C. Meier et al.
693
TABLE 1 T
m
Values of the T
7
(Arylamine-dG)T
7
Á dA
7
XdA
7
Hybrid
7 7
value (°C) against dA XdA
T
m
Oligonucleotide
X = C
X = G
X = T
X = A
TTTTTTT(dG)TTTTTTT
TTTTTTT(Tol-dG) TTTTTTT
TTTTTTT(ABP-dG) TTTTTTT
39.4
34.9
32.8
28.6
27.3
25.7
30.1
28.4
26.6
27.7
26.2
25.4
nucleotide was present in the opposite strand (T = 34.9°C and 32.8°C for the
m
toluidine and the aminobiphenyl-adduct, respectively). In the case of the toluidine
adduct, T values dropped further when dG or dA was placed opposite to the
m
adduct. Such an effect was also observed in the case of the reference oligo hybrid.
However, in the case of the aminobiphenyl adduct, the T was independent to the
m
nucleotide opposite to the adduct (Table 1).
Next, a mixed sequence 15-mer-oligonucleotide was used. Here, different
arylamine-adducts were incorporated in the middle while in the opposite strand
always dC was placed. The unmodified hybrid showed a T value of 51.6°C. T
m
m
values of the modified 15-mers were found to be 46.3°C (anilinyl-), 46.4°C
toluidinyl-), 45.0 (aminobiphenyl-), and 46.3°C (anisidinyl-adduct). Again, one
(
modification only was responsible for a considerable loss of thermal stability of the
duplex. Interestingly, the effect of the strong carcinogen aminobiphenyl was only
slightly stronger as compared to the borderline carcinogens aniline, toluidine,
and anisidine.
Further structural effects of the incorporation of the adducts into the
oligonucleotides were studied using CD-spectroscopy. However, no difference in
the CD-spectra was observed for the adduct-bearing oligonucleotides compared to
the unmodified case. All curves show shapes found previously for B-type DNA
duplexes. No difference between the aminobiphenyl adduct and the monocyclic
aromatic amines can be observed. The same CD properties have been found in
the case of the T (arylamine-dG)T Á dA XdA hybrids. Again, all showed typical
7
7
7
7
shapes for B-type DNA and no difference to the reference hybrid.
Finally, first experiments for primer-template extension studies have been
performed. As polymerase, the human enzyme pol b has been used. The adduct
has been incorporated into a 34-mer template strand. The primer ends directly
before the adduct. Thus, standing start conditions were used. Then, all four NTPs
were added separately. Only dCTP was incorporated opposite to the arylamine
adducts in all cases. However, a futher experiment was performed in which the
dC opposite to the arylamine adduct is already incorporated in the primer.
Thus, experiments under standing start + 1 conditions where done (Figure 2).
Surprisingly, in these experiments DNA pol b showed a considerable loss in fidelity!
Thus, errors seem to take place not at the opposite site of the adduct but one
step downstream. This effect was most apparent for the aminobiphenyl adduct but
was also observed in lower extent for the monocyclic aryl amines. This is an