Tethered naphthalene diimide intercalators enhance DNA triplex stability

Article abstract of DOI:10.1016/S0968-0896(01)00135-3

Naphthalene diimides function as effective intercalators and when tethered to the 5'-terminus of a pyrimidine-rich oligonucleotide can contribute significantly to the overall stabilization of DNA triplexes. This stabilization can be further enhanced by al

Full text of DOI:10.1016/S0968-0896(01)00135-3

Bioorganic & Medicinal Chemistry 9 (2001) 2329–2334
Tethered Naphthalene Diimide Intercalators
Enhance DNA Triplex Stability
Diego A. Gianolio and Larry W. McLaughlin*
Department of Chemistry, Merkert Chemistry Center, 2609 Beacon St., Boston College, Chestnut Hill, MA 02467, USA
Received 1 February 2001; accepted 12 April 2001
Abstract—Naphthalene diimides function as effective intercalators and when tethered to the 50-terminus of a pyrimidine-rich oli-
gonucleotide can contribute significantly to the overall stabilization of DNA triplexes. This stabilization can be further enhanced by
alterations to the linker tethering the DNA sequence and the intercalator. Less flexible linkers, and particularly one with a phenyl
ring present, appear to permit the stabilization afforded by the bound intercalator to be transferred more effectively to the three-
stranded complex. The conjugate containing the phenyl linker exhibits a TM value that is increased by 28 ꢀC relative to the uncon-
jugated triplex. That the linker itself contributes to the observed stabilization is clear since introduction of the phenyl linker
increases the observed TM by 11 ꢀC relative to a simple flexible linker. # 2001 Elsevier Science Ltd. All rights reserved.
Introduction
Tethering a ligand with the ability to provide added
binding interactions between the third strand and the
target duplex (or product triplex) can result in addi-
tional complex stabilization. Studies of this type have
been reported for tethered intercalators,14,15 groove
binding agents,16ꢁ18 and a variety of other ligands,19,20
all of which can provide significant, or at least moderate
enhancements in complex stability.
DNA triplexes were observed in the 1950s to result from
the disproportionation of polymers composed of poly A
and poly U1 and subsequently similar observations were
made for other polymer compositions.2,3 In 1987, Der-
van and coworkers4 and Helene and coworkers,5
demonstrated that purine-rich sequences of duplex
DNA could be intermolecularly targeted in a sequence-
specific manner with an oligonucleotide rich in pyr-
imidines. In the following year, similar duplex targeting
was effected using a purine-rich strand.6 A number of
studies have subsequently explored the possibility of
more generalized sequence targeting using a variety of
triplex motifs7,8 and including modified base
residues9ꢁ12 with varying levels of success.13
The tetracarboxylic diimides of naphthalene and per-
ylene are effective intercalating agents,21 and when
tethered to one or both termini of a third strand,22 the
center of the third strand,23 or in the center of a hairpin
triplex,24 they confer significant additional complex sta-
bility upon the DNA triplex. In the present work, we
have continued these studies to explore the nature of the
linker between the naphthalene diimide (NDI) inter-
calator and the oligonucleotide used in DNA triplex
formation.
DNA triplexes, composed of three strands of DNA are
generally less stable than corresponding duplexes and
this observation may in part result from the additional
charge–charge repulsive forces present when three
negatively charged DNA strands are brought together
in a single complex. A number of techniques, modified
nucleosides, and ligands have been examined to increase
the stability of three-stranded structures with a view to
developing effective anti-gene or anti-sense therapeutics.13
Results and Discussion
We have shown in previous studies that tethered NDI
intercalators can be effective in stabilizing triple helical
complexes formed from a third strand of DNA and a
target duplex. The originally reported NDI-intercalator
was tethered to one terminus of the third strand through
a 2-hydroxyethyl-2-ethoxy linker. In that case, dena-
turation of the complex appeared to occur in two distinct
*Corresponding author. Tel.:+1-617-552-3622; fax: +1-617-552-
2705; e-mail: larry.mclaughlin@bc.edu
0968-0896/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(01)00135-3

2330
D. A. Gianolio, L. W. McLaughlin / Bioorg. Med. Chem. 9 (2001) 2329–2334
steps when observed at 260 and 383 nm. Similar two-
step transitions are observed for the current linkers as
illustrated in Figure 1. The TM values obtained at
260 nm have been interpreted as representing the dis-
association of the third strand from the duplex while the
values obtained at 383 nm were interpreted as reflecting
the unstacking of the NDI residue from the duplex. The
mid-point of the triplex to duplex transition as mon-
itored at 260 nm does not correlate with the mid-point
of the corresponding NDI unstacking event as mon-
itored at 383 nm (Fig. 1). The fact that the values
obtained at 383 nm are substantially higher than those
observed at 260 nm (Fig. 1, Table 1) suggests a two-step
process with an intermediate complex that could be
characterized as having the NDI still intercalated into
the duplex, but with the third strand no longer bound to
the target. This intermediate state might in part be a
result of the length and flexibility of the linker. Mod-
ifications to the linker to reduce its flexibility might be
effective transferring the added stability of the inter-
calator to the entire complex and in doing so, enhance
the TM values measured at 260 nm.
was converted to a terminal phosphoramidite (Scheme
2). Addition of the desired NDI derivative containing a
free terminal hydroxyl group in the presence of tetrazole
facilitated the conjugation of the NDI derivatives to the
termini of the synthesized oligonucleotide strands.
While this reaction does not occur with the high yields
typical of nucleoside phosphoramidite couplings, the
conjugation can be achieved in yields of about 30%,
and in selected cases with yields of 50%. Conformation
of the purity and composition of the conjugates could
be accomplished by a combination of MALDI-TOF
spectrometry, UV spectroscopy, and HPLC analyses.
Table 1. TM value for NDI-conjugated triplexes
Sequence
Triplex TM values (ꢀC)
measured at 260 nm
NDI unstacking TM values
(ꢀC) measured at 383 nm
pH 6.4
pH 7.0
pH 6.4
Native
NDI
DMe–NDI
Phen–NDI
30
47
52
58
22
22
40
48
—
59
54
61
Two linkers were examined in this study, one was pre-
pared from 3-hydroxypropylamine containing gem-
dimethyl groups on the central carbon atom (Scheme 1)
and the second was derived from p-(2-hydroxyethyl)ani-
line. In both cases, the corresponding NDI derivatives
(Fig. 2) were obtained simply by heating equivalent
amounts of the linker, N,N-dimethylaminoethylamine,
and the tetracarboxylic dianhydride derivative of naph-
thalene. Although this approach also generates
significant amounts of both of the corresponding sym-
metrical diimides, in each case roughly a third of the
product mixture was the desired unsymmetrical diimide
that could easily be resolved by column chromatography.
The phosphoramidite derivatives of the diimides proved
to be quite unstable and could not be stored effectively
even at ꢁ20 ꢀC. Instead of employing the conventional
approach, we tethered the NDI derivatives to the 50-
terminus of the DNA strands using a reverse coupling
protocol. In this procedure, the 50-terminal DMT-pro-
tecting group was removed and the unmasked hydroxyl
The 15-mer polypyrimidine strand of the conjugates was
designed to target a 15-mer polypurine strand in a target
DNA duplex (see Table 1). The native triplex was
formed from the target duplex and the unconjugated
third strand, and this complex resulted in two thermally
induced transitions as we have reported previously. The
early transition occurred at 30 ꢀC (pH 6.4) or 22 ꢀC (pH
7.0) depending upon the pH of the solution, and the
later transition occurred at 71 ꢀC regardless of pH.22
The early transition was interpreted to reflect dissocia-
tion of the third strand from the duplex and the later
transition the denaturation of the duplex. When the
third strand tethers the NDI ligand through 2-hydroxy-
ethyl-2-ethoxy linker, these two transitions were
observed to increase by 17 and 13 ꢀC, respectively
(Table 1). Additionally, when the temperature versus
absorbance data is monitored at 383 nm, a single tran-
sition was observed to occur at 59 ꢀC (pH 6.4). The
transition observed at 383 nm was only present with the
NDI conjugates and was interpreted to reflect the
unstacking of the NDI chromophore from a position of
intercalation within the target duplex. The observation
that the unstacking TM (383 nm) and the triplex transi-
tion at pH 6.4 (260 nm) differ by some 12 ꢀC for the
simple NDI conjugate indicates that dissociation of the
conjugated third strand from the duplex occurs by at
least a two-step process.
We have examined TM effects for two additional com-
plexes with varying linkers tethering the NDI inter-
calator and the DNA third strand. The DMe–NDI
conjugate contains a linker reduced from six atoms to
four and additionally the central carbon of the linker is
gem-dimethyl substituted (Fig. 2); these modifications
should reduce the flexibility of the linker. The Phen–
NDI conjugate has a linker of about the same length as
Figure 1. Thermally induced transitions foe the DMe–NDI conjugate
measured at 260 nm (solid line) and 383 nm (dashed line). The mid-
points of the transitions (arrows) do not coincide suggesting a biphasic
transtion.

D. A. Gianolio, L. W. McLaughlin / Bioorg. Med. Chem. 9 (2001) 2329–2334
2331
Scheme 1. (i) DMT-Cl, Tos-Cl, Py; (ii) NaN3, DMF, 90 ꢀC; (iii) H2 Pd/C 10%, MeOH; (iv) isopropanol, Á; (v) 3% trichloroacetic acid/CH2Cl2.
the NDI conjugate, but incorporates a phenyl ring into
it. The TM of the triplex formed with the DMe con-
jugate was 52 ꢀC, some 5 ꢀC higher than that observed
with the NDI conjugate (Fig. 1), and 22 ꢀC higher than
the unconjugated triplex (Table 1). The unstacking TM
(383 nm) for the intercalator was reduced by 5 ꢀC rela-
tive to that observed for the NDI conjugate. By com-
parison, the TM for the Phen–NDI conjugate was
increased a further 6 ꢀC, a total of 11 ꢀC higher (pH 6.4)
than the NDI conjugate and some 28 ꢀC higher than the
unconjugated triplex (Table 1).
complex stabilization, as observed. The reduction in the
NDI unstacking TM may also reflect the shortened nat-
ure of the linker. With a shorter linker, the intermediate
complex, that with a denatured third strand and an
intercalated NDI may experience enhanced charge–
charge repulsion between the duplex and denatured
third strand. This effect could then reduce the NDI
unstacking TM.
A similar argument for increased overall stability, and a
reduced unstacking TM can be made for the Phen–NDI
conjugate since the introduction of the aromatic ring
into the linker should also reduce flexibility relative to
the NDI conjugate. This effect is likely present since the
triplex TM value for the Phen–NDI conjugate increases
by 11 ꢀC relative to the NDI conjugate and is 5 ꢀC
higher than the DMe–NDI conjugate (pH 6.4). How-
ever, the unstacking TM is also observed to increase
relative to the DMe–NDI conjugate, and is even slightly
higher than that of the NDI conjugate (Table 1). In fact,
this complex exhibits the smallest difference (3 ꢀC)
between the TM value obtained at 260 nm (58 ꢀC) and
the unstacking TM obtained at 383 nm (61 ꢀC). This
observation suggests that the Phen–NDI is not only
better able to intercalate into the duplex, but is also
better able to transfer that stabilizing effect to the bind-
ing by the third strand. In addition to reducing flex-
ibility in the linker, the presence of the phenyl ring in
the Phen–NDI conjugate could provide additional sta-
bilizing base stacking interactions. Simple model build-
ing suggests that the Phen–NDI ligand might be better
able to position itself such that it cold benefit from base
stacking interactions with all three base residues at the
final position in the triplex (Fig. 3). The additional base
stacking interactions resulting from the phenyl-based
linker would not only enhance the unstacking TM of the
ligand, but could provide an additional stabilizing force
for the DNA triplex itself. In a study of hairpin tri-
plexes, we have also observed that increasing the size of
the aromatic portion of the linker (from naphthalene to
perylene) enhanced complex stability presumably
through increased base stacking interactions with all
three residues of the terminal base triplet. Other studies
have also shown that ligands designed to interact with
all three residues of the base triplet are more effective in
triplex stabilization than simple DNA duplex inter-
calators.
The enhanced TM values for the DMe–NDI conjugate
may result from reduced linker flexibility as a result of
the shortened linker and the introduction of the dime-
thyl substituents. With reduced flexibility the linker may
permit the bound NDI to better stabilize the three-
stranded complex. Since the unstacking of the NDI
from the duplex occurs at a higher temperature than the
dissociation of the third-strand bases from the duplex, a
less flexible linker may better transfer the binding effects
of the intercalator and result in more effective overall
Figure 2. Structures of three naphthalene diimide based intercalators
with linkers based upon 2-hydroxy-2-ethoxyethylamine (NDI), 3-
hydroxy-2,2-dimethylpropylamine (DMe–NDI) and 2-hydroxy-
ethylaniline (Phen–NDI).

2332
D. A. Gianolio, L. W. McLaughlin / Bioorg. Med. Chem. 9 (2001) 2329–2334
standard. HRMS and LRMS (FAB) mass spectra were
obtained from the Mass Spectrometry Laboratory,
School of Chemical Sciences, University of Illinois,
Urbana, IL, USA. MALDI-TOF mass spectra were
obtained from the Mass Spectrometry Laboratory,
Merkert Chemistry Center, Boston College, MA, USA.
Rotary evaporations were performed under reduced
pressure utilizing Buchi systems. Thin layer chromato-
graphy (TLC) was performed on Silica Gel 60 F254 pre-
coated on aluminum sheets (EM Separations Technol-
ogy). Anhydrous solvents and starting materials were
purchased from Aldrich Chemical Company and used
without further purification. UV measurements were
obtained using a Beckman DU 640 spectrometer. Oli-
godeoxyribonucleotides were synthesized on an Applied
Biosystems 381A DNA Synthesizer. 20-Deoxyribo-
nucleotide phosphoramidites, fast deprotecting N-ace-
tyl-20-deoxycytidine phosphoramidite (Ac-dC), and 30
terminal nucleoside controlled pore glass support
(CPG), were purchased from Glen Research (Sterling,
VA, USA). Fast flow high performance liquid chroma-
tography (HPLC) was performed on a Waters 600
Pump Control Unit HPLC system, using Oligo R3 col-
umns (Perseptive Biosytems, Framingham, MA, USA)
with detection at 260 and 383 nm. Oligodeoxyr-
ibonucleotides were desalted with Econo-Pac 10DG
disposable chromatography columns (Bio Rad, Her-
cules, CA, USA). TM measurements were performed on
an AVIV Spectrometer, Model 14DS UV/Vis.
Scheme 2.
Conclusions
Selected ligands capable of stacking with the base tri-
plets present in three-stranded DNA triplexes can sig-
nificantly enhance the overall stability of the complex.
In the present study, the introduction of a NDI inter-
calator tethered to the DNA third strand with a phenyl-
containing linker results in a 28 ꢀC increase in the
observed TM value for the simple triplex. That the linker
itself can contribute to such dramatic increases in TM
values is clear from the observation that the presence of
the phenyl-containing linker enhances the observed TM
value by 11 ꢀC relative to a simple 2-hydroxyethyl-2-
ethoxy linker.
O-Dimethoxytrityl-O0-tosylneopentyl glycol (1). To 4.0 g
(0.04 mol) of neopentyl glycol, evaporated twice from
freshly distilled pyridine and under nitrogen atmo-
sphere, was added 50 mL of freshly distilled pyridine,
followed by 13.6 g (0.04 mol) of dimethoxytrityl chlo-
ride, 7.6 g (0.04 mol) of tosyl chloride, and 2.4 g
(0.02 mol) of dimethylaminopyridine. The mixture was
stirred at room temperature overnight. The solvent was
removed in vacuo until an oily residue was obtained, the
desired product was subsequently purified by silica gel
column chromatography, using a gradient of dichloro-
methane in hexane (0–40%) resulting in 11.2 g
(0.02 mol) of 1, 48% yield, as a yellowish oil. Rf
(chloroform): 0.68. 1H NMR (DMSO-d6): d 0.78 (s, 6H,
CH3); 2.38 (s, 3H, Tosyl–CH3); 2.70 (s, 2H, CH2); 3.73
(s, 6H, –OCH3); 3.83 (s, 2H, CH2); 6.83–7.74 (m, 13H,
Ar–H) ppm. 13C NMR (CDCl3): d 22.9, 23.2, 37.2, 56.4,
68.7, 76.9, 86.7, 114.0, 127.6, 128.7, 129.0, 129.1, 129.4,
130.8, 131.1, 131.3, 137.0, 145.5, 146.0, 159.3 ppm. HR-
MS calcd for C33H36O6S1 (M): 560.2233, found:
560.2233.
Experimental
NMR spectra were obtained on a Varian spectrometer
(400 MHz) and contained trimethylsilane as an internal
CAUTION: Azides are potentially explosive.28,29
O-Dimethoxytrityl-2-dimethyl-3-hydroxypropyl azide (2).
To 4.9 g (8.7 mmol) of 1 dissolved in 25 ml of dimethyl-
formamide (DMF), was added 1.4 g (22 mmol) of
sodium azide. The reaction flask was put on an oil bath
and heated up to 90 ꢀC, and the mixture was stirred for
2 days. The contents of the flask were then poured into
a separatory funnel and the desired product was
extracted with dichloromethane from water. The
organic phase was collected, dried on NaSO4 and fil-
Figure 3. Possible relationship of the Phen–NDI intercalator (bold)
and the terminal base triplet of a DNA triplex.

D. A. Gianolio, L. W. McLaughlin / Bioorg. Med. Chem. 9 (2001) 2329–2334
2333
tered. The solvent was removed in vacuo, and the
desired product was subsequently purified by silica gel
column chromatography, using a gradient of chloro-
form in hexane (0–40%) resulting in 3.2 g (7.5 mmol) of
2, 86% yield, as a yellowish oil. Rf (dichloromethane/hex-
dropwise 2 mL of 3% trichloroacetic acid in dichloro-
methane and the reaction was stirred for an additional 5
min. The volatiles were removed by rotary evaporation
and the desired product was purified by silica gel col-
umn chromatography using a gradient of methanol in
dichloromethane (0–5%). For 5, 130 mg (0.31 mmol) of
product was obtained (yield 76%) as a yellow solid. Rf
(methanol/dichloromethane, 1/9): 0.50; mp 244–247 ꢀC
1H NMR (DMSO-d6): d 1.01 (s, 6H, CH3), 2.35 (s, 6H,
N–CH3); 2.68 (t, J=7.2 Hz, 2H, CH2); 3.16 (s, 2H,
CH2); 4.19 (s, 2H, CH2); 4.36 (t, J=6.8 Hz, 2H, CH2);
8.78 (s, 4H, Ar-H) ppm. 13C NMR (CDCl3): d 24.6,
39.6, 39.9, 47.0, 58.1, 69.7, 128.0, 132.0, 132.5, 163.7,
165.2 ppm. HR-MS calcd for C23H26N3O5 (M+ H+):
424.1872, found: 424.1871.
1
ane, 1/1): 0.60. H NMR (CDCl3): d 0.97 (s, 6H, CH3);
2.89 (s, 2H, CH2); 3.34 (s, 2H, CH2); 3.82 (s, 6H, –OCH3);
6.86–7.49 (m, 13H, Ar–H) ppm. 13C N MR (CDCl): d
3
24.3, 38.1, 56.4, 61.1, 69.7, 86.7, 114.1, 127.7, 128.8, 129.2,
131.2, 137.3, 146.2, 159.3 ppm. HR-MS calcd for
C26H29N3O3 (M): 431.2209, found: 431.2208.
O-Dimethoxytrityl-2-dimethyl-3-hydroxypropyl amine (3).
The primary amine was obtained by hydrogenation of
the azide. 2.65 g (6.2 mmol) of 2 were put in a pressure
resistant glass bottle, dissolved in 25 mL of methanol,
and 0.9 g of Pd/C 10% w/w was added. The glass bottle
was filled with H2 to a pressure of 50 psi, and the mix-
ture was shaken vigorously overnight. The contents of
the glass bottle were filtered on Celite and washed with
methanol. The filtrate was collected, concentrated and
the desired product was subsequently purified by silica
gel column chromatography, using a gradient of
methanol in chloroform (0–5%) resulting in 1.3 g
(3.4 mmol) of 3, 54% yield, as an oil. Rf (methanol/di-
N-[4-(2-Hydroxyethyl)phen]-N0-[2-(N,N-dimethylamino)-
ethyl]-1,4,5,8-naphthalene tetracarboxylic diimide (6)
(Phen-NDI). The synthesis was performed following a
procedure similar to that described for 4.24,25 To 3.26 g
(12.2 mmol) of 1,4,5,8-naphthalenetetracarboxylic dia-
nhydride was added 2 g (14.6 mmol) of 4-aminopheny-
lethyl alcohol, 1.6 mL (14.6 mmol) of N,N-
dimethylethylenediamine, and 2.68 g (12.2 mmol) of zinc
acetate dihydrate in 50 mL of pyridine. The reaction
mixture refluxed overnight, the volatiles were removed
by rotary evaporation and the desired product was
purified by silica gel column chromatography using a
gradient of methanol in chloroform (0–4%). The
unsymmetrical diimide eluted as the ‘middle’ compound
between the two symmetric secondary products. For 6
1.12 g (2.44 mmol) of product was obtained (yield 20%)
as a light yellow solid. Rf (methanol/dichloromethane,
1/9): 0.41; mp 246–248 ꢀC. 1H NMR (DMSO-d6): d 2.21
(s, 6H, N–CH3); 2.54 (t, J=7.2 Hz, 2H, CH2); 2.81 (t,
J=7.2, 2H, CH2); 3.68 (q, J=6.8, 2H, CH2); 4.17 (t,
J=6.8 Hz, 2H, CH2); 4.74 (t, J=5.2 Hz, 1H, OH); 7.29–
7.38 (m, 4H, Ar–H); 8.67–8.68 (m, 4H, Ar–H) ppm. 13C
NMR (CDCl3): d 40.0, 40.2, 47.0, 58.2, 64.6, 127.6,
127.9, 129.5, 131.3, 132.1, 132.3, 163.8 ppm. HR-MS
calcd for C26H24N3O5 (M+ H+): 458.1716, found:
458.1717.
1
chloromethane, 1/19): 0.15. H NMR (CDCl3): d 0.90
(s, 6H, CH3); 2.58 (s, 2H, CH2); 2.82 (s, 2H, CH2); 3.79
(s, 6H, –OCH3); 6.81–7.45 (m, 13H, Ar–H) ppm. 13C
NMR (CDCl3): d 23.2, 49.9, 55.5, 70.8, 86.5, 113.5,
127.1, 128.2, 128.3, 130.3, 135.9, 144.9, 158.7 ppm.
N-(O-Dimethoxytrityl-2-dimethyl-3-hydroxypropyl)-N0-
[2-(N,N-dimethylamino)-ethyl]-1,4,5,8-naphthalene tetra-
carboxylic diimide (4) (DM e–NDI). The synthesis of
4 was performed following a general procedure for
the preparation of unsymmetrical naphthalene
21,22
diimides.
To 680 mg (2.6 mmol) of 1,4,5,8-naph-
thalenetetracarboxylic dianhydride was added 1.03 g of
3 and 280 mL (2.6 mmol) of N,N-dimethylethylenedia-
mine in 25 mL of isopropanol. The reaction mixture
refluxed overnight. The volatiles were removed by
rotary evaporation and the desired product was purified
by silica gel column chromatography using a gradient of
methanol in chloroform (0–3%). The unsymmetrical
diimide eluted as the ‘middle’ compound between the
two symmetric secondary products, obtained roughly in
a 1:2:1 ratio. For 4, 465 mg (0.64 mmol) of product was
obtained (yield 25%) as an orange solid. ꢀRf (methanol/
DNA synthesis
The native 25-mer target sequences, 50-d(GCGCGAAA-
GAAAAGAGAGAACCCGG)-30, 50-d(CCGGGTTCT-
CTCTTTTCTTTCGCGC)-30, and the control 15-mer
sequence, 50-d(TTTCTTTTCTCTCTT)-30 were pre-
pared by solid-phase DNA synthesis using conventional
protocols.26 The products were characterized by
MALDI-TOF mass spectrometry. For example, 50-
d(TTTCTTTTCTCTCTT)-30: calculated 15-mer: 4442,
found 15-mer: 4444.
1
dichloromethane, 1/19): 0.55; mp 96–99 C. H N MR
(CDCl3): d 1.05 (s, 6H, CH3), 2.35 (s, 6H, N–CH3); 2.68
(t, J=6.6 Hz, 2H, CH2); 2.98 (s, 2H, CH2); 3.74 (s, 6H,
–OCH3); 4.25 (s, 2H, CH2); 4.36 (t, J=6.6 Hz, 2H,
CH2); 6.65–7.38 (m, 13H, Ar–H); 8.65–8.73 (m, 4H, Ar-
H) ppm. 13C NMR (CDCl3): d 25.0, 30.0, 38.8, 46.0,
48.0, 55.4, 57.2, 71.4, 113.0, 126.7, 126.8, 127.0, 127.8,
128.6, 130.4, 131.1, 136.5, 158.4, 163.3, 163.6 ppm. HR-
MS calcd for C44H44N3O7 (M+ H+): 726.3179, found:
726.3182.
Introduction of the NDI intercalators at the 50 -termi-
nus The phosphoramidite derivatives of the synthesized
naphthalene diimides (NDIs) were difficult to store, or
to use effectively in coupling reactions owing to their
relatively rapid decomposition. The synthesis of the 15-
mers functionalized at the 50 terminus with the naph-
thalene intercalators was performed following the
described reverse coupling procedure.22,27
N-(2-dimethyl-3-hydroxypropyl)-N0-[2-(N,N-dimethyla-
mino)-ethyl]-1,4,5,8-naphthalene tetracarboxylic diimide
(5). To 290 mg (0.41 mmol) of 4 dissolved in 2 mL of
dichloromethane and stirred on ice for 5 min was added

2334
D. A. Gianolio, L. W. McLaughlin / Bioorg. Med. Chem. 9 (2001) 2329–2334
The oligomers were synthesized in the normal fashion
using the fast-deprotecting Ac-dC and dT phosphor-
amidites, and the DMT protecting group was removed
after assembly of the 15-mers. Subsequently 135 mL
(0.425 mmol) of 2-cyanoethyl tetraisopropylpho-
sphorbisamidite diluted in 300 mL of freshly distilled
acetonitrile was delivered to the columns to react with
the free 50-OH of the 15-mers over a 1 h period of time.
The columns were then flushed with acetonitrile for 1
min and then removed from the synthesizer and the
following step was performed manually. For the DMe-
NDI-15-mer, 22 mg (50 mmol) of 5, dissolved in 250 mL
of anhydrous dichloromethane, were delivered together
with 300 mL of a solution of sublimed 1H-tetrazole (as
the activating agent) in acetonitrile using a syringe. The
coupling reaction between the naphthalene moiety and
the terminal phosphoramidite was allowed to occur for
1 h. The column was then reconnected to the synthesi-
zer, rinsed with acetonitrile, and oxidized in the normal
manner. Phen-NDI-15mer, was prepared in a similar
fashion, using 22 mg of 6 (50 mmol), dissolved in 250 mL
of anhydrous DMF and allowing the coupling reaction
to occur overnight (ꢂ18 h).
greater than ꢄ1 ꢀC were not observed when using the
two procedures.
Acknowledgements
This work was supported by grants from the NSF
(MCB 0077667) and NIH (GM37065).
References and Notes
1. Felsenfeld, G.; Davies, D. R.; Rich, A. J. Am. Chem. Soc.
1957, 79, 2023.
2. Lipsett, M. N. Biochem. Biophys. Res. Comm. 1963, 11, 224.
3. Morgan, A. R.; Wells, R. D. J. Mol. Biol. 1968, 37, 63.
4. Moser, H. E.; Dervan, P. B. Science 1987, 238, 645.
5. LeDoan, T.; Perrouault, L.; Praseuth, D.; Habhoub, N.;
Decout, J. L.; Thoung, N. T.; Lhomme, J.; Helene, C. Nucleic
Acids Res. 1987, 15, 7749.
6. Cooney, M.; Czernuszewicz, G.; Postel, E. H.; Flint, S. J.;
Hogan, M. E. Science 1988, 241, 456.
7. Griffin, L. C.; Dervan, P. B. Science 1989, 245, 967.
8. Best, G. C.; Dervan, P. B. J. Am. Chem. Soc. 1995, 117, 1187.
9. Koh, J. S.; Dervan, P. B. J. Am. Chem. Soc. 1992, 114, 1470.
10. Koshlap, K. M.; Gillepsie, P.; Dervan, P. B.; Feigon, J. J.
Am. Chem. Soc. 1993, 115, 7908.
11. Stilz, H. U.; Dervan, P. B. Biochemistry 1993, 32, 2177.
12. Lehman, T. E.; Greenberg, W. A.; Liberles, D. A.; Wada,
C. K.; Dervan, P. B. Helv. Chimi. Acta 1997, 80, 2002.
13. Soyfer, V. N.; Potaman, V. N. Triple-Helical Nucleic
Acids; Springer: New York, 1996.
Upon completion of the DNA syntheses, the beads were
dried in vacuo and deprotected in 5 mL of 0.05 M
K2CO3 in methanol for 2 h at ambient temperature. The
pH was then adjusted to approximately 6.5 with 2%
acetic acid. The solutions were filtered to remove the
CPG beads, and concentrated in vacuo to remove
methanol. Purification of the conjugated oligonucleo-
tides from failed sequences and unconjugated 15-mer
was accomplished by fast flow HPLC (4.6ꢃ100 mm
column), starting with 0% B for 1 min, then using a
linear gradient 0–40% B over 6.0 min, and then 40% B
for 2 min (A: 50 mM triethylammonium acetate, adjus-
ted to pH=7.0 with glacial acetic acid; B: 50 mM tri-
ethylammonium acetate in 70% acetonitrile). The
desired sequences were collected as the latest eluting
peaks with a yield of roughly 30% each. These products
were then desalted (Sephadex G-10), and stored at
ꢁ20 ꢀC.
14. Collier, D. A.; Mergny, J. L.; Thuong, N. T.; Helene, C.
Nucleic Acids Res. 1991, 19, 4219.
15. Mergny, J. L.; Duval-Valentin, G.; Nguyen, C. H.; Per-
rouault, L.; Faucon, B.; Rougee, M.; Montenay-Garestier, T.;
Bisagni, E.; Helene, C. Science 1992, 256, 1521.
16. Lukhtanov, E. A.; Podyminogin, M. A.; Kutyavin, I. V.;
Meyer, R. B.; Gamper, H. B. Nucleic Acids Res. 1996, 24, 683.
17. Robles, J.; Rajur, S. B.; McLaughlin, L. W. J. Am. Chem.
Soc. 1996, 118, 5820.
18. Robles, J.; McLaughlin, L. W. J. Am. Chem. Soc. 1997,
119, 6014.
19. Letsinger, R. L.; Zhang, G.; Sun, D. K.; Ikeuchi, T.;
Sarin, P. S. Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 6553.
20. Tung, C. H.; Breslauer, K. J.; Stein, S. Nucleic Acids Res.
1993, 21, 5489.
MALDI-TOF MS calcd for 50 DMe–NDI-
TTTCTTTTCTCTCTT 30 (15-mer+): 4928, found: 4938.
21. Lokey, R. S.; Kwok, Y.; Guelev, V.; Pursell, C. J.; Hurley,
L. H.; Iverson, B. L. J. Am. Chem. Soc. 1997, 119, 7202.
22. Gianolio, D. A.; Segismundo, J. M.; McLaughlin, L. W.
Nucleic Acids Res. 2000, 28, 2128.
MALDI-TOF MS calcd for 50
Phen–NDI-
TTTCTTTTCTCTCTT 30 (15-mer+): 4962, found: 4972.
23. Gianolio, D. A.; McLaughlin, L. W. J. Am. Chem. Soc.
1999, 121, 6334.
24. Bevers, S.; O’Dea, T. P.; McLaughlin, L. W. J. Am. Chem.
Soc. 1998, 120, 11004.
25. Langhals, H.; Jona, W. Angew. Chem., Int. Ed. 1998, 37,
952.
26. Matteucci, M. D.; Caruthers, M. H. J. Am. Chem. Soc.
1981, 103, 3185.
27. Gianolio, D. A.; McLaughlin, L. W. Nucleosides &
Nucleotides 1999, 18, 1751.
28. Charette, A. B.; Wurz, R. P.; Ollevier, T. J. Org. Chem.
2000, 65, 9252.
29. Benari, L.; Calestani, G.; Nanni, D.; Spagnolo, P. J. Org.
Chem. 1998, 63, 4679.
Thermal melting analyses
TM values were obtained for complexes containing a
1:1:1 mixture of oligonucleotides at a concentration of
1 mM in 50 mM NaCl, 10 mM MgCl2 and 10 mM Pipes
(pH 6.4 and 7.0). Solutions were heated in 1 ꢀC steps
and absorbances were recorded after temperature
stabilization using an AVIV 14DS spectrophotometer.
Absorbance and temperature readings were plotted
using Igor Pro software. TM values were determined
from first order derivatives as well as graphically from
absorbance versus temperature plots. Differences