XPD is a subunit of TFIIH with 5'-3' helicase activity, which is essential for NER but not for transcription initiation. The XPD gene, located on chromosome 19q13.2, is the most complex and most interesting of the XP genes in terms of genotype-phenotype relationships. Mutations in this gene can result in a bewildering array of clinical outcomes, including XP (usually with neurological abnormalities), TTD, and also rare cases of combined features of XP and Cockayne syndrome (XP/CS), the CS-related disorder cerebro-oculofacioskeletal syndrome (COFS), and individuals with some features of both XP and TTD.

XPD, like XPB, is a subunit of TFIIH with helicase activity, although in the case of XPD, it is a 5'-3' helicase (83,84). Unlike XPB, XPD is not part of the central core of TFIIH, but it is somewhat more loosely associated and forms a link between the core and another three-component subcomplex, designated the CAK complex (85,86). Also in contrast to XPB, the helicase activity is not required for the transcription function of TFIIH. Destruction of the helicase activity by site-directed mutagenesis of the ATP-binding domain abolishes NER but has no effect on transcription (87). In assays for in vitro transcriptional activity of TFIIH, XPD had a stimulatory effect but was not essential (82,86). It thus appears that for the transcriptional function of TFIIH, the role of XPD is largely structural, being required to maintain the conformation and stability of the TFIIH complex. As a consequence, a large number of mutations are compatible with life and can result in the different clinical features depending on exactly how NER and transcription are differentially affected. More than 50 XPD mutations have been identified so far (Fig. 3). In contrast to its "supporting role" in transcription initiation, active XPD is absolutely required for NER. The helicase activity is needed for formation of the open complex by TFIIH (76,87).

Within the TFIIH complex, the C-terminal half of XPD interacts with the p44 subunit. This interaction with p44 results in a 10-fold stimulation of the helicase activity of XPD. Several mutations (located close to the C-terminus of XPD) identified in patients interfere with the interaction with p44. These mutations do not affect the basal helicase activity of XPD, but they abolish the p44-mediated stimulation (88). Structural studies using electron microscopy have revealed that the TFIIH core, together with XPD, form a ringlike structure with a central cavity, which could accommodate double-stranded DNA. XPD and XPB flank the p44 subunit, and the CAK proteins form a protuberance from the ring in the region of XPD/p44/XPB (89,90).

When the involvement of XPB and XPD in TFIIH was discovered, it was proposed that mutations in XPD that affect only NER might result in the clinical features of XP, which is considered to be a repair syndrome. On the other hand, if the mutations subtly affect the transcriptional activity, then the multisystem disorder TTD would result; that is, TTD is proposed to be a transcription syndrome (91). A prediction of these proposals is that mutations resulting in different clinical phenotypes would be located at different sites in the gene. Analysis of the different mutations found in the XPD gene in different patients has been difficult, because many patients are compound heterozygotes (92). However, the results of these analyses have supported the idea that each mutational site is disease specific; that is, the site of the mutation largely determines the clinical outcome (92) (see Fig. 3). Most of the mutations are located in the C-terminal third of the protein (92-95). Molecular modeling of the XPD protein on the structure of the Escherichia coli UvrB helicase shows that the C-terminal region that contains the majority of the mutations forms a separate structural domain (96).

Proof of the hypothesis that the site of the mutation in the XPD gene determines the phenotype comes from a mutant mouse generated by de Boer et al. (97). The mouse contained a single mutation in the XPD gene, which generated the change arg722trp—a mutation that had been found in several patients with TTD (see Fig. 3) but not in any with XP. The phenotype of the mouse was similar to


XPD Cell Uns COFS Cell Line

A TTD Cell Line Y XPD/TTÜ Cell Line XPD/CS Cell Line

XPD Cell Uns COFS Cell Line

A TTD Cell Line Y XPD/TTÜ Cell Line XPD/CS Cell Line

Figure 3 Spectrum of mutations found in the XPD gene. The XPD protein is shown as a bar with the seven helicase domains indicated. The flags denote the amino acid changes in different patients, with the number of symbols indicating the number of times that the particular allele has been identified in patients (i.e., two symbols for homozygotes, one for compound heterozygotes). Circles, XP; triangles, TTD; pentagons, XP/CS. Alterations considered to be completely inactivating (Ref. 92) are not shown. (Modified from Ref. 94.)

that of TTD in humans; namely, the hair was brittle and deficient in sulfur-rich matrix proteins. It went through cycles of loss and regrowth. Furthermore, like TTD individuals, the mice showed growth retardation, skin abnormalities, premature aging, and early death. To date, this is the only mouse with mutations in the XPD gene. Others are eagerly awaited.

Whereas the XP phenotype in XP-D patients is thought to result from reduction or abolition of helicase activity, the abnormal transcription thought to result in TTD appears to be caused by decreased stability of the TFIIH complex. Measurements of the levels of TFIIH in TTD cells show a substantial reduction in cellular TFIIH content in cultured fibroblasts (M. Stefanini, personal communication). In TTD patients with one specific mutation, arg658cys, the XPD protein appears to be temperature sensitive, with greatly decreased NER and transcrip-tional activity in cells cultured at elevated temperature (98). This correlates with the clinical phenotype, in that these patients tend to show hair loss in association with high temperatures due to fever.

Two severely affected XP-D individuals with unique mutations have the combined features of XP and CS. Cells from these patients are extremely sensitive to UV-induced lethality. On exposure to UV, these cells not only are unable to carry out NER, but they introduce aberrant breaks into the genomic DNA at sites distant from the UV-induced damage (99).

In general, the clinical features of XP and TTD are exclusive. Despite the very similar levels of NER, TTD patients do not develop the sunlight-induced skin pigmentation abnormalities and skin cancers characteristic of XP, and XP-D patients do not have sulfur-deficient brittle hair and other developmental abnormalities typical of TTD. The latter is explained relatively easily if the mutation affects only the repair capacity but not the transcription function of TFIIH. To account for the former, we have proposed that critical steps in the development of cancer and pigmentation abnormalities require fully active transcription. If transcription is impaired as a result of the XPD mutations, cancer does not ensue (100).

Recently, two XP-D patients have been identified that unusually do have some, but not all, of the features of XP and TTD, adding a further level of complexity to the range of clinical phenotypes (101). Finally, one patient with COFS, a disorder related to CS, has been found to be NER defective and to have a mutation in the XPD gene (102).

0 0

Post a comment