The XPF nuclease cleaves DNA on the 5' side of the damage. XP-F patients show mild clinical characteristics with skin sensitivity and no neurological abnormalities. Mutation analysis in several XP-F patients has identified only missense mutations in the responsible gene (110).

A heterodimer protein complex between the XPF protein and the ERCC1 gene product functions as the second structure-specific nuclease involved in the incision of the damaged DNA during NER. This complex formation between the XPF and ERCC1 proteins has been shown, both in vitro and in vivo, to be quite stable (111-115). XPF/ERCC1 cleaves the DNA 5' to the damage after XPG has incised on the 3' side (114). Recruitment of the XPF/ERCC1 protein complex to the 5' side of the damage is totally dependent on the XPA protein, since no relo-calization of the complex is observed at the damaged sites in locally irradiated XP-A cells (27). This is also consistent with in vitro observations that show specific interactions between XPF/ERCC1 and XPA (116,117). Positioning of the XPF/ERCC1 nuclease at the damage strand seems to be coordinated by RPA (63), possibly via interactions with the XPF protein (61,117).

As yet, there are no Xpf knockout mice, but Ercc1 knockout mice have been generated. These mice are viable, but they survive only for few weeks before dying of liver failure. They are extremely sick, having a severe phenotype resembling more of a syndrome of cell cycle arrest or premature aging than of a DNA-repair deficiency disorder (118,119). Cells from Ercc1 mice are characterized by increased levels of unrepaired lesions and double-stranded breaks leading to genomic instability. Since Xpa mice do not show any of these features, these findings imply that the ERCC1 protein—and most likely its protein partner, XPF—have an additional role. It has been proposed that this second role is an essential S phase-mediated recombination mechanism (120). This second role of XPF/ERCC1 protein complex might be the reason that the only mutations that have been identified so far are missense mutations in XPF, and there are no known patients that carry mutations that result in truncated form of either of the two proteins.

H. XP Variants

The XPV gene encodes DNA polymerase which can replicate DNA containing different types of damage. About 20% of patients with XP have no defect in NER, and their cells are only marginally sensitive to the killing action of UV. This class is designated XP variants. They have classic XP skin symptoms but no neurological degeneration. Despite the near-normal sensitivity to killing by UV, they are extremely hypermutable by UV (14). The defect in XP-V cells is in the ability to synthesize high molecular weight DNA during DNA replication following UV irradiation (5). This damage tolerance process is sometimes referred to as postrepli-cation repair or daughter-strand repair. The molecular basis for this defect has been established only recently. In vitro studies using cell extracts have shown that the defect is in the process of translesion synthesis (TLS), the ability to synthesize DNA directly past damaged nucleotides (6). This in vitro defect allows the protein defective in XP-V cells to be purified (121) and the gene cloned (7,8). The protein turns out to be a member of a new family of DNA polymerases and has been designated DNA polymerase ^ (poland the gene variously designated POLH, XPV, or hRAD30A (the latter by virtue of its sequence homology to Saccharomyces cerevisiae RAD30).

The members of this family of new polymerases (the Y-family) are related in sequence to each other but unrelated to classic DNA polymerases. They differ from classic polymerases in that they are distributive (unlike the highly processive replicative polymerases), and they have low fidelity when replicating undamaged templates (122,123). This low fidelity confers on them the ability to replicate past different types of DNA damage, in contrast to the highly faithful replicative poly-merases, which are completely blocked at sites of damage.

The ability of pol^ to replicate past different damaged structures has been examined in several laboratories. It can replicate a template containing CPDs with the same efficiency as undamaged templates (124). It also can carry out TLS past (in order of decreasing efficiency) 8-oxoguanine, AAF lesions, cisplatin adducts, and AP sites (124-127). Pol^ is not able to carry out TLS past a 6-4PP (124,128). Remarkably, when pol^ carries out TLS past lesions, in most cases, it inserts the correct bases (AA opposite the TT-CPD, C opposite AAF-G adducts, and CC opposite cisplatin-GG intrastrand cross links). This "accurate" replication is achieved first by a strong preference for the polymerase to insert the correct nu-cleotide, and second, by an inability to extend strands in which "incorrect" bases have been introduced. This double selection process provides a high degree of accuracy in copying damaged bases (124,129). It also most likely accounts for the elevated UV mutability of XP-V cells, which lack pol^. In these cells another polymerase with lower accuracy must carry out TLS.

The ability of pol^ to insert AA opposite TT-CPDs also can account for the different UV mutation spectra seen when comparing XP-V and normal cells. In normal and excision-defective XP cells, the majority of the induced mutations are C to T transitions, probably arising at sites of 6-4 PPs. In XP-V cells, however, the majority of the mutations are transversions and different mutations at A:T sites (130). These results indicate that, in the absence of pol^, a polymerase with decreased fidelity and different specificity carries out TLS at the sites of photoproducts.

The DNA polymerase Y-family shows a high degree of sequence conservation in the N-terminal half of the protein, which contains the active site. With pol^, the N-terminal 511 amino acids (aa) of the 713-aa protein are sufficient for full activity in vitro (124). The C-terminal part of the protein is involved in localization and presumably protein-protein interactions. Following UV irradiation, pol^ accumulates at foci in the nuclei, which represent replication factories where the replication forks are stalled at sites of DNA damage (131). The C-terminal 70 aa are required for nuclear localization, and a further 40 aa are required for relo-calization into replication foci (131). These relocalization sequences contain a putative C2H2 zinc finger and a putative proliferating cell nuclear antigen (PCNA) binding site.

Mutations in the POLH gene have been detected in XP-V patients. Many of these mutations result in severe truncation of the protein (7,8), but several missense mutations have been identified in the conserved N-terminal region (131a). In addition, three patients have truncations close to the C-terminus, which leave the TLS activity of the protein intact but prevent the protein from being localized in the nucleus (131).

The somatic hypermutation that is part of the generation of diversity in the immune response is thought to involve an error-prone DNA polymerase. Evidence for pol^ involvement in somatic hypermutation has come from an examination of the spectrum of mutations generated by somatic hypermutation in XP-V patients. Although the frequency of mutations in an immunoglobulin variable chain gene has been found to be unaltered, a significant change in the mutation spectrum has been shown, with a decrease in mutations at A:T sites and a concomitant rise at G:C sites (132). These data again suggest that functions of pol^ can be substituted by other polymerases with different specificity.

0 0

Post a comment