Carcinogenesis vol.25 no.3 pp.299±307, 2004
DOI: 10.1093/carcin/bgh013
COMMENTARY
Carcinogen-induced impairment of enzymes for replicative fidelity of DNA and the
initiation of tumours
Leon P.Bignold
Institute of Medical and Veterinary Science, Adelaide, South Australia and
Department of Pathology, University of Adelaide, SA 5005, Australia
Email: [email protected]
Not all carcinogens are mutagens, and many mutagens are
not carcinogens. Among related chemicals, small changes
of structure can markedly influence carcinogenic potency.
Many tumours are genetically unstable, but some, especially `benign' types, rarely exhibit `progression' or show
other evidence of genetic instability. Cells of particular
tumour types exhibit identifiable particular `sets' of phenotypic abnormalities (e.g. rapid growth, uniform nuclei,
little cytoplasm and occasionally production of adrenocorticotrophic hormone by anaplastic small-celled carcinoma
of the bronchus). Tumour cells pass their abnormalities on
to their daughter cells, indicating that a genomic alteration
probably underlies tumour formation. A possible mechanism, which might explain these phenomena is carcinogeninduced reduction of fidelity of replication of DNA
polymerase complexes during S phase of normal tissue stem
cells. A single `hit' by a reactive agent (chemical or physical) on one of the major enzymic sites (synthesis, proofreading, mismatch repairÐMMR) could cause multiple
sequence abnormalities in the length of DNA synthesized
by one DNA polymerase complex. Because this length
of DNA (half a replication `bubble') averages 15 000±
150 000 nucleotides, the affected DNA could include two
or more significant genomic elements (genes, especially for
tumour suppression, regulatory loci and other elements).
The particular mutant elements in the affected DNA could
then determine the `set' of phenotypic abnormalities exhibited by a resulting tumour. Non-genotoxic carcinogenicity,
non-carcinogenic mutagenicity, structure-dependent chemical carcinogenicity and the phenomenon of `sets' of
phenotypic abnormalities could thus be accommodated.
In experimental studies, the `hallmark pattern' of mutation
caused by this mechanism would be multiple mainly point
mutations clustered within the length of half a replication
`bubble'. Such a `hallmark pattern' of mutation might be
detectable in carcinogen-treated cell cultures by the use
of cycle-synchronized cultures, single cell subculturing,
restriction (endonuclease) fragment length analysis of the
clones and nucleotide sequencing of abnormal bands for
localization in the human genome. If the mechanism is
important to carcinogenesis generally, then non-carcinogenic mutagens should not cause the `hallmark pattern' of
mutations in either in vitro or in vivo systems. In human
tumour cells, the `hallmark pattern' of mutations may be
demonstrable in genetically stable human tumours, but
might well be lost or obscured by secondary mutations in
Abbreviation: RFL, restriction fragment length.
Carcinogenesis vol.25 no.3 # Oxford University Press; all rights reserved.
genetically unstable tumours. Among different cases of the
same type of human tumour, the clustered point mutations
might be tumour-type specific in their location in the genome, but vary case-to-case in the precise `points' mutated
in the cluster region. New assays for assessing the carcinogenic potential of environmental and synthetic substances
for human and animal populations may result. The hypothesis is not put forward to the exclusion of some established
mechanisms of carcinogenesis for particular human
tumours: for example, the `two-hit' mutational hypothesis
for retinoblastoma, the `multiple sequential mutational'
hypothesis for UV-induced lesions of the epidermis, and
the possibility of adduct-induced frameshift mutations by
some chemical carcinogens for experimental tumours.
Introduction
From the 1930s, when pure chemical carcinogens became
available (1), the investigation of the possibility of carcinogen-induced somatic mutation as the basis of tumours has
followed several lines.
First, it was found that some chemical carcinogens are mutagens (2,3). The methods of detection of mutations in early
studies were largely limited to chromosomal morphology
and/or limited phenotypic changes and the results were
thought unconvincing by some authors of the period (2,4,5).
Secondly, numerous attempts were made to establish particular `carcinogenic structural characteristics' among categories of chemical carcinogens, especially polycyclic
hydrocarbons, and aromatic amines (6±9). This search illuminated exquisite sensitivity of the carcinogenic potency of
some molecular species to structural changes. For example,
3,4-benzpyrene is a potent carcinogen while 1,2-benzpyrene
has only a weak carcinogenic effect (the difference between
the molecules being the position of one benzene ring) (7).
Similarly, 30 -methyl-N,N-dimethyl-4-aminoazobenzene is a
potent carcinogen, while 40 -methyl-dimethylaminoazobenzene
is a weak carcinogen (the difference between the two compounds being the position of one methyl group) (7). Specific
interactions of carcinogens with nucleotides have since been
established, for example between ultraviolet light and covalent
binding of pyrimidines (10) and between chemical carcinogens
and purines and pyrimidines (11), but no general relationships
between structure and carcinogenicity have been established
for any chemical group of carcinogens.
Thirdly, the actions of carcinogens were investigated in
relation to putative intracellular `targets'. Studies of protein
`targets' (7,9,12±14) showed that carcinogens or their active
metabolites (15) are reactive species and can bind to most
intracellular macromolecules, including cytosolic and nuclear
proteins. Ketterer (14) observed that transcription of chromatin
seemed particularly vulnerable to carcinogens, but no general
299
L.P.Bignold
mechanism based on interactions of carcinogens with proteins
emerged.
Attention turned to DNA as the `target' of carcinogens, especially after the discoveries that carcinogens can cause strand
breaks in DNA in cells (13,16,17) and that the inherited mutation of xeroderma pigmentosum is of the gene for an enzyme
associated with repair of damaged DNA (18). Farber (17) found
that most carcinogens (ethionine being one exception) induced
strand breaks in the DNA of hepatocytes. However, this
occurred whether the carcinogen caused tumour in the liver or
not (17). Farber (17) suggested that the basis of this cell typespecificity of action might be related to the observation that
strand breaks caused by agents that are carcinogenic in the liver
are only slowly repaired in that organ, while breaks caused by
non-hepatic carcinogens are rapidly repaired in this organ.
However, in the same period, it was noted in other studies
using the experimental hepatocarcinogenesis model (mice and
rats), that prior partial hepatectomy enables a single dose of
carcinogen to cause tumours in adult animals (19). Otherwise,
the carcinogen could only cause tumours when given as a
single dose to non-hepatectomized newborn animals, or when
given chronically in the diet to adult animals (19). Craddock
(19) commented that an event during cell replication might be
the critical point of action of the carcinogen.
In the 1970s, concerted study began of particular nucleotidecarcinogen binding products (`adducts') in the DNA of organs
of animals exposed to carcinogens, including alkylating
agents, polycyclic hydrocarbons, aromatic amines and mycotoxins (20±26). However, the biologic significance of adduct
formation has proved difficult to establish. Adducts are found
both in organs in which tumours form and in organs in which
tumours do not form, so that the detection of adducts in a tissue
does not necessarily indicate a specific tumourigenic risk for
that tissue (27). Single carcinogens, especially `bulky' ones
often cause a variety of adduct types (28), and the potencies of
particular types of adducts vary unpredictably by several
orders of magnitude (27). The possibility that particular types
of adducts might cause particular types of mutation (29) has
been investigated by adduct site-specific techniques, but without conclusive outcome. Loechler (28) indicates that the types
of mutation caused by adducts are influenced by experimental
parameters, including ratio of adducts to genome and the
nucleotide sequence context of the adduct. He also noted that
the types of mutation caused by adducts appear to vary with
cell type. No general relationship between carcinogenic
potency and tendency to cause particular types of mutations
has been established.
Genetic instability: not a feature of all tumours
Until the 1960s, tumours were thought to be composed of
genetically fairly uniform masses of cells which grow excessively possibly because of loss of gene(s) for regulation of cell
growth (30), and possibly having a disorder of `differentiation'
as the basis of at least some of their complex morphologic and
behavioural variability (8).
The concept that tumours are not genetically homogenous
came from work showing variability of metastatic properties
of cells cultured from single tumours in vivo (31,32), and this
phenomenon came to be termed `tumour cell heterogeneity'
(32±34). Busch in 1990 (34) noted that heterogeneity exists
with respect to virtually all features of cancers: growth,
300
invasiveness, metastasis, pigment production (in melanomas),
surface markers, chromosomal variants, markers of differentiation, secretion (of hormones and other products) and drug
resistance.
Meanwhile the work on strand breaks of DNA and Cleaver's
discovery of the basis of xeroderma pigmentosum (see above)
together with the renewed attention to chromosomal instability
in tumours (35,36) led to genetic instability being recognized
as the probable basis of both tumour cell heterogeneity and
tumour progression (32,33,37).
Subsequently, numerous mechanisms of induction of genetic
instability have been identified (38±44).
(i) Template DNA: exogenous genotoxic `adducts' (see
above); depurination, deamination and oxidation;
alkylation.
(ii) Nucleotides: imbalanced concentrations; inappropriate
analogues and enzymes normally degrading these.
(iii) Error-prone DNA polymerases (including proofreading
and MMR) and accessory proteins.
(iv) Presence of DNA polymerases which synthesize across
abnormal templates (`translesional synthesis') so that the
usual arrest of synthesis does not occur (44).
(v) Strand misalignments.
(vi) Abnormalities of DNA repair enzymes.
(vii) Abnormalities of cell cycle checkpoints (45).
(viii) Mechanisms affecting chromosomal integrity and
ploidy (46,47).
(ix) Epigenetic mechanisms, including DNA methylation
(48).
(x) Telomerase dysfunction (49).
(xi) Other mechanisms.
The importance of this list is that it indicates that a large
number of genes are required for preservation of the genome,
and correspondingly, that random mutational events are likely
to affect one of these genes (and hence cause genetic instability) more often than they will affect a single gene for a specific
phenotypic feature.
Loeb (50±53) was among the first to suggest that genetic
instability might be a necessary aspect of tumour formation,
and has proposed that normal cells acquire a `mutator phenotype' early in the neoplastic process. This is supported by
recent discoveries that tumour cells can contain up to 105
genomic events (54) (such numbers can only arise by enhanced
mutation rates among cells) and that genetic instability may
occur in tissue cells adjacent to tumours (55,56). Genetic
instability may also provide possible explanations of general
features of tumours, especially their cell-to-cell and focus-tofocus variability of cytologic and architectural morphology,
and the occasionally poor correlations of degrees of abnormal
morphologies with aggressive clinical behaviour (57).
Notwithstanding the foregoing, genetic instability may not be
an essential aspect of all tumours and not an essential initiating
step. Although genetic instability has been documented in many
malignant tumours (51±53,55,56), many human tumours
(mainly `benign' ones, for example, lipomas, neurilemomas,
seborrhoeic keratoses) show few abnormalities other than excess
growth, and do not show `progression' to suggest an unstable
genome. Some `low grade' malignant tumours, such as basal cell
carcinoma of epidermis and intestinal carcinoid tumour do not
undergo `progression'. Specific studies of large numbers of
`benign' and `low grade malignant' tumours for genetic
Fidelity of replication of DNA and carcinogenesis
instability have not been reported, and little evidence of the
phenomenon has been found in the limited studies of basal cell
carcinoma (58) and intestinal carcinoid (59) so far published.
Furthermore, experimental evidence has been published,
which weighs against genetic instability as the basis of all
tumour types. Many `immortal' or transformed cell lines
remain stable in culture for long periods of time and several
studies have shown that tumourigenic human cell lines are no
more susceptible to mutation than their non-tumourigenic
counterparts (60±63). In one report, extracts of transformed
cells showed no difference in replicative fidelity of DNA
compared with non-transformed cells (64).
Carcinogen-impaired DNA polymerase complexes as a
mechanism of the initiation of tumour formation
One mechanism of carcinogenesis, which can potentially
explain a variety of tumour phenomena is that the carcinogen
binds to enzymic sites of the DNA polymerase complex during
S phase of the cell cycle, and causes multiple mutations in
DNA. This idea was mentioned by Speyer (65) in 1965 and
discussed in some detail by Nelson and Mason (66) in 1972.
Loeb and co-workers in 1974 (49) discussed the hypothesis in
relation to early evidence of error-prone DNA polymerases in
tumour cells. The present author (67) drew attention to possible explanations provided by the hypothesis of phenomena
such as time delays in tumour development after exposure to
carcinogens, non-chemical promoters of carcinogenesis and
hormonal initiators and promoters of carcinogenesis.
The attraction of this notion is that if a single `hit' of
carcinogen on an enzymic site for DNA genomic fidelity
(base selection, proofreading or MMR) during S phase, then
multiple mutations could be inflicted on a length of nucleotide
sequence equal to a half replication `bubble' (15 000±150 000 nt)
(68) (Figure 1). There would be a significant chance of
`effective' (rather than `silent') mutations of the genomic elements in the affected length of DNA without any damage to the
remainder of the genome. The resulting phenotypic abnormalities of the cell would depend on the genomic composition of
the affected half replication `bubble', and two or more such
abnormalities would appear in a cell at the same time if two
genomic elements (genes and gene regulators) were present in
the affected length of DNA.
All of the relevant enzymic sites of the DNA polymerase
complex (for base selection, proofreading and mismatch
repair) are extremely dependent on precise structural determinants, perhaps corresponding to the precise structural characteristics required of carcinogens (see above). In particular, to
be a carcinogen, a molecule must derange the fidelity mechanism without arresting synthesis of DNA.
Furthermore, the affected DNA need not contain a gene
necessary for fidelity of replication of DNA for tumour to
form. Thus, if the affected half replication `bubble' included
only a tumour suppressor gene, the daughter cell would be only
hyperproliferative, and remain genetically stable. This then
could account for genetically stable tumours and cell lines
deriving from carcinogen-exposed cells in vivo and in vitro.
It can also be noted that a gene cluster containing genetic
stability genes, but not a growth suppressor could lead to
Fig. 1. A non-genotoxic mechanism by which a carcinogen could cause one or more mutations in one cell at the same time. If the half replication `bubble'
includes only a tumour-suppressor gene, only excessive growth might result. If the half replication `bubble' includes a gene supporting replication fidelity,
the daughter cell could be genetically unstable. More complex tumours could arise if other appropriate genes are present in the half replication `bubble'. N.b. The
carcinogen in the diagram is placed in the site of synthesis by the DNA polymerase, implying base selection is impaired. However, sites of proofreading or
mismatch repair could serve as alternative sites (see text).
301
L.P.Bignold
genetically unstable cells with little accumulation of cell numbers. This combination is that of `in-situ' tumours, which are
recognized in many tissues, including epidermis, bladder
epithelium and other tissues (69).
Experimental evidence for direct actions of carcinogens on
DNA polymerases
Much of the experimental basis for this notion derives from
studies of the loss of fidelity of replication of DNA induced
by carcinogenic metal ions such as those of cadmium, manganese and cobalt in cell-free systems. For general references see
(70±73) and for specific reviews see (74±81).
Sirover and Loeb in 1977 (82) showed that infidelity of
replication of DNA by DNA polymerase from avian myeloblastosis virus in the presence of Mn2 ‡ or Co2 ‡ is not likely to
be due to interactions of the ions either with the nucleotides
being added to the new strand of DNA during DNA synthesis
or with the template strand of DNA. Loeb, Sirover and
co-workers (83) in the same year suggested that induction of
infidelity of replication might be a useful test of carcinogenicity. Snow and co-workers (84,85) considered that chromium
ions and nickel ions might cause mutations and hence carcinogenesis by increasing the DNA polymerase processivity and
the rate of polymerase bypass of DNA lesions.
Pelletier and co-workers (86) used X-ray crystallography to
study the effects of Mg2 ‡ , Ca2 ‡ , Mn3 ‡ , Co2 ‡ , Cr3 ‡ and Ni2 ‡
on the active sites of human DNA polymerase beta. These
authors found that one way Mn2 ‡ may manifest its mutagenic
effect on polymerases is by promoting greater reactivity than
Mg2 ‡ at the catalytic site, thereby allowing the nucleotidyl
transfer reaction to take place with little or no regard to
instructions from a template. In addition, all metal ions tested,
with the exception of Mg2 ‡ , promoted a change in the sidechain position of aspartic acid 192, which is one of three highly
conserved active-site carboxylate residues.
Further evidence of non-genotoxic carcinogenesis has come
from a recent study of the mutagenicity of cadmium by Jin and
co-workers (87) who found that the mutagenic effects of cadmium ions in yeast cultures derive from the ability of the
cation to interfere with the fidelity of replication of DNA in
target cells. Jin et al. (87) found that large numbers of illegitimate nucleotide additions go uncorrected but the synthesis of
DNA is not blocked. They suggest that the precise site of the
impairment of the fidelity of replication is the MMR mechanism, because the basis that the pattern of uncorrected mismatches which they observed in Cd2 ‡ -exposed cells is
similar to the pattern in MMR-deficient strains of yeast.
Extracts of human cell lines showed direct inhibition of correction of heteroduplexes containing one base-loop (a function
of MMR enzymes) by Cd2 ‡ .
McMurray and Tainer in a commentary (88) mention many
mechanisms by which Cd2 ‡ can interfere with the functions of
proteins. These authors note that Cd2 ‡ may replace Zn2 ‡ in
tissues, and that perhaps the Cd2 ‡ ion disrupts a currently
unknown `zinc finger' structure of MMR sites.
In whole animal studies, Chan and Becker (89) reported that
DNA polymerase alpha isolated from rat livers after feeding
with N-2-fluorenylacetamide is significantly error-prone in
replication of synthetic polynucleotide templates compared
with the same enzyme from untreated controls. These authors
apparently did not obtain a similar result when a different
carcinogen (N-2-aminoacetamide) was used (90).
302
In contrast to the above, essentially negative results were
obtained in studies of the effects of carcinogens on cells in
culture described by Brucker, Loeb and Thielmann (91). These
authors applied carcinogens [Me(NO)(NO2 )Gdn and
MeNOUr] to living cells, and then extracted the DNA polymerase alpha±primase complexes from the cells and tested the
fidelity of replication of bacteriophage DNA (as template) in
cell-free systems. They found that the fidelity of DNA polymerase alpha±primase complexes was similar in carcinogentreated and -untreated cells. However, increased synthesis of
DNA occurred in carcinogen-treated cells compared with
untreated cells, which the authors thought might be due to a
carcinogen-induced alteration of an accessory protein of the
complex. The fidelity of replication by the DNA polymerase
was assessed in a cell-free system, however, so that a functional defect might have been present but undetected owing to
the assay conditions.
Discussion: comparison with other models
`Tumours' comprise a large number of lesions of the body,
which vary markedly in the degrees and natures of their morphological and behavioural abnormalities. For some particular
tumour types, there may be particular mechanisms. For example, for retinoblastoma, for which no role of any environmental
carcinogen is suspected, there is little reason to doubt
Knudson's `two (mutational) hit' hypothesis (92±94, and see
below). The `first hit' is inherited in most affected individuals,
and spontaneous mutations, being relatively common in
rapidly dividing tissues, can provide the `second hit'. As
another example, some chemical carcinogens may act principally by adducts (see Introduction), which induce frameshift
mutations. These mutations, by affecting lengths of DNA
between the site of an adduct and the site of termination of
DNA synthesis, could have a mutational impact similar to that
of impairment of DNA fidelity-of-replication sites. Yet again,
time-sequentially accumulated `point' mutations (`sequential
multi-hit' model, see below) may well be the mechanism by
which ultraviolet light (which occurs in humans as large numbers of exposures over decades) induces solar keratoses and
squamous celled carcinomas (both of which are often preceded
by an in situ phaseÐi.e. Bowen's disease).
However, in the consideration of claims that any particular
theory may have universal relevance, the following comments
may be relevant.
(i) Theories involving the direct mutagenic activity (genotoxicity) of carcinogens do not easily explain non-mutagenic
carcinogens, non-carcinogenic mutagens or the exquisite
structure-dependency of potency of chemical carcinogens.
Furthermore:
(a) Simple one-mutation hypotheses do not explain the
variability (heterogeneity) of tumours, because all
daughter cells after a mutation should be alike (5).
(b) The `two-hit' hypothesis of Knudson (93±95), and the
`sequential multi-hit' models [put forward on epidemiologic grounds by Nordling in 1953 (96), modified
by Armitage (97) and more recently espoused by
Vogelstein and Kinzler (98,99)] do not easily explain
tumours which are inducible experimentally by single
doses of chemical carcinogen [see Craddock (19) and
above], or the multiplicity of phenotypic abnormalities
in `sets' exhibited by many human tumours.
Fidelity of replication of DNA and carcinogenesis
(ii) Genetic instability, whether induced by direct mutation
of a genome-preserving gene, or by a non-genotoxic
mechanism, appears to provide a basis for morphologic
and behavioural heterogeneity of tumours (32±34,57).
However, the phenomenon may not serve as an initiation
event associated with all tumours, because many tumours
(especially benign ones) do not exhibit the features of
genetic instability. Furthermore, without a concurrent
mutation resulting in increased growth, genetic instability
is likely to result only in `in situ' cytologically abnormal
cells, rather than a hyperproliferative mass of cells.
(iii) Non-genotoxic mechanisms of carcinogenesis other than
direct action of carcinogen on DNA polymerases do not
readily offer explanation `sets' of phenotypic abnormalities (including occasional additional features) of
tumours. Specifically:
(a) Notions of abnormal `differentiation' (8,100±102)
(meaning either abnormal local specialization, or
abnormal local development), do not offer any explanation of phenotypic features which are unrelated to
differentiation (such as adrenocorticotrophic hormone
production by anaplastic small-celled carcinoma or
bronchus). Further, the hypotheses are vague, because
no particular biochemical target of carcinogens is
suggested.
(b) Recent suggestions concerning the roles of particular
proteins, including ligand-activated transcription factors (103) and peroxisome proliferators (104) and `gap
junctions' (105) as well as abnormalities of transcription (106) or translation (107) do not provide for the
immortal genetic abnormalities of tumours, and frequent genetic instability. Hypotheses that carcinogens
affect methylation of DNA and hence gene expression
(108±112) do not easily account for the permanently
inheritable nature of the change in the affected cell
without some early mutation occurring. A hypothesis,
which suggests that a mutation of gene(s) for an epigenetic mechanism as the first event of tumour formation, is not fundamentally an `epigenetic hypothesis'
but a mutational hypothesis, and subject to the difficulties mentioned above.
(c) The notion that carcinogens may act by generation of
endogenous genotoxic substances (113±116) is difficult to evaluate because the nature of all of the possible
ultimate endogenous genotoxins is unclear. No endogenous genotoxin or its precursor suggested so far is
characterized by exquisite structure-dependence of
their carcinogenic activation or actions.
Further investigation of carcinogen-induced impairment
of DNA fidelity of replication
For the purposes of investigation, the hypothesis that carcinogens may act on the fidelity of replication mechanisms of DNA
polymerase complexes suggests that a `hallmark pattern' of
mutation might result in the genome of daughter cells. In
experimental studies, the features of this pattern would be:
multiple mainly point mutations clustered within the length
of half a replication `bubble'. In human tumour cells (if the
pattern could be identified), the clusters might be tumour typespecific in their location in the human genome, but varying
case-to-case in the precise `points' mutated in the cluster
region. This pattern is distinct both from frameshift mutations,
in which the nucleotide sequence is preserved, but displaced
1 or 2 nt places in the DNA chain, and from random point
mutations throughout the genome.
At the present time, there is very little strong direct,
published experimental evidence to support the notion of
carcinogen-binding to DNA polymerases of local cells (see
above).
While various types of experiments have been reported,
without conclusive results, it is possible that the methodology
used has been insufficiently sensitive to detect the genomic
changes, which might have occurred. Especially since the
1970s, there have been numerous improvements of biochemical and molecular biologic techniques, which could now be
employed in re-visiting these basic experiments. These are
now mentioned according to their experimental type: cellfree, cell culture, whole animal and human lesional studies.
Cell-free studies
Some details of technique in the published studies involving
cell-free systems might possibly be improved. First the purification of the DNA polymerases may have involved use of
denaturing precipitants and dissociating agents [see Kornberg
and Tate (117)] (details of these preparative methods are not
available in many papers). Methods for producing functionally
superior (less denatured) DNA polymerase preparations are
now available (118).
Secondly, the conditions of incubation may have been suboptimal. For example, the addition of a protein such as serum
albumin can limit the non-specific damage of toxins to sensitive chemical structures.
Thirdly, the method of detecting illicit nucleotide additions
by these methods involved measuring the inclusion of illicit
labelled nucleotide in new chains synthesized on artificial
polynucleotide templates [for example C or G additions to a
chain synthesized on poly A-T (82,83)]. Currrently, short
`natural' template DNA could be used, and the product could
be nucleotide-sequenced by current standard automated methods and compared with the original template DNA.
It should be noted that cell-free systems lack the indirect
factors, which may affect replicative fidelity of DNA (see
section on genetic instability), so that their results cannot be
taken as conclusive evidence that the same effects of the
substances occur in vivo.
Cell cultures (Figure 2)
The major technical requirements of studies of the effects of
carcinogens on fidelity of replication of DNA in cell cultures
are seen to be:
(i) To accommodate possible cell type-specificity of effects
of some carcinogens and provide a stable genome with
which abnormal restriciton fragment length (RLF) bands
can be compared (see below).
(ii) To limit the cell toxicity side effect of the carcinogen.
(iii) To maximize the number of DNA polymerase complexes
relative to other proteins in the culture.
(iv) To detect the putative cluster of point mutations over a
range of 15 000±150 000 nt in the whole genome of the
daughter cells.
(v) If genotoxic substances are also to be tested, to distinguish
genomic changes produced by DNA-adducts from mutations caused by impaired DNA polymerase complexes.
303
L.P.Bignold
Fig. 2. Suggested scheme of investigation of hypothesis using cell cultures. Carcinogen is added to a synchronized cell culture during S phase. Single cell
subculturing and restriction (endonuclease) fragment length RFL analysis provides some subcultures with abnormal bands. Sequencing of the regions of the bands
from the various digests would show that they are paired. For each sub-culture the bands of each pair would be found to come from a region of high incidence
of mainly point mutations (the `hallmark pattern' of mutations) induced by a single impaired DNA polymerase complexÐsee Figure 1 and text. Different
subcultures would have different sites of the abnormal region, because there is no specificity of action of carcinogens on DNA polymerases according to which part
of the genome the polymerase synthesizes. Excess bands in the DNA of a clone would indicate a genetically unstable clone [see text and (67)]. In corresponding
studies of genetically stable human tumour tissue, sequencing of abnormal bands in comparison with the individual patient's normal cells may show tumour
type-specific location in the human genome, but varying case-to-case in the precise `points' mutated in the cluster region.
First, the genetically stable cell cultures used could be
as closely related as possible to the desired `target' cell
type. Primary explants of human and animal cells may be
useful.
Secondly, efforts to reduce cell toxicity should be made by
investigating the effects of cell-protective additives to the
culture system. Even cultures in connective tissue matrices,
such as of collagen could be considered.
Thirdly, in the published reports (see above), randomly
multiplying cultures were used. However, methods for synchronizing cell cultures have been available for many years
304
(119), and have been used occasionally in studies of experimental carcinogenesis [e.g. Meuller and co-workers (120)].
Synchronizing the target cell culture would have two methodological benefits. Because the carcinogen could be added at
the beginning and removed at the end of S phase, the exposure
of the cells to carcinogen for non-specific damage and cell
toxicity, would be limited. Also, because the proportion of the
target molecule (the DNA polymerase) relative to irrelevant
molecules to which the carcinogen might bind would be
increased, lower concentrations of carcinogen could be used,
and hence non-specfic cell toxicity reduced.
Fidelity of replication of DNA and carcinogenesis
Fourthly, to detect patterns of mutations, after the completion of the S phase, the culture could be single-cell subcultured, and these subcultures could be analysed for mutations
(when sufficiently multiplied) by restriction fragment length
(RFL) analysis in comparison with the genome of the original
cells. It is assumed the clustered multiple mutations would
cause change of a site of action of at least one restriction
enzyme. Batteries of such enzymes may be needed to demonstrate the mutant length of DNA. Once detected, abnormal
bands could be sequenced and then compared with the
human genome data for localization. Different clones should
have clusters of mutations at different loci in the genome. The
results of nucleotide sequencing should allow distinction both
from the `hallmark pattern' of mutation both from frameshift
mutations, and from genome-wide random mutations. The
dosage of carcinogen would be adjusted so that excessive
number of `hits' per cell do not make the results of subsequent
sequencing difficult to interpret.
The bacteria-based assay methods of Kunkel and co-workers
(118,121) might not be easily applicable to this type of
experiment.
Some subcultures are expected to show genetic instability
[due to the original impaired fidelity of replication by DNA
polymerase affecting genes for genetic instability in the daughter cell (67)] and such clones could be expected to have large
numbers of mutations throughout their genome, and correspondingly large numbers of abnormal bands of their DNA
after RFL.
Fifth, to limit numbers of adduct-derived mutations, nongenotoxic carcinogens (such as ethionine) could be tested,
because these would not cause possibly confusing concurrent
adduct-based mutations. For genotoxic carcinogens, the dose
should be kept as low as possible, to minimize these adductbased mutations.
Controls would consist of cell cultures treated with noncarcinogenic mutagens. If the direct impairment of DNA
mechanism is important, these agents will be found not to
cause the `hallmark pattern' of mutations.
Whole animal studies
For these studies, the general plan would be similar to that
described for cell culture experiments. In animal tumours, it
may be important to use an early carcinogen-induced neoplastic lesion, before genetic instability has created spurious additional mutations. An obvious animal experimental lesion is the
pre-malignant hepatic nodule of carcinogen-treated rats (17).
These lesions could be harvested, preferably at the earliest
possible time after administration of carcinogen, grown in
single cell sub-culture if insufficient in amount, and then
assessed by RFL analysis, sequencing and comparison with
the animal's genome as described above.
Human lesions
The same general plan may be usefully applied to human
tumours. Investigation of genomic abnormalities of `benign'
and other tumours, which do not show behavioural evidence of
genetic instability may be more appropriate than such studies
of genetically unstable tumours, because the former may more
clearly demonstrate the mutational `hall mark pattern' (multiple mainly point mutations clustered within the length of half a
replication `bubble', being tumour type specific in their location in the human genome but varying case-to-case in the
precise `points' mutated in the cluster region). It is possible
that once a tumour-type locus is identified, genomic changes in
various cases of tumours of this type could be analysed by PCR
using genome site-specific primers upstream of the affected
locus. The resultant product could be compared to the product
from the normal cells from the individual patient in whom the
tumour arose.
Conclusions
The somatic mutation theory of carcinogenesis has had a
chequered history even since pure carcinogens have become
available. The major difficulties have been the lack of explanations of non-genotoxic carcinogenicity, non-carcinogenic
mutagenicity and structure-dependence of carcinogenic
potency. Furthermore, there has been little suggested of mutational mechanisms to explain the occurrence of `sets' of phenotypic features (e.g. hyperproliferative and/or cytologically
abnormalities), which characterize the various tumour types.
The present genetic hypothesis offers a possible explanatory
basis for many of these observations, in particular, non-genotoxic
induction of tumours (which adduct-based mechanisms do not
easily provide) and the phenomenon of `sets' of phenotypic
abnormalities occurring in one cell at one time, which appears
to be the manner in which most tumour types arise.
Nevertheless, the hypothesis is not put forward to the exclusion of other established mechanisms of carcinogenesis, which
are believed to have particular roles in particular tumour
types. These mechanisms may also contribute (as does genetic
instability) to the general case-to-case and other variabilities
within tumour types.
The studies outlined above may lead to the development of
additional methods for the testing of possible carcinogenic
effects of non-genotoxic as well as possibly genotoxic environmental and synthetic substances for their carcinogenic risk
to human and animal populations.
Finally, more detailed study of the human genome for proximities of genes for tumorous phenomena, which occur in
common combination may well be valuable.
References
1. Kennaway,E.L. and Hieger,I. (1930) Carcinogenic substances and their
fluorescence spectra. Br. Med. J., i, 1044±1046.
2. Burdette,W.J. (1955) The significance of mutation in relation to the origin
of tumours: a review. Cancer Res., 15, 201±226.
3. Koller,P.C. (1957) The genetic component of cancer. In Raven R.W. (ed.)
Cancer, vol. 1. Butterworth & Co., London, pp. 335±403.
4. Berenblum,I. and Shubik,P. (1949) An experimental study of the initiating
stage of carcinogenesis and a re-examination of the somatic cell mutation
theory of cancer. Br. J. Cancer, 3, 109±118.
5. Willis,R.A. (1948) Pathology of Tumours. Butterworths, London.
6. Badger,G.M. (1954) Chemical constitution and carcinogenic activity. Adv.
Cancer Res., 2, 73±127.
7. S
uss,R., Kizel,V. and Scriber,J.D. (1973) Cancer Experiments and
Concepts. Springer-Verlag, New York.
8. Berenblum,I. (1958) The study of tumours in animals. In Florey,H.W.
(ed.) General Pathology, 2nd Edn. Lloyd-Luke, London, pp. 513±549.
9. Haddow,A. (1973) On the biological alkylating agents. Perspect. Biol.
Med., 16, 503±524.
10. Davies,M.J. (2003) Singlet oxygen-mediated damage to proteins and its
consequences. Biochem. Biophys. Res. Commun., 305, 761±770.
11. Dipple,A. (1995) DNA adducts of chemical carcinogens. Carcinogenesis,
16, 437±441.
12. Alexander,P. (1954) The reactions of carcinogens with macromolecules.
Adv. Cancer Res., 2, 1±72.
305
L.P.Bignold
13. Heidelberger,C. (1970) Chemical carcinogenesis, chemotherapy: cancer's
continuing core challengesÐG.H.A. Clowes Memorial Lecture. Cancer
Res., 30, 1549±1569.
14. Ketterer,B. (1980) Interactions between carcinogens and proteins. Br.
Med. Bull., 36, 71±78.
15. Miller,J.A. (1970) Carcinogenesis by chemicals: an overviewÐG.H.A.
Clowes memorial lecture. Cancer Res., 30, 559±576.
16. Legator,M.S. and Flamm,W.G. (1973) Environmental mutagenesis and
repair. Annu. Rev. Biochem., 42, 683±708.
17. Farber,E. (1973) CarcinogenesisÐcellular evolution as a unifying thread.
Presidential address. Cancer Res., 33, 2537±2550.
18. Cleaver,J.E. (1968) Defective repair replication of DNA in xeroderma
pigmentosum. Nature, 218, 652±656.
19. Craddock,V.M. (1971) Liver carcinomas induced in rats by single
administration of dimethylnitrosamine after partial hepatectomy. J. Natl
Cancer Inst. USA, 47, 899±907.
20. Curtiss,C.C. (1991) Chemical and physical carcinogenesis: advances and
perspectives for the 1990s. Cancer Res., 51 (suppl.), 5023s±5044s.
21. Kemminki,K., Dipple,A., Shuker,D.E.G., Kadlubar,F.F., Segerback,D. and
Bartsch,H. (eds) (1994) DNA adducts: identification and biological
significance. IARC Scientific Publications No. 125, IARC, Lyon.
22. Kaufmann,W.K. (1989) Pathways of human cell post-replication repair.
Carcinogenesis, 10, 1±11.
23. Lawley,P.D. (1994) From fluorescence spectra to mutational spectra, a
historical overview of DNA-reactive compounds. IARC Scientific
Publications No. 125, IARC, Lyon, pp. 3±22.
24. Seo,K.Y., Jelinsky,S.A. and Loechler,E.L. (2000) Factors that influence
the mutagenic patterns of DNA adducts from chemical carcinogens.
Mutat. Res., 463, 215±246.
25. Pages,V. and Fuchs,R.P. (2003) Uncoupling of leading- and lagging-strand
DNA replication during lesion bypass in vivo. Science, 300, 1300±1303.
26. Pages,V. and Fuchs,R.P. (2002) How DNA lesions are turned into
mutations within cells? Oncogene, 21, 8957±8966.
27. Nestmann,E.R., Bryant,D.W. and Carr,C.J. (1996) Toxicological significance of DNA adducts: summary of discussions with an expert panel.
Regul. Toxicol. Pharmacol., 24, 9±18.
28. Loechler,E.L. (1996) The role of adduct site-specific mutagenesis in
understanding how carcinogen-DNA adducts cause mutations: perspective, prospects and problems. Carcinogenesis, 17, 895±902.
29. Singer,B. and Eissigman,J.M. (1991) Site-specific mutagenesis: retrospective and prospective. Carcinogenesis, 12, 949±955.
30. Lockhart-Mummery,J.P. (1934) The Origin of Cancer. J. & A. Churchill,
London.
31. Greene,H.S. (1957) Heterotransplantation of tumors. Ann. N. Y. Acad. Sci.,
69, 818±829.
32. Fidler,I.J. (1978) Tumor heterogeneity and the biology of cancer invasion
and metastasis. Cancer Res., 38, 2651±2660.
33. Heppner,G.H. and Miller,F.R. (1998) The cellular basis of tumor
progression. Int. Rev. Cytol., 177, 1±56.
34. Busch,H. (1990) The final common pathway of cancer. Cancer Res., 50,
4830±4838.
35. Sandberg,A.A. and Hossfeld D.K. (1970) Chromosomal abnormalities in
human neoplasia. Annu. Rev. Med., 21, 379±408.
36. Ohno,S. (1971) Genetic implication of karyological instability of
malignant somatic cells. Physiol. Rev., 51, 496±526.
37. Nowell,P.C. (1976) The clonal evolution of tumor cell populations.
Science, 194, 23±28.
38. Loeb,L.A. and Kunkel,T.A. (1982) Fidelity of DNA synthesis. Annu. Rev.
Biochem., 51, 429±457.
39. Cheng,K.C. and Loeb,L.A. (1993) Genomic instability and tumor
progression: mechanistic considerations. Adv. Cancer Res., 60, 121±156.
40. Coleman,W.B. and Tsongalis,G.J. (1995) Multiple mechanisms account
for genomic instability and molecular mutation in neoplastic transformation. Clin. Chem., 41, 644±657.
41. Minnick,D.T. and Kunkel,T.A. (1996) DNA synthesis, errors, mutators
and cancer. Cancer Surv., 28, 3±20.
42. Cheng,K.C. and Loeb,L.A. (1997) Genomic stability and instability: a
working paradigm. Curr. Top. Microbiol. Immunol., 221, 5±18.
43. Pihan,G.A. and Doxsey,S.J. (1999) The mitotic machinery as a source of
genetic instability in cancer. Semin. Cancer Biol., 9, 289±302.
44. Kunkel,T.A. and Bebenek,K. (2000) DNA replication fidelity. Annu. Rev.
Biochem., 69, 497±529.
45. Cahill,D.P., Lengauer,C., Yu,J. et al. (1998) Mutations of mitotic
checkpoint genes in human cancers. Nature, 392, 300±303.
46. Miyagawa,K. (1998) Genetic instability and cancer. Int. J. Haematol.,
67, 3±14.
306
47. Holloway,S.L., Poruthu,J. and Scata,K. (1999) Chromosome segregation
and cancer. Exp. Cell Res., 253, 308±314.
48. Jones,P.A. and Laird,P.W. (1999) Cancer epigenetics comes of age.
Nature Genet., 21, 163±167.
49. Hahn,W.C. (2003) Role of telomeres and telomerase in the pathogenesis of
human cancer. J. Clin. Oncol., 21, 2034±2043.
50. Loeb,L.A., Springgate,C.F. and Battula,N. (1974) Errors in DNA
replication as a basis of malignant changes. Cancer Res., 34, 2311±2321.
51. Loeb,L.A. (2001) A mutator phenotype in cancer. Cancer Res., 61,
3230±3239.
52. Loeb,K.R. and Loeb,L.A. (2000) Significance of multiple mutations in
cancer. Carcinogenesis, 21, 379±385.
53. Jackson,A.L. and Loeb,L.A. (2001) The contribution of endogenous
sources of DNA damage to the multiple mutations in cancer. Mutat. Res.,
477, 7±21.
54. Stoler,D.L., Chen,N., Basik,M., Kahlenberg,M.S., Rodriguez-Bigas,M.A.,
Petrelli,N.J. and Anderson,G.R. (1999) The onset and extent of genomic
instability in sporadic colorectal tumor progression. Proc. Natl Acad. Sci.
USA, 96, 15121±15126.
55. Larson,P.S., de las Morenas,A., Bennett,S.R., Cupples,L.A. and
Rosenberg,C.L. (2002). Loss of heterozygosity or allele imbalance in
histologically normal breast epithelium is distinct from loss of heterozygosity or allele imbalance in co-existing carcinomas. Am. J. Pathol.,
161, 283±290.
56. Hittelman,W.N. (2001) Genetic instability in epithelial tissues at risk of
cancer. Ann. N. Y. Acad. Sci., 952, 1±12.
57. Bignold,L.P. (2002) The mutator phenotype/clonal selection hypothesis
can provide a basis for the complex and variable morphological and
behavioural features of solid tumour malignancies. Cell. Mol. Life Sci., 59,
950±958.
58. Stefanini,M., Lagomarsini,P., Berardesca,E., Borroni,G., Rabbiosi,G. and
Nuzzo,F. (1988) Normal sensitivity to mutagens, spontaneous chromosome breakage and mutation frequency in nevoid basal cell carcinoma
syndrome. Arch. Dermatol. Res., 280 (suppl.), S19±23.
59. D'adda,T., Pizzi,S., Azzoni,C., Bottarelli,L., Crafa,P., Pasquali,C.,
Davoli,C., Corleto,V.D., Fave,G.D. and Bordi,C. (2002) Different patterns
of 11q allelic losses in digestive endocrine tumors. Hum. Pathol., 33,
322±329.
60. Wittenkeller,J.L., Storer,B., Bittner,G. and Schiller,J.H. (1997)
Comparison of spontaneous and induced mutation rates in an immortalized
human bronchial epithelial cell line and its tumorigenic derivative.
Oncology, 54, 335±341.
61. Sager,R. (1988) Mutation rates and mutational spectra in tumorigenic cell
lines. Cancer Surv., 7, 325±333.
62. Kendal,W.S. and Frost,P. (1986) Metastatic potential and spontaneous
mutation rates: studies with two murine cell lines and their recently
induced metastatic variants. Cancer Res., 46, 6131±6135.
63. Yamashina,K. and Heppner,G.H. (1985) Correlation of frequency of
induced mutation and metastatic potential in tumor cell lines from a single
mouse mammary tumor. Cancer Res., 45, 4015±4019.
64. Boyer,J.C., Thomas,D.C., Maher,V.M., McCormick,J.J. and Kunkel,T.A.
(1993) Fidelity of DNA replication by extracts of normal and malignantly
transformed human cells. Cancer Res., 53, 3270±3275.
65. Speyer,J.F. (1965) Mutagenic DNA polymerase. Biochem. Biophys. Res.
Commun., 21, 6±8.
66. Nelson,R.L. and Mason,H.S. (1972) An explicit hypothesis for chemical
carcinogenesis. J. Theor. Biol., 37, 197±200.
67. Bignold,L.P. (2003) Initiation of genetic instability and tumour formation:
a review and hypothesis of a nongenotoxic mechanism. Cell. Mol. Life
Sci., 60, 1107±1117.
68. Alberts,B., Bray,D., Lewis,J., Raff,M., Roberts,K. and Walter,P. (eds)
(2002) Molecular Biology of the Cell, 4th Edn. Garland Publishing Inc.,
New York.
69. Carter,R.L. (ed) (1984) Precancerous States. Oxford University Press,
Oxford.
70. IARC (1990) Monographs on the Evaluation of Carcinogenic Risks to
Humans. Chromium, Nickel and Welding, Vol. 49. IARC Scientific
Publications, IARC, Lyon.
71. IARC (1992) Cadmium in the Human Environment: Toxicity and
Carcinogenicity, Vol. 118. IARC Scientific Publications, IARC, Lyon.
72. IARC (1990) Monographs on the Evaluation of Carcinogenic Risks to
Humans. Beryllium, Cadmium, Mercury and Exposures in the Glass Manufacturing Industry, Vol. 58. IARC Scientific Publications, IARC, Lyon.
73. Costa,M., Zoroddu,M.A., Suk,W.A., Thompson,C. and Rossman,T. (eds)
(2002) Molecular mechanisms of metal toxicity and carcinogenicity.
Environ. Health Perspect. Supplements, 110 (suppl. 5).
Fidelity of replication of DNA and carcinogenesis
74. Jennette,K.W. (1981) The role of metals in carcinogenesis: biochemistry
and metabolism. Environ. Health Perspect., 40, 233±252.
75. Sunderman,F.W. (1984) Recent advances in metal carcinogenesis. Ann.
Clin. Lab. Sci., 14, 93±122.
76. Sky-Peck,H.H. (1986) Trace metals and neoplasia. Clin. Physiol.
Biochem., 4, 99±111.
77. Coogan,T.P., Latta,D.M., Snow,E.T. and Costa,M. (1989) Toxicity and
carcinogenicity of nickel compounds. Crit. Rev. Toxicol., 19, 341±384.
[Erratum published in: Crit. Rev. Toxicol., 20, 135].
78. Snow,E.T. (1992) Metal carcinogenesis: mechanistic implications.
Pharmacol. Ther., 53, 31±65.
79. Klein,C.B., Snow,E.T. and Frenkel,K. (1998) Molecular mechanisms in
metal carcinogenesis: role of oxidative stress. In Aruoma,O.I. and
Halliwell,B. (eds) Molecular Biology of Free Radicals in Human
Diseases. OICA International, London, pp. 79±137.
80. Hartwig,A. (1998) Carcinogenicity of metal compounds: possible role of
DNA repair inhibition. Toxicol. Lett., 102±103, 235±239.
81. Hartwig,A. (2001) Zinc finger proteins as potential targets for toxic metal
ions: differential effects on structure and function. Antioxid. Redox.
Signal., 3, 625±634.
82. Sirover,M.A. and Loeb,L.A. (1977) On the fidelity of DNA replication.
Effect of metal activators during synthesis with avian myeloblastosis virus
DNA polymerase. J. Biol. Chem., 252, 3605±3610.
83. Loeb,L.A., Sirover,M.A., Weymouth,L.A., Dube,D.K., Seal,G.,
Agarwal,S.S. and Katz,E. (1977) Infidelity of DNA synthesis as related to
mutagenesis and carcinogenesis. J. Toxicol. Environ. Health., 2, 1297±1304.
84. Snow,E.T. (1994) Effects of chromium on DNA replication in vitro.
Environ. Health Perspect. Suppl., 102 (suppl. 3), 41±4.
85. Chin,Y.E., Snow,E.T. and Christie,N.T. (1994) The stimulatory effect of
nickel chloride on DNA replication in human HeLa cells and Escherichia
coli. Carcinogenesis, 15, 1013±1016.
86. Pelletier,H., Sawaya,M.R., Wolfle,W., Wilson,S.H. and Kraut,J.A. (1996)
Structural basis for metal ion mutagenicity and nucleotide selectivity in
human DNA polymerase beta. Biochemistry, 35, 12762±12777.
87. Jin,Y.H., Clark,A.B., Slebos,R.J., Al-Refai,H., Taylor,J.A., Kunkel,T.A.,
Resnick,M.A. and Gordenin,D.A. (2003) Cadmium is a mutagen that acts
by inhibiting mismatch repair. Nature Genet., 34, 326±329.
88. McMurray,C.T. and Tainer,J.A. (2003) Cancer, cadmium and genome
integrity. Nature Genet., 34, 239±241.
89. Chan,Y.J. and Becker,F.F. (1979) Decreased fidelity of DNA polymerase
activity during N-2-fluorenylacetamide hepatocarcinogenesis. Proc. Natl
Acad. Sci. USA, 76, 814±818.
90. Becker,F.F. (1981) Presidential address. Recent concepts of initiation and
promotion in carcinogenesis. Am. J. Pathol., 105, 3±9.
91. Brucker,A., Loeb,L.A. and Thielmann,H.W. (1990) DNA polymerase
alpha-primase complexes from carcinogen-treated Chinese hamster ovary
cells. Cancer Res., 50, 6894±6901.
92. Knudson,A.G. (1996) Hereditary cancer: two hits revisited. J. Cancer Res.
Clin. Oncol., 122, 135±140.
93. Knudson,A.G. (2000) Chasing the cancer demon. Annu. Rev. Genet.,
34, 1±19.
94. Knudson,A.G. (2001) Two genetic hits (more or less) to cancer. Nature
Rev. Cancer, 1, 157±162.
95. Knudson,A.G. (2002) Cancer genetics. Am. J. Med. Genet., 111, 96±102.
96. Nordling,C.O. (1953) A new theory of the cancer-inducing mechanism.
Br. J. Cancer, 7, 68±72.
97. Armitage,P. (1985) Multistage models of carcinogenesis. Environ. Health
Perspect., 63, 195±201.
98. Vogelstein,B. and Kinzler,K.W. (1993) The multistep nature of cancer.
Trends Genet., 9, 138±141.
99. Kinzler,K.W. and Vogelstein,B. (1996) Lessons from hereditary colorectal cancer. Cell, 87, 159±170.
100. Fiala,S. (1968) The cancer cell as a stem cell unable to differentiate.
Neoplasma, 15, 607±622.
101. Pierce,G.B. (1977) Relationship between differentiation and carcinogenesis. J. Toxicol. Environ. Health, 2, 1335±1342.
102. Sell,S. (1993) Cellular origin of cancer: dedifferentiation or stem cell
maturation arrest? Environ. Health Perspect., 101 (suppl. 5), 15±26.
103. Green,S. (1992) Nuclear receptors and chemical carcinogenesis. Trends
Pharmacol. Sci., 13, 251±255.
104. Yeldandi,A.V., Rao,M.S. and Reddy,J.K. (2000) Hydrogen peroxide
generation in peroxisome proliferator-induced oncogenesis. Mutat. Res.,
448, 159±177.
105. Trosko,J.E. and Chang,C.C. (2001) Mechanism of up-regulated gap
junctional intercellular communication during chemoprevention and
chemotherapy of cancer. Mutat Res., 480±481, 219±229.
106. Cox,P.M. and Goding,C.R. (1991) Transcription and cancer. Br. J.
Cancer, 63, 651±662.
107. Clemens,M.J. and Bommer,U.A. (1999) Translational control: the cancer
connection. Int. J. Biochem. Cell Biol., 31, 1±23.
108. Counts,J.L. and Goodman,J.I. (1994) Hypomethylation of DNA: an
epigenetic mechanism involved in tumor promotion. Mol. Carcinogen.,
11, 185±188.
109. Tycko,B. (2000) Epigenetic gene silencing in cancer. J. Clin. Invest.,
105, 401±407.
110. Jones,P.A. and Laird,P.W. (1999) Cancer epigenetics comes of age.
Nature Genet. 21, 163±167.
111. Baylin,S.B. and Herman,J.G. (2000) DNA hypermethylation in tumorigenesis: epigenetics joins genetics. Trends Genet., 16, 168±174.
112. Ehrlich,M. (2002) DNA methylation in cancer: too much, but also too
little. Oncogene, 21, 5400±5413.
113. Lutz,W.K. (1990) Endogenous genotoxic agents and processes as a basis
of spontaneous carcinogenesis. Mutat. Res., 238, 287±295.
114. Pratt,I.S. and Barron,T. (2003) Regulatory recognition of indirect
genotoxicity mechanisms in the European Union. Toxicol. Lett.,
140±141, 53±62.
115. Lau,S.S., Yoon,H.S, Patel,S.K., Everitt,J.I., Walker,C.L. and Monks,T.J.
(2001) Mutagenicity and carcinogenicity of biological reactive intermediates derived from a `non-genotoxic' carcinogen. Adv. Exp. Med.
Biol., 500, 83±92.
116. Davidson,J.F., Guo,H.H. and Loeb,L.A. (2002) Endogenous mutagenesis
and cancer. Mutat. Res., 509, 17±21.
117. Kornberg,A. and Tate,T.A. (1992) DNA Replication, 2nd Edn. W. H. Freeman
and Co., New York, pp. 197±198.
118. Bebenek,K. and Kunkel,T.A. (1995) Analyzing fidelity of DNA
polymerases. Methods Enzymol., 262, 217±232.
119. Nais,A.H. and Fox,M. (1971) Synchronization of mammalian cells with
respect to the mitotic cycle. Cell Tissue Kinet., 4, 375±398.
120. Mueller,G.C., Kensler,T.W. and Kajiwara,K. (1978) Mechanisms of
DNA and chromatin replication: possible targets for cocarcinogenesis.
In Slaga,T.G., Sivak,A. and Boutwell,R.K. Carcinogenesis, Vol 2,
Mechanisms of Tumour Promotion. Raven Press, New York,
pp. 79±90.
121. Boyer,J.C., Bebenek,K. and Kunkel,T.A. (1996) Analyzing the fidelity
of reverse transcription and transcription. Methods Enzymol., 275,
523±537.
Received August 20, 2003; revised October 19, 2003;
accepted October 28, 2003
307