Carcinogens are generally mutagens, which is understandable given that tumo
r cells grow uncontrollably because they have mutations in critical genes i
nvolved in growth control. Carcinogens often induce a complex pattern of mu
tations (e.g., GC-->TA, GC-->AT, etc.). These mutations are thought to be i
nitiated when a DNA polymerase encounters a carcinogen-DNA adduct during re
plication. In principle, mutational complexity could be due to either a col
lection of different adducts each inducing a single kind of mutation (Hypot
hesis la), or a single adduct inducing different kinds of mutations (Hypoth
esis Ib). Examples of each are discussed. Regarding Hypothesis Ib, structur
al factors (e.g., DNA sequence context) and biological factors (e.g., diffe
ring DNA polymerases) that can affect the pattern of adduct mutagenesis are
discussed. This raises the question: how do structural and biological fact
ors influence the pattern of adduct mutagenesis. For structural factors, th
ree possibilities are considered: (Hypothesis 2a) a single conformation of
an adduct giving rise to multiple mutations -- dNTP insertion by DNA polyme
rase being influenced by (e.g.) the surrounding DNA sequence context; (Hypo
thesis 2b) a variation on this ("dislocation mutagenesis"); or (Hypothesis
2c) a single adduct adopting multiple conformations, each capable of giving
a different pattern of mutations. Hypotheses 2a, 2b and 2c can each in pri
nciple rationalize many mutational results, including how the pattern of ad
duct mutagenesis might be influenced by factors, such as DNA sequence conte
xt. Five lines of evidence are discussed suggesting that Hypothesis 2c can
be correct for base substitution mutagenesis. For example, previous work fr
om our laboratory was interpreted to indicate that [+ta]-B[a]P-N-2-dG in a
5'-CGG sequence context (G115) could be trapped in a conformation giving pr
edominantly G-->T mutations, but hearing caused the adduct to equilibrate t
o its thermodynamic mixture of conformations, leading to a decrease in the
fraction of G-->T mutations. New work is described suggesting that [+ta]-B[
a]P-N-2-dG at G115 can also be trapped predominantly in the G-->A mutationa
l conformation, from which equilibration can also occur, leading to an incr
ease in the fraction of G-->T mutations. Evidence is also presented that th
e fraction of G-->T mutations is higher when [+ta]-B[a]P-N-2-dG at G115 is
in ss-DNA (similar to 89%) vs. ds-DNA (similar to 66%), a finding that can
be rationalized if the mixture of adduct conformations is different in ss-
and ds-DNA. In summary, the factors affecting adduct mutagenesis are review
ed and five lines of evidence that support one hypothesis (2c: adduct confo
rmational complexity can cause adduct mutational complexity) are discussed.
(C) 2000 Elsevier Science B.V. All rights reserved.