The many faces of Cockayne syndrome

The challenges intrinsic to the maintenance of genetic information are revealed when one surveys the growing list of human disorders caused by defects in the repair of damage to DNA. Although these diseases are extremely rare, their effects are often devastating for patients and their families.

One such rare autosomal disease, Cockayne syndrome (CS), can be caused by mutations in two genes, CKN1 and ERCC6, located on chromosomes 5 and 10, respectively. There are two complementation groups of CS: CS-A patients have mutations in CKN1, whereas CS-B is caused by mutations in ERCC6 (also known as CSB). In this issue of PNAS, Horibata et al. (1) describe a mutation in the ERCC6/CSB gene that gives rise to a different sun-sensitive, DNA-repair-deficient disease, UV-sensitive syndrome (UV^sS).

UV-sensitive syndrome comprises at least two complementation groups.

Repair of photoproducts and many other “bulky” DNA lesions is carried out by the nucleotide excision repair (NER) pathway (2). NER of certain lesions has two operational modes: a general mechanism that operates throughout the genome, termed global genomic NER (GG-NER), and a distinct mode, transcription-coupled NER (TC-NER), which requires a subset of NER proteins plus additional factors and is targeted to transcriptionally active regions of the genome. Another pathway, base excision repair, operates on damage to bases induced by reactive oxygen species, ionizing radiation, and most alkylating agents. Xeroderma pigmentosum (XP) is the most prevalent human DNA repair deficient syndrome, with nearly 1,000 patients worldwide. There are seven complementation groups of XP with deficiencies in NER, named XP-A through XP-G; patients in the eighth group, XP variant, have normal NER but are deficient in translesion synthesis of DNA containing certain lesions. XP is characterized by sun sensitivity, abnormal skin pigmentation, a 1,000-fold increase in cancer in sun-exposed tissues, and progressive neurological degeneration in 20% of the patients. Patients with CS, totaling >180, are also sun-sensitive and DNA-repair-deficient as elaborated below, but, in contrast with XP, they are not cancer-prone. CS patients generally acquire a characteristic physical appearance of premature aging: loss of adipose tissue with resulting sunken eyes and beaked nose, prominent ears, thinning of the skin and hair, stooped posture, dwarfism, hypogonadism, mental retardation, and joint contractures. The pathology often reveals cataracts, dental caries, hearing loss, osteoporosis, demyelination in the central and peripheral nervous systems, calcification in the cortex and basal ganglia, and severe neuronal loss. Besides mutations in the CKN1 and ERCC6 genes, certain mutations in three XP genes (XPB, XPD, and XPG) also result in CS-like features in addition to those of XP. The seminal discovery that CS cells had normal NER but were unable to recover RNA synthesis after treatment with UV light indicated that CS cells lacked a repair mechanism dedicated to the removal of photoproducts from active genes (3), as was shown many years later (4). UV^sS is a photosensitive syndrome with abnormal pigmentation but no enhanced skin cancer; growth, development, and life span appear normal. At the cellular level, the responses of CS and UV^sS cells to UV irradiation are indistinguishable: hypersensitivity to UV light, defective recovery of RNA synthesis, normal GG-NER, hypersensitivity to accumulation of the tumor promoter p53 after low UV doses, and defective TC-NER of UV-induced cyclobutane pyrimidine dimers (5). Initially, all known UV^sS cases had been assigned to one complementation group distinct from XP or CS. Horibata et al. (1) demonstrate that one UV^sS patient, UV^s1KO, carries a homozygous null mutation in the CSB gene. Correction of UV sensitivity when normal CSB cDNA was introduced into UV^s1KO cells and genetic complementation of defective recovery of RNA synthesis after UV irradiation confirmed that the photosensitive phenotype of UV^s1KO was due to the mutation in CSB, whereas Kps3, a UV^sS cell line from an unrelated patient, contains wild-type CSB. Thus, UV^sS comprises at least two complementation groups, defined by mutations in CSB and in another as-yet-unidentified gene.

What Causes CS?

CSB is a 1,493-aa protein with several conserved motifs typical of the SNF-2 family of RNA and DNA helicases, although no helicase activity has been reported to date. The mutation found by Horibata et al. (1) implies that a severely truncated CSB protein containing only the N-terminal 76 aa would be generated in UV^s1KO cells. Western blot analysis with antibodies to either the N- or C-terminal regions of the protein failed to reveal any CSB polypeptides in UV^s1KO cells. These provocative findings lead to a surprising conclusion: Patients with no detectable CSB fare better than patients with mutated CSB. Clearly, intact CSB protein is required for TC-NER, but there appears to be a second activity involved in the developmental failure of CS patients. This concept is supported by the absence of CS-like phenotypes in some XP patients, like XP-A, who also are defective in TC-NER. What this activity might be has been the subject of many discussions and reviews; models include a defect in basal transcription (6), a defect in repair of oxidative damage overall and/or in active genes (7, 8), a defect in transcription-dependent gene activation by demethylation of CpG islands (9), and excessive cell death by apoptosis induced by arrested transcription complexes in tissues undergoing intense metabolic activity (10). Each of these hypotheses when taken alone fails to explain how CS can be caused by mutations in five genes with different cellular functions, why most CS children appear normal at birth, and the progressive nature of the disease. One scenario (11) combines some of the ideas described above: Defective transcription-coupled DNA repair (TCR) could lead to accumulation of arrested transcription complexes at endogenously induced lesions, with progressive depletion of RNA polymerase and transcription insufficiency, while apoptosis and failure in the up-regulation of genes result in severely underdeveloped tissues and organs. Some of these models imply that CS cells are defective in the repair of endogenously induced DNA lesions, particularly in active genes. Which lesions that are induced by reactive metabolites could be subject to TCR? The base damage 8-oxoguanine requires CSB for its removal when it is located within a transcribed sequence in a shuttle vector (8); other candidate lesions include cyclode-oxyadenosine and products of lipid peroxidation, like malondialdehyde, which are repaired by the NER pathway and have been shown to arrest transcription (12–14).

Mutations Versus Phenotypes

As noted previously, CS can result from mutations in five different genes: CSA, CSB, XPB, XPD, and XPG. Conversely, mutations in CSB cause at least four different diseases, as depicted in Fig. 1. The clinical classification of CS patients into three general groups is based on the severity of symptoms and age of onset, rather than on the position or the type of mutation. CS type I or “classical CS” affects ≈80% of the patients and can develop from early to late childhood with life expectancy into the second and third decades; CS type II or “connatal CS” is evident at birth, with a life expectancy of 6–7 years; CS type III is assigned to individuals with some late-onset features of CS, with mostly normal growth and development (15). An important conclusion from the work by Horibata et al. (1), noted above, is that carrying a truncated or mutated version of CSB can be highly deleterious, whereas the complete absence of the protein results in mild phenotypes. It will be interesting to learn the locations and types of mutation in other CS-B patients with UV^sS-like symptoms, such as a Japanese case described in ref. 16. There are no genetic data from patients with the less severe symptoms in CS type III, except for a case in which an interstitial deletion in chromosome 10 may overlap or completely encompass the CSB gene in 10q11 (17). It is tempting to speculate on a direct correlation between the severity of symptoms and the amount and/or length of truncated CSB products. Thus, UV^sS with the mildest phenotype and no detectable protein would represent one end of the spectrum, and the other end would include the severest cases of CS type II, with mutations affecting important (but not essential) functions.

Fig. 1.

Schematic representation of mutations in the CSB protein. The bars indicate postulated functional domains: green, acidic domain; red, nuclear-localization signals/casein kinase II phosphorylation sites; black, helicase motifs I–VI in the SNF-2 domain; and white, nucleotide-binding fold. Localization of mutations and resulting syndromes are indicated by flags: CS, Cockayne syndrome; UV^SS, UV-sensitive syndrome; DS-C, DeSanctis–Cacchione syndrome; COFS, cerebro-oculo-facio-skeletal syndrome. This figure was adapted from data and diagrams in refs. 15 and 19–22.

Cellular Roles of CSB

What could be the function of CSB that can be repressed by a truncated or malformed protein and is spared in the absence of the protein? Binding of defective CSB could block the protein–protein or protein–nucleic acid interactions required for any of the several pathways mediated by CSB. For example, it was reported that RNA polymerase II is hypophosphorylated when bound to the promoter (RNAPIIa) and hyperphosphorylated during elongation (RNAPIIo). When RNAPIIo is arrested at a lesion, RNAPIIa is depleted and initiation of transcription is repressed. Recovery of RNAPIIa is observed in normal cells but not in CSB cells; thus, CSB may play a role in “shuttling” RNAPIIo to RNAPIIa. Other mechanisms that require functional CSB include the ubiquitination of stalled RNAPIIo, targeting it for proteosomal degradation, chromatin remodeling, recruitment of factors (including CSA) to the nuclear matrix after UV irradiation, and effective incision of base damage in vivo and in vitro (ref. 15 and references therein).

There remain many unanswered questions regarding the mechanism of TCR, the role of CSB in different cellular transactions, and the correlations between mutations in the gene and clinical syndromes. Animal models and in vitro assays provide important insights into the effects of certain mutations on physiological and molecular mechanisms, but the ultimate answers depend on the wealth of information to be gleaned from medical records and biological samples obtained from patients. A workshop entitled “Cockayne Syndrome and Related Disorders of DNA Repair and Transcription: From Bench to Bedside and Back,” held in Landsdowne, VA, in May 2004 was the first to incorporate basic scientists, clinicians, and patient representatives. The proceedings of that workshop (18) highlight the importance of integrating these diverse groups involved in elucidating mechanisms and finding cures for rare diseases.