6620 J. Am. Chem. Soc., Vol. 120, No. 26, 1998
Communications to the Editor
{5c2‚poly(dA) and 5c‚poly(dA)}. Analysis of Figure 2 also
indicates that the DNmt triplex with RNA is more stable than
the corresponding DNmt‚DNA complex [triplex mp ) 61 °C for
poly(rA) vs mp ) 32 °C for poly(dA)]. Under identical
conditions, for solutions which contained 5c and either poly(dG)
or poly(dC), no hyperchromic shift at 260 nm was observed
between ca. 5-93 °C (Figure 2).25 From these preliminary results,
DNmt appears to bind with DNA and RNA with specificity in
forming hybrid duplex and triplex structures.
5c has a significantly greater affinity for poly(dA) and poly-
(rA) than does thymidyl DNA,26 and the effect of ionic strength
on duplex stability is quite pronounced. As expected,13,23,24,27 we
find that the ionic strength has an opposite effect on the Tm values
of DNmt hybrids with DNA (Figure 2) or RNA as compared to
DNA complexes with DNA or RNA since electrostatic interac-
tions are attenuated by increasing salt concentration. DNmt
compounds are stable under the acidic and basic conditions
required for DNA synthesis14,28-31 and would be expected to be
stable in vivo due to the absence of phosphodiester linkages.
In conclusion, we show in these preliminary investigations that
the attractive forces between negatively charged DNA or RNA
and positively charged DNmt contribute significantly to the
stability of heteroduplex and -triplex structures formed between
these species. DNmt, with positively charged {-NHC(dSMe+)-
NH-} linkages, has a lessened electrostatic attraction to DNA
and RNA when compared to the positively charged DNG with
guanido {-NHC(dNH2+)NH-} linkages. DNmt duplex and
triplex structures with DNA or RNA, however, are much more
stable at physiological ionic strength than the corresponding
structures with DNA and RNA as the sole components while less
stable than the guanido-linked thymidines that form even stronger
duplex and triplex structures. Further studies, involving the use
of mismatch sequences and complementary binding studies with
mixed sequences would help us define the line between the
binding enhancement due to electrostatic attractions of the
backbones vs the specificity of the sequence due to hydrogen
bonding. The increased hydrophobicity due to the alkyl group
should considerably alter the cell diffusion and uptake of these
oligos. Combined with the fact that these oligonucleotides can
be synthesized with relative ease, have an achiral backbone
linkage, would be stable to enzymatic hydrolysis due to the lack
of a phosphodiester linkage, have increased binding due to the
oppositely charged backbone and a alkyl group, which can be
altered to control the hydrophobic/electrostatic interactions,
DNmts should serve as a promising lead in the development of
new antisense therapeutics.
Figure 1. Job plot of poly(rA) (5.3 × 10-5 M) and 5c in 0.15 mM
K2HPO4 and 0.15 M KCl, pH 7.5.
Figure 2. Plots of A260 vs T (°C) for (5c) annealed to poly(dA) (+) at
µ ) 0.05 (KCl), poly(dA) (O) at µ ) 0.25 (KCl), poly(rA) (0), poly-
(dC) ([), and poly(dG) (×) at pH 7.5 (0.001 M K2HPO4 buffer) and an
ionic strength of 0.25 (KCl). The concentration of each of the oligo-
nucleotides was 2.17 × 10-5 M in bases. The ratio of 5c to polynucleotides
was 2:1. The change in absorbance (260 nm) with increasing temperature
was monitored using a Cary 1-E spectrophotometer.
the oligomer. The thiourea-linked thymidyl oligomers 4a-c were
successfully converted to compounds 5a-c by methylation of
the thiourea linkages to methylisothiouronium salts in excess
methyl iodide followed by deprotection with acetic acid and
purified on a preparative Alltech SCX cation exchange column
employing 1.50 M ammonium acetate buffer, pH 7.0, as the
mobile phase. The purity of the sample was further confirmed
by running it on an analytical cation exchange column with 1.5
M guanidine HCl as the eluant (see the Supporting Information).
Compounds 5a, 5b, and 5c have retention times of 9.8, 14, and
18 min, respectively, consistent with the presence of three, four,
and five positive charges.
Acknowledgment. This work was supported by the National Institute
of Health and the Office of Naval Research.
To investigate the interaction of 5c with polynucleotides, we
constructed UV continuous variation plots at different ionic
strengths, wavelengths, and temperatures. The plots (Figure 1)
for interaction of 5c with poly(rA) and poly(dA) always show
maximum hypochromicity at an approximate 1:2::A:Tmt ratio
which indicates triple-helix formation in a TmtApTmt triad,
consistent with mixing curves obtained with other positively
charged oligonucleotides.13,22-24
The slow thermal melting and cooling profiles (0.2 deg/min)
of 5c bound to poly(dA) show pronounced hysteresis at pH 7.0
and at an ionic strength of 0.15 (see the Supporting Information),
further evidence of the triple-helical binding of positively charged
5c to polynucleotides.23,24 In the thermal denaturation analysis
of 5c bound to poly(dA), plots of absorbance at 260 nm (A260) vs
temperature exhibit two distinct inflections (Figure 2). On this
basis, we assign the plots in Figure 2 to represent denaturation
curves of triple- and double-helical structures of 5c with ssDNA
Supporting Information Available: Hysteresis curves of 5c and poly-
(dA) with specific conditions for melting and annealing, HPLC chro-
matogram, mass spectrum, and UV absorption spectrum of compound
5c (3 pages, print/PDF). See any current masthead page for ordering
information and Web access instructions.
JA980629Q
(25) Little hyperchromic shift was observed with polyribonucleosides when
melted with 5c under similar conditions.
(26) The duplex has a melting point of >85 °C in 10-3 M K2HPO4 with
poly(dA), whereas Tp-5 (5 mer of DNA) does not show any appreciable
binding below an ionic strength of 0.12; Tg-5 (DNG) on the other hand, has
an even higher duplex melting temperature of >95 °C.
(27) Letsinger, R. L.; Singman, C. N.; Histand, G.; Salunkhe, M. J. Am.
Chem. Soc. 1988, 110, 4470-4471.
(28) Rasmussen, C. R.; Villani, F., Jr.; Reynolds, B. E.; Plampin, J. N.;
Hood, A. R.; Hecker, L. R.; Nortey, S. O.; Hanslin, A.; Costanzo, M. J.;
Howse, R. M.; Molinari, A. J. Synthesis 1988, 460-466.
(29) Wu, T.; Ogilvie, K. K.; Pon, R. T. Nucleic Acids Res. 1989, 17 (9),
3501-3517.
(30) Dempcy, R. O.; Browne, K. A.; Bruice, T. C. Proc. Natl. Acad. Sci.
U.S.A. 1995, 92, 6097-6101.
(31) Chaix, C.; Molko, D.; Teoule, R. Tetrahedron Lett. 1989, 30 (1), 71-
74.
(23) Blasko, A.; Dempcy, R. O.; Minyat, E. E.; Bruice, T. C. Biochemistry
1997, 36, 7821-7831.
(24) Blasko, A.; Dempcy, R. O.; Minyat, E. E.; Bruice, T. C. J. Am. Chem.
Soc. 1996, 118, 7892-7899.