Читать книгу Replicating And Repairing The Genome: From Basic Mechanisms To Modern Genetic Technologies - Kenneth N Kreuzer - Страница 36

The above picture of an efficient replication fork that proceeds smoothly around the chromosome from origin to terminus is incomplete. Replication forks can encounter a wide variety of damaged DNA bases, DNA breaks, chemical crosslinks between the strands or between the DNA and a protein, proteins that are bound tightly though noncovalently, and other structures that challenge the process. As we will discuss in later chapters, efficient repair systems greatly reduce the burden of many of these lesions, but these systems are imperfect, leaving lesions in the path of the replisome. Recent research has shown that bacteria have multiple pathways that allow completion of replication in spite of the kinds of DNA lesions mentioned earlier. While the existence and basic mechanisms of these pathways have been uncovered, we have a long way to go to understand many important details, such as which pathways are used for which lesions, how the interplay between the pathways is regulated, and how increased levels of damage alter the efficiency and balance of these pathways. We can also expect to find new pathways as research continues, and there are hints that DNA repair pathways described in later chapters might also be directly triggered by replication fork blockage.

We already discussed one situation, involving damage on the lagging-strand template, in which the replisome is able to essentially bypass the lesion and continue replication (Section 3.4). Surprisingly, recent evidence suggests that a similar pathway can occur with leading-strand damage. Biochemical experiments showed that when a leading-strand polymerase encounters a blocking lesion, primase is able to synthesize a new primer on the leading strand ahead of the blocking lesion and the clamp loader is able to load a new clamp (Figure 3.2B). This series of events primes a new stretch of DNA synthesis on the leading strand, leaving a short gap of ssDNA behind. As with the lagging-strand damage pathway, the gap and blocking lesion still need to be dealt with by a subsequent repair event of some kind, but at least the chromosome can complete its replication.

Recall that E. coli has a special protein that loads the replicative helicase at the replication origin and prevents inappropriate loading elsewhere (Section 3.4; also see Chapter 5). The existence of a second helicase-loading pathway in E. coli was therefore a longstanding mystery. This pathway, which involves several proteins, was discovered many decades ago as being required for the replication of a bacterial virus called ϕX174. Surprisingly, however, these proteins were not needed for replication from the bacterial replication origin. Why would E. coli encode proteins that are unnecessary for its own replication but allow the replication of a lethal virus? It took a couple of decades to solve this mystery, but the answer revealed that DNA replication is much more robust and flexible than previously thought.

Scientists discovered that E. coli uses these proteins to salvage replication when it has been disturbed or blocked in various ways. Depending on the exact combination of proteins involved, four distinct “replication restart” pathways have been defined. Mutations that inactivate the proteins needed for replication restart are nearly lethal to E. coli, implying that most rounds of DNA replication must invoke one or more restart pathways. The detailed mechanisms and protein requirements of these pathways are quite complex and beyond the scope of this chapter. Also, the exact situations when the different pathways are used are still being worked out. Some of the pathways seem to provide an alternative to restart replication when either the leading- or lagging-strand polymerase is blocked. It is not yet clear how the choice is made between these replication restart pathways and the re-priming pathways discussed just above (which don’t require additional proteins).

Several additional replication-rescue pathways utilize proteins and mechanisms involved in DNA repair and recombination pathways, and these will be described in more detail in subsequent chapters that highlight DNA repair and recombination. A brief overview follows as a teaser for these later chapters.

DNA breaks are very serious lesions that must be repaired for cells to maintain their intact genomes and ultimately to survive, and it is therefore no surprise that cells have mechanisms to repair DNA breaks. A single-strand break or nick would seem to be rather innocuous, because it can be easily repaired by DNA ligase. However, ligase cannot repair a nick unless it has one 5′ phosphate and one 3′ hydroxyl end, and some do not. What happens when a DNA replication fork encounters a nick that cannot be repaired by ligase, or even a sealable nick that just escaped ligation? As the replicative helicase drives through the region of the nick, the arm with the nick is broken off, which seems a rather disastrous event. As we will see in Chapter 11, a special version of a DNA-break-repair pathway can essentially stitch this broken arm back onto its partner and reconstitute a functioning replication fork. This pathway involves reloading the replicative helicase using some of the restart proteins mentioned earlier.

Excision repair pathways replace damaged bases with the correct undamaged base, but these pathways require that the damage is located within a duplex region of DNA (Chapter 10). Again, what happens if the process of replication gets ahead of the process of excision repair of a damaged base? This would be particularly prone to occur when cells suffer a high level of DNA damage from exposure to radiation and/or chemical DNA–damaging agents, overwhelming the excision repair system. When the replication fork reaches such a damaged base, the polymerase is unable to insert a complementary base for certain forms of damage. By then, however, it is too late for excision repair to operate! The damaged base is no longer in a duplex region — the replicative helicase has already unwound this region of the DNA (Figure 3.6, top).

A remarkable kind of gymnastics can come to the rescue — the replication fork can actually back up by a process called replication fork regression (Figure 3.6A). During regression, the two newly synthesized strands anneal with each other and the branch point of the fork is zipped backward. Fork regression is driven by special DNA helicases, which have additional roles in DNA repair and/or recombination. Obviously, the process of fork regression must be very carefully regulated or else replication would turn into a hopelessly complicated process. Getting back to the DNA damage that caused the problem in the first place, the process of fork regression has driven the damage back into a duplex region, and thereby allows excision repair to occur successfully and fix the problem (Figure 3.6B). Presumably, excision repair is somehow coupled to this process of regression, but the details are not yet clear. Following excision repair, the fork regression must be reversed (Figure 3.6C), and the normal replication fork restarted, and again one assumes that these are all well-coordinated events.

Fork regression can be used to achieve another remarkable outcome, inserting the correct complementary base opposite a damaged base that itself cannot serve as a template for correct base pairing. This can occur in the situation where the leading-strand polymerase has been blocked by a damaged template base, but lagging-strand replication has proceeded a short distance further on. Fork regression can again extrude the two newly synthesized strands into a (mostly) duplex region (Figure 3.7A). In this case, the 3′ end of the new leading strand, which had been blocked by the damage prior to regression, is properly base paired with the new and longer laggingstrand product. Notice that the sequence of the new lagging-strand product would be identical to the sequence of the old leadingstrand template, because it is “the complement of the complement.” Thus, the lagging-strand product can serve as an accurate template for leading-strand synthesis! DNA polymerase extension of the 3′ end of the leading strand, even for a few bases, has the effect of circumventing or bypassing the original damage (Figure 3.7B). When the regression process is reversed, the leading-strand product has now been extended past the damage, with the correct base opposite the damage (Figure 3.7C). The fork can restart and complete the replication of the genome. The damaged base remains, but the cell can complete its round of cell division and now has another cell cycle to properly repair the damage, for example, by excision repair.

Figure 3.6.Replication fork regression allows repair of blocking template lesions. The two newly synthesized strands at the replication fork are complementary to each other and can thereby anneal to each other and allow the fork to “back up” (step A). Once the lesion is restored into a region of duplex DNA, repair pathways described later in this book can repair the damage (step B). Reversal of the regression restores a normal replication fork structure (step C), and the process of replication can then resume.

Figure 3.7.Error-free bypass of a blocking lesion via replication fork regression. After blockage of the leading-strand polymerase, replication fork regression places the blocked 3′ end in a duplex with the lagging-strand product (step A). This allows accurate incorporation of several base residues by DNA polymerase (step B). Reversal of regression (step C) then allows resumption of normal DNA synthesis, with the blocking lesion bypassed.

Importantly, fork regression-driven bypass occurs in an accurate, mutation-free manner. A different and very important form of lesion bypass involves specialized DNA polymerases, and will be covered in Chapter 12. Overall, it is clear that numerous sophisticated pathways have evolved, allowing the completion of bacterial DNA replication even when the DNA is damaged.