A series of nonpolar thymidine analogues of increasing size: DNA base pairing and stacking properties

Article abstract of DOI:10.1021/jo048061t

(Chemical Equation Presented) We describe the properties in DNA of a set of five nonpolar nucleoside mimics in which shape is similar but size is increased gradually. The compounds vary in the size of their exocyclic substituents, which range from hydroge

Full text of DOI:10.1021/jo048061t

A Series of Nonpolar Thymidine Analogues of Increasing Size:
DNA Base Pairing and Stacking Properties
Tae Woo Kim and Eric T. Kool*
Department of Chemistry, Stanford University, Stanford California 94305-5080
kool@stanford.edu
Received November 1, 2004
We describe the properties in DNA of a set of five nonpolar nucleoside mimics in which shape is
similar but size is increased gradually. The compounds vary in the size of their exocyclic substituents,
which range from hydrogen to iodine, and are designed to test the steric effects of nucleosides,
nucleotides, and DNA in biological systems in a systematic way. We describe the conversion of
toluene, 2,4-difluorotoluene, 2,4-dichlorotoluene, 2,4-dibromotoluene, and 2,4-diiodotoluene deoxy-
ribosides into suitably protected phosphoramidite derivatives and their incorporation into synthetic
DNAs. Studies of their behavior in the context of hexamer and dodecamer duplexes were carried
out, with comparison to natural thymine. Thermal melting data with compounds in 5′ dangling
positions showed that all five compounds stack more strongly than thymine, and all the dihalo-
substituted cases stack more strongly than the unsubstituted toluene case. Stacking correlated
with surface area and hydrophobicity, both of which increase across the series. In base-pairing
studies, all five compounds showed destabilized pairing opposite natural bases (relative to thymine-
adenine pairing), as expected. Notably, pairing among the nonpolar base analogues was considerably
more stable, and some of the pairs involving the largest analogues showed stability equal to that
of a natural thymine-adenine pair. The results establish the base pairing properties of a potentially
useful new series of biochemical probes for DNA-protein interactions and also identify a set of
new, stable hydrophobic base pairs for designed genetic pairing systems.
Introduction
all Watson-Crick hydrogen-bonding groups in nucleoside
analogues, preparing several “nonpolar nucleoside isos-
teres”, which maintain the steric size and shapes of
natural nucleobases but lack polar functionality.3,4 Ex-
amples included 4-methylazabenzimidazole, an adenine
mimic,5 and 2,4-difluorotoluene, a thymine mimic.6
Nucleoside analogues that lack specific hydrogen-
bonding groups have proven useful in a number of
biological contexts for probing the physical and chemical
importance of such electrostatically charged moieties. For
example, Strazewski and Tamm reported over two de-
cades ago the synthesis of pyrimidine analogues lacking
one of three hydrogen bonding groups and investigated
their substrate abilities with DNA polymerase enzymes.1
In another example, McLaughlin reported nucleobases
with single functional groups deleted and described their
properties in pairing in DNA.2 Taking this approach to
its logical limit, we described the strategy of removing
Nonpolar nucleoside isosteres have proven useful in
probing the recognition of DNA by other nucleic acids3
(2) McLaughlin, L. W.; Wilson, M.; Ha, S. B. Compr. Nat. Prod.
Chem. 1999, 7, 251.
(3) (a) Schweitzer, B. A.; Kool, E. T. J. Org. Chem. 1994, 59, 7238.
(b) Schweitzer, B. A.; Kool, E. T. J. Am. Chem. Soc. 1995, 117, 1863.
(4) Kool, E. T.; Morales, J. C.; Guckian, K. M. Angew. Chem., Int.
Ed. 2000, 39, 990.
(5) Morales J. C.; Kool, E. T. J. Am. Chem. Soc. 1999, 121, 2323.
(6) Guckian, K. M.; Kool, E. T. Angew. Chem., Int. Ed. 1997, 36,
2825.
(1) Strazewski, P.; Tamm, C. Angew. Chem., Int. Ed. Engl. 1990,
29, 36.
10.1021/jo048061t CCC: $30.25 © 2005 American Chemical Society
Published on Web 02/16/2005
2048
J. Org. Chem. 2005, 70, 2048-2053

Thymidine Analogues of Increasing Size in DNA
and in studies of the physical origins of DNA curvature.7
Biophysical studies have shown that thymine and ade-
nine isosteres destabilize DNAs in which they are
substituted, unless they are in a terminal position, in
which case they can be strongly stabilizing, due to their
avid stacking with natural DNA bases.4,8 Structural
studies have shown that, despite the destabilization when
present at nonterminal locations, thymine and adenine
mimics show essentially the same structures as the
natural congeners. For example, difluorotoluene was
shown to occupy a Watson-Crick-like position opposite
adenine in a 12mer DNA duplex,9 even though several
lines of evidence showed that the fluoroaromatic com-
pound does not measurably form hydrogen bonds in
aqueous solution.4
The results of replication with isosteres have led to the
hypothesis that, for many high-fidelity enzymes, steric
exclusion effects may be dominant in the selective
replication of matched base pairs at the exclusion of
mismatched ones.20 In such a model, high fidelity re-
quires a tight active site to reject sterically mismatched
nucleotides and to enforce correct overall Watson-Crick-
like geometry in the incipient pair.21 Interestingly, recent
studies of sterically altered deoxyribose have added
support to this idea at the level of the DNA backbone.22
Since steric effects are widely believed to be crucial to
biological selectivity in enzymatic systems, it would be
useful to have chemical tools to probe such effects in a
systematic way. With this in mind, we conceived a new
series of thymine analogues; 2,4-diHydrogentoluene (H),
diFluorotoluene (F), 2,4-dichLorotoluene (L), 2,4-diBro-
motoluene (B), and 2,4-diIodotoluene (I) in which the size
is varied systematically by replacing the oxygen nucleo-
base substituents with hydrogen, fluorine, chlorine,
bromine, and iodine (Figure 1). Since the oxygens are the
main protruding groups of thymine on its Watson-Crick
edge, this replacement has the effect of maintaining an
approximate shape of T while gradually increasing size
by about one Angstrom across the series. An early report
described the synthesis of the five deoxyribosides in this
series, and established their sugar ring conformational
preferences, which are virtually the same as those of
thymidine.23 For applications in biological recognition
studies, it is important to evaluate the behavior of these
molecular probes in DNA, in the absence of enzymes.
Here we describe the derivatization of these compounds
for incorporation into oligonucleotides, the characteriza-
tion of DNA strands containing them, and the evaluation
of their pairing and stacking properties in the double
helix.
Nonpolar nucleoside mimics have also been increas-
ingly useful of late in the study of protein-DNA and
enzyme-DNA recognition. Studies have been reported
with purine and pyrimidine mimics in a number of DNA
repair enzymes, including MutY,10 fpg,11 MutS, and
homologues,12 and in polypurine tract recognition by HIV
reverse transcriptase.13 Those studies have shed light on
the relative importance of hydrogen bonding and steric
interactions to these enzymes’ biochemical activities. In
addition to this, nonpolar nucleoside isosteres have
proven broadly useful in the study of DNA replication
by a wide variety of polymerase enzymes. Such nonpolar
analogues were first reported in 1997 to act as surpris-
ingly strong substrates for DNA polymerase I,14,15 leading
to the conclusion that at least some replicative DNA
polymerases function well in the synthesis of a base pair
without Watson-Crick hydrogen bonds. This has since
been confirmed by a number of studies of varied poly-
14-16
merase enzymes in vitro5,
and recently in living
bacterial cells as well.17 The discovery of the lack of a
hydrogen-bonding requirement in replication has led to
the design of other nonisosteric DNA base pairs for
expansion of the genetic information-encoding system.18,19
Experimental Section
Synthesis of Modified Nucleoside Phosphoram-
idites. The C-glycoside series (dH, dL, dB, dI) was
prepared as described previously,23 while one compound
(dF) is now commercially available as the phosphor-
amidite derivative. 5′-Tritylation of the other four com-
(7) (a) Maki, A. S.; Kim, T. W.; Kool, E. T. Biochemistry 2004, 43,
1102. (b) Maki, A. S.; Brownewell, F. F.; Liu, D.; Kool, E. T. Nucleic
Acids Res. 2003, 31, 1059.
(8) (a) Guckian, K. M.; Schweitzer, B. A.; Ren R. X.-F.; Sheils C. J.;
Tahmassebi, D. C.; Kool, E. T. J. Am. Chem. Soc. 2000, 122, 2213. (b)
Guckian, K. M.; Schweitzer, B. A.; Ren, R. X.-F.; Sheils, C. J.; Paris,
P. L.; Tahmassebi, D. C.; Kool, E. T. J. Am. Chem. Soc. 1996, 118,
8182.
(9) Guckian, K. M.; Krugh, T. R.; Kool, E. T. Nat. Struct. Biol. 1998,
5, 954.
(10) (a) Chepanoske, C. L.; Langelier, C. R.; Chmiel, N. H.; David,
S. S. Org. Lett. 2000, 2, 1341. (b) Chmiel, N. H.; Golinelli, M.-P.;
Francis, A. W.; David, S. S. Nucleic Acids Res. 2001, 29, 553.
(11) Francis, A. W.; Helquist, S. A.; Kool, E. T.; David, S. S. J. Am.
Chem. Soc. 2003, 125, 16235.
(12) (a) Schofield, M. J.; Brownewell, F. E.; Nayak, S.; Du, C.; Kool,
E. T.; Hsieh, P. J. Biol. Chem. 2001, 276, 45505. (b) Drotschmann, K.;
Yang, W.; Brownewell, F. E.; Kool, E. T.; Kunkel T. A. J. Biol. Chem.
2001, 276, 46225.
(13) Rausch, J. W.; Qu, J.; Yi-Brunozzi, H. Y.; Kool, E. T.; Le Grice
S. F. J. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 11279.
(14) Moran, S.; Ren, R. X.-F.; Rumney, S.; Kool, E. T. J. Am. Chem.
Soc. 1997, 119, 2056.
(15) Moran, S.; Ren, R. X.-F.; Kool, E. T. Proc. Natl. Acad. Sci. U.S.A.
1997, 94, 10506.
(16) (a) Matray, T. J.; Kool, E. T. J. Am. Chem. Soc. 1998, 120, 6191.
(b) Morales, J. C.; Kool, E. T. J. Am. Chem. Soc. 2000, 122, 1001. (c)
Dzantiev, L.; Alekseyev, Y. O.; Morales, J. C.; Kool, E. T.; Romano, L.
J. Biochemistry 2001, 40, 3215. (d) Morales, J. C.; Kool, E. T.
Biochemistry 2000, 39, 12979.
(17) Delaney, J. C.; Henderson, P. T.; Helquist, S. A.; Morales, J.
C.; Essigmann, J. M.; Kool, E. T. Proc. Natl. Acad. Sci. U.S.A. 2003,
100, 4469.
(18) (a) McMinn, D. L.; Ogawa, A. K.; Wu, Y.; Liu, J.; Schultz, P.
G.; Romesberg, F. E. J. Am. Chem. Soc. 1999, 121, 11585. (b) Ogawa,
A. K.; Wu, Y.; McMinn, D. L.; Liu, J.; Schultz, P. G.; Romesberg, F. E.
J. Am. Chem. Soc. 2000, 122, 3274. (c) Wu, Y.; Ogawa, A. K.; Berger,
M.; McMinn, D. L.; Schultz, P. G.; Romesberg, F. E. J. Am. Chem. Soc.
2000, 122, 7621. (d) Tae, E. L.; Wu, Y.; Xia, G.; Schultz, P. G.;
Romesberg, F. E. J. Am. Chem. Soc. 2001, 123, 7439. (e) Yu, C.; Henry,
A. A.; Romesberg, F. E.; Schultz, P. G. Angew. Chem., Int. Ed. 2002,
41, 3841. (f) Henry, A. A.; Yu, C.; Romesberg, F. E. J. Am. Chem. Soc.
2003, 125, 9638. (g) Henry, A. A.; Olsen, A. G.; Matsuda, S.; Yu, C.;
Geierstanger, B. H.; Romesberg, F. E. J. Am. Chem. Soc. 2004, 126,
6923.
(19) (a) Piccirilli, J. A.; Krauch, T.; Moroney, S. E.; Benner S. A.
Nature 1990, 343, 6233. (b) Lutz, M. J.; Held, H. A.; Hottiger, M.;
Hubscher, U.; Benner, S. A. Nucleic Acids Res. 1996, 24, 1308. (c)
Hirao, I.; Harada, Y.; Kimoto, M.; Mitsui, T.; Fujiwara, T.; Yokoyama,
S. J. Am. Chem. Soc. 2004, 126, 13298. (d) Mitsui, T.; Kitamura, A.;
Kimoto, M.; To, T.; Sato, A.; Hirao, I.; Yokoyama, S. J. Am. Chem.
Soc. 2003, 125, 5298.
(20) (a) Kool, E. T. Annu. Rev. Biophys. Biomol. Struct. 2001, 30, 1.
(b) Kool, E. T. Curr. Opin. Chem. Biol. 2000, 4, 602.
(21) Kool, E. T. Annu. Rev. Biochem. 2002, 71, 191.
(22) (a) Strerath, M.; Cramer, J.; Restle, T.; Marx, A. J. Am. Chem.
Soc. 2002, 124, 11230. (b) Summerer, D.; Marx, A. J. Am. Chem. Soc.
2002, 124, 910. (c) Summerer, D.; Marx, A. Angew. Chem., Int. Ed.
2001, 40, 3693.
(23) Kim, T. W.; Kool, E. T. Org. Lett. 2004, 6, 3949.
J. Org. Chem, Vol. 70, No. 6, 2005 2049

Kim and Kool
FIGURE 1. Structures of the thymidine nucleoside analogues (top), designed to have gradually increasing steric demand, compared
with thymidine at right. Below them are space-filling models of the base analogues with methyl groups at the point of attachment
to deoxyribose, with PM3-calculated electrostatic potentials (Spartan ′02, Wavefunction, Inc.) mapped onto the van der Waals
surfaces (electrostatic scale: -50 to +30).
SCHEME 1a
purities were checked by analytical reversed-phase HPLC
(Supporting Information, Figure S3).
The oligomers were quantitated by absorbance at 260
nm. Molar extinction coefficients were calculated by the
nearest neighbor method.25 Values for oligonucleotides
containing nonnatural residues were obtained in the
following way: The molar extinction coefficients for each
of the new nucleosides were measured at 260 nm in
methanol because of their low water solubility. The molar
extinction coefficients for dH, dF, dL, dB, and dI were
found to be 250, 1000, 250, 500, and 3900 M-1 cm-1
,
respectively. The individual extinction coefficients for all
the bases in a given oligomer were summed. Since in the
most cases the content of nonnatural residues in the
sequences is low, this estimation method is unlikely to
generate large errors in concentration.
a Conditions: (a) 4,4′-dimethoxytrityl chloride, DIPEA, pyridine;
(b) 2-cyanoethyl N,N-diisopropylchlorophosphoramidite, DIPEA,
CH2Cl2.
Thermal Denaturation Studies. After the samples
were prepared in melting buffer (1 M NaCl, 10 mM
phosphate buffer (pH 7.0) with 0.1 mM EDTA), they were
heated to 95 °C and allowed to slowly cool at a rate of 1
°C/min to 5 °C. The melting studies were carried out in
Teflon-stoppered 1 cm path length quartz cells on a
Varian Cary 1 UV-vis spectophotometer equipped with
thermoprogrammer. Absorbance was monitored at 280
nm for stacking and at 260 nm for base pairing. In all
cases the complexes displayed sharp, apparently two-
state transitions. The data were analyzed by the melt
curve processing program, MeltWin v. 3.0. Melting tem-
peratures (Tm) were determined by computer-fit of the
first derivative of absorbance with respect to 1/T. Un-
certainty in Tm is estimated at (0.5 °C based on repeti-
tions of experiments. Free-energy values were derived
by two methods: (1) The denaturation data was computer-
fitted with an algorithm employing linear sloping base-
lines, using the two-state approximation for melting. Fits
were excellent, with ø-2 values of 10-6 or better. (2) Van’t
Hoff thermodynamic parameters were derived from
linear plots of 1/Tm vs ln[CT] by measuring Tm as a
function of concentration. Close agreement was seen with
pounds was carried out in pyridine in the presence of
N,N-diisopropylethylamine, giving the corresponding 5′-
ethers (1a-4a) in yields of 45-51%. The 3′-phosphor-
amidite species were prepared using standard methods,
yielding the desired phosphorylated species (1b-4b) in
50-74% (Scheme 1).
Oligodeoxynucleotide Synthesis. Phosphoramidite
derivatives we prepared as described above. The phos-
phoramidite derivative of analogue dF was purchased
from Glen Research. DNA oligonucleotides were synthe-
sized on an Applied Biosystems 394 synthesizer using
standard â-cyanoethylphosphoramidite chemistry. Self-
complementary oligomers with 5′-dangling ends (X) were
synthesized in DMT-off mode and purified by reversed-
phase HPLC (SB-C18 column, 0 to n %, 30 min gradient
CH3CN/ 50 mM triethylammonium acetate (TEAA) (pH
7.5), n ) 18, 20, 25, 28, 33, 37, 45 for X ) none, T, H, F,
L, B, I). Oligomers for base pairing were synthesized in
DMT-on mode. Sequences containing only natural bases
were purified by Poly-Pack II (Glen Research, cat. no.
60-3100-10), and the sequences with a nonnatural base
were purified by reversed-phase HPLC (SB-C18 column,
0-50%, 20 min gradient CH3CN/50 mM TEAA (pH 7.5)).
The post-HPLC detritylation and precipitation followed
the standard protocol.24 All oligonucleotides with a non-
natural base were characterized by MALDI mass spec-
trometry (Supporting Information, Table S4), and their
(24) Current Protocols in Nucleic Acid Chemistry; Beaucage, S. L.,
Bergstrom, D. E., Glick, G. D., Jones, R. A., Eds.; John Wiley & Sons:
New York, 2002; p 10.5.4.
(25) Beaucage, S. L.; Caruthers, M. H. Terahedron Lett. 1981, 22,
1859.
2050 J. Org. Chem., Vol. 70, No. 6, 2005

Thymidine Analogues of Increasing Size in DNA
TABLE 1. Melting Data and Thermodynamic Parameters for Stacking for Thymine and Nonpolar Thymine Analogues,
As Measured by Dangling End Thermal Denaturation Studies with Self-Complementary Strands (5′-XCGCGCG)a
b
dangling
residue (X)
Tm
∆Tm
(°C)
∆H° (kcal/mol,
van’t Hoff)
∆S° (cal/K‚mol,
van’t Hoff)
∆G°37 (kcal/mol,
∆G°37
(kcal/mol, fits)
∆∆G° stacking per base
(°C)
van’t Hoff)
(kcal/mol)
none
T
38.6
46.0
49.7
52.4
52.3
54.3
53.8
-
-41.6 ( 5.0
-54.4 ( 4.1
-55.4 ( 2.0
-54.5 ( 0.9
-52.7 ( 1.2
-62.0 ( 2.1
-55.7 ( 4.4
-109 ( 16
-146 ( 13
-147 ( 6
-143 ( 3
-138 ( 4
-165 ( 7
-146 ( 14
-7.8 ( 0.2
-9.1 ( 0.1
-9.7 ( 0.1
-10.1 ( 0.0
-10.0 ( 0.1
-10.8 ( 0.1
-10.4 ( 0.2
-7.8 ( 0.1
-9.1 ( 0.1
-9.7 ( 0.0
-10.1 ( 0.0
-10.1 ( 0.0
-10.2 ( 0.0
-10.2 ( 0.1
-
7.4
0.65
0.95
1.15
1.10
1.50
1.30
H
F
11.1
13.8
13.7
15.7
15.2
L
B
I
a Conditions: 1 M NaCl, 10 mM phosphate buffer (pH 7.0) with 0.1 mM EDTA, monitored at 280 nm. b 5.0 µM DNA strand concentration
for Tm value shown.
the results from curve-fitting, indicating that the two-
state approximation is a reasonable one for these two
sequences.
Results
Sequence Design. We incorporated all five phos-
phoramidites into three new sequence contexts each for
further study of pairing and stacking properties. The first
is a self-complementary sequence designed to form a
hexamer duplex, with the nonnatural base analogue
overhanging (dangling) at the 5′ ends (Table 1). This
same sequence context has been used previously in
several studies of DNA base stacking.8 The second and
third contexts comprise a pair of 12mer sequences that
form duplexes with well-behaved two-state behavior.3b
One strand is pyrimidine-rich, while the other is purine-
rich; we substituted each analogue at a central position
of each strand to test two widely varied sequence contexts
for pairing effects.
Base Stacking. DNA base stacking stability has been
correlated with the size of compounds when mono-, bi-,
tri-, and tetracyclic base analogues were compared.8 To
test this with the present series, which varies in size by
a smaller amount, we examined the short self-comple-
mentary oligonucleotides by thermal denaturation ex-
periments. The self-complementary duplexes contained
the modified nucleosides as a 5′ overhanging base. The
data are given in Table 1, and examples of denaturation
curves and van’t Hoff plots are given in Supporting
Information, Figure S1. The results show that all five
FIGURE 2. Base pairing of nonpolar thymine analogues (H,
compounds stabilized the duplex more than natural
thymidine does in the same position. The least stabilizing
of the series was the smallest, the toluene case (H).
Interestingly, the remaining four larger compounds (F,
L, B, I) showed only small differences in their stabiliza-
tion of this duplex, with a small increase in stabilization
with the increase in size overall.
Because stacking propensities of DNA bases and
analogues have been most strongly correlated with
surface area and hydrophobicity,8 we calculated these
parameters for the base analogues and their deoxyribo-
side derivatives. The data are shown in Supporting
Information, Table S1. Results showed that the surface
area of the base analogues increased in a regular way
from 154 to 201 Å2 as the series proceeded from the
smallest substituents (H) to the largest (I). Calculated
log P (octanol-water) values were determined for the free
deoxyribosides of the series; these also increased mono-
tonically as size increased. The smallest log P value (for
dH) was 1.25, and the largest (for dI) was 3.97. Thus
F, L, B, I) opposite the natural bases in the center of a 12-
base pair duplex, evaluated by thermal melting temperature
(Tm). Nonnatural base analogues were substituted in a pyrim-
idine-rich strand (A) or a purine-rich strand (B).
these nucleosides are expected to be strongly hydrophobic
compared to thymidine nucleoside (calculated log P )
-1.29), which is a relatively polar, water-soluble com-
pound.
Base Pairing. Pairing studies of the nonpolar series
were then carried out in a 12 base pair duplex; results
are summarized in Figure 2 (see also the Supporting
Information, Table S2). The data showed that all five
compounds are quite destabilizing (compared to T) when
paired opposite natural bases. Results were the same
regardless of whether the analogues were in the pyrimi-
dine- or purine-rich strand (Figure 2A,B). For each
analogue, pairing stability was nearly the same regard-
less of which natural base was opposite it; however, all
five compounds showed a small but significant preference
J. Org. Chem, Vol. 70, No. 6, 2005 2051

Kim and Kool
across the series is probably a desirable property for a
number of biophysical and mechanistic applications, such
as probing protein-DNA interactions and enzyme active
sites. This keeps nonsteric properties as equal as possible
as the effects of these compounds are compared, allowing
for more confident conclusions about steric influences
alone.
With natural bases as partners, the members of this
series behaved remarkably similarly in pairing ability
despite the changes in size. Stability was nearly the same
regardless of which natural base was opposite the
compounds. It is interesting, however, that all five
compounds showed a small but significant preference for
adenine as partner. We attribute this to two known
properties of adenine: first, it is the strongest stacking
of the four natural bases.8-a,26 Probably for the same
reason, adenine is known to be the most stabilizing of
the natural bases when the partner lacks a base alto-
gether (an abasic site).27 The second property is the
relatively weak solvation of adenine compared to the
other three bases.28 If a nonpolar base is placed opposite
adenine, partially desolvating it, this would have less
energetic cost than if any other base were desolvated.
Looking at the trend in the series, there was a slightly
more pronounced preference for adenine exhibited by the
largest members of the series. One possible explanation
for this is the greater hydrophobicity of the larger
analogues, which might lead to a stronger preference for
the least well solvated natural base.
The earliest studies of nonpolar-nonpolar base ana-
logue pairs reported that nonpolar-nonpolar pairs were
less destabilizing than pairs of nonpolar bases opposite
the polar natural bases,3b and this same effect is observed
here as well. This selective pairing of nonpolar com-
pounds can be attributed to avoidance of desolvation costs
when pairing opposite polar compounds, as well as to
increased surface area of hydrophobic contact. Notably,
some of the pairs in this series are not destabilizing to
DNA, and a few of the largest pairs, such as B-I and
I-I, were as stabilizing as a natural base pair. Thus,
these last pairs can be added to a small but growing list
of nonpolar aromatic DNA base replacements that are
stable and selective in DNA, but which are also orthogo-
nal, pairing poorly with natural bases.18,19,29 Such proper-
ties may one day find use in designed genetic pairing
systems.
FIGURE 3. Pairing of hydrophobic nucleosides with them-
selves near the center of a 12-base pair duplex, as measured
by thermal melting temperature.
for adenine as partner. The data also showed that,
although all analogues were destabilizing opposite the
natural bases, there was a general lessening of the
destabilization as the analogue increased in size. Of the
five nonpolar compounds, the diiodo case was the least
destabilizing when paired opposite natural bases.
Finally, the entire set of nonnatural pairs was tested,
evaluating various combinations of compounds (H, F, L,
B, I) paired against each other. The data showed a wide
range of thermal stabilities for the duplexes containing
nonpolar pairs (Figure 3 and Supporting Information,
Table S3). Most of the nonnatural pairs were destabiliz-
ing relative to a T-A base pair in the same context; the
most destabilizing nonpolar pairs involved the smallest
base analogues. The H-H and F-H pairs were among
these (Figure 3). Notably, however, some pairs, involving
L, B, I in the pyrimidine strand and B, I in the purine
strand, were not as destabilizing, and a few (L-I, B-I,
and I-I in particular) were equally as stabilizing as a
thymine-adenine pair in this context.
A more detailed evaluation of the thermodynamics of
duplexes containing the smaller set of self-pairs was
carried out by combined use of melt curve fitting and
van’t Hoff analysis (Table 2 and Supporting Information,
Figure S2). The data revealed the I-I self-pair to be the
most stable, and the H-H pair to be the least stable,
with a large difference of 13 °C in Tm and 2.7 kcal/mol
between these dodecamer duplexes containing only a
single pair difference. Both the Tm and free energy values
for the I-I-containing duplex were identical to the values
for the same duplex containing a T-A pair at that site
(Supporting Information, Table S2).
Experimental Section
General Procedure for Preparation of 5′-O-Tritylated
â-C-Nucleosides. 1′,2′-Dideoxy-â-1′-(3-methylphenyl)-D-ribo-
furanose (1a) (0.5 mmol) was coevaporated with dry pyridine
twice and dissolved in pyridine (3 mL). To the solution were
Discussion
Since these nonpolar analogues lack the ability to
undergo Watson-Crick hydrogen bonding, their stabili-
zation of DNA (in cases where stabilization does occur)
must arise instead from favorable stacking and/or the
associated changes in solvation. The current data show
that the halogenated compounds increase in their stack-
ing ability by a small amount as size increases, a trend
that is generally expected since surface area and
hydrophobicityswhich are strong predictors of stacking
in DNA8s increase as well. However, since the size
change is relatively small across the series, the difference
in stacking is moderate. A similar magnitude of stacking
(26) Allawi, H. T.; SantaLucia, J., Jr. Nucleic Acids Res. 1998, 26,
2694.
(27) (a) Gelfand, C. A.; Plum, G. E.; Grollman, A. P.; Johnson, F.;
Breslauer, K. J. Biochemistry 1998, 37, 7321. (b) Pompizi, I.; Haberli,
A.; Leumann, C. J. Nucleic Acids Res. 2000, 28, 2702.
(28) Gao, J. Biophys. Chem. 1994, 51, 253.
(29) (a) Matsuda, S.; Henry, A. A.; Schultz, P. G.; Romesberg, F. E.
J. Am. Chem. Soc. 2003, 125, 6134. (b) Berger, M.; Luzzi, S. D.; Henry,
A. A.; Romesberg, F. E. J. Am. Chem. Soc. 2002, 124, 1222. (c) Berger,
M.; Ogawa, A. K.; McMinn, D. L.; Wu, Y.; Schultz, P. G.; Romesberg,
F. E. Angew. Chem., Int. Ed. 2000, 39, 2940. (d) Berger, M.; Wu, Y.;
Ogawa, A. K.; McMinn, D. L.; Schultz, P. G.; Romesberg, F. E. Nucleic
Acids Res. 2000, 28, 2911. (e) Brotschi, C.; Leumann, C. J. Angew.
Chem., Int. Ed. 2003, 42, 1655.
2052 J. Org. Chem., Vol. 70, No. 6, 2005

Thymidine Analogues of Increasing Size in DNA
TABLE 2. Stabilities of X-X Self-Pairs of Nonpolar Thymidine Analogues (H, F, L, B, I), As Measured by Melting
Temperature (Tm) and Free Energy (∆G°) for the Duplexesa
5′-CTTTTCXTTCTT
3′-GAAAAGXAAGAA
∆H° (kcal/mol,
van’t Hoff)
∆S° (cal/K‚mol,
van’t Hoff)
∆G°37 (kcal/mol,
-∆G°37
(kcal/mol, fits)
Tmb (°C)
van’t Hoff)
H‚H
F‚F
L‚L
B‚B
I‚I
29.4
31.9
38.0
40.8
42.1
-54.3 ( 4.6
-65.4 ( 7.8
-73.3 ( 11.2
-73.0 ( 3.5
-69.0 ( 3.5
-152 ( 15
-187 ( 26
-208 ( 36
-205 ( 11
-192 ( 6
-7.1 ( 0.1
-7.2 ( 0.2
-8.5 ( 0.3
-9.3 ( 0.1
-9.5 ( 0.0
-7.0 ( 0.1
-7.3 ( 0.1
-8.5 ( 0.1
-9.3 ( 0.1
-9.7 ( 0.1
a Conditions: 1 M NaCl, 10 mM phosphate buffer (pH 7.0) with 0.1 mM EDTA, monitored at 260 nm. b 2.5 µM concentration for each
DNA strand for Tm value. For reference, Tm ) 42.4 °C for X1 ) T, X2 ) A and Tm ) 42.4 °C for X1 ) A, X2 ) T.
added diisopropylethylamine (0.13 mL, 1.5 equiv) and 4,4′-
dimethoxytrityl (DMT) chloride (170 mg, 1.0 equiv). The
mixture was stirred at room temperature for 2 h, and then an
additional portion of DMTCl (42 mg, 0.25 equiv) was added to
the solution. After 1 h, the reaction mixture was quenched with
methanol (5 mL) and volatiles were removed in vacuo. The
residue was loaded onto a silica gel column (preequilibrated
with 5% triethylamine in hexane) and eluted (30:10:1 hexane/
ethyl acetate/triethylamine).
phoramidite (1b): 1H NMR (CDCl3, ppm) 7.50-7.47 (m, 2H),
7.39-7.35 (m, 4H), 7.28-7.18 (m, 6H), 7.09-7.07 (m, 1H),
6.82-6.79 (m, 4H), 5.13 (app quintet, 1H, J ) 6.0 Hz), 4.53-
4.50 (m, 1H), 4.23 (br, 1H), 3.85-3.66 (m, 2H), 3.78 (app d,
6H, J ) 3.0 Hz, OCH3), 3.63-3.58 (m, 2H), 3.35-3.29 (m, 1H),
3.26-3.22 (m, 1H), 2.61 (app t, 1H, J ) 6.5 Hz), 2.45 (app t +
d, 1H, J ) 7.0 (for triplet), 2.0 (for doublet) Hz), 2.41-2.31
(m, 1H), 2.31 (s, 3H, ArCH3), 2.08-2.01 (m, 1H), 1.20-1.15
(m, 8H), 1.08 (app d, 4H, J ) 6.5 Hz); 13C NMR (CDCl3, ppm)
158.7, 145.2, 141.9, 138.2 (d), 136.4 (m), 130.4 (d), 128.6 (d),
128.0, 127.0 (t), 123.5, 117.8 (d), 113.3, 86.3 (d), 85.9 (d),80.6
(d), 76.5 (d), 76.2 (d), 64.5 (d), 58.6 (d), 55.5 (d), 43.5 (m), 31.9,
24.8 (m), 22.9, 21.7, 20.6 (m), 14.4; 31P NMR (CDCl3, ppm,
external H3PO4 standard) 148.2, 148.5; HRMS (FAB+, NBA
matrix) calcd mass 733.3382 for [C42H51N2O6P + Na], found
733.3378.
1′,2′-Dideoxy-â-1′-(3-methylphenyl)-5′-O-(4,4′-dimethoxy-
1
trityl)-D-ribofuranose (1a): H NMR (CDCl3, ppm) 7.47 (d,
2H, J ) 7.2 Hz), 7.36 (dd, 4H, J ) 8.8, 2.4 Hz), 7.29-7.15 (m,
6H), 7.08 (d, 1H, J ) 7.2 Hz), 6.82 (d, 4H, J ) 8.8 Hz), 5.14
(dd, 1H, J ) 10, 5.6 Hz), 4.42 (br, 1H), 4.07-4.04 (m, 1H),
3.78 (s, 6H), 3.37-3.34 (m, 1H), 3.29-3.25 (m, 1H), 2.30 (s,
3H), 2.23 (ddd, 1H, J ) 13, 5.6, 2.0 Hz), 2.10-2.03 (m, 1H);
13C NMR (CDCl3, ppm) 158.6, 145.1, 141.96, 138.2, 136.3,
130.3, 128.4, 128.1, 127.0, 126.9, 123.4, 113.3, 86.5, 86.4, 80.3,
74.9, 64.7, 55.4, 44.1, 21.7; HRMS (FAB+, NBA matrix) calcd
mass 533.2303 for [C33H34O5 + Na], found 533.2288.
Acknowledgment. This work was supported by the
U.S. National Institutes of Health (GM072705). T.W.K.
acknowledges support from a KOSEF postdoctoral fel-
lowship.
General Procedure for Preparation of 3′-O-Phos-
phoramidites. 1′,2′-Dideoxy-â-1′-(3-methylphenyl)-5′-O-(4,4′-
dimethoxytrityl)-D-ribofuranose (0.4 mmol) was dissolved in
dry dichloromethane (5 mL), and to this were added diisopro-
pylethylamine (0.28 mL, 4 equiv) and 2-cyanoethyl N,N-
diisopropylchlorophosphoramidite (0.13 mL, 1.5 equiv). The
reaction mixture was stirred at room temperature for 3 h, and
then all volatiles were removed in vacuo. The residue was
loaded onto a silica gel column (preequilibrated with 5%
triethylamine in hexane) and eluted (60:20:1 hexane/ethyl
acetate/triethylamine).
Supporting Information Available: Thermodynamic
data (melting curves, van’t Hoff plots, Tm data) and calculated
physical properties. Characterization of compounds 2a,b, 3a,b
1
and 4a,b and H NMR spectra of 5′-O-tritylated â-C-nucleo-
sides and 3′-O-phosphoramidites. HPLC chromatograms and
MALDI MS of unnatural oligonucleotides. This material is
available free of charge via the Internet at http://pubs.acs.org.
1′,2′-Dideoxy-â-1′-(3-methylphenyl)-5′-O-(4,4′-dimethoxy-
trityl)-D-ribofuranose cyanoethyl N,N-diisopropylphos-
JO048061T
J. Org. Chem, Vol. 70, No. 6, 2005 2053