Solid-Phase Synthesis of Positively Charged Deoxynucleic Guanidine (DNG) Tethering a Hoechst 33258 Analogue: Triplex and Duplex Stabilization by Simultaneous Minor Groove Binding

Article abstract of DOI:10.1021/ja031557s

Deoxynucleic guanidine (DNG), a DNA analogue in which positively charged guanidine replaces the phosphodiester linkages, tethering to Hoechst 33258 fluorophore by varying lengths has been synthesized. A pentameric thymidine DNG was synthesized on solid phase in the 3′ → 5′ direction that allowed stepwise incorporation of straight chain amino acid linkers and a bis-benzimidazole (Hoechst 33258) ligand at the 5′-terminus using PyBOP/HOBt chemistry. The stability of (DNA)₂·DNG-H triplexes and DNA. DNG-H duplexes formed by DNG and DNG-Hoechst 33258 (DNG-H) conjugates with 30-mer double-strand (ds) DNA, d(CGCCGCGCGCGCGAAAAACCCGGCGCGCGC)/d(GCGGCGCGCGCGCTTTTTGGGCCGCGCGCG), and single-strand (ss) DNA, 5′-CGCCGCGCGCGCGAAAAACCCGGCGCGCGC-3′, respectively, has been evaluated by thermal melting and fluorescence emission experiments. The presence of tethered Hoechst ligand in the 5′-terminus of the DNG enhances the (DNA)₂·DNG-H triplex stability by a ΔT_m of 13 °C. The fluorescence emission studies of (DNA)₂·DNG-H triplex complexes show that the DNG moiety of the conjugates bind in the major groove while the Hoechst ligand resides in the A:T rich minor groove of dsDNA. A single G:C base pair mismatch in the target site decreases the (DNA)₂·DNG triplex stability by 11 °C, whereas (DNA)₂·DNG-H triplex stability was decreased by 23 °C. Inversion of A:T base pair into T:A base pair in the center of the binding site, which provides a mismatch selectively for DNG moiety, decreases the triplex stability by only 5-6 °C. Upon hybridization of DNG-Hoechst conjugates with the 30-mer ssDNA, the DNA·DNG-H duplex exhibited significant increase in the fluorescence emission due to the binding of the tethered Hoechst ligand in the generated DNA·DNG minor groove, and the duplex stability was enhanced by ΔT_m of 7 °C. The stability of (DNA)₂·DNG triplexes and DNA·DNG duplexes is independent of pH, whereas the stability of (DNA)₂·DNG-H triplexes decreases with increase in pH.

Full text of DOI:10.1021/ja031557s

Published on Web 03/06/2004
Solid-Phase Synthesis of Positively Charged Deoxynucleic
Guanidine (DNG) Tethering a Hoechst 33258 Analogue:
Triplex and Duplex Stabilization by Simultaneous Minor
Groove Binding
Putta Mallikarjuna Reddy and Thomas C. Bruice*
Contribution from the Department of Chemistry and Biochemistry, UniVersity of California
at Santa Barbara, Santa Barbara, California 93106
Received December 5, 2003; E-mail: tcbruice@bioorganic.ucsb.edu
Abstract: Deoxynucleic guanidine (DNG), a DNA analogue in which positively charged guanidine replaces
the phosphodiester linkages, tethering to Hoechst 33258 fluorophore by varying lengths has been
synthesized. A pentameric thymidine DNG was synthesized on solid phase in the 3′ f 5′ direction that
allowed stepwise incorporation of straight chain amino acid linkers and a bis-benzimidazole (Hoechst 33258)
ligand at the 5′-terminus using PyBOP/HOBt chemistry. The stability of (DNA)2‚DNG-H triplexes and DNA‚
DNG-H duplexes formed by DNG and DNG-Hoechst 33258 (DNG-H) conjugates with 30-mer double-
strand (ds) DNA, d(CGCCGCGCGCGCGAAAAACCCGGCGCGCGC)/d(GCGGCGCGCGCGCTTTTTGGGC-
CGCGCGCG), and single-strand (ss) DNA, 5′-CGCCGCGCGCGCGAAAAACCCGGCGCGCGC-3′,
respectively, has been evaluated by thermal melting and fluorescence emission experiments. The presence
of tethered Hoechst ligand in the 5′-terminus of the DNG enhances the (DNA)2‚DNG-H triplex stability by
a ∆Tm of 13 °C. The fluorescence emission studies of (DNA)2‚DNG-H triplex complexes show that the
DNG moiety of the conjugates bind in the major groove while the Hoechst ligand resides in the A:T rich
minor groove of dsDNA. A single G:C base pair mismatch in the target site decreases the (DNA)2‚DNG
triplex stability by 11 °C, whereas (DNA)2‚DNG-H triplex stability was decreased by 23 °C. Inversion of
A:T base pair into T:A base pair in the center of the binding site, which provides a mismatch selectively for
DNG moiety, decreases the triplex stability by only 5-6 °C. Upon hybridization of DNG-Hoechst conjugates
with the 30-mer ssDNA, the DNA‚DNG-H duplex exhibited significant increase in the fluorescence emission
due to the binding of the tethered Hoechst ligand in the generated DNA‚DNG minor groove, and the duplex
stability was enhanced by ∆Tm of 7 °C. The stability of (DNA)2‚DNG triplexes and DNA‚DNG duplexes is
independent of pH, whereas the stability of (DNA)2‚DNG-H triplexes decreases with increase in pH.
Introduction
Sequence-specific targeting of ss and ds/DNA using duplex-
and triplex-forming oligonucleotides offers a promising anti-
sense/antigene strategy to control the regulation of gene
expression,1 site-directed mutagenesis,2 and gene repair.3,4
Although the DNA oligonucleotides bind with high specificity,
the triplex complexes formed are thermodynamically less stable
than the duplex complexes. This is partially due to the charge
repulsion resulting from bringing together the three polyanionic
DNA strands. Therefore, the ideal antisense/antigene agents
should have high affinity for DNA while still maintaining
fidelity of recognition, stability toward nucleases, and efficient
membrane permeability. To enhance the affinity of the duplex-
Figure 1. Chemical structures of DNA and DNG.
and triplex-forming oligonucleotide sequences for the target ss/
dsDNA sequences under physiological conditions, various
approaches have been taken to develop novel chemically
modified nucleotides.5 One approach is the incorporation of
neutral internucleoside linkages that eliminate mutual repulsions
between the negatively charged phosphodiester backbones.6-10
(1) (a) Giovannangeli, C.; Helene, C. Antisense Nucleic Acid Drug DeV. 1997,
7, 413. (b) Giovannangeli, C.; Perrouault, L.; Escude, C.; Thuong, N.;
Helene, C. Biochemistry 1996, 35, 10539. (c) Maher, L. J.; Dervan, P. B.;
Wold, B. Biochemistry 1992, 31, 70.
(2) (a) Wang, G.; Levy, D. D.; Seidman, M. M.; Glazer, P. M. Mol. Cell.
Biol. 1995, 15, 1759. (b) Wang, G.; Seidman, M. M.; Glazer, P. M. Science
1996, 271, 802.
(3) Broitman, S.; Amosova, O.; Dolinnaya, N. G.; Fresco, J. R. J. Biol. Chem.
1999, 274, 21763.
(4) Johnson, M. D., III; Fresco, J. R. Chromosoma 1999, 108, 181.
(5) (a) Uhlmann, E.; Peyman, A. Chem. ReV. 1990, 90, 543. (b) Milligan, J.
F.; Matteucci, M. D.; Martin, J. C. J. Med. Chem. 1993, 36, 1923. (c)
Kurreck, J. Eur. J. Biochem. 2003, 270, 1628.
(6) De Mesmaeker, A.; Waldener, A.; Lebreton, J.; Hoffmann, P.; Fritsch, V.;
Wolf, R. M.; Freier, S. M. Angew. Chem., Int. Ed. Engl. 1994, 33, 226.
(7) Levis, J. T.; Butler, W. O.; Tseng, B. Y.; Ts’o, P. O. P. Antisense Res.
DeV. 1995, 5, 251.
9
3736
J. AM. CHEM. SOC. 2004, 126, 3736-3747
10.1021/ja031557s CCC: $27.50 © 2004 American Chemical Society

Synthesis of DNG Tethering a Hoechst 33258 Analogue
A R T I C L E S
Another approach is to replace the entire phosphodiester
backbone, such as in the cases of peptide nucleic acid (PNA),11
phosphonic ester nucleic acids (PHONA),12 and nucleic acid
analogue peptide (NAAP).13 Recent studies have shown that
the introduction of positively charged groups at multiple sites
in the backbone,14,15 sugar,16 or base17 can produce stable
duplexes and triplexes.18 Our approach is to replace the
phosphodiester linkages with positively charged achiral guani-
dinium groups. Incorporation of positively charged guanidinium
groups (Figure 1) in the place of negatively charged phospho-
diester linkages greatly enhances the oligonucleotide complex
stability through charge-charge interactions.19
A number of well-characterized small molecule ligands such
as intercalators,20 polyamines,21 polyamides,22 and fluorescent
dyes23 are known to interact with duplex and triplex DNA and
enhance stability by providing additional interactions. Conjuga-
tion of these ligands to duplex- and triplex-forming oligonucle-
otides exhibited further enhanced stability of the duplex and
triplex complexes.24 Among these ligands, the bis-benzimidazole
derivatives bind with high affinity in the minor groove of
double-stranded B-DNA with a strong preference for A:T base
pairs. This binding results in enhanced helix stabilization and
also a tremendous enhancement in the observed fluorescence
emission of the ligand.25 These fluorescence properties have
been useful in a variety of applications such as determination
of A:T base pair content in DNA samples,26 determination of
Figure 2. Pentameric thymidyl DNG and Hoechst 33258-tethered DNGs.
cell numbers,27 and chromosomal sorting.28 The best-known
compound in the bis-benzimidazole family, the Hoechst 33258
dye’s assay sensitivity is approximately 1 ng/mL, and the
quantum yield effects are sensitive enough to detect one target
cell in a million mixed cell population.29 Furthermore, the
Hoechst 33258 dye is capable of crossing the cellular and
nuclear membranes and stain as fluorescent DNA and chromo-
somes.28 We have recently taken advantage of these character-
istics of Hoechst 33258 ligand and developed novel tripyrrole-
Hoechst conjugates.30 These conjugates are capable of passing
through NIH 3T3 cell membrane and inhibit a DNA-TF
complex formation by binding to its nuclear DNA targets.31
Although the DNA oligonucleotides conjugated to minor
groove binding ligands exhibited increased duplex and triplex
stability,22,23 the oligonucleotides must be of a reasonable length
to form stable duplexes and triplexes. We have recently shown
that a short sequence of DNG having positively charged
guanidinium backbone can form stable complexes by electro-
static attractions.19 Additionally, the guanidinium linkages are
resistant to nucleases,32 and also the positive charges of the
backbone may give rise to cell membrane permeability through
electrostatic attraction of the oligonucleotide to the negatively
charged phosphates of the cell surface. Complex stabilization
by groove binding as well as membrane permeability of Hoechst
derivatives suggests that tethering a fluorophore such as Hoechst
(8) Waldener, A.; De Mesmaeker, A.; Lebreton, J.; Fritsch, V.; Wolf, R. M.
Synlett 1994, 57.
(9) Morvan, F.; Sanghvi, Y. S.; Perbost, M.; Vasseur, J.-J.; Bellon. L. J. Am.
Chem. Soc. 1996, 118, 255.
(10) Wang, J.; Matteucci, M. D. Bioorg. Med. Chem. Lett. 1997, 7, 229.
(11) (a) Nielsen, P. E. Pure Appl. Chem. 1998, 70, 105. (b) Uhlmann, E.;
Peyman, A.; Breipohl, G.; Will, D. W. Angew. Chem., Int. Ed. 1998, 37,
2796.
(12) Peyman, A.; Uhlmann, E.; Wagner, K.; Augustin, S.; Breipohl, G.; Will,
D. W.; Schafer, A.; Wallmeier, H. Angew. Chem., Int. Ed. Engl. 1996, 35,
2636.
(13) Fujii, M.; Yoshida, K.; Hidaka, J. Bioorg. Med. Chem. Lett. 1997, 7, 637.
(14) Michel, T.; Debart, F.; Vasseur, J.-J. Tetrahedron Lett. 2003, 44, 6579.
(15) (a) Skibo, E. B.; Xing, C. Biochemistry 1998, 37, 15199. (b) Chaturvedi,
S.; Horn, T.; Letsinger, R. L. Nucleic Acids Res. 1996, 24, 2318. (c)
Letsinger, R. L.; Singman, C. N.; Histand, G.; Salunkhe, M. J. Am. Chem.
Soc. 1988, 110, 4470.
(16) Hichman, D. T.; Tan, T. H. S.; Morral, J.; King, P. M.; Cooper, M. A.;
Micklefield, J. Org. Biomol. Chem. 2003, 1, 3277.
(17) Ueno, Y.; Mikawa, M.; Matsuda, A. Bioconjugate Chem. 1998, 9, 33.
(18) For review see: Fox, K. R. Curr. Med. Chem. 2000, 7, 17.
(19) (a) Reddy, P. M.; Bruice, T. C. Bioorg. Med. Chem. Lett. 2003, 13, 1281.
(b) Linkletter, B. A.; Szabo, I. E.; Bruice, T. C. Nucleic Acid Res. 2001,
29, 2370. (c) Linkletter, B. A.; Szabo, I. E.; Bruice, T. C. J. Am. Chem.
Soc. 1999, 121, 3888. (d) Browne, K. A.; Dempcy, R. O.; Bruice, T. C.
Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 7051. (e) Dempcy, R. O.; Browne,
K. A.; Bruice, T. C. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 6097. (f)
Dempcy, R. O.; Browne, K. A.; Bruice, T. C. J. Am. Chem. Soc. 1995,
117, 6140.
(20) Thoung, N. T.; Asseline, U.; Moneney-Garestier, T. In Oligonucleotides.
Anti-Sense Inhibitors of Gene Expression; Cohen, J. S., Ed.; CRC Press:
Boca Raton, FL, 1989; pp 25-52.
(25) (a) Zimmer, C.; Wahnert, U. Prog. Biophys. Mol. Biol. 1986, 47, 31. (b)
Lootiens, F. G.; Regenfuss, P.; Zechel, A.; Dumortier, L.; Clegg, R. M.
Biochemistry 1990, 29, 9029.
(21) Lemaitre, M.; Bayard, B.; Lebleu, B. Proc. Natl. Acad. Sci. U.S.A. 1987,
84, 648.
(26) (a) Weisblum, B.; Haenssler, E. Chromosoma 1974, 46, 255. (b) Coming,
D. E. Chromosoma 1975, 52, 229. (c) Sterzel, W.; Bedford, P.; Eisenbrand,
G. Anal. Biochem. 1985, 147, 462. (d) Araki, T.; Yamamoto, A.; Yamada,
M. Histochemistry 1987, 87, 331. (e) Karawajew, L.; Rudchenko, S.;
Wlasik, T.; Trakht, I. J. Immunol. Methods 1990, 129, 277.
(27) (a) Downs, T. R.; Wilfinger, W. W. Anal. Biochem. 1983, 131, 538. (b)
Adams, C. J.; Storrie, B. Histochem. Cytochem. 1981, 29, 326.
(28) (a) Holmquist, G. Chromosoma 1975, 49, 333. (b) Arndt-Jovin, D. J.; Jovin,
T. M. Cytometry 1990, 11, 80. (c) Frau, S.; Bernadou, J.; Meunier, B. Bull.
Soc. Chim. Fr. 1996, 133, 1053.
(22) (a) Szewczyk, J.; Baird, E. E.; Dervan, P. B. Angew. Chem., Int. Ed. Engl.
1996, 35, 1487. (b) Sinyakov, A. N.; Lokhov, S. G.; Kutyavin, I. V.;
Gamper, H. B.; Meyer, R. B. J. Am. Chem. Soc. 1995, 117, 4995.
(23) (a) Robles, J.; Rajur, S. B.; McLaughlin, L. W. J. Am. Chem. Soc. 1996,
118, 5820. (b) Wiederholt, K.; Rajur, S. B.; Giuliano, J., Jr.; O’Donnell,
M. J.; McLaughlin, L. W. J. Am. Chem. Soc. 1996, 118, 7055. (c) Robles,
J.; McLaughlin, L. W. J. Am. Chem. Soc. 1997, 119, 6014. (d) Rajur, S.
B.; Robles, J.; Wiederholt, K.; Kuimelis, R. G.; McLaughlin, L. W. J. Org.
Chem. 1997, 62, 523.
(24) Helix-stabilizing ligands: (a) Mergny, J. L.; Duval-Valentin, G.; Nguyen,
C. H.; Perrouault, L.; Faucon, B.; Rougee, M.; Montenay-Garestier, T.;
Bisagni, E.; Helene C. Science 1992, 256, 1681. (b) Lee, J. S.; Latimer, L.
J. P.; Hampel, K. J. Biochemistry 1993, 32, 5591. (c) Wilson, W. D.;
Tanious, F. A.; Mizan, S.; Yao, S.; Kiselyov, A. S.; Zon, A. S.; Strekowski,
G. L. Biochemistry 1993, 32, 10614. (d) Fox, K. R.; Polucci, P.; Jenkins,
T. C.; Neidle, S. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 7887. (e) Escude,
C.; Nguyen, C. H.; Kukreti, S.; Janin, Y.; Sun, J. S.; Bisagni, E.; Garestier
T.; Helene, C. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 3591.
(29) Lee, B. R.; Haseman, D. B.; Reynolds, C. P. Cytometry 1989, 10, 256.
(30) (a) Satz, A. L.; Bruice, T. C. J. Am. Chem. Soc. 2001, 123, 2469. (b) Satz,
A. L.; Bruice, T. C. Bioorg. Med. Chem. 2002, 10, 241. (c) Reddy, P. M.;
Jindra, P. T.; Satz, A. L.; Bruice, T. C. J. Am. Chem. Soc. 2003, 125, 7843.
(31) (a) White, C. M.; Satz, A. L.; Bruice, T. C.; Beerman, T. A. Proc. Natl.
Acad. Sci. U.S.A. 2001, 98, 10590. (b) White, C. M.; Satz, A. L.; Gawron,
L. S.; Bruice, T. C.; Beerman, T. A. Biochim. Biophys. Acta 2001, 1574,
100.
(32) Barawkar, D. A.; Bruice, T. C. Proc. Natl. Acad. Sci. U.S.A. 1998, 95,
11047.
9
J. AM. CHEM. SOC. VOL. 126, NO. 12, 2004 3737

A R T I C L E S
Reddy and Bruice
Figure 3. Proposed models for DNA‚DNG-H duplex and (DNA)2‚DNG-H triplex complexes formed by DNA and DNG-Hoechst 33258 conjugates. The
ssDNA and dsDNA are represented in yellow, DNG is represented in magenta, Hoechst dye is represented in spheres, and the linker is represented in blue.
33258 to the 5′-terminus of the positively charged DNG could
further enhance the stability of DNA‚DNG duplex and (DNA)2‚
DNG triplex complexes by simultaneous minor groove binding.
Furthermore, these DNG-Hoechst conjugates might be better
able to cross the cellular and nuclear membranes than the
unconjugated DNG sequences and bind to its nuclear DNA
targets. More importantly, these conjugates, which would grasp
dsDNA through both major and minor groove binding, are also
promising for development of transcription factor inhibitors.31,33
In this study, we describe the synthesis of the modified
monomers for the DNG solid-phase synthesis (SPS) and the
Hoechst 33258 derivative containing a linker that allows
covalent attachment to the 5′-terminus of the DNG sequence.
We report on the covalent conjugation of such agents to DNG,
the duplex and triplex stabilization, and the fluorescent properties
resulting from simultaneous minor groove binding of these
DNG-Hoechst 33258 conjugates (1-3, Figure 2).
benzimidazole ligand to the polycationic DNG strand could
greatly increase the already strong binding of DNG to DNA.
DNG-Hoechst 33258 conjugates should bind sequence-specif-
ically to a dsDNA target site to form a local (DNA)2‚DNG-H
triplex in which the DNG strand would be located in the major
groove and the Hoechst ligand would occupy the minor groove
of the target duplex (Figure 3). On the other hand, the same
DNG-Hoechst conjugates can also form stable DNA‚DNG-H
duplexes by binding sequence-specifically to the targeted
ssDNA. The tethered Hoechst moiety can further stabilize the
DNA/DNG duplex by folding back into the generated minor
groove, thereby providing additional binding interactions (Figure
3). For a (DNA)2‚DNG-H triplex to benefit from both modes
of binding, the linker used for tethering the ligand to the DNG
sequence must be long enough to reach the minor groove from
the 5′-terminus of the DNG third strand in the major groove by
traversing the phosphoribose backbone. However, in the case
of DNA‚DNG-H duplex, a shorter linker may suffice to fold
back and permit the tethered ligand to reach the minor groove
of DNA‚DNG duplex. Also, the minor groove of DNA‚DNG
duplex will not be the same as that of DNA‚DNA duplex.35 To
evaluate these considerations, we have synthesized DNG-
Hoechst 33258 conjugates with 11 and 18 atom linkers
(compounds 2 and 3 respectively).
The synthetic strategy we have developed for the DNG SPS
involves the coupling of 3′-Fmoc-protected thiourea in the
presence of HgCl2/TEA with the corresponding 5′-NH2 of the
growing oligo chain on long chain alkylamine-derivatized
controlled pore glass (CPG).19a,b The DNG-Hoechst 33258
conjugates (2 and 3) were also synthesized on solid phase by
stepwise incorporation of an amino acid straight chain linker
and a Hoechst acid into the 5′-NH2 of the DNG using PyBOP/
HOBt chemistry. To facilitate the stepwise synthesis of DNG
and DNG-Hoechst conjugates, the required monomers 4, 8,
10, and 15 were synthesized as described in Schemes 1-4 .
Synthesis of Monomers. The loading monomer, 5′-mono-
methoxytritylamino-2′,5′-dideoxythymidine was synthesized
Results and Discussion
In the antisense technology for controlling translation, a
designed oligonucleotide binds sequence-specifically to the
mRNA by Watson-Crick hydrogen bonds.5 Antisense oligo-
nucleotide sequences generate a minor groove similar to dsDNA
upon hybridization to the complementary target sequence. On
the other hand, in the antigene technology for controlling gene
transcription, a designed oligonucleotide binds to the major
groove of the dsDNA by Hoogsteen or reverse Hoogsteen
hydrogen bonds and forms a local triple helix.34 In DNA triplex,
in which the third strand occupies the major groove, the minor
groove remains largely unencumbered. A variety of studies have
shown that tethering a minor groove binding ligand such as
Hoechst 33258 or polyamides to the 5′-terminus of the duplex-
and triplex-forming strand enhances the duplex and triplex
stability by binding simultaneously in the minor groove.22,23 On
the basis of these results and the strong affinity of DNG toward
target DNA, we envisaged the possibility that tethering a bis-
(33) Arya, D. P.; Willis, B. J. Am. Chem. Soc. 2003, 125, 12398.
(34) Thuong, N. T.; Helene, C. Angew. Chem., Int. Ed. Engl. 1993, 32, 666.
(35) Luo, J.; Bruice, T. C. J. Am. Chem. Soc. 1998, 120, 1115.
9
3738 J. AM. CHEM. SOC. VOL. 126, NO. 12, 2004

Synthesis of DNG Tethering a Hoechst 33258 Analogue
A R T I C L E S
Scheme 1. CPG Loaded 5′-Modified Monomer
was determined spectrophotometrically from the amount of
MMTr cation released.
The 3′,5′-diamino protected thymidyl building block 8,
required for the coupling reaction to make the guanidinium
backbone, was accomplished from 3′-OH inverted 519a,37
(Scheme 2). The 3′-OH inverted thymidine 5 was readily
converted into 3′,5′-dimesyl derivative 6 in quantitative yield
by reacting with mesyl chloride in pyridine. Treatment of 6 with
LiN3 in DMF at 80 °C resulted in its 3′,5′-diazido derivative,
which on further reduction with 10% Pd/C gave 3′,5′-diamino-
2′,3′,5′-trideoxythymidine 7.19a,38 Selective protection of 5′- and
3′-NH2 groups of 7 with acid-labile MMTr and base-cleavable
Fmoc groups gave 8. Initially, the 5′-NH2 was selectively
protected by reacting with 1 equiv of MMTrCl in the presence
of TEA, and then the 3′-NH2 was protected by the addition of
Fmoc-NCS39 to afford the desired coupling monomer 8 (Scheme
2).
Scheme 2 a
To achieve the stepwise synthesis of DNG-Hoechst conju-
gates with different linker lengths, we chose 10 and 15 as
building blocks (Schemes 3 and 4). The 6-monomethoxytrityl-
amino-hexanoic acid (10) was prepared easily, in ∼95% yield,40
through the reaction of commercially available 6-aminohexanoic
acid (9) with MMTrCl in dry pyridine (Scheme 3). The Hoechst
acid 15 was synthesized as described in Scheme 4. Benzyl 4-(4-
formylphenoxy)butanoate (12) was synthesized in 98% yield
from benzyl 4-bromobutyrate41 (11) by reacting with 4-hy-
droxybenzaldehyde in the presence of Cs2CO3 in N,N-dimethyl-
acetamide (DMA). Condensation of 12 with ortho-diamine 1342
in nitrobenzene at 130 °C gave benzyl ester of Hoechst acid 14
in 87% yield. Further, hydrogenation at 50 psi using 10% Pd/C
conveniently afforded 15 in quantitative yield.
a Reagents and conditions: (a) CH3SO2Cl 5.0 equiv, pyridine, 0 °C.
Room temperature, overnight. (b) LiN3 20.0 equiv, DMF, 80 °C, 4 h. (c)
10% Pd/C, H2, EtOH, 2 h. (d) MMTrCl 1.0 equiv, TEA 2.0 equiv, DCM,
2 h. (e) Fmoc-NCS, 1.2 equiv, DCM, room temperature, 2 h.
Scheme 3 a
a Reagents and conditions: (a) MMTrCl 1.2 equiv, pyridine, room
temperature, 6 h.
DNG (1) Solid-Phase Synthesis. The pentameric thymidyl
DNG 1 was synthesized in a stepwise manner on solid-phase
using CPG-derivatized 4 and 3′,5′-protected monomer 8 (Scheme
5). The synthesis proceeds in a 3′ f 5′ direction that is
compatible with the cleavage conditions used in the DNA SPS.
Coupling of 8 with MMTr deprotected 4 for the formation of
the guanidinium linkage was accomplished in the presence of
HgCl2 and TEA. Treatment of 8 with HgCl2 in the presence of
TEA converts the 3′-Fmoc-protected thiourea into an intermedi-
ate carbodiimide,19a,b which reacts in situ with the 5′-NH2 of
CPG-loaded monomer to provide an Fmoc-protected guani-
dinium linkage (Scheme 5). This Fmoc protecting group remains
in place on the guanidinium linkage until the end of the SPS
when it is readily removed during cleavage of the oligomer from
the CPG. The unreacted 5′-NH2 sites were blocked by capping
reaction, rendering them inert toward further chain extension.
The terminal 5′-MMTr group was removed, and the coupling
yield for this step was 99% as determined by UV absorbance
of the released MMTr cation. The whole coupling cycle
(coupling/capping/deprotection) was repeated three more times
Scheme 4 a
a Reagents and conditions: (a) 4-hydroxybenzaldehyde, Cs2CO3, anhy-
drous DMA, 100 °C, 15 h. (b) Nitrobenzene, 130 °C, 24 h. (c) 10% Pd/C,
H2, EtOH, 4 h.
(37) Pathak, A. K.; Pathak, V.; Seitz, L. E.; Tiwari, K. N.; Akhtar, M. S.;
Reynolds, R. C. Tetrahedron Lett. 2001, 42, 7755.
(38) Herdewijn, P.; Balzarini, J.; Pauwels, R.; Janssen, G.; Van Aerschot, A.;
De Clercq, E. Nucleosides Nucleotides 1989, 8, 1231.
(39) Kearney, P. C.; Fernandez, M.; Flygare, J. A. J. Org. Chem. 1998, 63,
196.
(40) Berube, G.; Richardson, V. J.; Ford, C. H. J. Synth. Commun. 1991, 21,
931.
(41) Baba, A.; Kawamura, N.; Makino, H.; Ohta, Y.; Taketomi, S.; Sohda, T.
J. Med. Chem. 1996, 39, 5176.
from 5′-amino-2′-deoxythymidine36 and loaded on to the CPG
solid support via succinyl linker (Scheme 1) to afford the CPG-
derivatized monomer 419a (see Supporting Information). The
unreacted CPG amine sites were covered by capping with acetic
anhydride/TEA, and then the 5′-MMTr was deprotected with
3% DCA in DCM solution. The loading yield, 39.6 µmol/g,
(42) (a) Sadat Ebrahimi, S. E.; Bibby, M. C.; Fox, K. R.; Douglas, K. T. Anti-
Cancer Drug Des. 1995, 10, 463. (b) Argentini, M.; Dos Santos, D. F.;
Weinreich, R.; Hansen, H.-J. Inorg. Chem. 1998, 37, 6018.
(36) Horwitz, J. P.; Tomson, A. J.; Urbanski, J. A.; Chua, J. J. Org. Chem.
1962, 27, 3045.
9
J. AM. CHEM. SOC. VOL. 126, NO. 12, 2004 3739

A R T I C L E S
Reddy and Bruice
Scheme 5. Solid-Phase Synthetic Scheme for Pentameric
Scheme 6. Solid-Phase Synthetic Scheme for DNG-Hoechst
Thymidyl DNGa
Conjugatesa
a Reagents and conditions: (a) Capping: (CH3CO)2O, TEA, DMF, 10
min. (b) Deprotection: 3% DCA in DCM, 1 min. (c) Coupling: monomer
8, HgCl2, TEA, DMF, 2 h, then 20% PhSH in DMF, 1 min. (d) Methanolic
ammonia, room temperature, 2 h. (e) 3% DCA in DCM, 1 min.
a Reagents and conditions: (a) Capping: (CH3CO)2O, TEA, DMF, 10
min. (b) Deprotection: 3% DCA in DCM, 1 min. (c) Coupling: monomer
10, PyBOP, HOBt, DIPEA, DMF, 12 h. (d) Ht acid 15, PyBOP, HOBt,
DIPEA, DMF, 24 h. (e) 0.1 M NaOH in 4:1 MeOH/H2O, 1 h. Ht ) Hoechst
33258; g ) guanidinium group.
with 8 to afford the desired DNG pentamer 17. The coupling
yield in each cycle was 95-99%, and the overall yield of the
5′-MMTr-protected oligomer was therefore expected to be
∼85%. The DNG pentamer was cleaved from CPG using a
methanolic ammonia solution. The Fmoc protection on guani-
dinium linkages was also removed in the same step. The crude
“trityl on” product was purified on reverse-phase HPLC (C8
column) using solvent A (0.1 M TEAA buffer, pH 7.0) and a
gradient of solvent B (acetonitrile) 5% f 80% in 30 min. ESI/
TOF+ analysis of MMTr-protected 1 exhibited desired peaks
at m/z 787.87 (M + 2H) and 525.58 (M + 3H); calcd 787.85
(M + 2H) and 525.56 (M + 3H) for C74H91N23O17. The 5′-
MMTr of oligomer was deprotected with 3% DCA in DCM
solution and precipitated with excess of ether. The precipitated
product was collected by centrifugation, and analysis by RP-
HPLC using the same column and solvent system revealed
single peak for 1. ESI/TOF+ analysis exhibited expected peaks
at m/z 1302.72 (M + H) and 651.85 (M + 2H); calcd 1302.58
(M + H) and 651.79 (M + 2H) for C54H75N23O16.
Synthesis of DNG-Hoechst Conjugates 2 and 3. The novel
DNG-Hoechst conjugates 2 and 3 were synthesized on solid-
phase by stepwise addition of 10 and 15 to the 5′-terminus of
the thymidyl pentamer 17 using PyBOP/HOBt chemistry
(Scheme 6). After capping any unreacted 5′-NH2 sites of 17,
the 5′-MMTr was removed to facilitate attachment of the amino
acid straight chain linker through amide bond formation. The
coupling reaction of 10 with MMTr-removed 17 was ac-
complished in the presence of PyBOP, HOBt, and DIPEA in
DMF to afford 18. Unreacted sites were capped with acetic
anhydride/TEA, and then the MMTr substituent was removed
using 3% DCA in DCM. The removal of MMTr on linker took
a little longer than that of the 5′-MMTr of oligomer. The beads
were treated with DCA solution for 1 min, filtered off, and
washed with DCM, and the DCA treatment was repeated twice
more. This procedure completely removes the MMTr moiety.
The extent of the linker coupling reaction was determined to
be 100% by UV absorbance of the released MMTr cation. The
CPG was split into two portions, and one portion was coupled
to 15 using PyBOP, HOBt, and DIPEA to afford the DNG-
Hoechst conjugate 20 with an 11-atom linker. For the second
portion, the coupling/capping/deprotection reaction cycle with
10 was repeated one more time before Hoechst acid 15 was
finally coupled to afford the conjugate 21 with an 18-atom
linker. The CPG beads of conjugates 20 and 21 fluoresced under
long-wave UV light, indicating the successful addition of the
Hoechst acid 15 onto the linker. Conjugates 20 and 21 were
cleaved from CPG under very mild cleavage conditions. Direct
assault with NH4OH or methanolic ammonia solution resulted
in the cleavage of conjugates from CPG, as well as the cleavage
of linker between DNG and Hoechst ligand. Hence, we
performed the deprotection and cleavage of DNG-Hoechst
conjugates using 0.1 M NaOH solution in 4:1 methanol/water
at room temperature for 1h, in which a majority of the conjugate
was intact as determined by HPLC. The Fmoc protection on
the guanidinium groups was also removed in the same step.
After cleavage, the solution was desalted using NAP-10 columns
and HPLC grade water as the solvent. The eluted solutions were
turbid, possibly due to the aggregation of the conjugate. The
Hoechst 33258 is known to aggregate in aqueous solutions at
∼30 µM concentrations.25b Addition of 0.01% TFA in water
afforded a clear solution that fluoresces under long-wave UV.
Purification of crude conjugates was accomplished on RP-HPLC
(C8 column) using solvent A (0.1% TFA in water) and a gradient
of solvent B (acetonitrile) 5% f 80% in 30 min. Both
conjugates 2 and 3 were eluted as broad peaks at ∼14 min.
ESI/TOF+ analysis of conjugate 2 exhibited desired peaks at
m/z 954.57 (M + 2H), 636.70 (M + 3H), and 477.79 (M +
4H); calcd 954.45 (M + 2H), 636.63 (M + 3H), and 477.72
(M + 4H) for C89H114N30O19. Conjugate 3 exhibited peaks at
9
3740 J. AM. CHEM. SOC. VOL. 126, NO. 12, 2004

Synthesis of DNG Tethering a Hoechst 33258 Analogue
A R T I C L E S
Table 1. Thermal Melting (Tm) and Fluorescence Emission Values
of Triplexes Formed by dsDNA and DNG-Hoechst 33258
Conjugatesa
Tm
(
°C) at pH
XXXXX
YYYYY
DNG
6.0
7.0
7.0
8.0
b
entry
conjugate
(0.1 M)
(0.1 M)
(0.7 M)
(0.1 M)
F450
1
AAAAA
TTTTT
AAAAA
TTTTT
AAAAA
TTTTT
AAGAA
TTCTT
AAGAA
TTCTT
AAGAA
TTCTT
AATAA
TTATT
AATAA
TTATT
AATAA
TTATT
1
2
3
1
2
3
1
2
3
38
53
52
-
-
-
-
-
-
-
36
49
49
25
26
26
26
43
44
ntc
32
35
46
46
-
-
-
-
-
-
-
-
246
248
-
2
3
42
41
-
-
-
-
-
-
-
4
Figure 4. Normalized triplex Tm curves for 30-mer dsDNA and dsDNA
+ 1 equiv of 1 or 2 or 3 in 10 mM KHPO4 buffer, pH 7.0, containing 100
mM KCl.
5
34
6
34
m/z 1011.04 (M + 2H), 674.35 (M + 3H), and 506.26 (M +
4H); calcd 1010.99 (M + 2H), 674.32 (M + 3H), and 505.99
(M + 4H) for C95H125N31O20.
7
-
8
254
262
Thermal Melting (Tm) Stability of (DNA)2‚DNG-H Tri-
plex Complexes. The stability of triplexes formed by DNG and
DNG-Hoechst conjugates was assessed by examining absor-
bance vs temperature plots. We chose a 30-mer DNA duplex,
d(CGCCGCGCGCGCGAAAAACCCGGCGCGCGC)/d(GCG-
GCGCGCGCGCTTTTTGGGCCGCGCGCG), containing the
pentameric AAAAA/TTTTT tract at the center as the target site.
The sequence on either side of the target site is composed solely
of G:C base pairs, such that the triplex formation by DNG and
minor groove-binding by Hoechst ligand could occur only at
the A5/T5 tract. The binding by Hoechst 33258 ligand in a
dsDNA minor groove requires at least a four base pair A:T rich
site.43-45 The pentameric thymidyl DNG (1) forms a stable
(DNA)2‚DNG triplex (Tm ) 36 °C) at the dsDNA target site
due to charge-charge interactions. Furthermore, the Hoechst
33258-tethered DNG (conjugates 2 and 3) exhibited significant
triplex stabilization compared to the DNG lacking a pendant
Hoechst dye (Figure 4). Conjugates 2 and 3 formed stable
(DNA)2‚DNG-H triplexes and exhibited a Tm of 49 °C and a
∆Tm of +13 °C compared with (DNA)2‚DNG triplex formed
by DNG 1, in a 10 mM potassium phosphate buffer, pH 7.0,
containing 100 mM KCl (Table 1). The presence of a five base
pair A:T rich site in the target dsDNA (entries 1-3, Table 1)
provides effective binding site for the tethered ligand. Previous
reports indicate that binding of Hoechst 33258 dye somewhat
destabilizes the DNA triplex structure while stabilizing duplex.46
However, our results show that covalent conjugation of Hoechst
33258 dye with triplex-forming DNG stabilizes both (DNA)2‚
DNG-H triplex and target duplex due to the simultaneous
binding of the DNG-Hoechst 33258 conjugate in the major
and minor grooves of the target dsDNA site. These results are
consistent with the results exhibited by DNA-Hoechst conju-
9
10
AGAGA 1 or 2 or 3
TCTCT
nfd
a
Tm studies were carried out in 10 mM KHPO4 buffer, pH 6.0, 7.0, or
8.0 containing 100 or 700 mM KCl as mentioned. Tm values were
determined by first derivative analysis, and standard deviations are (1 °C.
The target dsDNA Tm values are 84-86 °C (pH 7.0 and 0.1M KCl);
however, increased Tm (90-92 °C) values were observed in triplex state
due to the binding of Hoechst moiety in the minor groove or salt
b
concentration. Values represent the fluorescence emission enhancement
(in arbitrary units) at 450 nm for the 50 nM triplex complexes in 10 mM
KHPO4 buffer pH 7.0 containing 100 mM KCl. No triplex observed.d
c
No fluorescence enhancement.
gates.23 The change in the tether length from 18 to 11 atoms
did not have any significant effect on triplex Tm (entries 2 and
3, Table 1), which indicates that the 11-atom linker is sufficient
to traverse the backbone, permitting the Hoechst ligand to reside
deep into the A:T rich minor groove. A significant decrease in
triplex Tm values was observed with an increase in salt
concentration (Table 1), which is attributed to a reduction in
the electrostatic attraction between the oppositely charged
backbones. Upon increasing the salt concentration from 100 mM
to 700 mM KCl, while maintaining pH 7.0, the (DNA)2‚DNG
triplex Tm was decreased by 4 °C, whereas the (DNA)2‚DNG-H
triplex Tm was decreased by 6-7 °C. Conversely, the Tm of
dsDNA was increased as expected (Figure 5). The stability of
(DNA)2‚DNG triplex complex formed by DNG (1) and dsDNA
is almost independent of pH. However, the stabilities of (DNA)2‚
DNG-H triplex complexes formed by conjugates 2 and 3 with
dsDNA are slightly decreased with the increase of pH (entries
2 and 3, Table 1). In (DNA)2‚DNG-H triplex complexes, the
ligand may be binding more effectively in the dsDNA minor
groove at lower pH due to the protonation of benzimidazoles47
and interaction with the phosphodiester backbone. As anticipated
from previous studies of bis-benzimidazoles,47 the conjugates
2 and 3 exhibited significant increase in fluorescence emission
with decreasing pH 8 to 6. The dependence of binding of 2 and
(43) (a) van Dyke, M. W.; Hertzberg, R. P.; Dervan, P. B. Proc. Natl. Acad.
Sci. U.S.A. 1982, 79, 5470. (b) Harshman, K. D.; Dervan, P. B. Nucleic
Acids Res. 1985, 13, 4825.
(44) Portugal, J.; Waring, M. J. Biochim. Biophys. Acta 1988, 949, 158.
(45) (a) Abu-Daya, A.; Brown, P. M.; Fox, K. R. Nucleic Acids Res. 1995, 23,
3385. (b) Fox, K. R.; Waring, M. J. Nucleic Acids Res. 1984, 12, 9271.
(46) (a) Durand, M.; Thuong, N. T.; Maurizot, J. C. Biochimie 1994, 76, 181.
(b) Kim, H.-K.; Kim, J.-M.; Kim, S. K.; Rodger, A.; Norden, B.
Biochemistry 1996, 35, 1187.
(47) Gorner, H. Photochem. Photobiol. 2001, 73, 339.
9
J. AM. CHEM. SOC. VOL. 126, NO. 12, 2004 3741

A R T I C L E S
Reddy and Bruice
Figure 5. Normalized triplex Tm curves formed by 2 µM 30-mer dsDNA
and 1 equiv of 3 in 10 mM KHPO4 buffer, pH 7.0, containing 0.1 M KCl
and 0.7 M KCl.
Figure 7. Change in fluorescence emission spectrum with temperature of
a 50 nM solution of triplex (formed from 30-mer dsDNA and conjugate 3)
in 10 mM KHPO4 buffer, pH 7.0, containing 100 mM KCl.
for DNG 1, whereas the (DNA)2‚DNG-H triplex Tm values
decreased by 23 °C (entries 2 and 5 or 3 and 6, Table 1) for
conjugates 2 and 3, respectively. However, inversion of an A:T
base pair into a T:A base pair in the center of the binding site
(entries 7-9, Table 1) decreased the (DNA)2‚DNG triplex Tm
by 10 °C (entries 1 and 7, Table 1), whereas the (DNA)2‚
DNG-H triplex Tm values decreased only 5-6 °C (entries 2
and 8 or 3 and 9, Table 1). It is noteworthy that both DNG (1)
and DNG-Hoechst conjugates (2 and 3) exhibited essentially
the same triplex Tm (25-26 °C, Table 1) values for the G:C
mismatched dsDNA sequence. On the other hand, the DNG-
Hoechst conjugates (2 and 3) exhibited enhanced triplex Tm
(42-43 °C, Table 1) for the T:A mismatched dsDNA sequence.
Hoechst 33258 has a stronger affinity for A:T- than G:C-
containing dsDNA sequences. Incorporation of a single G:C
base pair in the binding site generates a mismatch sequence for
both DNG and Hoechst moieties. The inability of the tethered
Hoechst ligand binding tightly in the mismatched minor groove
prevents the stabilization of the triplex structure. As a result,
triplexes formed by both DNG and DNG-Hoechst conjugates
with G:C mismatch-containing dsDNA sequence exhibited
similar Tm values (entries 4-6, Table 1). Inversion of an A:T
base pair into a T:A base pair selectively generates a mismatch
sequence for the DNG moiety, but serves as a favorable binding
site for the Hoechst moiety. Therefore, the tethered Hoechst
moiety is able to bind tightly in the A:T rich minor groove,
which enhances the (DNA)2‚DNG-H triplex Tm by 16 to 17
°C (entries 7 and 8 or 7 and 9). With two G:C base pair
mismatches (entry 10, Table 1) there is no triplex formation.
Fluorescence Characteristics of (DNA)2‚DNG-H Triplex
Complexes. In addition to the enhanced triplex stability, as
determined by Tm studies, the fluorescence emission of DNG-
Hoechst conjugates was increased greatly on complexation with
the target dsDNA. When excited at 345 nm, the (DNA)2‚
DNG-H triplexes formed by conjugates 2 and 3 emit a broad
fluorescent signal centered at 450 nm (Figure 7). This signal is
consistent with the dsDNA/Hoechst 33258 complexes, which
generally emit a broad fluorescence signal centered at 445
nm.49-51 To confirm that the observed fluorescence signal was
due to binding of tethered Hoechst 33258 fluorophore in the
minor groove of target dsDNA, we examined the temperature
effects of the fluorescence emission spectra. The fluorescence
Figure 6. pH vs ∆G25° (standard free energy of binding) plot. The ∆G25°
values for (DNA)2‚DNG-H formation were calculated from Tm curves.47
3 to dsDNA is best examined by a plot of the standard free
energy of complex formation (∆G25°) vs pH (Figure 6). The
∆G25° values for (DNA)2‚DNG-H triplex formation were
calculated from Tm curves.48 Essentially, the same plot serves
for both conjugates 2 and 3. If the equilibrium constant for
triplex formation required a single proton, the plot would be
linear with a slope of -1. The apparently linear plot has a slope
of -0.42 between pH 6 and 8. The Hoechst moiety has pKa
values at 5.5 and 8.5. Thus, the pH -∆G25° profile tends to
flatten somewhat between pH 6 and 8. We may conclude that
the diprotonated species binds in the minor groove better than
the monoprotonated species.
To analyze the sequence selectivity in dsDNA recognition
by triplex forming DNG and DNG-Hoechst conjugates, we
have investigated the triplex formation between DNG conjugates
(1-3) and dsDNA containing one or two mismatched base pairs
in the target site. As can be seen from triplex Tm values in Table
1, a single base pair mismatch (entries 4-9) in the center of
the target dsDNA site exhibited a dramatic decrease in the triplex
stability. Incorporation of a G:C base pair mismatch in the center
of the target dsDNA site (entries 4-6, Table 1) decreased the
(DNA)2‚DNG triplex Tm by 11 °C (entries 1 and 4, Table 1)
(48) (a) Marky, L. A.; Breslauer, K. J. Biopolymers 1987, 26, 1601. (b) Gralla,
J.; Crothers, D. M. J. Mol. Biol. 1973, 78, 301.
9
3742 J. AM. CHEM. SOC. VOL. 126, NO. 12, 2004

Synthesis of DNG Tethering a Hoechst 33258 Analogue
A R T I C L E S
Table 2. Thermal Melting (Tm) and Fluorescence Emission Values
of Duplexes Formed by DNA and DNG-Hoechst Conjugatesa
d
entry
DNA
DNG conjugate
Tm
(°
C)
F475
1
2
3
4
5
6
7
8
9
10
11
12
13
14
AAAAA
AAAAA
AAAAA
AACAA
AACAA
AACAA
CAAAA
CAAAA
CAAAA
AAAAC
AAAAC
AAAAC
30-merb
1
2
3
1
2
3
1
2
3
1
2
3
2
3
43
45
45
35
36
36
38
39
39
36
38
38
-
24
26
-
nfe
nfe
-
nfe
nfe
-
nfe
nfe
50c
50c
278
272
30-merb
a
Tm studies were carried out at 6 µM concentrations in 10 mM KHPO4
buffer, pH 7.0, containing 100 mM KCl. Tm values were determined by
first derivative analysis, and standard deviations are (1 °C. The DNA
Figure 8. Fluorescence change vs temperature plot of a 50 nM triplex
(formed by conjugate 3 and 30-mer dsDNA) in 10 mM KHPO4 buffer, pH
7.0, containing 100 mM KCl. Inset: absorbance vs temperature plot for
the same triplex.
b
oligomers are all pentamers unless stated otherwise. 30-mer ssDNA
c
sequence is 5′-CGCCGCGCGCGCGAAAAACCCGGCGC GCGC-3′. Tm
d
determined from fluorescence vs temperature plot. Values represent the
fluorescence emission enhancement (in arbitrary units) at 475 nm for the
560 nM duplex complexes in 10 mM KHPO4 buffer, pH 7.0, containing
100 mM KCl. No fluorescence enhancement observed.
signal decreases with increase in temperature, and the change
becomes dramatic in the DNA duplex Tm (∼84 °C) range
(Figure 7). An examination of fluorescence vs temperature plot
(Figure 8) for this (DNA)2‚DNG-H triplex complex reveals
that there is only a marginal decrease in the fluorescence
emission until ∼70 °C, that is even after the DNG third strand
was separated from the major groove (triplex Tm ) 49 °C).
However, the fluorescence emission dropped tremendously upon
reaching the duplex DNA Tm (∼84 °C) range, which indicates
that the Hoechst fluorophore is binding solely in the Watson-
Crick minor groove of the target dsDNA and not anywhere else
in the (DNA)2‚DNG-H triplex region.
e
The fluorescence emission of the DNG-Hoechst conjugates
increased by 246-248 (arbitrary) units upon hybridization with
the dsDNA containing the five A:T base pair binding site (entries
2 and 3, Table 1). Introduction of a G:C base pair mismatch in
the center of the target dsDNA sequence (entries 4-6, Table
1) exhibited fluorescence emission enhancement to only 34 units
due to the inhibition of Hoechst fluorophore’s binding. No
fluorescence emission enhancement was observed in the pres-
ence of two G:C base pair mismatches in the binding site (entry
10, Table 1). This indicates that Hoechst fluorophore of
conjugates 2 and 3 is not binding in the G:C mismatch-
containing minor groove site and is therefore unable to enhance
the (DNA)2‚DNG triplex stability as observed in the triplex Tm
study (entries 4-6 and 10, Table 1). However, the triplex
complexes formed by T:A mismatch-containing sequence
exhibited fluorescence emission enhancement to 254-262 units
(entries 8 and 9, Table 1) similar to that of triplex complexes
formed by nonmismatch sequence (entries 2 and 3, Table 1).
Furthermore, the temperature vs fluorescence behavior of these
triplex complexes is consistent with the triplex complexes
formed by nonmismatch dsDNA sequence. These observations
support the assertion (Figure 3) that complexation of DNG-
Hoechst 33258 conjugates with dsDNA target site results in
the formation of a local (DNA)2‚DNG-H triplex in which the
Figure 9. Normalized duplex Tm curves formed by ssDNA and 1 equiv of
1 or 2 or 3 in 10 mM KHPO4 buffer, pH 7.0, containing 100 mM KCl.
DNG strand occupies the major groove and Hoechst ligand binds
in the Watson-Crick minor groove simultaneously.
Thermal Melting (Tm) Stability of DNA‚DNG-H Duplex
Complexes. The DNA‚DNG duplex is stabilized by electrostatic
attractions.19 The generated minor groove of DNA‚DNG duplex
is slightly different35 from that of the minor groove of dsDNA,
and the binding properties of Hoechst ligand are unknown. To
determine the effect of tethered Hoechst ligand on the stability
of DNA‚DNG duplex by binding in the generated minor groove,
we have examined thermal melting characteristics when the
DNG is lacking or tethering a pendant Hoechst ligand. The
duplex formation and stability was monitored by temperature
vs absorbance at 260 nm. The complementary and mismatch
DNA sequences chosen for this study and the observed Tm
values are summarized in Table 2.
The DNA‚DNG duplex formed by pentameric thymidyl DNG
(1) and complementary pentameric adenyl DNA exhibited a Tm
of 43 °C, whereas the DNA‚DNG-H duplexes formed from
pentameric adenyl DNA and DNG-Hoechst conjugate (2 or
3) exhibited a Tm value of 45 °C (∆Tm ) +2 °C) (Figure 9).
(49) (a) Loontiens, F. G.; Regenfuss, P.; Zechel, A.; Dumortier, L.; Clegg, R.
M. Biochemistry 1990, 29, 9029. (b) Loontiens, F. G.; McLaughlin, L.
W.; Diekmann, S.; Clegg, R. M. Biochemistry 1991, 30, 182.
(50) Haq, I.; Ladbury, J. E.; Chowdhry, B. Z.; Jenkins, T. C.; Chaires, J. B. J.
Mol. Biol. 1997, 271, 244.
(51) Bostock-Smith, C. E.; Searle, M. S. Nucleic Acids Res. 1999, 27, 1619.
9
J. AM. CHEM. SOC. VOL. 126, NO. 12, 2004 3743

A R T I C L E S
Reddy and Bruice
Figure 10. Possible self-dimers of 30-mer ssDNA sequence 5′-CGC-
CGCGCGCGCGAAAAACCCGGCGCGCGC-3′.
This marginal increase, when compared with that of 30-mer
(DNA)2‚DNG-H triplex (∆Tm of +13 °C, Table 1), indicates
very weak binding by Hoechst ligand in the minor groove of
pentameric DNA‚DNG duplex. Our molecular dynamics studies
of a DNA‚DNG duplex revealed that the charge-charge
attractions between oppositely charged backbones reduce the
width of the DNA‚DNG minor groove by 0.7 Å.35 Under the
experimental conditions, the piperazine ring of the Hoechst dye
carries a positive charge,23 as does the guanidinium group. The
narrower minor groove of DNA‚DNG duplex and charge-
charge repulsion effects may not allow the Hoechst dye,
possessing a bulky and nonplanar N-methylpiperazine ring, to
penetrate deep into the floor of the minor groove for effective
binding.
As we have seen, complexation of DNG-Hoechst conjugates
2 or 3 with the 30-mer ssDNA sequence, 5′-CGCCGCGCGCGC-
GAAAAACCCGGCGCGCGC-3′, is accompanied by consider-
able enhancement in the observed fluorescence emission (see
fluorescence characteristics). The presence of a long sequence
of G:C base pairs on either side of the target adenyl site provides
a flexible DNA‚DNG minor groove in which the tethered
Hoechst fluorophore can fit. As a result, the fluorescence
emission was enhanced by 278 units. However, this 30-mer
sequence did not aid in determining the DNA‚DNG-H duplex
Tm value because of the long self-dimers (Figure 10), formed
by G:C sequences of the ssDNA, and whose Tm profile was
more dominate. However, the analysis of fluorescence vs
temperature plot (see duplex fluorescence) revealed the DNA‚
DNG-H duplex Tm ≈ 50 °C (∆Tm ) ∼7 °C; see entries 1 and
14, Table 2).
To analyze the sequence specificity, we have investigated
the duplexes between DNG conjugates (1-3) and complemen-
tary DNA containing one mismatched base either at the end of
the sequence or in the center (Table 2). Incorporation of a single
cytidine mismatch in the complementary DNA sequence
exhibited significant decrease in DNA‚DNG duplex Tm values
(Table 2), and an internal mismatch exhibited a much more
pronounced effect (∆Tm ) -8 °C, entries 1 and 4) than a
terminal mismatch19c (∆Tm ) -5 °C, entries 1 and 7). The
duplexes formed by DNG conjugates 1-3 with complementary
DNA sequences were independent of pH over the range 6.0-
8.0. The tethered Hoechst derivative did not exhibit any
significant effect on the DNA‚DNG duplex stability at either
lower or higher pH because of its poor binding.
Figure 11. Fluorescence change vs temperature plot for duplex formed
by conjugate 3 and 30-mer ssDNA sequence, 5′-CGCCGCGCGCGC-
GAAAAACCCGGCGCGCGC-3′) in 10 mM KHPO4 buffer, pH 7.0,
containing 100 mM KCl.
will be formed and the fluorescence signal of the Hoechst dye
should increase if it can fold back and bind in the generated
minor groove. When excited at 345 nm, the DNA/DNG-
Hoechst conjugate (2 or 3) complexes (entries 2 and 3, Table
2) emitted a weak (24-unit increase) fluorescence signal centered
at 475 nm, suggesting the Hoechst fluorophore’s inability to
bind effectively in the minor groove of pentameric DNA‚DNG.
However, upon complexation with 30-mer ssDNA sequence,
5′-CGCCGCGCGCGCGAAAAACCCGGCGCGCGC-3′, the
fluorescence emission of conjugates 2 and 3 was increased
greatly, by 278 units at 475 nm (see Supporting Information).
This signal is red-shifted by 25 nm with respect to the triplex
fluorescence signal, which emits a broad signal centered at 450
nm (Figure 7). However, the intensity of the emission signal
was ∼10-fold less than that of the triplex fluorescence emission
signal at the same concentrations. This weak binding of
fluorophore may be attributed to charge-charge repulsions
between protonated piperazine ring and guanidinium groups and/
or differences in the shape of the minor groove.
To confirm that the observed fluorescence emission was a
result of binding of the Hoechst 33258 fluorophore in the DNA‚
DNG minor groove, we studied the temperature vs fluorescence
emission characteristics of DNA‚DNG-H duplex formed by
30-mer ssDNA and conjugate 3. A plot of fluorescence vs
temperature revealed that the relationship is sigmoidal (Figure
11). Analysis of Figure 11 shows a midpoint in fluorescence
emission occurring at ∼50 °C, which is higher by ∼5-7 °C
than the Tm values (43-45 °C, entries 1-3, Table 2) of duplexes
formed by pentameric DNA and DNG-Hoechst conjugates.
These results support that the Hoechst ligand is able to bind in
the DNA‚DNG minor groove and provides enhanced duplex
stability.
Conclusions
Positively charged thymidyl DNG, an analogue of DNA, was
synthesized on solid phase in a 3′ f 5′ direction, and Hoechst
33258 ligand was tethered via 11 and 18 atom linkers to the
5′-terminus of DNG using PyBOP/HOBt chemistry. We have
shown that hybridization of DNG-Hoechst conjugates to the
target dsDNA enhances the (DNA)2‚DNG-H triplex stability
through simultaneous minor groove binding by tethered Hoechst
Fluorescence Characteristics of DNA‚DNG-H Duplex
Complexes. Upon complexation of conjugates 2 and 3 with
complementary pentameric ssDNA, a DNA‚DNG-H duplex
9
3744 J. AM. CHEM. SOC. VOL. 126, NO. 12, 2004

Synthesis of DNG Tethering a Hoechst 33258 Analogue
A R T I C L E S
a 2.5 mL quartz cell thermally isolated with a water jacket. Solutions
at concentrations of 50 nM were excited at 345 nm, and emissions
were monitored between 360 and 600 nm. Temperature vs fluorescence
spectra were obtained by controlling the temperature with a recirculating
water bath. Emission spectra were recorded with λex ) 345 nm.
33258 fluorophore. Furthermore, when provided with a flexible
minor groove, the Hoechst fluorophore is able to bind in the
DNA‚DNG minor groove and further enhances the duplex
stability and fluorescence emission significantly. The nuclease
resistant guanidine linkages and the tethered Hoechst fluoro-
phore are expected to enhance the cellular and nuclear mem-
brane permeability of these DNG conjugates. Investigation of
the cell invasion and possible antisense/antigene characteristics
of these conjugates is in progress.
Synthesis: 9-(3,5-Di-O-mesyl-2,3,5-trideoxy-â-D-threo-pentofura-
nosyl)thymidine (6). Mesyl chloride (1.55 mL, 20 mmol) was added
dropwise via syringe to an ice-cold solution of 536 (0.97 g, 4 mmol) in
dry pyridine (20 mL) under argon atmosphere. The reaction mixture
was allowed to reach room temperature over ∼4 h and was then stirred
overnight. The pyridine was rotoevaporated, and the residue was
partitioned between chloroform and water. Aqueous layer was extracted
with 3 × 100 mL of chloroform. The combined chloroform layer was
dried, treated with charcoal, and rotoevaporated. The product was pure
and was used for further reactions without purification, yield 1.52 g,
Experimental Section
Materials. Unless otherwise noted, all solvents and reagents were
obtained from Aldrich and used without further purification. TLCs were
carried out on commercially available flexible TLC silica gel (Silica
gel 60 F254) plates purchased from Selecto Scientific, and compounds
on TLC were visualized using shortwave UV light. Silica gel (pore
size: 32-63 Å, Selecto Scientific) was used for flash column
chromatography. All NMR spectra were recorded on a Varian 400 MHz
instrument. HRMS (ESI/TOF+) mass spectral analysis was performed
on a Micromass Q-Tof-2 quadrapole instrument. Long chain alkylamine
CPG (pore size: 500 Å; mesh size: 80-120) was obtained from Sigma
and soaked in and washed with dry DMF before use. DNA oligomers
were purchased as prepurified from the Biomedical Resource Center
at UCSF.
NAP-10 (Sephadex G25 DNA grade) columns were purchased from
Amersham Biosciences. HPLC grade acetonitrile and water were
purchased from Fisher Scientific, and triethylammonium acetate
(TEAA) buffer was purchased from Aldrich. Reverse-phase HPLC was
performed on a Hewlett Packard 1050 instrument equipped with a
quaternary solvent delivery system and a diode array detector. Alltech
Macrosphere 300 Å, C8, silica 7 µm, 250 mm × 10 mm preparative
reverse-phase column was used. UV detector was set at 260 nm for
DNG or 260 and 343 nm for DNG-Ht. A gradient from 90% f 20%
eluent A (0.1% TFA in water) and 10% f 80% eluent B (acetonitrile)
over 30 min with a flow rate of 3.0 mL/min was used.
DNA, DNG, and DNG-Ht Conjugates Extinction Coefficients.
The DNA and DNG (1) extinction coefficients at 260 nm were
determined according to the nearest neighbor method, and for the
conjugates (2 and 3) the extinction coefficient of the Hoechst acid (15)
at 260 nm (20800 A260 units mM-1) was added to the calculated value
of DNG. However, the concentrations of conjugates solutions were
determined using the extinction coefficient (46400 A347 units M-1) of
Hoechst acid 15.
Thermal Melting (Tm) Studies. All Tm experiments were carried
out in 10 mM potassium phosphate buffer, pH 6.0, 7.0, or 8.0 containing
100 or 700 mM KCl, as mentioned, at oligomer concentrations of 2
µM (for triplex) and 6 µM (for duplex). The solutions were annealed
by heating to 95 °C using a heating block and allowing to cool slowly
to reach room temperature before being stored at 4 °C. Absorbance
(260 nm) vs temperature values were obtained on a Cary 100 Bio UV/
vis spectrophotometer equipped with a temperature programmable
cellblock. Data points between 5 and 95 °C were taken for every 1 °C,
with a temperature ramp of 0.5 °C/min. Tm temperatures were calculated
by first-derivative analysis and also by direct graphical analysis of the
absorbance vs temperature plot to determine the midpoint of the
transition. Both techniques gave values that were within the experi-
mental error ((1 °C) for the analysis. Also, we have carried out a
concentration-dependent Tm study to rule out the possibility of
intramolecular duplex formation.
1
(96%). H NMR (DMSO-d6, 400 MHz): δ 1.79 (s, 3H, -CH3), 2.31
(m, 1H, 2′-H), 2.87 (m, 1H, 2′′-H), 3.26 (s, 3H, mesyl CH3), 3.32 (s,
3H, mesyl CH3), 4.37 (m, 1H, 3′-H), 4.48 (m, 1H, 5′H), 4.55 (m, 1H,
5′′-H), 5.38 (m, 1H, 4′-H), 6.20 (m, 1H, 1′-H), 7.45 (s, 1H, -CH-),
11.39 (s, 1H, -NH-). 13C NMR (DMSO-d6, 400 MHz): 12.34, 36.85,
37.68, 38.08, 67.79, 78.66, 79.11, 83.08, 109.88, 135.43, 150.46, 163.68.
HRMS (ESI/TOF+) m/z: 399.0528 (M + H); calcd 399.0532 (M +
H) for C12H18N2O9S2.
3,5-Diamino-2,3,5-trideoxythymidine (7). Lithium azide (3.43 g,
70 mmol) was suspended in a solution of 6 (1.39 g, 3.5 mmol) in dry
DMF (35 mL) and stirred at 100 °C for 4 h. The reaction mixture was
cooled to room temperature, and DMF was rotoevaporated under
vacuum. Residue was partitioned between chloroform and water.
Aqueous layer was extracted with 3 × 100 mL chloroform. The
combined chloroform layer was dried and rotoevaporated. The residue
was purified on silica gel column using 0-5% methanol in dichlo-
romethane to yield 0.74 g (73%) of pure 3,5-diazido-2,3,5-trideoxy-
1
thymidine. H NMR (DMSO-d6, 400 MHz): δ 1.76 (s, 3H, -CH3),
2.47 (m, 1H, 2′-H), 2.59 (m, 1H, 2′′-H), 3.47 (m, 1H, 5′-H), 3.58 (m,
1H, 5′′-H), 4.37 (m, 1H, 3′-H), 5.27 (m, 1H, 4′-H), 5.88 (t, J 4, 1H,
1′-H), 7.60 (s, 1H, -CH-). HRMS (ESI/TOF+) m/z: 293.1093 (M +
H); calcd 293.1111 (M + H) for C10H12N8O3. To a solution of 3,5-
diazido-2,3,5-trideoxythymidine (0.65 g, 2.25 mmol) in 95% ethanol
(50 mL) was added 50 mg of 10% Pd/C and hydrogenated at 50 psi
for 4 h. The solution was filtered through Celite, and the filtrate was
rotoevaporated to dryness under reduced pressure. The residue was
further dried overnight under high vacuum to give a quantitative yield
1
of the pure compound 7. H NMR (DMSO-d6, 400 MHz): δ 1.79 (s,
3H, -CH3), 2.01 (m, 1H, 2′-H), 2.13 (m, 1H, 2′′-H), 2.82 (m, 1H,
5′-H), 2.89 (m, 1H, 5′′-H), 3.37 (m, 1H, 4′-H), 3.48 (m, 1H, 3′-H),
6.10 (t, J ) 7 Hz, 1H, 1′-H), 7.64 (s, 1H, -CH-). 13C NMR (DMSO-
d6, 400 MHz): 12.16, 39.84, 42.61, 52.13, 83.14, 86.52, 109.53, 136.53,
150.46, 163.80. HRMS (ESI/TOF+) m/z: 241.1293 (M + H); calcd
241.1301 (M + H) for C10H16N4O3.
5′-Monomethoxytritylamino-3′-(N′-9-fluorenylmethoxycarbon-
ylthiouria)-2′,3′,5′-trideoxythymidine (8). Monomethoxytrityl chloride
(1.73 g, 5.6 mmol) in dry dichloromethane (10 mL) was added dropwise
to a solution of 7 (1.35 g, 5.6 mmol) and triethylamine (1.56 mL, 11.2
mmol) in dry DCM (100 mL) at 0 °C. After being stirred at room
temperature for 4 h, TLC showed the completion of the reaction.
Reaction mixture was diluted with another 100 mL of DCM and washed
with water (2 × 100 mL). The DCM layer was dried and rotoevaporated
to solid. The crude product was purified on a silica gel column using
0-1% methanol in DCM containing 0.5% TEA to yield 2.26 g (91%)
of pure 5′-momomethoxytritylamino-3′-amino-2′,3′,5′-trideoxythymi-
1
Fluorescence Emission Studies. Fluorescence spectra were obtained
on a Perkin-Elmer LS50B fluorophotometer equipped with a constant-
temperature water bath set at 25 °C. All measurements were performed
with the following parameters: slit width, Ex/Em ) 10 nm/10 nm;
high sensitivity; high speed 500 nm/s. Solutions in 10 mM potassium
phosphate buffer pH 7.0 containing 100 mM KCl were introduced into
dine. H NMR (DMSO-d6, 400 MHz): δ 1.70 (s, 3H, -CH3), 2.01
(m, 1H, 2′-H), 2.14 (m, 1H, 5′-H), 2.20 (m, 1H, 5′′-H), 2.35 (m, 1H,
2′′-H), 2.63 (m, 1H, 3′-H), 3.60 (m, 1H, 4′-H), 3.72 (s, 3H, -OCH3),
6.08 (t, J ) 6 Hz, 1H, 1′-H), 6.84 (d, J ) 9 Hz, 2H, Ar-H), 7.18 (t,
2H, J ) 7 Hz, Ar-H), 7.29 (m, 6H, Ar-H and -CH-), 7.40 (m, 5H,
Ar-H). 13C NMR (DMSO-d6, 400 MHz): 12.21, 40.31, 45.72, 52.74,
9
J. AM. CHEM. SOC. VOL. 126, NO. 12, 2004 3745

A R T I C L E S
Reddy and Bruice
54.98, 69.77, 83.19, 85.89, 109.40, 113.07, 126.09, 127.75, 128.31,
129.65, 136.01, 137.82, 146.31, 150.42, 157.41, 163.76, 178.96. HRMS
(ESI/TOF+) m/z: 513.2483 (M + H); calcd 513.2502 (M + H) for
C30H32N4O4. The above product (1.53 g, 3 mmol) was dissolved in dry
DCM (30 mL), and 9-fluorenylmethoxycarbonyl isothiocyanate39
(Fmoc-NCS) (0.87 g, 3.1 mmol) was added portion wise at room
temperature. After the solution was stirred for 2 h at room temperature,
TLC (20:1 DCM/methanol) indicated the completion of the reaction.
The solvent was rotoevaporated, and the residue was redissolved in 5
mL of DCM containing 10% methanol and precipitated by adding
excess hexanes with vigorous stirring. The precipitate was filtered and
dried to get pure product 8 as white solid, yield 2.1 g (89%). 1H NMR
(DMSO-d6, 400 mHz): δ 1.74 (s, 3H, T-CH3), 2.26 (m, 1H, 2′-H),
2.37 (m, 2H, 5′ and 5′′-H), 2.53 (m, 1H, 2′′-H), 2.87 (m, 1H, 4′-H),
3.69 (s, 3H, -OCH3), 4.02 (m, 1H, 4′-H), 4.22-4.44 (m, 4H, Fmoc
-CH- and -CH2-), 5.12 (m, 1H, 3′-NH-), 6.16 (t, J ) 6 Hz, 1H,
1′-H), 6.81 (d, J ) 9 Hz, 2H, Ar-H), 7.15 (t, 2H, J ) 7 Hz, Ar-H),
7.22-7.45 (m, 6H, Ar-H), 7.65 (s, 1H, T-CH), 7.81-7.92 (m, 5H,
Ar-H), 10.03 (d, J ) 7, 1H, -NH-Fmoc), 11.35 (s, 1H, 5′-NH), 11.55
(s, 1H, T-NH). 13C NMR (DMSO-d6, 400 MHz): 12.51, 39.84, 46.01,
54.92, 67.35, 69.69, 81.98, 83.20, 109.68, 113.06, 120.09, 121.37,
125.55, 126.05, 127.21, 127.68, 127.81, 128.25, 128.91, 129.59, 136.34,
137.41, 137.66, 139.40, 140.71, 142.56, 143.27, 146.18, 150.33, 150.41,
153.25, 157.37, 163.72, 179.56. HRMS (ESI/TOF+) m/z: 794.2996
(M + H); calcd 794.3012 (M + H) for C46H43N5O6S.
6-Monomethoxytritylamino-hexanoic Acid (10). Monomethoxy-
trityl chloride (3.71 g, 12 mmol) in dry pyridine (25 mL) was added
dropwise to a solution of 9 (1.31 g, 10.0 mmol) in dry pyridine (50
mL) at room temperature. After addition, the reaction mixture was
stirred at room temperature overnight. Pyridine was rotoevaporated to
dryness, and the residue was partitioned between ether and water. The
aqueous layer was extracted with ether twice more (2 × 100 mL), and
the combined ether layer was dried and rotoevaporated. The residue
was purified by silica gel column chromatography using 0-2%
methanol in DCM, yield 3.2 g (79%). 1H NMR (DMSO-d6, 400
mHz): δ 1.26 (m, 2H, -CH2-), 1.43 (m, 4H, 2× -CH2-), 1.94 (t, J
) 7, 2H, -CH2COOH), 2.16 (t, J ) 7, 2H, -NH-CH2-), 3.71 (s,
3H, -OCH3), 6.84 (dd, J ) 9, 3, 2H, Ar-H), 7.16 (t, J ) 7, 2H, Ar-
H), 7.27 (m, 6H, Ar-H), 7.38 (dd, J ) 8, 2, 4H, Ar-H). 13C NMR
(DMSO-d6, 400 MHz): 25.19, 27.17, 30.47, 34.30, 43.90, 55.60, 70.51,
113.61, 126.53, 128.28, 128.95, 130.21, 138.85, 147.28, 157.94, 175.19.
HRMS (ESI/TOF+) m/z: 404.2201 (M + H) and 426.2037 (M + Na);
calcd 404.2225 (M + H) and 426.2045 (M + Na) for C26H29NO3.
Benzyl 4-(4-Formylphenoxy)butanoate (12). To a solution of
4-hydroxybenzaldehyde (3.3 g, 27 mmol) and benzyl 4-bromobutyrate41
(11, 7.71 g, 30 mmol) in dry DMA (50 mL) was added Cs2CO3 (10.74
g, 33 mmol), and the solution was stirred at 100 °C for 15 h. The
reaction mixture was then cooled to room temperature, and the solid
was filtered off. DMA was rotoevaporated completely under high
vacuum, residue was dissolved in dry DCM (250 mL), and insoluble
material was filtered off. The DCM solution was then washed with
water (100 mL), 2N NaOH solution (2 × 100 mL), and brine (100
mL). The DCM layer was dried with Na2SO4, treated with activated
charcoal, and rotoevaporated. The residue was further dried under high
vacuum overnight to afford pure compound 12, yield 7.9 g (98%). 1H
NMR (CDCl3, 400 MHz): δ 1.17 (m, 2H, -CH2-), 2.60 (t, 2H, J )
7, -CH2-), 4.09 (t, 2H, J ) 6, -CH2-), 5.14 (s, 2H, -CH2-), 6.96
(dd, 2H, J ) 9, 2, Ar-H), 7.35 (m, 5H, Ar-H), 7.82 (dd, 2H, J ) 9,
2, Ar-H). 13C NMR (400 MHz, CDCl3): δ 24.58, 30.80, 66.62, 67.20,
114.91, 128.46, 128.52, 128.80, 130.14, 132.20, 135.99, 164.01, 173.01,
191.04. HRMS (ESI/TOF+) m/z: 299.1274 (M + H); calcd 299.1283
(M + H) for C18H18O4.
precipitate the product. The precipitated crude product was collected
by filtration and then purified by silica gel flash column chromatography
using 9:1 ethyl acetate/methanol containing 0.1% triethylamine. Yield
1.04 g (87%), Rf ) 0.2 (silica, 7:3 ethyl acetate/methanol containing
1
0.1% of triethylamine mixture). H NMR (400 MHz, DMSO-d6): δ
2.05 (m, 2H, -CH2-), 2.28 (s, 3H, CH3-NR2), 2.51-2.59 (m, 6H,
piperazine -CH2- and -CH2C(O)-), (3.14 (s, 4H, piperazine -CH2-
), 4.10 (t, J ) 6, 2H), 5.13 (s, 2H, Ar-CH2-), signals detected between
6.9 and 8.4 ppm are due to Ar protons, 6.93 (m, 2H), 7.11 (d, J ) 7,
2H), 7.37 (m, 5H), 7.60 (m, 2H), 8.00 (m, 1H), 8.14 (dd, 2H, J ) 9,
3), 8.35 (s, 1H); HRMS (ESI/TOF+) m/z: 601.2911 (M + H); calcd
601.2927 (M + H) for C36H36N6O3.
Synthesis of Ht Acid (15). Compound 14 (1.02 g, 1.7 mmol) was
dissolved in ethanol (100 mL), and 250 mg of 10% Pd/C was added
and hydrogenated on a hydrogenator for 5 h under 50 psi. The solution
was filtered and retoevaporated under reduced pressure. The residue
was redissolved in DCM containing 10% methanol, rotoevaporated,
and dried under high vacuum overnight. The product 15 was pure, and
1
the yield was quantitative. H NMR (DMSO-d6, 400 MHz): δ 1.98
(m, 2H, -CH2-), 2.26 (s, 3H, CH3-NR2), 2.42 (t, J ) 7, 2H, -OCH2-
), 2.52 (bs, 4H, piperazine -CH2-), 3.13 (bs, 4H, piperazine -CH2-
), 4.07 (t, J ) 6, 2H, -CH2-COOH), signals detected between 6.9
and 8.4 ppm are due to Ar protons, 6.92 (m, 1H), 7.13 (d, J ) 9, 2H),
7.46 (m, 1H), 7.65 (m, 1H), 8.00 (m, 1H), 8.15 (d, 2H, J ) 9), 8.28
(m, 1H). HRMS (ESI/TOF+) m/z: 511.2439 (M + H) and 533.2267
(M + Na); calcd 511.2457 (M + H) and 533.2277 (M + Na) for
C29H30N6O3.
Solid-Phase Synthesis of DNG 1. The solid-phase synthesis of 1
was accomplished using long chain alkylamine controlled pore glass
(CPG). The 5′- modified monomer19a was loaded on to the CPG as its
succinyl derivative 4 by adopting the literature procedure52 (see
Supporting Information). The unloaded amine sites on CPG were
terminated by capping with acetic anhydride/TEA, and then 5′-MMTr
was deprotected with 3% DCA in DCM solution. The loading yield,
39.6 µmol/g, was determined spectrophotometrically from the amount
of MMTr cation released. Synthesis of pentameric thymidyl DNG 1
was started on 15 µmol scale to accomplish conjugates 2 and 3 also
on a 5 µmol scale each. A solution of 8 (59 mg, 75 µmol, 5 equiv) in
1 mL of DMF was poured over the beads. Then 1 mL of a 200 mM
HgCl2 solution and 1 mL of a 250 mM TEA solution in DMF were
added quickly and simultaneously via two syringes. A thick white
precipitate was formed immediately. The tube was capped tightly and
agitated at room temperature for 2 h. The solution was filtered off,
and beads were washed with DMF until all the visible precipitate has
been removed. However, the CPG beads were darkened due to the black
precipitate (HgS) formed in the reaction. A solution of 20% thiophenol
in DMF (5 mL) was poured over the beads and agitated for 1 min to
remove any black HgS precipitate. Finally, beads were washed with
copious amounts of DMF followed by 1% TEA in DMF, and the
coupling reaction was repeated two more times to increase the coupling
yield of 16. After the third coupling (16), the whole cycle of capping/
deprotection/coupling was repeated three more times to get the desired
pentameric DNG 17, which was then deprotected/cleaved from CPG.
Before cleaving the DNG oligomer from the CPG, one-third of the
beads (∼5 µmol) were separated from the SPS tube, washed with
methanol, followed by DCM, and dried under vacuum. The dried beads
were then transferred into a vial, and methanolic ammonia (5 mL) was
poured over the beads. The vial was capped tightly and agitated at
room temperature for 2 h. The supernatant solution was filtered and
lyophilized to get white residue of 5′-MMTr protected DNG oligomer.
The crude trityl-on product was purified on reverse-phase HPLC using
5 f 80% gradient of acetonitrile in 100 mM TEAA buffer, pH 7.0,
Synthesis of Benzyl Ester of Hoechst Acid (14). A mixture of
aldehyde 12 (596 mg, 2 mmol) and diamine 1342 (640 mg, 2 mmol) in
nitrobenzene (20 mL) was stirred at 130 °C for 24 h. The solution was
cooled to room temperature, and an excess of hexanes was added to
(52) Atkinson, T.; Smith, M. Solid-Phase Synthesis of Oligodeoxyribonucleotides
by the Phosphite-Triester Methods. In Oligonucleotide Synthesis:
A
Practical Approach; Gain, M. J., Ed.; Oxford University Press: New York,
1990; pp 35-48.
9
3746 J. AM. CHEM. SOC. VOL. 126, NO. 12, 2004

Synthesis of DNG Tethering a Hoechst 33258 Analogue
A R T I C L E S
and characterized by ESI mass spectrometry. To the trityl-on DNG
was added 3% DCA in DCM (1 mL) and agitated for 1 min. Excess of
hexanes was added to precipitate the trityl-off DNG 1. The solvents
were decanted after centrifugation, and the product was dried and
analyzed by RP-HPLC.
Solid-Phase Synthesis of DNG-Hoechst 33258 Conjugates (2 and
3). After making pentameric DNG 17, the unreacted 5′-NH2 sites of
2/3 CPG (∼10 µmol) were capped with acetic anhydride/TEA. The
linker 10 and Hoechst acid 15 were added stepwise on to 17 in the
presence of PyBOP/HOBt. Both Hoechst-tethered DNGs, 2 and 3, were
synthesized on 5 µmol scale each as described below (Scheme 6).
(a) Capping. Acetic anhydride (1 mL, 200 mM) and TEA (1 mL,
250 mM) solutions in DMF were added to CPG beads (17), and the
mixture was agitated for 10 min. The solutions were filtered off, and
the beads were washed thoroughly with DMF and DCM, followed by
DMF.
(b) Deprotection. The 5′-MMTr groups of 17 were cleaved by
treatment with DCA solution. A solution of 3% DCA in DCM (3 mL)
was poured over the beads, and the mixture was agitated for 1 min.
The solution was filtered, and beads were washed with 3% DCA
solution until no more yellow color was present in the filtrate. The
combined filtrate was made up to a known volume, and yield was
determined from UV absorbance. Before proceeding to the next
coupling reaction, the CPG beads were washed thoroughly with DMF,
DCM, and DMF, followed by 1% TEA in DMF solution.
(c) Coupling of Linker 10. Compound 10 (40 mg, 0.1 mmol, 10
equiv), PyBOP (53 mg, 0.1 mmol, 10 equiv), and HOBt (13 mg, 0.1
mmol, 10 equiv) were added to the MMTr-removed CPG (17) beads,
and 3 mL of DMF was poured over the beads. DIPEA (175 µL, 1.0
mmol) was then added, the tube was capped tightly, and the mixture
was agitated for 12 h at room temperature. The solution was filtered
off, and beads were washed with copious amounts of DMF, DCM,
DMF, and finally with 1% TEA in DMF. The coupling reaction was
repeated once more to increase the coupling yield, and the unreacted
5′-NH2 sites were finally capped with Ac2O/TEA.
(d) Coupling of Hoechst Acid 15. Compound 15 (12.8 mg, 25 µmol,
5 equiv), PyBOP (27 mg, 50 µmol, 10 equiv), and HOBt (6.6 mg, 50
µmol, 10 equiv) were added to the MMTr-removed CPG (18 or 19)
beads. Then DMF (3 mL) and DIPEA (175 µL, 1.0 mmol) were added,
the tube was capped tightly, and the mixture was agitated for 24 h at
room temperature. The solution was filtered off, and beads were washed
with copious amount of DMF and methanol, followed by DCM, and
dried under vacuum. The beads were fluoresced under longwave UV,
indicating the successful addition of Hoechst acid.
(e) Cleavage and Deprotection. The dried CPG beads of 20 or 21
were transferred to a vial, and 0.1 M NaOH solution in 4:1 methanol/
water (5 mL) was poured over the beads. The vial was capped tightly,
and the mixture was agitated at room temperature for 1 h. The
supernatant solution was pipetted out and desalted using NAP-10
(Sephadex G25) columns and HPLC grade water. The solution was
diluted with 0.1% TFA in water (1 mL) and purified on reverse-phase
HPLC using 0.1% TFA in water and gradient of acetonitrile (5 f 80%
over 30 min). Both products 2 and 3 were eluted as broad peaks at
∼14 min.
Acknowledgment. This research work was supported by a
grant (5R37DK09171-39) from the National Institutes of Health.
We acknowledge valuable conversations with Hemavathi Challa
and Istvan Szabo.
Supporting Information Available: Experimental details of
4, figures showing ESI/TOF+ mass spectra of DNG conjugates
1-3, UV spectrum of 3 and 15, and temperature vs fluorescence
emission change of DNA‚DNG-H duplex (PDF). This material
is available free of charge via the Internet at http://pubs.acs.org.
JA031557S
9
J. AM. CHEM. SOC. VOL. 126, NO. 12, 2004 3747