-labeled anti-dig Antibody), and the QDot, were visualized and presented individually or as a merged image. Each dig tract was classified as a rightward or leftward moving fork depending on the location of the origin. If the tract ended within 2 pixels (~ 0.3 μm) of the QDot, it was considered arrested. Cartoons depicting each type of collision, the expected outcome based on the 3′ to 5′ ssDNA translocation model (checkmarks), representative examples of the raw data, and the frequency of each outcome are included. A hypothetical model in which a dsDNA translocase bypasses a lagging strand roadblock is presented in Figure S4.

Figure 5. Model for replication initiation

Model for helicase activation in which MCM2-7 encircles dsDNA in G1 and ssDNA in S phase. See text for details.