A R T I C L E S
Lewis et al.
Table 4. Thermodynamic Parameters for Hairpin Formationa
blue-shifted in aqueous solution, whereas the fluorescence
spectrum is red-shifted (Figure 1). These changes are similar
to those observed by Whitten and co-workers for stilbene
aggregates in Langmuir-Blodgett (LB) films and attributed to
exciton coupling in the H-aggregates.12
We assume that the stilbene diethers form laminar aggregates,
similar to those of known bolaamphiphiles possessing aromatic
cores and pendant hydroxyalkyl groups.13 The fluorescence
quantum yields of the stilbene aggregates in aqueous solution
are lower than those of the isolated molecules in methanol
solution; however, the fluorescence decay times are longer in
aqueous solution (Table 1). The increased lifetimes can be
attributed to an increased barrier for torsion about the stilbene
double bond in the stilbene diether aggregates. The stilbene
aggregates display remarkable thermal stability. Fluorescence
melting studies indicate that the transition from aggregate to
monomer occurs above 75 °C.
Hairpins with T:A Base Pairs. The thermal dissociation
profiles and CD spectra of conjugates 2-9 indicate that they
form hairpin structures with T:A base pairing between their
polyT and polyA arms. Their CD spectra (Figure 5a) show
negative and positive peaks near 250 and 283 nm, characteristic
of B-DNA.14 Even conjugate 2, which can form only two T:A
base pairs, displays significant hyperchromism with a derivative
at 33 °C (Figure 4) and a room-temperature CD spectrum
(Figure 5a) similar to those of its homologues 3-5. Increasing
the number of T:A base pairs in the series 2-5 results in an
increase in the value of Tm (Table 2) and sharpening of the CD
spectra (Figure 5a).
The Tm values of the hairpins 2-5 (Table 2) are higher than
those of the analogous stilbenediamide-linked hairpins (Chart
1) reported by Letsinger and Wu.2 Thus, the stilbene diether
linkers form the most stable synthetic polyA-linker-polyT
hairpins reported to date. To our knowledge, no DNA or RNA
oligonucleotides have been reported to form hairpin structures
with short poly(T:A) stems.15 The exceptional stability of the
Sd2-linked hairpins is, no doubt, related to the compact structure
of the linker, as observed in the crystal structure of hairpin 1 in
which the average plane-to-plane separation between the stilbene
linker chromophore and adjacent G:C base pair is 3.25 Å,
somewhat shorter than the 3.4 Å average base stacking distance
in B-form DNA.4 The π-stacking interaction can account for
the red-shifted absorption of the linker chromophore (Figure
3) and the weak negative circular dichroism observed between
300 and 350 nm (Figure 5a). In view of the crystal structure of
hairpin 1 it is somewhat surprising that the Sd2-linked hairpins
display neither a strong induced CD spectrum nor a Cotton
effect.16 One possible explanation for the absence of strong
induced circular dichroism is that the dihedral angle between
−∆G°298
kcal/mol
,
−∆H°,
kcal/mol
−∆S°,
cal/(mol‚K)
hairpin
solvent
Tm,b °C
4
0.1 M NaCl
water
54.0
50.8
37.6
26.1
48.3
48.8
53.4
2.5
1.5
1.2
0.08
1.5
2.8
2.5
23.2
18.4
29.2
22.9
21.1
37.4
28.5
69.5
56.9
93.8
76.3
65.7
116
30% EtOH-water
30% EtOH-water
0.1 M NaCl
0.1 M NaCl
0.1 M NaCl
c-4
8
13
14
87.1
a Determined from absorption data using the method of ref 19. b Data
from Table 2.
the transition moments of the stilbene and adjacent base
chromophores is near 0 or 180°, in which case the amplitude
of the Cotton effect will vanish.17
The Tm values for hairpins 4 and 6, which possess four A:T
base pairs, and 5 and 7, which possess six A:T base pairs, are
similar (Table 2). Thus, increasing the length of the alkane
chains in the stilbene diether linkers from two to three
methylenes has little effect on the thermal stability of the hairpin.
In contrast the Tm value for hairpin 8, which possesses a Sd4
linker, is lower than that of either 4 or 6; however, its CD
spectrum is similar to that of 4. The minimized structure of 8
(Figure 2d) possesses a gauche-gauche turn in one of the
tetramethylene linkers. The Tm value for hairpin 9, which
possesses a Sd4 linker, is also significantly lower than that of
either 5 or 7, which possess Sd2 or Sd3 linkers, respectively.
The Tm values of 8 and 9 are remarkably similar to those of the
corresponding Sa-linked hairpins, which have the same number
of atoms separating the two hydroxyl groups (Chart 1).2 Thus,
it appears that the geometry of the linker rather than its donor
or acceptor properties determines the thermal stability of these
synthetic hairpins. It is interesting to note that the melting
temperatures of DNA hairpins with polyT loops decrease with
the length of the loop.18
Thermodynamic parameters for formation of DNA hairpins
can be readily obtained from analysis of the absorption spectral
data using the method of Marky and Breslauer19 in cases where
the low- and high-temperature plateaus of the melting curve
can be determined. This is not possible for conjugates 2 and 3
for which the melting transition occurs over a broad temperature
range (Figure 4) or for conjugates with melting temperatures
over 60 °C for which the high-temperature plateau is not
resolved. Thermodynamic parameters for conjugates 4, c-4, and
8 are reported in Table 4. The values of Tm for 4 decrease, and
∆G becomes less favorable, as expected, as the solvent is
changed from 0.1 M NaCl to water or ethanol-water.15
Reducing the salt concentration results in a less favorable ∆H°,
which is partially compensated by a more favorable ∆S°.
Addition of ethanol makes ∆H° more favorable and ∆S° less
favorable, as previously reported by Hickey and Turner20 for
RNA duplex formation. The lower melting temperature of c-4
versus 4 in 30% ethanol-water is largely a consequence of less
favorable entropy, in accord with the distorted hairpin structure
of c-4 (Figure 2c). The lower melting temperature and less
(11) (a) Saltiel, J.; Sun, Y.-P. In Photochromism, Molecules and Systems; Duerr,
H., Boaus-Laurent, H., Eds.; Elsevier: Amsterdam, 1990; pp 64-164. (b)
Waldeck, D. H. Chem. ReV. 1991, 91, 415-436.
(12) (a) Whitten, D. G. Acc. Chem. Res. 1993, 26, 502-509. (b) Song, X.;
Geiger, C.; Vaday, S.; Perlstein, J.; Whitten, D. G. J. Photochem. Photobiol.,
A 1996, 102, 39-45.
(13) (a) Hentrich, F.; Tschierske, C.; Diele, S.; Sauere, C. J. Mater. Chem. 1994,
4, 1547-1558. (b) Ko¨lbel, M.; Beyersdorff, T.; Cheng, X. H.; Tschierske,
C.; Kain, J.; Diele, S. J. Am. Chem. Soc. 2001, 123, 6809-6818.
(14) Bloomfield, V. A.; Crothers, D. M.; Tinoco, I. Nucleic Acids. Structures,
Properties, and Functions; University Science Books: Sausalito, CA, 2000.
(15) Turner, D. H. In Nucleic Acids. Structures, Properties, and Functions;
Bloomfield, V. A., Crothers, D. M., Tinoco, I., Eds.; University Science
Books: Sausalito, CA, 2000.
(17) Harada, N.; Nakanishi, K. Circular Dichroic Spectroscopy; University
Science Books: Mill Valley, CA, 1983.
(18) Hilbers, C. W.; Haasnoot, C. A. G.; de Bruin, S. H.; Joordens, J. J. M.;
Van Der Marel, G. A.; Van Boom, H. H. Biochemie 1985, 67, 685-695.
(19) Marky, L. A.; Breslauer, K. J. Biopolymers 1987, 26, 1601-1620.
(20) Hickey, D. R.; Turner, D. H. Biochemistry 1985, 24, 2086-2094.
(16) Ardhammar, M.; Kurucsev, T.; Norde´n, B. In Circular Dichroism; Berova,
N., Nakanishi, K., Woody, R. W., Eds.; Wiley: New York, 2000.
9
12170 J. AM. CHEM. SOC. VOL. 124, NO. 41, 2002