Replicating And Repairing The Genome: From Basic Mechanisms To Modern Genetic Technologies. Kenneth N Kreuzer
brings us back to where we started and completes the cycle (Figure 2.5). During replication of the T7 genome, which is roughly 40,000-base pairs long, this lagging-strand cycle occurs about 50 times on average. Because of the repeated cycle whereby the lagging-strand loop grows and shrinks, this mode of replication has been described as the “trombone model.”
Several features of the trombone model are worthy of note (Figure 2.5). First, during much of the cycle, two different regions of the lagging-strand template are single stranded — part of the loop (a segment that was just unwound by the helicase) and the segment between the active lagging-strand polymerase and the previous Okazaki fragment. Second, the loop starts off as a very small single-stranded segment and grows to include a large stretch each of single- and double-stranded DNA. The duplex DNA within the loop consists of the portion of the current Okazaki fragment that has been completed. Third, each of the single-stranded regions is coated with the ssDNA-binding protein (not shown in Figure 2.5), and so this protein must rapidly exchange and re-equilibrate on the strands as replication proceeds. Fourth, the lagging-strand polymerase remains associated with the replisome for multiple rounds of Okazaki fragment synthesis, rather than exchanging after each cycle as would be expected with the splayed-out replication fork in Figure 2.3. Experimental evidence for this conclusion includes the finding that an ongoing replication reaction continues many rounds of Okazaki fragment synthesis after being diluted extensively, to the point where any remaining free DNA polymerase in solution could not bind to the complex. Fifth, replication of the leading and lagging strands occurs at essentially the same rates, reflecting coordinated synthesis of the two strands.
Figure 2.5. The trombone model of DNA replication. The figure depicts one round of Okazaki fragment synthesis, with only the two DNA polymerases and replicative helicase shown for simplicity. Newly synthesized DNA is in red, and RNA primers (also red) are numbered. Note that the overall structures at the top and bottom are identical, except that one additional Okazaki fragment has been synthesized and the leading strand is extended accordingly. See text for detailed discussion of the steps in this model.
The looping model raises an important question — what happens if DNA polymerization on one of the strands is stalled or blocked? Does the polymerase on the other strand continue synthesizing DNA? This would create a dangerous situation, with a long stretch of ssDNA on the strand with the blocked polymerase and a complete uncoupling of the two DNA polymerases of the replisome. This question has been approached with a very special replication substrate using the T7 replication system (see “How did they test that?” at the end of this chapter). Polymerase extension on the lagging-strand template was blocked by incorporation of a special “chain-terminating” dideoxynucleotide only on that strand. Dramatically, synthesis of both the leading and lagging strands were strongly inhibited, demonstrating that the leading-strand polymerase halts when the lagging-strand polymerase is inhibited. Thus, the replisome is highly coordinated and the polymerases on the two strands somehow communicate with each other to maintain this coordination.
While a looped lagging strand is likely to be generally applicable to the process of DNA replication throughout biology, some of the Figures in the remainder of this book will show the “old-fashioned” splayed-out version for simplicity. Keep this in mind when you consider those simplified figures.
2.6Structural model for the T7 replisome
A recent culmination of structural approaches provides a dramatic and informative three-dimensional model for the T7 replisome. One key study determined the structure of a complex of the helicase/primase with T7 DNA polymerase in the absence of DNA, using X-ray crystallography and other methods (Figure 2.6A). The helicase/primase was in a heptameric form, consistent with the DNA-free form discussed above. Somewhat surprisingly, the complex contained not two but three copies of the DNA polymerase! The authors speculated that the third DNA polymerase might serve as a spare, available to exchange with one of the actively synthesizing polymerases if the latter became blocked or disabled.
More recently, a method called cryo-electron microscopy (cryo-EM) was used to determine the structure of the T7 replisome complexed with DNA that mimics a replication fork (Figure 2.6B and cartoon in Figure 2.6C). In this case, the helicase/primase was in a hexameric form, as expected when bound to ssDNA (see above). As in the previous study using X-ray crystallography, three DNA polymerases were bound to the helicase/primase complex. In the cryo-EM study, the two active polymerases (leading and lagging strand) could be identified, and the third polymerase did appear as a spare, showing no engagement with a DNA primer/template.4
This stunning model of an active replication complex has several notable features that illuminate the replication process. Starting at the parental DNA, the separation point of the parental duplex is actually within the leading-strand polymerase, not the helicase! This came as a surprise to those who expected that the unwinding enzyme would be at the frontline of an unwinding reaction. However, recall that unwinding by the helicase is greatly accelerated when it is in complex with polymerase, and the two proteins clearly work in concert with one another (see above).
Figure 2.6.Current models for the structure of the T7 replisome. Two orientations of an X-ray crystal structure of the T7 replisome in the absence of DNA are shown in panel A. The central core is a heptameric helicase (varied colors) with three associated DNA polymerase/thioredoxin complexes (labeled Pol and Trx, respectively). Recall that the T7 helicase forms a heptamer rather than a hexamer in the absence of DNA. The image on the right provides a rotated view showing the underside with three of the primase domains (labeled Pri) extruding from the core. The structures are from the RCSB PDB (www.rcsb.org) of PDB ID 5IKN (Wallen et al., 2017). The replisome structure in panel B, containing replication fork DNA, was derived from a combination of structural methods. The hexameric form of the helicase is complexed with three DNA polymerases, one engaged with leading strand, one with lagging strand, and one “apo” form that is not engaged with DNA. The leading-strand polymerase and helicase collaborate in unwinding the parental DNA, and the template strand for laggingstrand synthesis threads through the helicase before engaging DNA polymerase. This image is reproduced from Gao et al. (2019), with permission from the American Association for the Advancement of Science; permission conveyed by Copyright Clearance Center, Inc. Panel C provides a cartoon tracing how the replisome structure in panel B relates to the trombone model of Figure 2.5 and the helicase movement pathway in Figure 2.4B.
The leading-strand template DNA, just past the unwinding point, enters the active site of the leading-strand polymerase (top DNA polymerase in Figure 2.6B). This leaves only a few nucleotides of ssDNA between the unwinding point and the point of leading-strand extension on that template strand.
In contrast, the lagging-strand template DNA bends away from the leading-strand polymerase, threads through the helicase/primase complex (12 nucleotides of DNA; see above), and emerges in the vicinity of the lagging-strand polymerase (bottom left polymerase in Figure 2.6B). As it exits the helicase/primase, the DNA traverses a protein-free region before entering the lagging-strand DNA polymerase, presumably allowing the formation and loss of the lagging-strand loops during active DNA replication.
Each of the three bound DNA polymerases contacts two primase domains in this replisome structure. Recall that an alternative form of helicase/primase lacking a portion of the primase domain is also produced in vivo. Future studies will undoubtedly