Multiple Pathways of Replication Restart

Figure 7 Model Replication Forks Used for Substrates to Detect Replication Fork Helicase Loading. If a fork helicase is loaded, the parental strands will be unwound.

Because the key to the establishment of a replication fork is the loading of DnaB onto the lagging-strand template, we first assessed the ability of various combinations of restart primosomal proteins to load DnaB to the lagging-strand template of model replication fork structures formed from oligonucleotides, as scored by unwinding of the fork (Figure 7). Using this type of assay, we have determined that there are 2 biochemically distinct pathways of DnaB loading: 1 that requires PriA, PriB, DnaT, DnaB, and DnaC and 1 that requires PriC, DnaB, and DnaC. A systematic assessment of the effect of the disposition of the nascent strands at the fork demonstrated that these 2 pathways had different substrate preferences: the PriA system prefers substrates where the 3-end of the leading strand is at the fork junction of the template strands and is inhibited as the 3-end is moved away, creating a gap.

	[Enlarge Image] Figure 8 The PriA and PriC replication fork helicase loading systems respond in inverse fashion to gaps in the nascent leading strand. Unwinding of the fork substrate (slow-moving band) is scored by the appearance of the fast-moving single-stranded DNA product.

The PriC system responds in just the opposite fashion: loading is activated when there is a gap between the nascent leading strand and the fork and inhibited as the gap is closed (Figure 8). These observations suggested that the PriA system prefers, as a substrate, stalled forks where the leading strand is ahead of the lagging strand; whereas the PriC system prefers situations where the lagging strand is ahead of the leading strand. Both possibilities presumably occur in the cell, although there has been no direct assessment of this particular issue.

Figure 9 A new substrate for studying replication fork restart. Denaturing alkaline agarose gels separate the long nascent leading strand from the shorter nascent Okazaki fragments.

To examine the replisome-loading activity of these 2 pathways, we developed a new linear template that allows us to dictate the nature of the nascent leading- and lagging-strands at the fork (Figure 9). It consists of a 7 kb-long, linear duplex region with two 38 nt non-homologous ss tails at one end forming a fork structure. We can therefore anneal to the fork any combination of nascent strands by adding appropriate oligonucleotides. The products of the reaction are analyzed on a denaturing alkaline gel where the 7 kb leading strand is easily separated from the Okazaki fragments, which have a median size of about 1 kb.

Both the PriA- and PriC-dependent systems can utilize this template to form replisomes, and their response is identical to what we observed with respect to DnaB loading in the helicase assays: the preferred substrate is one where there is no gap on the nascent leading strand for the PriA system; whereas the PriC system prefers a gap in the nascent leading strand. We are currently investigating the role of Rep in the PriC system, where we suspect that it unwinds the nascent lagging-strand to provide a landing pad for DnaB during replisome assembly. We are also developing methodology to place DNA damage at specific locations in the linear template, thereby allowing us to analyze in precise detail the consequences of a collision between a replisome and template DNA damage.