7-deaza-7-nitro dGTP, and 5-hydroxy dUTP) that can be
incorporated into DNA via the polymerase chain reaction
(PCR).6 One of these, 5-hydroxy-dU, was less efficient than
the others as a substrate for Taq polymerase, resulting in
difficulty with amplification of certain sequences. It was
therefore desirable to replace 5-hydroxy-dU with a new “T”
analogue having improved performance in PCR. To this end,
we designed and synthesized a new thymidine (T) analogue,
5-amino-2′-deoxyuridine (5-amino-dU). Like 5-hydroxy-dU,
5-amino-dU is roughly isosteric with T. However, whereas
the C4 carbonyl group of T is available for hydrogen bonding
to a protein, the C4 carbonyl of the analogue is tied up in an
intramolecular hydrogen bonding interaction with the vicinal
5-amino group. Furthermore, the polar 5-amino group of the
analogue would be less able to engage in hydrophobic
contacts than the 5-methyl group of T. For these reasons,
we suspected the analogue would interfere with the estab-
lishment of T-specific protein contacts. Finally, we thought
that the electron-rich 5,6-enamine moiety of 5-amino-dU
might be susceptible to attack by permanganate, a prereq-
uisite for DNA strand cleavage.
analysis (data not shown). Encouraged by the results of these
tests, we converted 5-amino-dU to the corresponding 5′-
triphosphate, using the standard procedure8 (Scheme 1). The
analogue proved to be a relatively poor substrate for chemical
phosphorylation, but we could easily purify the desired
product by anion-exchange chromatography, so we pro-
ceeded without any further optimization of the synthesis.
To assess the biochemical characteristics of 5-amino-dU,
we used the analogue triphosphate in Taq-based PCR
amplification of a 209 base pair segment of a plasmid
carrying the E. coli ada promoter sequence. Unlike 5-hy-
droxy-dUTP, 5-amino-dUTP supported robust DNA poly-
merization even under conditions of uniform incorporation,
when every T in the sequence is replaced by the analogue
(data not shown). To determine the ratio of 5-amino-dUTP/
dTTP that affords statistical incorporation of a single
5-amino-dU moiety per DNA strand, we radiolabeled one
of the two PCR primers and carried out a series of PCR
reactions using different ratios. The PCR products were then
treated with 100 µM potassium permanganate followed by
10% aqueous piperidine (a detailed procedure is provided
in the Supporting Information). Following evaporation and
resuspension in loading dye, the fragments were analyzed
5-Amino-dU was prepared by treatment of 5-bromo-dU
with liquid ammonia7 (Scheme 1). We found that the
Scheme 1. Synthesis of 5-Amino-dUTPa
a Reagents and conditions: (a) liquid NH3, 55-60 °C, 44 h
(65%); (b) POCl3, -10 °C, 3 h; (c) tributylammonium pyrophos-
phate, 0 °C, 2 min; (d) 1.5 M triethylammonium bicarbonate, pH
7.7, rt, 1 h (overall isolated yield 10% for steps b-d).
chromophore of 5-amino-dU was completely destroyed by
brief exposure to aqueous potassium permanganate and
piperidine (data not shown), suggesting the analogue would
be selectively cleavable in DNA. Importantly, the nucleoside
analogue was found to withstand 30 cycles of thermal cycling
in Taq polymerase reaction buffer, as judged by HPLC
Figure 1. A comparison of TDI footprints generated by 5-amino-
dU with those generated by 5-hydroxy-dU. (A) Phosphorimage
renderings of TDI footprints of both analogues made by the Ada
protein on the ada promoter. The sequence is typed next to the
image, and the T residues showing interference (T12 and T13) are
in red. The lane denoted by “•” represents the pattern of the cleaved
starting DNA pool, and the lane labeled B corresponds to cleaved
protein-bound DNA. (B) Densitometric analysis: blue, cleaved
starting DNA pool; red, cleaved protein-bound DNA.
(5) A scheme depicting the overall procedure for TDI footprinting,
accompanied by a detailed experimental procedure, is available in the
Supporting Information.
(6) Storek, M. J.; Ernst, A.; Verdine, G. L. Nat. Biotechnol. 2002, 20,
183-186.
(7) Barawkar, D. A.; Ganesh, K. N. Bioorg. Med. Chem. Lett. 1993, 3,
347-352.
3868
Org. Lett., Vol. 4, No. 22, 2002