© 2005 Nature Publishing Group
60. Bajenaru, M. L. et al. Optic nerve glioma in mice requires
astrocyte Nf1 gene inactivation and Nf1 brain
heterozygosity. Cancer Res. 63, 8573–8577 (2003).
61. Bajenaru, M. L., Garbow, J. R., Perry, A., Hernandez, M. R.
& Gutmann, D. H. Natural history of neurofibromatosis 1-
associated optic nerve glioma in mice. Ann. Neurol. 57,
62. Reilly, K. M., Loisel, D. A., Bronson, R. T., McLaughlin, M. E.
& Jacks, T. Nf1;Trp53 mutant mice develop glioblastoma
with evidence of strain-specific effects. Nature Genet. 26,
63. Ridley, A. J., Paterson, H. F., Noble, M. & Land, H. Ras-
mediated cell cycle arrest is altered by nuclear
oncogenes to induce Schwann cell transformation.
EMBO J. 7, 1635–1645 (1988).
64. Dasgupta, B., Li, W., Perry, A., Gutmann, D. H. Glioma
formation in neurofibromatosis 1 reflects preferential
activation of K-RAS in astrocytes. Cancer Res. 65,
65. King, D., Yang, G., Thompson, M. A. & Hiebert, S. W.
Loss of neurofibromatosis-1 and p19ARF cooperate to
induce a multiple tumor phenotype. Oncogene 21,
66. Reilly, K. M. et al. Susceptibility to astrocytoma in mice
mutant for Nf1 and Trp53 is linked to chromosome 11
and subject to epigenetic effects. Proc. Natl Acad. Sci.
USA 101, 13008–13013 (2004).
67. Kim, J. Y. et al. Medulloblastoma tumorigenesis diverges
from cerebellar granule cell differentiation in patched
heterozygous mice. Dev. Biol. 263, 50–66 (2003).
68. Goodrich, L. V., Milenkovic, L., Higgins, K. M. & Scott, M. P.
Altered neural cell fates and medulloblastoma in mouse
patched mutants. Science 277, 1109–1113 (1997).
69. Singh, S. K. et al. Identification of human brain tumour
initiating cells. Nature 432, 396–401 (2004).
70. Singh, S. K. et al. Identification of a cancer stem cell in
human brain tumors. Cancer Res. 63, 5821–5828
71. Woerner, B. et al. in Ninth Annual Meeting of the Society
for Neuro-Oncology (ed. Bigner, D. D.) 323 (Duke Univ.
Press, Toronto, Ontario Canada, 2004).
72. Dasgupta, B., Yi, Y., Chen, D. Y., Weber, J. D. &
Gutmann, D. H. Proteomic analysis reveals
hyperactivation of the mTOR pathway in NF1-associated
human and mouse brain tumors. Cancer Res. 65,
73. Oliver T.G. et al. Loss of patched and disruption of
granule cell development in a pre-neoplastic stage
of medulloblastoma. Development 132, 2425-2439
The authors thank our Neurofibromatosis Center Group at
Washington University, especially A. Perry, J. Weber,
M. Watson and J. Garbow, as well as R. Wechsler-Reya for
critical reading of the manuscript. D.H.G. is supported by
grants from the Department of Defense, National Institute of
Neurological Disorders and Strokes, and James S. McDonnell
Foundation. J.R. is a scholar of the Child Health Research
Center of Excellence in Developmental Biology at Washington
University School of Medicine and receives additional support
from the National Institute of Child Health and Human
Development, The American Cancer Society and Hope Street
Kids. We also acknowledge the generous support from
Schnuck Markets, Inc.
Competing interests statement
The authors declare no competing financial interests.
The following terms in this article are linked online to:
CDNK2A | NF1
National Cancer Institute: http://cancer.gov
National Cancer Institute: http://cancer.gov
leukaemia | pheochromocytoma
neurofibromatosis type 1
Access to this interactive links box is free online.
Cancer in xeroderma pigmentosum
and related disorders of DNA repair
James E. Cleaver
Abstract | Nucleotide-excision repair
diseases exhibit cancer, complex
developmental disorders and
neurodegeneration. Cancer is the hallmark
of xeroderma pigmentosum (XP), and
neurodegeneration and developmental
disorders are the hallmarks of Cockayne
syndrome and trichothiodystrophy. A
distinguishing feature is that the DNA-
repair or DNA-replication deficiencies of
XP involve most of the genome, whereas
the defects in CS are confined to actively
transcribed genes. Many of the proteins
involved in repair are also components
of dynamic multiprotein complexes,
transcription factors, ubiquitylation
cofactors and signal-transduction
networks. Complex clinical phenotypes
might therefore result from unanticipated
effects on other genes and proteins.
Xeroderma pigmentosum (XP) is the arche-
type of an expanding family of nucleotide-
excision repair (NER) diseases that includes
XP itself, the XP variant (XP-V), Cockayne
syndrome (CS), cerebro oculofacial skel-
etal syndrome (COFS), a mild ultraviolet
(UV)-light-sensitive syndrome, trichothi-
odystrophy (TTD), and some diseases with
combined symptoms of XP/CS and XP/
TTD1. These diseases have complex over-
lapping symptoms associated with cancer,
developmental delay, immunological defects,
neurodegeneration, retinal degeneration and
ageing. They represent a continuum with
skin cancer alone at one extreme and neuro-
degeneration and developmental disorders at
the other BOX 1.
The total number of genes directly
involved in NER is estimated to be around
40 REF. 2, but only about a dozen of these
genes have been found to be deregulated in
NER-related human diseases. The remainder
represent either essential genes that would
be lethal if mutated or that might have such
mild clinical disease that they blend into
the population of ‘sun-sensitive’ individu-
als. The molecular defects in repair and the
clinical symptoms of XP as compared with
CS and TTD enable us to understand how
certain pathways for processing DNA dam-
age lead to cancer, whereas others lead to
more complex disorders. Before the NER
genes were cloned and identified, patients
were assigned to complementation groups,
such as XP-A or XP-D, by cell-fusion
experiments1. Once the mutated genes
were identified and sequenced, they were
named after the complementation group
with which they associate.
Many reviews have described the
molecular mechanisms of NER and the
amazing varieties of damage recogni-
tion and DNA manipulation that are
required for the recognition and removal
of DNA damage represented by the
UV-light-induced dipyrimidine pho-
toproducts, the 5–5, 6–6 cyclobutane
pyrimidine dimers (CPDs) and the pyri-
midine–pyrimidone dimers, and other
large carcinogen or oxidative products in
DNA3–6 (FIG. 1). I want to summarize these
mechanisms briefly in terms of the function
of the XP genes and address broad general
questions about the comparison between
XP and CS. These include two contrasting
questions, the considerations of which are
particularly informative. Why do patients
with mutations in genes specific for global
genomic repair (GGR) or bypass replication
predominantly develop cancer but rarely
neurodegeneration? And why do patients
with mutations in genes specific for repair
of damage in transcribed genes not get
cancer but suffer from developmental and
The question of how DNA damage is rec-
ognized can be viewed from two different
perspectives: one question asks ‘which pro-
teins interact directly with damaged bases
in DNA’, and the other asks ‘which protein
acts first in this damage-response pathway’
(FIG. 1). Many of the NER proteins interact
directly with damaged DNA, including XPE,
XPC–HR23B, XPA, replication protein A
(RPA) and the transcription factor TFIIH3
TABLE 1. Which proteins act first depends
on the transcriptional activity of the DNA.
564 | JULY 2005 | VOLUME 5
© 2005 Nature Publishing Group
Actively transcribed genes are repaired more
rapidly than non-transcribed regions through
transcription-coupled repair (TCR). Damage
appears to be recognized by the arrest of RNA
polymerase II (Pol II) that is relieved through
an assembled protein complex that includes
two proteins, CSA and CSB7. Damage to
non-transcribed regions of the genome is
recognized by binding of XPE and XPC
protein complexes through GGR3–5,8. These
recognition events in TCR and GGR initiate
a common pathway through which the DNA
is unwound (the XPB and XPD helicase com-
ponents of TFIIH), XPA binds and presents
docking sites for the structure-specific nucle-
ases (XPG and XPF–ERCC1) and the damage
is excised and replaced (FIG. 1).
Global damage recognition
Both XPE and XPC exist as heterodimeric
proteins: XPE as a damage-specific DNA
binding protein 2 (DDB2)–DDB1 (p48–p127)
dimer9; XPC as an XPC–HR23B dimer10.
Patients are found only with mutations in
DDB2 or XPC, not in their partner genes11.
These two main components of the GGR
pathway, XPC and XPE, activate ubiquityla-
tion and proteolytic pathways, and interact
with the proteasome system BOX 2.
The DNA-binding protein XPE. XP-E
patients are mildly affected — their cells
carry out near-normal levels of NER12,13.
Because of difficulties of quantifying small
differences in DNA repair (unscheduled
DNA synthesis), there was initial confusion
about the assignment of cells to this group.
This has been clarified and it is now gen-
erally agreed that XP-E represents a group
of patients with mutations in the DDB2
component of the heterodimer12.
XPE binding to a damaged site facilitates
the binding of XPC, and seems to suppress
the mutagenicity rather than the toxicity
of damage14,15. The DDB1 component is
present in excess over its partner, and is
predominantly cytoplasmic but translocates
to the nucleus after UV-light irradiation16.
Mutations in DDB2 also prevent the accu-
mulation of DDB1 in the nucleus17. DDB2
expression is induced by UV light through
p53-dependent transcription in human cells
(but not mouse or hamster cells)9,18, which
provides a partial explanation for early
observations that excision repair (GGR) is
low in mouse cell lines19.
The global damage recognition protein
XPC. The XPC–HR23B complex10,20 is
the main early damage detector and, with
XPE, serves to stabilize the low-affinity
XPA and RPA proteins that subsequently
bind at damaged sites21,22. The XPE subunit
DDB2 seems to facilitate XPC transloca-
tion within the nucleus after irradiation23.
In the absence of XPC, the residual NER is
biased towards transcribed regions, but the
total amount of residual repair appears to
involve large stretches of DNA, greater than
the amount represented by the active genes
alone24,25 TABLE 1.
Highly distorting lesions provide suffi-
cient opening of the DNA strands so, XPC
can be dispensed with in cell-fee systems
of NER26,27. The XPC–HR23B complex is
also associated with CEN2, a protein that
stabilizes XPC and that is involved in cen-
trosome duplication, indicating a direct
link between GGR and cell division28. The
constitutive level of expression of XPC is
controlled by p53, and can be induced by
UV-light irradiation, enhancing GGR29.
The initial damage-recognition mechanism
for TCR is the stalled RNA Pol II itself30,31.
Arrested RNA Pol II is phosphorylated
on its carboxy-terminal domain and,
subsequently, is polyubiquitylated by a
mechanism that involves CSA and CSB32–34
BOX 2 . Two interpretations of the role of
ubiquitylation have been proposed. One
indicates that ubiquitylation marks the
RNA Pol II molecules for degradation,
leaving the active genes accessible for repair
and the resumption of transcription7,34. The
other indicates that a blocked RNA Pol II
does not need to be degraded to facilitate
repair35, that the ubiquitylation is not
essential for TCR36, and the ubiquitylation
linkage through Lys63 on RNA Pol II is a
signal for activation of cellular processes
including repair37. The arrested RNA Pol
II caused by TCR deficiency is a strong
trigger for apoptosis in fibroblasts, but not
necessarily for other cell types38,39.
Association of CSB with the transcription
elongation complex enhances elongation
of nascent RNA40 and might also stabilize
a blocked RNA Pol II REF. 41, therefore
increasing the efficiency with which RNA
Pol II negotiates refractory secondary struc-
tures or T-rich regions in template DNA42
BOX 1 . CSB can block retrograde move-
ment of a blocked RNA Pol II33, possibly
by its ability to wrap the DNA43. CSB is a
member of the ATP-dependent SWI2–SNF2
chromatin-remodelling family and binds to
DNA as a dimer. In the presence of ATP,
CSB actively wraps the DNA around itself,
and following ATP hydrolysis releases the
Box 1 | Cockayne syndrome and endogenous DNA damage
Cockayne syndrome (CS) is an autosomal-recessive disease characterized by cachectic
dwarfism, retinopathy, microcephaly, ganglial calcification, deafness, neural defects,
retardation of growth and development after birth, and sun sensitivity. It is not characterized
by cancer1,135. The clinical presentation ranges from the very mild ultraviolet (UV)-light-
sensitive syndrome (UVs) syndrome to a range classified as CS types I, II and III, and the
severe neonatal lethal cerebro oculofacial skeletal syndrome135. Neurological defects in
xeroderma pigmentosum (XP) patients include diminished deep-tendon reflexes,
sensorineural deafness, peripheral neuropathy, walking difficulties and progressive mental
deterioration. These XP symptoms are sufficiently different from those in CS such that
patients who have XP with neurological disease (typically XPA) can be discriminated
clinically from those with XP plus CS. XP patients have cerebral atrophy and primary
neuronal degeneration (in the grey matter), whereas CS patients have dysmyelination (in the
white matter), retinal and Purkinje cell loss, and growth retardation, but without loss of
Mutations in the two CS genes, CSA and CSB, at present show no obvious clustering within
the sequences that can be correlated with the disease. A null mutation in CSB has been
identified in one mild UVs patient, whereas point mutations in CSB can be more severe136, and
sometimes the same mutation can give quite different symptoms137. CSA and CSB are required
for enhanced repair of actively transcribed genes and for ubiquitylation of RNA polymerase II
after UV-light-induced damage TABLE 1.
Trichothiodystrophy (TTD) is another nucleotide-excision repair (NER)-defective disease
characterized by sulphur-deficient brittle hair with characteristic tiger tail banding when
observed in polarized light, and by skin photosensitivity. It is not characterized by increased
pigmentation or cancer47. TTD is a complex neuroectodermal disorder with a common
deficiency in synthesis of high-sulphur matrix proteins that results in growth retardation,
neurological abnormalities and sulphur-deficient hair and nails. TTD is caused by a subset of
mutations in XPB, XPD or TFB5, all components of the TFIIH transcription factor that reduce
the overall transcription capacity of differentiated cells47,50.
NATURE REVIEWS | CANCER
VOLUME 5 | JULY 2005 | 565
© 2005 Nature Publishing Group
Global genomic repair and replication
– cancer but no CNS disorder
– CNS disorder but no cancer
XPE–XPC–HR23B CSA–CSB–RNA Pol II
XPG 3' nuclease
XPF–ERCC1 5' nuclease
Common pathway — both CNS disorder and cancer
of damaged DNA — cancer
but no disorder
DNA43. Irreversibly blocked RNA Pol II is
degraded by the transcription release fac-
tor IIS42. CSB deficiency might therefore
affect the transcription efficiency of many
other unrelated genes through its action on
RNA Pol II.
CSA is a cofactor for an SCF-type ubiq-
uitin ligase (a SKP1–Cullin–F-box mul-
tisubunit protein, with up to 60 different
F-box proteins in human cells; BOX 2)44 and
translocates to the nuclear matrix after UV-
light irradiation and associates with RNA
Pol II by a process that requires CSB45. CSA
deficiency might therefore affect the func-
tions and half-lives of those proteins that
are substrates for SCF-mediated ubiquityla-
tion, but might be functionally unrelated to
XPA is common to GGR and TCR
The XPA protein freely diffuses in the cell
nucleus and binds to a damaged site that
has been previously marked by XPE and
XPC8 and unwound by the action of TFIIH3.
The single-strand-binding protein RPA
also binds at this site, and to XPA, with
a DNA footprint that corresponds to the
repair-patch size. XPA anchors the nuclease
XPF–ERCC1 and activates the nuclease activ-
ity of XPG8, but is then released following
binding of the excision nucleases4.
The core DNA-binding region of XPA,
amino acids Met98 to Phe219 encoded by
exons 2 through 5, contains a zinc finger
and a positively charged cleft at its carboxy
region, into which single-stranded dam-
aged DNA can bind. Mutations in this
core region generally give rise to severe
symptoms, but mutations in the C-terminal
exon 6 are generally milder1,11. Some muta-
tions, especially those near splice sites,
can be leaky and permit sufficient normal
protein to be made to minimize clinical
symptoms. Mutation of an arginine at the
C-terminal end of the core region (Arg207)
to an uncharged glycine or lysine created
an XPA revertant that no longer repaired
CPDs in non-transcribed regions, but
retained repair of more accessible damage
such as [6-4] photoproducts and CPDs in
transcribed regions46. This XPA revertant
might have a reduced binding to DNA
owing to the loss of the positive charge on
Remodelling — the TFIIH components
The transcription factor TFIIH is a 10-mem-
ber protein complex involved in transcrip-
tion initiation and transcript elongation.
The protein components are involved in
the regulation of gene expression and the
cell cycle, as well as in DNA repair. Two
components, XPB and XPD, are helicases
that serve to unwind the DNA, and TFIIH
might in the process serve to distinguish
damaged strands from undamaged strands3
TABLE 1. XP-B patients are exceedingly
rare, being found only in three families
to date, whereas XP-D patients are more
common. The XPD gene is particularly
complex, because mutations are associated
with multiple disease outcomes including
XP, XP/CS and TTD47–49. Deletions of XPD
are embryonic lethal, and only missense
mutations cause clinical phenotypes49. One
mutation, R683W, is exclusively associated
with XP-D, whereas others — R112H, R658H
and R722W — are found in TTD patients.
Certain mutations in XPB and XPD, or in a
third component, p8, regulate quantitative
levels of TFIIH and are associated with the
short brittle hair disorder (due to a lack of
sulphur and cystine because of impaired syn-
thesis of high-sulphur matrix protein) that is
a diagnostic hallmark of TTD50 BOX 1.
Removal of damage
The structure-specific endonucleases XPG
and XPF–ERCC1 are directed to the damaged
site by XPA and make the 3′ and 5′ incisions
either side of the damaged site51,52. Previous
helicase activity of TFIIH is required before
loading of XPF–ERCC1 REF. 53. These nucle-
ases are also involved in repair of DNA–DNA
crosslinks and homologous recombination
where single-strand–double-strand DNA
junctions are found, similar to those in an
opened excision site51,54. XPF–ERCC1 is also
a component of the telomeric TRF2 complex
and removes the 3′ overhang from uncapped
telomeres55. Complex phenotypes involving
multiple organ systems could therefore be
caused by the involvement of these nucleases
in multiple pathways of gene expression,
recombination and chromosome stability.
XPG makes the first incision, 3′ to dam-
age, and patients with mutations in XPG are
generally the more severely affected, showing
both XP and XP/CS symptoms51. Most muta-
tions are truncations, resulting in very low
levels of repair. XPG also activates a glycosy-
lase involved in repair of oxidative damage51.
One mild case of XP-G has been reported in
which an alternative splicing event elimi-
nated the TFIIH-interacting region of the
XPG protein, but retained other functions56.
Figure 1 | Nucleotide-excision repair showing the two main branches of transcription-coupled
repair and global genomic repair, their convergence on a common pathway, and the main
genes involved with various steps along the pathways. The initial damage response in transcription-
coupled repair is mediated by the coupling factors CSA and CSB associated with the RNA polymerase II
transcription elongation complex. The initial damage response in global genomic repair is mediated by the
damage-recognition factors XPE (xeroderma pigmentosum, complementation group E) and XPC.
Following damage recognition, the damaged site is remodelled by the helicase activities of XPB and XPD
and binding of XPA and replication protein A (RPA). The nucleases XPG and XPF–ERCC1 then cut the
damaged oligonucleotide region either side of the damage and this is excised. The repair patch is
synthesized by proliferating-cell nuclear antigen (PCNA) and replicative polymerase (Pol) δ. Damage that is
not excised must be replicated by the action of Pol η. The precise order of some of these factors is still the
subject of active research, and might not necessarily be invariant under all conditions. CNS, central
566 | JULY 2005 | VOLUME 5
© 2005 Nature Publishing Group
The symptoms of XP-F patients are mild,
with considerable levels of residual repair
and very rare cases of neurodegeneration57
TABLE 1, and mutations in ERCC1 have not
yet been reported. This is probably because
mutations that would inactivate either pro-
tein would be embryonic or neonatal lethal
in humans, as they are in mice, or cause
very severe symptoms such as those seen
in COFS patients58.
Replication of damaged DNA — Pol η η
DNA replication is a crucial process by
which many proteins are coordinated to
pass on genetic information in a stable,
accurate fashion to succeeding progeny.
The process is even more complex when
damaged DNA is to be processed, muta-
tions minimized and genomic stability
maintained59. An impressive array of
checkpoints delays cell proliferation to per-
mit time for repair to be completed. Loss
of replication control by mutation or loss
of one or more checkpoint or replication
components is a major contributor to the
genetic instability leading to chromosomal
and gene copy number changes associated
with cancer60 and is the underlying cause of
the XP-V disease.
XP-V patients have mutations in a low-
fidelity Y-family DNA polymerase, Pol η, that
is required for bypass replication of CPDs and
other lesions61–63 TABLE 1; BOX 3. The enzyme
has a catalytic region at the 5′ end of the
gene and signal sequences at the 3′ end for
translocation to arrested DNA replication
sites and binding to proliferating-cell nuclear
antigen (PCNA)64,65. Pol η is also involved in
somatic hypermutation during B-cell immu-
noglobulin class-switch recombination66–69.
One of the distinctive features of XP-V
cells has been their increased sensitivity
to toxic effects of high concentrations of
caffeine after UV-light irradiation70. One
interpretation attributes the sensitization to
the interference of caffeine with members of
the phosphatidylinositol-3-kinase-like family
of proteins, ATM, ATR and DNA-depend-
ent protein kinase, which are involved in the
DNA-damage response71, but the mechanism
is still obscure. The presence of p53 is required
for sensitization, because p53-null XP-V cells
show a greatly increased sensitivity to UV
light with no enhancement by caffeine72.
XP-V cells accumulate in the S-phase
after UV-light-induced damage, indicating
that Pol η is required for UV-light-damaged
cells to pass through the S-phase check-
point73. During this arrest, XP-V cells dem-
onstrate increased DNA-fork breakage that
might signify chromosome instability74,75.
The stability of these arrested replication
forks is strongly dependent on the DNA-
binding capacity of p53 for DNA junctions
and termini76,77, on the single-strand-bind-
ing protein RPA78, and on p53-dependent
repression of RAD51 recombination79.
Recovery of DNA replication after UV-light-
induced damage then triggers a range of
Table 1 | Genes affected in xeroderma pigmentosum/Cockayne syndrome
Gene ChromosomeSize (aa)Residual
11p11-12 427>50 Damage binding,
3p25.1940 5–20 Damage binding
5q12.1396100 Ubiquitylation E3
13q32-33 1186 <23ʹ nuclease
XPV (Pol η) 6p21713100
Table compiled from data in REF. 160.
Box 2 | Ubiquitylation and the response to ultraviolet-light damage
Many of the major pathways of the DNA-damage response involve protein modification and
degradation by ubiquitylation82,138. Ubiquitin is a 76-residue polypeptide that is conjugated to
target protein substrates to modify their functions (mono-ubiquitylation) or mark them for
degradation through the proteasomal system (poly-ubiquitylation)139–141. The ubiquitylation
and de-ubiquitylation of a substrate involves four classes of enzyme: a ubiquitin-activating
enzyme (E1), a ubiquitin-conjugating enzyme (E2), a ubiquitin protein ligase (E3) and a
de-ubiquitylation enzyme (E4). Many of these are themselves complex multicomponent protein
systems44. The E3 ligases are primarily responsible for conferring substrate specificity and there
are many more E3 enzymes other than ubiquitin-associated proteins139–141. A deficiency in an
enzyme on the ubiquitylation pathway, especially E3 but occasionally earlier enzymes, can
present unpredictable phenotypes that depend on the particular substrate proteins targeted by
the system139,141. E1 activating enzyme and the 19S and 20S proteasome subunits regulate both
global genomic repair (GGR) and transcription-coupled repair (TCR)82.
The XPE heterodimer has a role both as a component of an SCF-type ubiquitylation E3 ligase
and as a target for ubiquitylation itself44,142. Damage-specific DNA binding protein 1 (DDB1),
part of the XPE heterodimer, also acts as a cofactor for E3 ubiquitylation ligases for both DDB2
and CSA44. The HR23B partner of the XPC–HR23B damage-recognition complex interacts with
XPC by its carboxy-terminal end, but its amino-terminal region has ubiquitin-like regions143,144.
These regions interact with other ubiquitylated proteins and with proteasome subunits to inhibit
protein degradation138,144–147. Recruitment of XPC to damaged sites also requires the function of
the 20S proteasome subunit.
The replication of damaged DNA involves mono-ubiquitylation, and the initial enzymes on
the post-replication repair pathway — RAD6 and RAD18 — are ubiquitylation enzymes148 that
might be responsible for modifying proliferating-cell nuclear antigen to recruit the bypass
polymerases η and ι REF. 65 BOX 3.
Ubiquitylation has consequently emerged as a new important factor in modulating the
activity of repair enzymes along each of the main branches of the ultraviolet-light response,
GGR, TCR and bypass replication, and might represent novel targets for modifying repair
activities in therapies for repair-deficient diseases.
NATURE REVIEWS | CANCER
VOLUME 5 | JULY 2005 | 567
© 2005 Nature Publishing Group
recombination processes, involving both
homologous recombination that is depend-
ent on RAD51, and fork breakage associ-
ated with the MRE11–RAD50–NBS1 and
γH2AX74,75,80 complex. Enhanced replica-
tion-fork instability could be an important
contributor to genomic instability that
presages cancer. XP-V patients are clini-
cally very similar to XP-C patients, indicat-
ing that failure of either global repair or of
DNA bypass replication contribute equiva-
lently to genomic instability and cancer
incidence1,60,61 BOX 3.
Tissue-specific modifications of NER
Many of the tissues involved in the clini-
cal presentation of XP and related diseases
are partially or terminally differentiated,
which can affect their repair capacity. The
precise mechanisms of DNA repair that
operate in different terminally differenti-
ated tissues are unclear. Some differenti-
ated cells of blood and neural origin, for
example, appear to strongly downregulate
GGR, leaving residual repair in transcribed
regions of the genome81. Keratinocytes, in
contrast, upregulate GGR relative to TCR
and show an altered apoptotic and p53
response to UVB light that is independ-
ent of TCR38. The residual repair in neural
tissues is subtly different from TCR in
proliferating cells. The transcribed and
non-transcribed strands of active genes
appear to be repaired initially at similar
rates (named differentiation-associated
repair), in contrast to the enhanced repair
on the transcribed strand usually associated
with TCR in proliferating cells81. A recent
study has shown that a temperature-sensi-
tive mutation in the E1 activation enzyme
of the ubiquitylation system BOX 2 can
downregulate both GGR and TCR, which
might be one mechanism for a global
control of NER in differentiated tissues82.
The testis is also inherently repair defi-
cient, lacking XPA, both excision nucleases,
XRCC1 and Pol η, perhaps explaining the
sensitivity of testicular cancer to treatment
by chemotherapeutic agents that are DNA
damaging (such as cisplatin)83,84.
Cancer and chromosomal instability
Cancer is the characteristic disease of most XP
patients, but not of TCR-defective CS patients
or TTD patients, despite their sun sensitivity.
Although the absence of repair in XP can be
readily linked to cancer through increased
UV-light-induced mutagenesis, the lack of
cancer in CS remains an enigma. Studies with
UV-light-damaged episomal plasmids show
that CS cells specifically fail to repair CPDs,
and that both XP and CS cells show increased
mutagenesis85,86, indicating that a difference in
mutagenesis levels does not strongly discrimi-
nate between these two diseases. However,
mutagenesis in episomal vectors might not
involve the full spectrum of damage responses,
such as apoptosis, cell-cycle checkpoints and
chromosomal instability, that occur when
whole cells are damaged.
The increased mutagenesis seen in both XP
and CS cells might therefore be necessary but
not sufficient for carcinogenesis. It has been
inferred that that XP-related cancers should
not involve chromosome instability and aneu-
ploidy because of their high gene mutation
rates87. Indeed, mutation rates are increased
in TP53 REF. 88 and patched (PTC) genes89
in XP tumours, but experimental evidence for
genome copy number changes or aneuploidy
has yet to be obtained. But the absence of can-
cer in CS patients compared with XP patients,
both of whom have similarly increased muta-
tion rates in response to UV light, indicates
that additional chromosome instability might
actually be required for malignant transfor-
mation in XP. Chromosomal instability has
been considered as crucial in carcinogenesis
for other tissues60. Chromosomal instability in
XP, in contrast to CS, might therefore repre-
sent a global genome response when GGR or
bypass replication is defective. The absence of
such increased genomic instability from CS
cells, which have a normal global damage
response, would minimize the development
of cancer in TCR-defective cells despite their
increased point-mutation rates.
The clinical spectrum of disorders presented
by XP and related diseases clearly indicated
early on that more than just DNA repair was
impaired1. Alternative targets for repair
might be generated by endogenous oxida-
tive damage in specific tissues like brain90.
However, once a link between repair and
transcription was established, these diseases
could be envisaged as also being tran-
scription-defective diseases. More recent
discoveries that certain repair proteins are
associated with ubiquitylation44,91 indicate
that the secondary modification of proteins
Box 3 | The Y-family polymerases
DNA photoproducts are blocks to the replicative DNA polymerases (Pols) α, δ and ε,
which cannot accommodate large distortions, such as DNA photoproducts or adducts, in
their active sites149. Replicative bypass of these photoproducts is achieved by damage-
specific Y-family polymerases that have larger active sites for nucleotide binding and
relaxed substrate specificity150,151. The Y-family polymerases have unique properties of
accommodating large adducts within their active sites for replicating past such damage,
with a wide range of substrate specificity, albeit with a significant error rate that can be as
high as 1–2%150–154. Three Y-family polymerases have been identified in the mammalian
genome — Pol η, ι and κ150.
Mutations in Pol η are associated with a subset of xeroderma pigmentosum (XP) patients
classified as the XP variants (XP-V; TABLE 1) 61–63. Pol η specifically replicates past ultraviolet
(UV)-light photoproducts and other large adducts. Pol ι has a poorer capacity for
replication of UV-light damage, and Pol κ seems unable to replicate UV-light damage but
might replicate other damage. The base specificity for Pol η and Pol ι indicate that Pol η
might exercise a preference for replicating T-containing photoproducts, and Pol ι a
preference for C-containing sites. Other polymerases, especially Pol κ and Pol ζ, can extend
a DNA chain from mismatched termini, leading to a two-step process of bypassing
damage155,156. The loss of Pol η function results in increased UV-light-induced mutagenesis
because of the need to recruit other, less accurate replication processes157.
Y-family polymerases need to be excluded from replication forks except when needed,
otherwise the results would be catastrophic to the cells owing to their low fidelity. Several
mechanisms regulate access to the replication forks. Replication arrest might facilitate
increased access to blocked forks. The polymerases only translocate to replication foci after
irradiation70,84, and proliferating-cell nuclear antigen is modified by mono-ubiquitylation to
facilitate polymerase binding65. Since Pol η acts distributively, extending the nascent DNA
chain by only a few bases across from photoproducts, before the replicative polymerases
resume chain extension, the absolute error frequency will be low compared with replication on
extended regions of undamaged DNA.
Several early studies on nucleotide-excision repair (NER) showed a small but consistently
greater DNA strand break frequency associated with excision in XP-V cells than normal
cells158,159. This could now be re-interpreted to indicate that Pol η might have a small role in the
re-synthesis step of NER, perhaps when the parental strand also contains a photoproduct or
other refractory sequence.
568 | JULY 2005 | VOLUME 5
© 2005 Nature Publishing Group
Global damage NER
DNA replication arrest (XP-V)Genomically unstable cells
(XPA, TFIIH, XPG, XPF)
Neuronal apoptosisCNS disorder
RNA Pol II arrest
Genomically stable cells
Cancer induction by direct defects in global responses
Neurodegeneration by apoptosis and transcription arrest
for signalling and degradation is also impor-
tant. A significant number of the clinical
symptoms of these diseases might therefore
be distantly related to their repair defects,
but depend on the targets for disruption of
gene expression, protein stability and signal
transduction. The protein products of XPB
and XPD, for example, are involved in
basal and hormonal-induced transcription
through the AB domain of nuclear receptors
such as retinoic acid and oestrogen receptors
and in transcription of MYC92,93.
Neurodegeneration and developmental
disorders are major features of the NER dis-
orders BOXES 1,2. Patients with mutations
in genes unique to GGR (for example, XPC
and XPE) and DNA replication (for exam-
ple, Pol η) have fewest clinical symptoms
outside of cancer. Patients with mutations
in components of TCR (CSA and CSB) and
common elements (XPA, XPB, XPD, XPF
and XPG) TABLE 1; FIG. 1 show very com-
plex clinical symptoms involving the central
nervous system and other organ systems1.
Three of these genes — XPB, XPD and
XPG — give rise to combined symptoms of
XP/TTD and XP/CS, according to the par-
ticular mutations carried by the patients11.
Mutations in XPD have relatively similar
levels of repair deficiencies, but elicit a wide
range of clinical disorders ranging across
the complete spectrum of XP, CS and TTD
that are influenced by transcription capac-
ity94,95. Reductions in levels of TFIIH due to
certain mutations in several components
— XPB, XPD or TFB5 — result in the TTD
disorder50,95, indicating this disease can be
caused by limiting amounts of TFIIH in dif-
ferentiating keratinocytes and hair-follicle
cells, and other cell types, and can even be
associated with β-thallasaemia96, a genetic
disorder of anaemia.
Neurodegeneration might be ascribed
to the greater relative importance of a
TCR-like repair in differentiated brain
cells — transcription-blocking lesions
that persist due to a failure of TCR could
trigger an apoptotic response that would
be pathological30,31,39. The most obvious
source for damage would be the high level
of oxidative metabolism that occurs in
the brain and produces lesions that cause
transcription arrest90,97. Active transcrip-
tion can be estimated to involve only about
1–2% of total genomic DNA24,25, indicating
that the lethal (apoptotic) signal gener-
ated by a failure of TCR of active genes30,39
must be about 50 to 100 times as potent
per unit of repairable DNA, as is the signal
for lethal events from a failure of GGR.
This potent apoptotic signal might be a
source for the pathological consequences
in CS and cause cell loss from non-divid-
ing tissues such as brain and retina (FIG. 2).
Apoptosis that removed damaged cells
from the skin would conversely prevent
UV-light carcinogenesis, especially in CS
cells38. Oxidative damage has been reported
in the brains of repair-deficient patients98.
Some tissues that degenerate in CS appear
to be unusually sensitive to oxygen levels,
including the Purkinje and retinal cells99–101.
Defects in protein processing would make
CS a disease similar in principle to other
neurodegenerative disorders such as
Alzheimer’s, Parkinson’s, Lou-Gehrig’s and
prion diseases that involve ubiquitylation
and defects in protein degradation102–104.
There has been steady progress in making
mouse knockout strains that reflect the
human DNA-repair defects105. In gen-
eral, these display the cancer symptoms
associated with repair defects, and some
additional phenotypes of severe neonatal
death that would be missed in humans.
Thus, Xpa–/–, Xpc–/– and Xpe–/– mice all
show increased skin cancer from UV
light or chemical carcinogens, but so do
Csa–/–, Csb–/– and Ttd–/– mice. Xpd–/– and
Xpb–/– mice are embryonic lethal, consist-
ent with the human, in which only point
mutations are viable, and a mouse mim-
icking a human TTD mutation is viable.
Xpg–/–, Xpf–/– and Ercc1–/– strains all have
severe developmental disorders.
Human XP or CS heterozygotes appear
to have no clearly demonstrable symptoms;
however, some (but not all) Xpc+/– and
Xpe+/– and other heterozygous knockouts
show increased cancer incidence after
UVB or carcinogen exposure106,107. The
variability points to additional roles in skin
carcinogenesis from strain backgrounds
and modifier genes. A comparison between
Xpc–/– and Csb–/– mice on the same genetic
background showed that although cancer
incidence from UVB light was increased in
both, Xpc–/– mice were more sensitive than
Csb–/– mice108. These outcomes indicate that
failure of GGR is more strongly associated
with cancer predisposition, and failure of
TCR is associated with acute effects such
as immunosuppression and erythema
(UV-light-induced skin irritation).
But there are other areas of major differ-
ence between human and mice: neurode-
generation, especially, is much milder and
more difficult to develop in mice105,108,109.
The Xpa–/– knockout strain, despite corre-
sponding to one of the more severe groups
of human patients, shows no neurological
disorder throughout its lifespan110–112. Several
examples have shown that reducing overall
repair capacity by crossing repair-deficient
mouse strains with Xpa–/– mice that have
no NER can increase the severity of neu-
rodegenerative phenotypes. These include
Csb–/– mice crossed with Xpa–/– mice113;
Ttd–/– mice crossed with Xpa–/– mice114; and
Figure 2 | Mechanisms by which cancer or neurodegeneration could be caused by defects in
global genomic repair and bypass replication or by defects in transcription-coupled repair.
Defects in the common pathway can give rise to both sets of symptoms according to individual gene
functions. Damage to the whole genome (predominantly non-transcribed regions) is processed by XPE
(xeroderma pigmentosum, complementation group E) and XPC, other nucleotide-excision repair (NER)
components and polymerase (Pol) η, and defects could result in genomically unstable cells that, under
selection, could result in malignant transformation. Damage to transcriptionally active genes causes
transcription arrest, and defects in transcription-coupled repair might not cause a high degree of genomic
instability, but might enhance apoptosis that causes neuronal loss in the brain. CNS, central nervous
system; XP-V, xeroderma pigmentosum variant.
NATURE REVIEWS | CANCER
VOLUME 5 | JULY 2005 | 569
© 2005 Nature Publishing Group
Xpg–/– mice crossed with Xpa–/– mice115. But
crossing Xpc–/– mice with Xpa–/– mice does
not enhance the cancer susceptibility or cause
the development of neurological symptoms105.
The repair function of Xpc must therefore be
channelled completely within the functions of
Xpa, whereas the functions of the other genes
involve downstream end points additional to
those determined by Xpa.
TFIIH is an essential transcription factor,
so complete knockouts of these are lethal
at the 2-cell embryo stage105. Engineering
a human TTD mutation in the mouse Xpd
gene recapitulated most of the human
disease except for an observed increase in
cancer susceptibility in the mouse, and a
potential role for XPD-mediated DNA repair
in ageing116,117. The knockout of Xpg reflects
the severe disorders seen in human XP-G
patients and has been used to define the
last 360 amino acids as crucial for nuclease
function and growth retardation118. Similarly,
knockouts of either Ercc1 REF. 119 or Xpf58
show severe neonatal-lethal phenotypes.
Mouse strains with defective Y-family
polymerases are all viable. Deletion of a large
5ʹ region of the gene was not viable. Targeting
exons 4 or 8 in the catalytic region at the 5ʹ
end of the gene120 or exon 11 at the 3ʹ end of
the gene that interacts with PCNA has pro-
duced viable fertile mice (R.R. Laposa and
C.J.E., unpublished observations). An inacti-
vating mutation in the Pol ι gene has occurred
spontaneously in the mouse 129 strain, a
strain often used as a source of embryonic
stem cells for generation of knockout mice
by homologous recombination121.
Generalized defects in oxidative repair
generally do not cause the kind of neurolog-
ical symptoms associated with CS. Neither
mice defective in repair of 8-oxo-guanine99,122
nor mice with defective Xrcc1 (which is cru-
cial for most base-excision repair123,124) show
neurodegenerative symptoms. However,
oxidative stress has been suggested as a
mechanism for the loss of Purkinje cells
in repair-deficient Xpg–/– mice101 and in
Atm-null mice100, with stimulation of inap-
propriate activation of the cell cycle125. The
offending lesions that cause endogenous
DNA damage relevant to neurodegenera-
tion might therefore be a subset of chemical
modifications caused by reactive oxidative
species such as lipid peroxidation products97
or cyclo-dA90 (a cyclic oxidative product
of adenine formed in DNA) that act as
transcription-blocking lesions subject
to TCR, rather than single base lesions
like 8-oxo-guanine that are repaired by
The cancer incidence in Csa–/–, Csb–/–and
Ttd–/– mice could be an artefact, caused by
several peculiarities of NER in the mouse
genome. Generally, mouse cells require
fewer steps for malignant transformation
than human cells, and contain longer tel-
omeres126,127. Additionally, DDB2 in mouse
cells is not upregulated by UV light, as in
human cells, because the promoter contains
a mutated inactive p53 response element13,
which might contribute inherent genomic
instability above that in human cells.
The mouse models of NER-defective dis-
eases have therefore become an extremely
informative set of strains with many features
that recapitulate the human condition, but
with some features that could be unique to
the mouse. These strains will surely increase
in value not only for basic understanding
of these diseases but also as platforms to
develop and test therapies that will aid the
Diagnosis, treatment and cure?
The NER diseases present major challenges
in diagnosis and in patient care. UV-light
sensitivity and DNA-repair deficiency can
be used for patient and prenatal diagno-
sis129. These procedures, though easy for
a DNA-repair laboratory, are specialized
techniques and a more general approach
would be direct DNA sequencing available
to most clinical genetic diagnostic laborato-
ries. No strong hot spots have been identi-
fied in the general population, so each gene
and mutation has to be identified directly.
A problem then lies in what should be
done with the information prognostically,
especially in light of available treatment.
Evidence for sun sensitivity and cancer
predisposition in XP launches patients and
their families into radical lifestyle changes
of extreme solar protection. But the muta-
tions in XP and CS genes are not sufficiently
correlated directly with clinical symptoms
to predict with certainty the future progress
of disease, especially when the mutation is
novel and the patient young.
Unique cures are at present unavail-
able, and standard clinical care is the norm,
though some approaches show promise.
Bacterial repair enzymes in liposomes have
been delivered to the skin and show reduc-
tion of UV-light-induced damage130, and
chemoprevention with retinoids showed
promise, but with severe side effects131. An
XPC gene expressed from an adenovirus vec-
tor injected into the skin of an XPC mouse
strain reduces skin carcinogenesis128.
XP was first identified as a DNA-repair-
deficient human disease approximately
35 years ago with a somewhat overconfi-
dent statement: ‘‘Patients with xeroderma
pigmentosum develop fatal skin cancers
when exposed to sunlight, and so the
failure of DNA repair must be related
to carcinogenesis’’132. This has proven a
true, but over-simplified statement. Here,
I have attempted to provide a perspective
on the relationships between molecular
understanding of the NER diseases and the
general nature of their clinical symptoms.
Many case-by-case analyses still need to be
done for individual patients according to
their particular combination of mutations.
Major frontiers remain. One is the eluci-
dation of the structure of each of the NER
proteins and polymerases133,134, together
with an understanding of how they inter-
act and enter into molecular regulatory
networks. A final frontier is to relate, in
ways that will ultimately benefit patients,
Box 4 | Clinical and cellular characteristics unique to GGR or TCR
Xeroderma pigmentosum groups C, E, V
The clinical characteristics of these diseases include sun sensitivity, freckling, skin cancers and
corneal damage. Individuals with XPE are very mildly affected. The clinical characteristics of
XPC and XPV are indistinguishable, and the symptoms are moderate to severe. Most cases are
neurologically normal, though there have been some very rare cases of neurological
The cellular characteristics include ultraviolet (UV)-light sensitivity. Individuals with XPE
and XPV show slight sensitivity, and individuals with XPC are more sensitive but highly
variable. The damage induced by UV light causes reduced DNA repair or DNA replication.
The clinical characteristics of Cockayne syndrome include sun sensitivity, growth failure,
absence of subcutaneous fat, sunken facial features, short life expectancy, progressive
neurological impairment, deafness, retinal degeneration, dysmyelination, and brain and
The cellular characteristics include ultraviolet (UV)-light sensitivity. The damage induced by
UV light causes reduced transcription-coupled repair.
570 | JULY 2005 | VOLUME 5
© 2005 Nature Publishing Group
the clinical features of the NER diseases
to the developing map of the purturbed
XP patients with mutations in the GGR
recognition genes, XPC and XPE, or in
Pol η are at high risk for solar-induced skin
cancer, but rarely, if ever, have significant
neurodegenerative disorders BOX 4, whereas
the situation for CS is the converse1. This
suggests to me that an element of cancer
induction in these diseases involves a glo-
bal destabilization of the genome leading
to increased copy number changes and
aneuploidy, whereas neurodegeneration and
other symptoms are more specifically related
to targets of gene transcription and protein
modification (FIG. 2).
James E. Cleaver is at the Auerback Melanoma
Laboratory, Room N431, UCSF Cancer Center,
Box 0808, Room N431, UCSF Cancer Center,
University of California,
San Francisco, California, 94143-0808, USA.
1. Bootsma, D., Kraemer, K. H., Cleaver, J. E. &
Hoeijmakers, J. H. J. in The Genetic Basis of Human
Cancer (eds. Vogelstein, B. & Kinzler, K. W.) 245–274
(McGraw-Hill, New York, 1998).
Wood, R. D., Mitchell, M., Sgouros, J. & Lindahl, T.
Human DNA repair genes. Science 291, 1284–1289
Batty, D. P. & Wood, R. W. Damage recognition in
nucleotide excision repair of DNA. Gene 241, 193–204
Riedl, T., Hanaoka, F. & Egly, J. M. The comings and
goings of nucleotide excision repair factors on damaged
DNA. EMBO J. 22, 5293–5303 (2003).
Sancar, A., Lindsey-Boltz, L. A., Unsal-Kacmaz, K. &
Linn, S. Molecular mechanisms of mammalian DNA
repair and the DNA damage checkpoints. Annu. Rev.
Biochem. 73, 39–85 (2004).
Balajee, A. S., Dianova, I. & Bohr, V. A. Oxidative
damage-induced PCNA complex formation is efficient in
xeroderma pigmentosum group A but reduced in
Cockayne syndrome group B cells. Nucleic Acids Res.
27, 4476–4482 (1999).
Svejstrup, J. Q. Rescue of arrested RNA polymerase II
complexes. J. Cell Sci. 116, 447–451 (2003).
Volker, M. et al. Sequential assembly of the nucleotide
excision repair factors in vivo. Mol. Cell 8, 213–224
Nichols, A. F. et al. Human damage-specific DNA binding
protein p48. Characterization of XPE mutations and
regulation following UV irradiation. J. Biol. Chem. 275,
10. Masutani, C. et al. Purification and cloning of a nucleotide
excision repair complex involving the xeroderma
pigmentosum group C protein and a human homologue
of yeast RAD23. EMBO J. 13, 1831–1843 (1994).
11. Cleaver, J. E., Thompson, L. H., Richardson, A. S. &
States, J. C. A summary of mutations in the UV-sensitive
disorders: xeroderma pigmentosum, Cockayne
syndrome, and trichothiodystrophy. Hum. Mut. 14, 9–22
12. Itoh, T., Linn, S., Ono, T. & Yamaizumi, M. Reinvestigation
of the classification of five cell strains of xeroderma
pigmentosum group E with reclassification of three of
them. J. Invest. Dermatol. 114, 1022–1029 (2000).
13. Tang, J. & Chu, G. Xeroderma pigmentosum
complementation group E and UV-damaged DNA-
binding protein. DNA Repair 1, 601–616 (2002).
14. Hwang, B. J., Ford, J. M., Hanawalt, P. C. & Chu, G.
Expression of the p48 xeroderma pigmentosum gene is
p53-dependent and is involved in global genome repair.
Proc. Natl Acad. Sci. USA 96, 424–428 (1999).
15. Tang, J. Y., Hwang, B. J., Ford, J. M., Hanawalt, P. C. &
Chu, G. Xeroderma pigmentosum p48 gene enhances
global genomic repair and suppresses UV-induced
mutagenesis. Mol. Cell 5, 737–744 (2000).
16. Liu, W. et al. Nuclear transport of human DDB protein
induced by ultraviolet light. J. Biol. Chem. 275, 21429–
17. Shiyanov, P. et al. The naturally occurring mutants of DDB
are impaired in stimulating nuclear import of the p125
subunit and E2F1-activated transcription. Mol. Cell Biol.
19, 4935–4943 (1999).
18. Tan, T. & Chu, G. p53 Binds and activates the xeroderma
pigmentosum DDB2 gene in humans but not mice. Mol.
Cell Biol. 22, 3247–3254 (2002).
19. Painter, R. B. & Cleaver, J. E. Repair replication,
unscheduled DNA synthesis, and the repair of
mammalian DNA. Radiat. Res. 37, 451–466 (1969).
20. Shivji, M. K., Eker, A. P. & Wood, R. D. DNA repair defect
in xeroderma pigmentosum group C and complementing
factor from HeLa cells. J. Biol. Chem. 269, 22749–22757
21. Sugasawa, K. et al. Xeroderma pigmentosum group C
protein complex is the initiator of global nucleotide
excision repair. Mol. Cell 2, 223–232 (1998).
22. Wood, R. D. DNA damage recognition during nucleotide
excision repair in mammalian cells. Biochimie 81, 39–44
23. Wang, Q. E., Zhu, Q., Wani, G., Chen, J. & Wani, A. A.
UV radiation-induced XPC translocation within chromatin
is mediated by damaged-DNA binding protein, DDB2.
Carcinogenesis 25, 1033–1043 (2004).
24. Consortium, I. H. G. S. Finishing the euchromatic sequence
of the human genome. Nature 431, 931–945 (2004).
25. Boon, W. M. et al. A comparative analysis of transcribed
genes in the mouse hypothalamus and neocortex reveals
chromosomal clustering. Proc. Natl Acad. Sci. USA 101,
26. Wood, R. D. Nucleotide excision repair in mammalian
cells. J. Biol. Chem. 272, 23465–23468 (1997).
27. Sancar, A. DNA excision repair. Annu. Rev. Biochem. 65,
28. Araki, M. et al. Centrosome protein centrin 2/caltractin 1
is part of the xeroderma pigmentosum group C complex
that initiates global nucleotide excision repair. J. Biol.
Chem. 276, 18665–18672 (2001).
29. Ford, J. M. & Hanawalt, P. C. Expression of wild-type p53
is required for efficient global genomic nucleotide excision
repair in UV-irradiated human fibroblasts. J. Biol. Chem.
272, 28073–28080 (1997).
30. Proietti De Santis, L. et al. Transcription coupled repair
efficiency determines the cell cycle progression and
apoptosis after UV exposure in hamster cells. DNA
Repair 1, 209–223 (2002).
31. D’Errico, M. et al. Apoptosis and efficient repair of DNA
damage protect human keratinocytes against UVB. Cell
Death Differ. 10, 754–756 (2003).
32. Bregman, D. B. et al. UV-induced ubiquitination of RNA
polymerase II: a novel modification deficient in Cockayne
syndrome cells. Proc. Natl Acad. Sci. USA 93,
33. Lee, K. B., Wang, D., Lippard, S. J. & Sharp, P. A.
Transcription-coupled and DNA damage-dependent
ubiquitination of RNA polymerase II in vitro. Proc. Natl
Acad. Sci. USA 99, 4239–4244 (2002).
34. Yang, L. J., Jiang, H. & Rangel, K. M. RNA polymerase II
stalled on a DNA template during transcription elongation
is ubiquitinated and the ubiquitination facilitates
displacement of the elongation complex. Int. J. Oncol.
22, 683–689 (2003).
35. Tremeau-Bravard, A., Riedl, T., Egly, J. M. & Dahmus, M. E.
Fate of RNA polymerase II stalled at a cisplatin lesion.
J. Biol. Chem. 279, 7751–7759 (2004).
36. Lommel, L., Bucheli, M. E. & Sweder, K. S.
Transcription-coupled repair in yeast is independent from
ubiquitylation of RNA pol II: implications for Cockayne’s
syndrome. Proc. Natl Acad. Sci. USA 97, 9088–9092
37. Lee, K. B. & Sharp, P. A. Transcription-dependent
polyubiquitination of RNA polymerase II requires lysine
63 of ubiquitin. Biochemistry 43, 15223–15229
38. D’Errico, M. et al. Differential role of transcription-coupled
repair in UVB-induced response of human fibroblasts and
keratinocytes. Cancer Res. 65, 432–438 (2005).
39. Ljungman, M. & Zhang, F. Blockage of RNA polymerase
as a possible trigger for u. v. light-induced apoptosis.
Oncogene 13, 823–831 (1996).
40. Selby, C. P. & Sancar, A. Cockayne syndrome group B
protein enhances elongation by RNA polymerase II. Proc.
Natl Acad. Sci. USA 94, 11205–11209 (1997).
41. Hara, R., Selby, C. P., Liu, M., Price, D. H. & Sancar, A.
Human transcription release factor 2 dissociates RNA
polymerases I and II stalled at a cyclobutane thymine
dimer. J. Biol. Chem. 274, 24779–24786 (1999).
42. Shilatifard, A. Factors regulating the transcriptional
elongation activity of RNA polymerase II. FASEB J. 12,
43. Beerens, N., Hoeijmakers, J. H., Kanaar, R., Vermeulen, W.
& Wyman, C. The CSB protein actively wraps DNA.
J. Biol. Chem. 280, 4722–4729 (2005).
44. Groisman, R. et al. The ubiquitin ligase activity in the
DDB2 and CSA complexes is differentially regulated by
the COP9 signalosome in response to DNA damage. Cell
113, 357–367 (2003).
45. Kamiuchi, S. et al. Translocation of Cockayne syndrome
group A protein to the nuclear matrix: possible relevance
to transcription-coupled DNA repair. Proc. Natl Acad.
Sci. USA 99, 201–206 (2002).
46. McDowell, M., Nguyen, T. & Cleaver, J. E. A single-site
mutation in the XPAC gene alters photoproduct
recognition. Mutagenesis 8, 155–161 (1993).
47. Itin, P. H., Sarasin, A. & Pittelkow, M. R.
Trichothiodystrophy: update on the sulfur-deficient brittle
hair syndromes. J. Am. Acad. Dermatol. 44, 891–920
48. Mondello, C. et al. Molecular analysis of the XP-D-gene in
Italian families with patients affected by
trichothiodystrophy and xeroderma pigmentosum
group D. Mutat. Res. 314, 159–165 (1994).
49. Broughton, B. C., Steingrimsdottir, H., Weber, C. A. &
Lehmann, A. R. Mutations in the xeroderma
pigmentosum group D DNA repair/transcription gene in
patients with trichothiodystrophy. Nature Genet. 7,
50. Giglia-Mari, G. et al. A new, tenth subunit of TFIIH is
responsible for the DNA repair syndrome trichothiodystrophy
group A. Nature Genet. 36, 714–719 (2004).
51. Clarkson, S. G. The XPG story. Biochimie 85, 1113–1121
52. Gaillard, P. H. & Wood, R. D. Activity of individual ERCC1
and XPF subunits in DNA nucleotide excision repair.
Nucleic Acids Res. 29, 872–879 (2001).
53. Mone, M. J. et al. In vivo dynamics of chromatin-
associated complex formation in mammalian nucleotide
excision repair. Proc. Natl Acad. Sci. USA 101,
54. Niedernhofer, L. J. et al. The structure-specific
endonuclease Ercc1–Xpf is required to resolve DNA
interstrand crosslink-induced double-strand breaks. Mol.
Cell. Biol. 24, 5776–5787 (2004).
55. Zhu, X. D. et al. ERCC1/XPF removes the 3ʹ overhang
from uncapped telomeres and represses formation of
telomeric DNA-containing double minute chromosomes.
Mol. Cell 12, 1498–1498 (2003).
56. Thorel, F. et al. Definition of a short region of XPG necessary
for TFIIH interaction and stable recruitment to sites of UV
damage. Mol. Cell. Biol. 24, 10670–1080 (2004).
57. Sijbers, A. M. et al. Homozygous R488W point mutation
in the XPF gene of a patient with xeroderma
pigmentosum and late-onset neurologic disease.
J. Invest. Dermatol. 110, 832–836 (1998).
58. Tian, M., Shinkura, R., Shinkura, N. & Alt, F. W. Growth
retardation, early death, and DNA repair defects in mice
deficient for the nucleotide excision repair enzyme XPF.
Mol. Cell. Biol. 24, 1200–1205 (2004).
59. Toueille, M. & Hubscher, U. Regulation of the DNA
replication fork: a way to fight genomic instability.
Chromosoma 113, 113–125 (2004).
60. Nowak, M. A. et al. The role of chromosomal instability
in tumor initiation. Proc. Natl Acad. Sci. USA 99,
61. Cleaver, J. E. Xeroderma pigmentosum: variants with
normal DNA repair and normal sensitivity to ultraviolet
light. J. Invest. Dermatol. 58, 124–128 (1972).
62. Johnson, R. E., Kondratick, C. M., Prakash, S. &
Prakash, L. hRAD30 mutations in the variant form of
xeroderma pigmentosum. Science 264, 263–265 (1999).
63. Masutani, C. et al. The XPV (xeroderma pigmentosum
variant) gene encodes human DNA polymerase η. Nature
399, 700–704 (1999).
64. Kannouche, P. et al. Localization of DNA polymerases µ
and ι to the replication machinery is tightly coordinated in
human cells. EMBO J. 22, 1223–1233 (2003).
65. Kannouche, P. L., Wing, J. & Lehmann, A. R. Interaction
of human DNA polymerase µ with monoubiquitinated
PCNA: a possible mechanism for the polymerase switch
in response to DNA damage. Mol. Cell 14, 491–500
66. Faili, A. et al. Induction of somatic hypermutation in
immunoglobulin genes is dependent on DNA polymerase ι.
Nature 419, 944–947 (2002).
67. Franklin, A., Milburn, P. J., Blanden, R. V. & Steele, E. J.
Human DNA polymerase η, an A-T mutator in somatic
hypermutation of rearranged immunoglobulin genes, is a
NATURE REVIEWS | CANCER
VOLUME 5 | JULY 2005 | 571
© 2005 Nature Publishing Group
reverse transcriptase. Immunol. Cell Biol. 82, 219–225
68. Zeng, X. et al. DNA polymerase η is an A-T mutator in
somatic hypermutation of immunoglobulin variable
genes. Nature Immunol. 2, 537–541 (2001).
69. Zeng, X., Negrete, G. A., Kasmer, C., Yang, W. W. &
Gearhart, P. J. Absence of DNA polymerase η reveals
targeting of C mutations on the nontranscribed strand in
immunoglobulin switch regions. J. Exp. Med. 199,
70. Kannouche, P. et al. Domain structure, localization, and
function of DNA polymerase η, defective in xeroderma
pigmentosum variant cells. Genes Dev. 15, 158–172
71. Sarkaria, J. N. et al. Inhibition of ATM and ATR kinase
activities by the radiosensitizing agent, caffeine. Cancer
Res. 59, 4375–4382 (1999).
72. Cleaver, J. E. et al. Increased UV sensitivity and
chromosomal instability related to p53 function in the
xeroderma pigmentosum variant. Cancer Res. 59,
73. Laposa, R. R., Feeney, L. & Cleaver, J. E. Recapitulation
of the cellular xeroderma pigmentosum-variant
phenotype using short interfering RNA for DNA
polymerase H. Cancer Res. 63, 3909–3912 (2003).
74. Limoli, C. L., Giedzinski, E., Morgan, W. F. & Cleaver, J. E.
Polymerase η deficiency in the XP variant uncovers an
overlap between the S phase checkpoint and double
strand break repair. Proc. Natl Acad. Sci. USA 97,
75. Limoli, C. L., Giedzinski, E., Bonner, W. M. & Cleaver, J. E.
UV-induced replication arrest in the xeroderma
pigmentosum variant leads to double strand breaks,
γ-H2Ax formation, and Mre11 relocalization. Proc. Natl
Acad. Sci. USA 99, 233–238 (2002).
76. Lee, S., Cavallo, L. & Griffith, J. Human p53 binds
Holliday junctions strongly and facilitates their cleavage.
J. Biol. Chem. 272, 7532–7539 (1997).
77. Jayaraman, J. & Prives, C. Activation of p53 sequence-
specific DNA binding by short single strands of DNA
requires the p53 C-terminus. Cell 81, 1021–1029
78. Dutta, A., Ruppert, J. M., Aster, J. C. & Winchester, E.
Inhibition of DNA replication factor RPA by p53. Nature
365, 79–82 (1993).
79. Buchhop, S. et al. Interaction of p53 with the human
rad51 protein. Nucleic Acids Res. 25, 3868–3874 (1997).
80. Limoli, C. L., Giedzinski, E. & Cleaver, J. E. Alternative
recombination pathways in UV-irradiated XP Variant
cells. Oncogene 7 Mar 2005 (doi:1038/
81. Neuspikel, T. & Hanawalt, P. C. Terminally differentiated
human neurons repair transcribed genes but display
attenuated global DNA repair and modulation of repair
gene expression. Mol. Cell. Biol. 20, 1562–1570
82. Wang, Q. E. et al. Cellular ubiquitination and proteasomal
functions positively modulate mammalian nucleotide
excision repair. Mol. Carcinog. 42, 53–64 (2005).
83. Welsh, C. et al. Reduced levels of XPA, ERCC1 and XPF
DNA repair proteins in testis tumor cell lines. Int. J.
Cancer 110, 352–361 (2004).
84. Thakur, M. et al. DNA polymerase η undergoes alternative
splicing, protects against UV sensitivity and apoptosis,
and suppresses Mre11-dependent recombination. Genes
Chromosom. Cancer 32, 222–235 (2001).
85. Barrett, S. F., Robbins, J. H. Tarone, R. E. & Kraemer, K. H.
Evidence for defective repair of cyclobutane dimers with
normal repair of other photoproducts in a transcriptionally
active gene transfected into Cockayne syndrome cells.
Mutat. Res. 255, 281–291 (1991).
86. Parris, C. H. & K. H. Kraemer. Ultraviolet-light induced
mutations in Cockayne syndrome cells are primarily
caused by cyclobutane dimer photoproducts while repair
of other photoproducts is normal. Proc. Natl Acad. Sci.
USA 90, 7260–7264 (1993).
87. Rajagopalan, H. & Lengauer, C. Aneuploidy and cancer.
Nature 432, 338–341 (2004).
88. Spatz, A., Giglia-Mari, G., Benhamou, S. & Sarasin, A.
Association between DNA repair-deficiency and high
level of p53 mutations in melanoma of xeroderma
pigmentosum. Cancer Res. 61, 2480–2486 (2001).
89. Bodak, N. et al. High levels of patched gene mutations in
basal-cell carcinomas from patients with xeroderma
pigmentosum. Proc. Natl Acad. Sci. USA 96, 5117–5122
90. Kuraoka, I. et al. Removal of oxygen free-radical-induced
5ʹ,8-purine cyclodeoxynucleosides from DNA by the
nucleotide excision-repair pathway in human cells. Proc.
Natl Acad. Sci. USA 97, 3832–3837 (2000).
91. Weeda, G. et al. The XPB subunit of repair/transcription
factor TFIIH directly interacts with SUG1, a subunit of the
26S proteasome and putative transcription factor.
Nucleic Acids Res. 25, 2274–2283 (1997).
92. Schaeffer, L. et al. DNA repair helicase: a component of
BTF2 (TFIIH) basic transcription factor. Science 260,
93. Keriel, A., Stary, A., Sarasin, A., Rochette-Egly, C. & Egly, J.-M.
XPD mutations prevent TFIIH-dependent transactivation
by nuclear receptors and phosphorylation of RARα. Cell
109, 125–135 (2002).
94. Taylor, E. M. et al. Xeroderma pigmentosum and
trichothiodystrophy are associated with different
mutations in the XPD (ERCC2) repair/transcription gene.
Proc. Natl Acad. Sci. USA 94, 8658–8663 (1997).
95. Dubaele, S. et al. Basal transcription defect discriminates
between xeroderma pigmentosum and
trichothiodystrophy in XPD patients. Mol. Cell 11,
96. Viprakasit, V. et al. Mutations in the general transcription
factor TFIIH result in β-thalassaemia in individuals with
trichothiodystrophy. Hum. Mol. Genet. 10, 2797–2802
97. Cline, S. D., Riggins, J. N., Tornaletti, S., Marnett, L. J. &
Hanawalt, P. C. Malondialdehyde adducts in DNA arrest
transcription by T7 RNA polymerase and mammalian
RNA polymerase II. Proc. Natl Acad. Sci. USA 101,
98. Hayashi, M. et al. Oxidative stress and disturbed glutamate
transport in hereditary nucleotide repair disorders. J.
Neuropathol. Exp. Neurol. 60, 350–356 (2001).
99. de Waard, H. et al. Cell type-specific hypersensitivity to
oxidative damage in CSB and XPA mice. DNA Repair 2,
100. Chen, P. et al. Oxidative stress is responsible for deficient
survival and dendritogenesis in purkinje neurons from
ataxia-telangiectasia mutated mutant mice. J. Neurosci.
Res. 23, 11453–11460 (2003).
101. Sun, X. Z., Harada, Y. N., Takahashi, S., Shiomi, N. &
Shiomi, T. Purkinje cell degeneration in mice lacking the
xeroderma pigmentosum group G gene. J. Neurosci.
Res. 64, 348–354 (2001).
102. Ciechanover, A. & Brundin, P. The ubiquitin proteasome
system in neurodegenerative diseases: sometimes the
chicken, sometimes the egg. Neuron 40, 427–446
103. Lopez Salon, M., Morelli, L., Castano, E. M., Soto, E. F. &
Pasquini, J. M. Defective ubiquitination of cerebral
proteins in Alzheimer’’s disease. J. Neurosci. Res. 62,
104. von Coelln, R., Dawson, V. L. & Dawson, T. M. Parkin-
associated Parkinson’s disease. Cell Tissue Res. 318,
105. Friedberg, E. C., Meira, L. B. & Cheo, D. L. Database of
mouse strains carrying targeted mutations in genes
affecting cellular responses ot DNA damage. Version 2.
Mutat. Res. 407, 217–226 (1998).
106. Itoh, T., Cado, D., Kamide, R. & Linn, S. DDB2 gene
disruption leads to skin tumors and resistance to
apoptosis after exposure to ultraviolet light but not a
chemical carcinogen. Proc. Natl Acad. Sci. USA 101,
107. Cheo, D. L. et al. Ultraviolet B radiation-induced skin
cancer in mice defective in the Xpc, Trp53, and Apex
(HAP1) genes: genotype-specific effects on cancer
predisposition and pathology of tumors. Cancer Res. 60,
108. Berg, R. J. et al. Impact of global genome repair versus
transcription-coupled repair on ultraviolet
carcinogenesis in hairless mice. Cancer Res. 60, 2858–
109. van der Horst, G. T. et al. UVB radiation-induced cancer
predisposition in Cockayne syndrome group A (Csa)
mutant mice. DNA Repair 1, 143–157 (2002).
110. Berg, R. J. W., Vries, A. D., Steeg, H. V. & Gruijl, F. R. D.
Relative susceptibilities of XPA knockout mice and their
heterozygous and wild-type littermates to UVB-induced
skin cancer. Cancer Res. 57, 581–584 (1997).
111. De Vries, A. et al. Increased susceptibility to ultraviolet-B
and carcinogens of mice lacking the DNA excision repair
gene XPA. Nature 377, 169–173 (1995).
112. Tanaka, K. et al. UV-induced skin carcinogenesis in
xeroderma pigmentosum group A (XPA) gene-knockout
mice with nucleotide excision repair-deficiency. Mutat.
Res. 477, 31–40 (2001).
113. Murai, M. et al. Early postnatal ataxia and abnormal
cerebellar development in mice lacking Xeroderma
pigmentosum Group A and Cockayne syndrome Group B
DNA repair genes. Proc. Natl Acad. Sci. USA 98, 13379–
114. de Boer, J. et al. Premature aging in mice deficient in
DNA repair and transcription. Science 296, 1276–1279
115. Shiomi, N. et al. Severe growth retardation and short life-
span of double mutant mice lacking Xpa and exon 15 of
Xpg. DNA Repair 4, 351–357 (2005).
116. de Boer, J. et al. A mouse model for the basal
transcription/DNA repair syndrome trichothiodystrophy.
Mol. Cell 1, 981–990 (1998).
117. Mitchell, J. R., Hoeijmakers, J. H. & Niedernhofer, L. J.
Divide and conquer: nucleotide excision repair battles
cancer and ageing. Curr. Opin. Cell Biol. 15, 232–240
118. Shiomi, N. et al. Identification of the XPG region that
causes the onset of Cockayne syndrome by using Xpg
mutant mice generated by the cDNA-mediated knock-in
method. Mol. Cell. Biol. 24, 3712–3719 (2004).
119. Prasher, J. M. et al. Reduced hematopoietic reserves in
DNA interstrand crosslink repair-deficient Ercc1–/– mice.
EMBO J. 24, 861–871 (2005).
120. Delbos, F. et al. Contribution of DNA polymerase η to
immunoglobulin gene hypermutation in the mouse.
J. Exp. Med. 201, 1191–1196 (2005).
121. McDonald, J. P. et al. 129-derived strains of mice are
deficient in DNA polymerase ι and have normal
immunoglobulin hypermutation. J. Exp. Med. 198,
122. Osterod, M. et al. A global DNA repair mechanism
involving the Cockayne syndrome B (CSB) gene product
can prevent the in vivo accumulation of endogenous
oxidative DNA base damage. Oncogene 21, 8232–8239
123. Tebbs, R. S. et al. Requirement for the Xrcc1 DNA base
excision repair gene during early mouse development.
Dev. Biol. 208, 513–529 (1999).
124. Tebbs, R. S., Thompson, L. H. & Cleaver, J. E. Rescue
of Xrcc1 knockout mouse embryo lethality by
transgene complementation. DNA Repair 2, 1405–
125. Yang, Y. & Herrup, K. Loss of neuronal cell cycle control
in ataxia-telangiectasia:a unified disease mechanism.
J. Neuroscience 25, 2522–2529 (2005).
126. Rangarajan, A., Hong, H. J., Gifford, A. & Weinberg, R. A.
Species and cell type-specific requirements for cellular
transformation. Cancer Cell 6, 171–183 (2004).
127. Blackburn, E. H. Structure and function of telomeres.
Nature 350, 569–573 (1991).
128. Marchetto, M. C., Muotri, A. R., Burns, D. K., Friedberg, E. C.
& Menck, C. F. Gene transduction in skin cells: preventing
cancer in xeroderma pigmentosum mice. Proc. Natl
Acad. Sci. USA 101, 17759–17764 (2004).
129. Cleaver, J. E., Volpe, J. P. G., Charles, W. C. & Thomas, G. H.
Prenatal diagnosis of xeroderma pigmentosum and
Cockayne syndrome. Prenat. Diagn. 14, 921–928 (1994).
130. Yarosh, D. et al. Effect of topically applied T4
endonuclease V in liposomes on skin cancer in
xeroderma pigmentosum: a randomised study.
Xeroderma Pigmentosum Study Group. Lancet 357,
131. Kraemer, K. H., DiGiovanna, J. J. & Peck, G. L.
Chemoprevention of skin cancer in xeroderma
pigmentosum. J. Dermatol. 19, 715–718 (1992).
132. Cleaver, J. E. Defective repair replication in xeroderma
pigmentosum. Nature 218, 652–656 (1968).
133. Daughdrill, G. W. et al. Chemical shift changes provide
evidence for overlapping single-stranded DNA- and
XPA-binding sites on the 70 kDa subunit of human
replication protein A. Nucleic Acids Res. 31, 4176–4183
134. Ling, H., Boudsocq, F., Woodgate. R. & Yang, W. Crystal
structure of a Y-family DNA polymerase in action. a
mechanism for error-prone and lesion-bypass replication.
Cell 107, 91–102 (2001).
135. Nance, M. A. & Berry, S. A. Cockayne syndrome:
review of 140 cases. Am. J. Med. Genet. 42, 68–84
136. Horibata, K. et al. Complete absence of Cockayne
syndrome group B gene product gives rise to
UV-sensitive syndrome but not Cockayne syndrome.
Proc. Natl Acad. Sci. USA 101, 15410–15415 (2004).
137. Colella, S., Nardo, T., Botta, E., Lehmann, A. R. &
Stefanini, M. Identical mutations in the CSB gene
associated with either Cockayne syndrome or the
DeSanctis-cacchione variant of xeroderma
pigmentosum. Hum. Mol. Genet. 9, 1171–1175
138. Lommel, L., Chen, L., Madura, K. & Sweder, K. The 26S
proteasome negatively regulates the level of overall
genomic nucleotide excision repair. Nucleic Acids Res.
28, 4839–4845 (2000).
572 | JULY 2005 | VOLUME 5
© 2005 Nature Publishing Group
139. Semple, C. A., Group., R. G. & Members., G. The
comparative proteomics of ubiquitination in mouse.
Genome Res. 13, 1389–1394 (2003).
140. Hochstrasser, M. Ubiquitin-dependent protein
degradation. Ann. Rev. Genet. 30, 403–409 (1996).
141. Wilkinson, K. D. Ubiquitination and deubiquitination:
targeting of proteins for degradation by the proteasome.
Semin. Cell Dev. Biol. 11, 141–148 (2000).
142. Chen, Z., Zhang, Y., Douglas, L. & Zhou, P. UV-damaged
DNA-binding proteins are targets of CUL-4A-mediated
ubiquitination and degradation. J. Biol. Chem. 276,
143. Schauber, C. et al. Rad23 links DNA repair to the ubiquitin/
proteasome pathway. Nature 391, 715–718 (1998).
144. Chen, L., Shinde, U., Ortolan, T. G. & Madura, K.
Ubiquitin-associated (UBA) domains in Rad23 bind
ubiquitin and promote inhibition of multi-ubiquitin chain
assembly. EMBO Rep. 2, 933–938 (2001).
145. Gillette, T. G. et al. The 19S complex of the proteasome
regulates nucleotide excision repair in yeast. Genes Dev.
15, 1528–1539 (2001).
146. Russell, S. J., Reed, S. H., Huang, W., Friedberg, E. C. &
Johnston, S. A. The 19S regulatory complex of the
proteasome functions independently of proteolysis in
nucleotide excision repair. Mol. Cell 3, 687–695 (1999).
147. van Laar, T., van der Eb, A. J. & Terleth, C. A role for
Rad23 proteins in 26S proteasome-dependent protein
degradation? Mutat. Res. 499, 53–61 (2002).
148. Hoege, C., Pfander, B., Moldovan, G. L., Pyrowolakis, G.
& Jentsch, S. RAD6-dependent DNA repair is linked to
modification of PCNA by ubiquitin and SUMO. Nature
419, 135–141 (2002).
149. Steitz, T. A. DNA polymerases: structural diversity and
common mechanisms. J. Biol. Chem. 274,
150. Ohmori, H. et al. The Y-family of DNA polymerases. Mol.
Cell 8, 7–8 (2001).
151. Trincao, J. et al. Structure of the catalytic core of
S. cerevisiae DNA polymerase η: implications for
translesion synthesis. Mol. Cell 8, 417–426 (2001).
152. McCulloch, S. D. et al. Preferential cis-syn thymine dimer
bypass by DNA polymerase η occurs with biased fidelity.
Nature 428, 97–100 (2004).
153. Johnson, R. E., Washington, M. T., Prakash, S. &
Prakash, L. Fidelity of human DNA polymerase η. J. Biol.
Chem. 275, 7447–7450 (2000).
154. Matsuda, T., Bebenek, K., Masutani, C., Hanoaka, F. &
Kunkel. T. A. Low fidelity DNA synthesis by human DNA
polymerase η. Nature 404, 1011–1013 (2000).
155. Johnson, R. E., Washington, M. T., Haracska, L.,
Prakash, S. & Prakash, L. Eukaryotic polymerases ι and
ζ act sequentially to bypass DNA lesions. Nature 406,
156. Woodgate, R. Evolution of the two–step model for
UV-mutagenesis. Mutat. Res. 485, 83–92 (2001).
157. Wang, Y. C., Maher, V. M. & McCormick, J. J. Xeroderma
pigmentosum variant cells are less likely than normal cells
to incorporate dAMP opposite photoproducts during
replication of UV-irradiated plasmids. Proc. Natl Acad.
Sci. USA 88, 7810–7814 (1991).
158. Cleaver, J. E. Sensitivity of excision repair in normal
human, xeroderma pigmentosum variant and
Cockayne’s syndrome fibroblasts to inhibition by
cytosine arabinoside. J. Cell Physiol. 108, 163–173
159. Dingman, T. & Kakunaga, T. DNA strand breaking and
rejoining in response to ultraviolet light in normal and
xeroderma pigmentosum cells. Int. J. Radiat. Biol. 30,
160. Cleaver, J. E. & Mitchell, D. L. in Cancer Medicine Vol. 1
(eds. Kufe, D. W. et al.) 303–311 (BC Becher Inc.,
Hamilton, Ontario, 2003).
This article was written with support from a grant from the
National Institutes of Environmental Health Sciences, and gener-
ous gifts from the Xeroderma Pigmentosum Society and the
Luke O’Brien Foundation.
Competing interests statement
The author declares no competing financial interests.
The following terms in this article are linked online to:
Entrez Gene: http://www.ncbi.nlm.nih.gov/entrez/query.
CEN2 | CSA | CSB | DDB1 | DDB2 | HR23B | p53 | PCNA |
RAD51 | XPA | XPB | XPC | XPD | XPF
XP-A | XP-B | XP-C | XP-D | XP-E | XP-F | XP-G
Allelic variations of the XP genes: www.xpmutations.org
Access to this interactive links box is free online.
Killing time for cancer cells
Shoshana Klein, Frank McCormick and Alexander Levitzki
Abstract | As the signalling pathways that
control cellular proliferation and death
are unravelled, a range of targets have
emerged as candidates for molecular
cancer therapy. For their survival, cancer
cells depend on a few highly activated
signalling pathways; inhibition of these
pathways has a strong apoptotic effect
and can lead to tumour regression. But
drugs that exploit this weakness, such
as imatinib, have not cured patients:
withdrawal of the drug leads to disease
recurrence, and sustained treatment leads
to the emergence of drug-resistant clones.
Can cancer be cured, or will it have to be
controlled as a chronic disease?
Cancer is a disease of miscommunication.
The uncontrolled proliferation of cancer
cells is one side of the coin. Impaired sig-
nalling for apoptosis (programmed cell
death) is the other. Paradoxically, cancer
cells are less robust than normal cells. The
traditional regimens for treating cancer,
chemotherapy and radiotherapy, take
advantage of the increased sensitivity of
cancer cells to DNA or microtubule dam-
age. In recent years, specific molecular
targets have been identified as selective
treatments for cancer cells.
In healthy cells, a myriad of interacting
signalling pathways provide redundancy,
spreading out the cost of damage to a single
pathway. Cancer cells accumulate assorted
mutations in oncogenes and tumour-sup-
pressor genes, rendering a few signalling
pathways overactive. Other mutations lead
to the elimination of redundant signalling
pathways. Cancer cells can be particularly
sensitive to inhibition of the remaining
hyperactive pathways (REF. 1 and T. Geiger
and A.L., unpublished observations; FIG. 1).
Many components of these signalling
pathways are kinases, frequently tyrosine
kinases. Specific tyrosine-kinase inhibi-
tors have been developed, which arrest the
proliferation of cancer cells2. Therefore, the
strategy that has been developed over the
past 15 years is to pinpoint the survival fac-
tors of a tumour and, by inhibiting those
factors, specifically arrest proliferation and/
or induce apoptosis of the cancer cells3.
This approach of ‘signal-transduction
therapy’4 has been validated in the clinic.
The spectacular success in the treatment
of early chronic myelogenous leukaemia
(CML) with imatinib (Glivec) was greeted
with euphoria. But is imatinib actually the
exception that proves the concept, but not
the rule? Early CML is unusual in that it
has a single survival factor — BCR–ABL
— and, in the chronic phase, the disease
is relatively homogeneous. Most tumours,
especially solid tumours, are dependent on
several pathways and are heterogeneous. It
is therefore unlikely that a single therapeutic
medium will eliminate a cancerous growth.
It will be necessary to develop appropriate
combinations of treatment. Other signal-
transduction inhibitors have also found
their way into the clinic, and many more
are in preclinical studies (FIG. 2). By assess-
ing to what degree these drugs are fulfilling
their promise, we can redirect strategies for
combating cancer. Experience has shown
that cancer can be eradicated only if it is
caught early. The new drugs might allow
us to ‘tame’ more advanced cancer into a
controlled, quiescent state.
Targets for signal-transduction therapy
Even though each cancer is expected to have
its own spectrum of signature mutations,
some aberrations in signalling appear in a
broad range of cancers. These are attrac-
tive targets for drug development, because
they should be widely applicable. We shall
highlight a few pertinent cases.
Targeting cell-division pathways to arrest
tumour growth. The canonical growth-
factor-mediated signalling pathway is
well known (FIG. 2). The pathway can be
hyperactivated by mutation at several
NATURE REVIEWS | CANCER
VOLUME 5 | JULY 2005 | 573