A minimal repeat length (usually more than 30 repeats) of defined purity is a critical factor for repeat expansion. Mutations in many genes can influence repeat expansion or contractions; however, deficiencies in replication, repair, or recombination functions are not required, a priori, for repeat instability. In one instance, a d(CTG) ■ d(CAG) repeat integrated at a specific site in one mouse showed many different rates of instability in different cell types, unrelated to the state of cell proliferation. In addition, when cells from different tissues from this mouse were cultured the propensity for instability persisted (Gomes-Pereira et al. 2001). In contrast, cells from other mice with repeats integrated at different locations did not show this variation (Fortune et al. 2000). Complex cis- and trans-acting factors effecting these differences are only beginning to be revealed.
Given the clear role for replication in repeat instability, as evidenced by differences in stability as a function of orientation with respect to the direction of replication (Kang et al. 1995a; Maurer et al. 1996; Freudenreich et al. 1997; Miret et al. 1997; Hashem et al. 2002; Cleary et al. 2002; Pani-grahi et al. 2002), and a role for transcription (Bowater et al. 1997; Mochmann and Wells 2004), the distance and orientation of repeats with respect to replication origins might be a critical cis-acting factor in repeat instability. In mammalian cells and yeast genetically defined replication control regions, or replicators, overlap with biochemically defined replication initiation zones, or origins (DePamphilis 2003; Schwob 2004; Gilbert 2004; Aladjem and Fanning 2004). Though generally more expansive and less well-defined in terms of structural and functional modules than those of their yeast counterparts, several mammalian replicators have been identified (Dijkwel et al. 1991; Little et al. 1993; Aladjem et al. 1995; Kobayashi et al. 1998; Liu et al. 2003; Aladjem 2004; Paixao et al. 2004). The initiation zone neighboring the hamster DHFR gene encompasses more than 55 kb of DNA comprising multiple start sites firing with different efficiency in a cell population. Replication also initiates at multiple sites within the human endogenous ribosomal RNA, 0-globin, and c-myc origins (Little et al. 1993; Malott and Leffak 1999; Liu et al. 2003; Aladjem 2004), and a zone of initiation accompanies translocation of the DHFR, 0-globin and c-myc replicator elements to ectopic sites (Malott and Leffak 1999; Altman and Fanning 2001; Liu et al. 2003; Aladjem 2004). Origin specification results from the poorly understood interplay of sequence-directed DNA structures, histone and non-histone protein binding, and epigenetic modification of chromatin (Anglana et al. 2003; Debatisse et al. 2004; Schwob 2004; Gilbert 2004; Danis et al. 2004). Differential origin specification in murine cells provides one explanation for the variable repeat instability observed in mice and cultured cells (Fortune et al. 2000; Gomes-Pereira et al. 2001).
Numerous recent reviews have discussed the cis effects of origin proximity and the direction of replication of the repeat tracts on repeat instability (Mirkin and Smirnova 2002; Mirkin 2004, 2005; Cleary and Pearson 2005). Because the location and utilization of initiation sites within human origins can vary, the strand of the repeat tract that constitutes the leading or lagging strand can also vary. This has been termed the "ori switch" model (Mirkin and Smirnova 2002). This is shown in Fig. 9, panel A, as alternative directions of fork progression, defined by the site of replication initiation within an origin. Opposite polarities of replication are analogous to reversing the orientation of repeats in bacteria or yeast, with respect to their defined origins. DNA repeats that exhibit differential DNA secondary structure stabilities [e.g., d(CAG)„ ■ d(CTG)„ or d(CCTG)„ ■ d(CAGG)„], may behave in this way because structures may form in the lagging strand, given initiation from one side of the repeat (Fig. 9, panel A, 1), but not in the leading strand, given replication from the other direction (Fig. 9, panel A, 3). As discussed already, repeat instability varies depending on the direction of replication in virtually all experimental systems examined. In addition, the distance is variable between a DNA repeat and alternate potential initiation sites within an origin zone (Fig. 9, panel A, 2). This has been termed the "ori shift" model (Mirkin and Smirnova 2002). Using the SV40 viral replication origin system, Cleary et al. (2002) showed that expansions were favored when replication initiated 103 bp 3' of a d(CTG)79 tract, but that deletions predominated when initiation occurred 230 or 536 bp away. To complicate matters, both deletions and duplications were observed when initiation took place 667 bp away.
Several factors may be important for the ori shift (Mirkin and Smirnova 2002), also called the "fork shift" (Cleary and Pearson 2005), model. First, the neighborhood of the viral SV40 origin could be unusual in that the DNA structure, torsional stress, chromatin organization, and amount of single-stranded DNA may be unusual or unique directly adjacent to the site of replication initiation (termed replication initiation site in Fig. 9, panel A). Whether the SV40 viral replication origin is representative of the more complex human origins is not known. Unlike human replication forks, the SV40 replication fork contains T-antigen, the initiator protein and a potent replica-tive helicase (Borowiec et al. 1990). It is not known whether polymerases working in conjunction with endogenous human helicases will act differently during unwinding of DNA repeats. In vitro evidence indicates that the SV40 replication fork does not require ORC-dependent prereplicative complex formation, minichromosome maintenance, Cdc45, ATR proteins, or other factors that may assist in replicating alternative DNA structures (Waga and Stillman 1998). Moreover, the exact positioning of the 3' or the 5' end of the Okazaki fragment within the repeat tract and the length of the repeat tract may have significant consequences for repeat instability, as described by Richards and Sutherland (1994). As evident from Fig. 5, a hairpin flap would have a good opportunity to form if the 5' end of the Okazaki fragment began within the repeat tract (Fig. 9, panel B, model 1, fragment set 2; model 2, fragment set 2,3). This is because the opportunity for DNA secondary structure formation at the 5' end is greater than at the 3' end, which is bound by the polymerase. Another factor is the length of the repeat tract with respect to the size of the Okazaki fragment. As the length of the repeat tract becomes longer than the length of the Okazaki fragment (approximately 140 nt), the number of nicks to be ligated increases, and this may increase the probability of structure formation and repeat instability (Richards and Sutherland 1994) (Fig. 9, panel B). The sequences of repeats may have important consequences for Okazaki fragment initiation given preferred sites for RNA synthesis (Cleary and Pearson 2005). In many DNA repeats, only one strand might easily support generation of primers by RNA primase. Thus, depending on the direction of replication, the forks could become unbalanced with the generation of an unusually long tract of single-stranded DNA in the lagging template strand. In summary, experimental evidence suggests that in addition to repeat sequence and length, the spatial relationship between
A Fig. 9 Cis effects of replication on repeat instability: location and proximity of the origin and positioning of Okazaki fragments. In metazoan cells replication origins encompass regions ranging from about 2 kb to as much as 55 kb DNA. Within this region the initiation DNA synthesis requires unwinding of the DNA and binding of helicases and polymerase a/primase to lay down the RNA primer for extension by DNA polymerases. This occurs at multiple sites within the origin where the selection of specific sites for initiation maybe a stochastic process. Site utilization maybe different in different cell types. This variation can influence repeat instability. A Positional effects of the replication initiation site (RIS) on repeat instability. Replication is shown starting from RIS 1, RIS 2, or RIS 3, in parts 1,2, and 3, respectively. The DNA repeat tract is denoted by the shaded section; an unstable situation is denoted by the gradient of shading over the unstable strand. See text for details. B Positional effects of Okazaki fragments on repeat instability. The relative localization of an Okazaki fragment can vary with respect to a DNA repeat tract, as shown. Model 1 shows three different positions of Okazaki fragments across a repeat tract, where the length of the tract is shorter than the Okazaki fragment. In set 1, the 3' end of the middle fragment is positioned within the repeat. In set 2, the 5' end of the leftmost fragment is positioned within the repeat, and in set 3, the Okazaki fragment straddles the repeat. It is not known if in a cell population only one set or multiple sets of positions will occur. Nevertheless, the number of nicks that need to be ligated within a repeat tract would range from 0 to 1. As the repeat tract lengthens the probability of nicks falling within the repeat tract increases as shown in models 2 and 3
a DNA repeat and its origin of DNA replication may be critically important in determining repeat instability. Understanding all the factors that govern instability will require additional investigation.
Was this article helpful?