Hoechst 33258 tethered by a hexa(ethylene glycol) linker to the 5'-termini of oligodeoxynucleotide 15-mers: Duplex stabilization and fluorescence properties

Article abstract of DOI:10.1021/jo9618536

A fluorescent Hoechst 33258 derivative has been prepared in which a hexa(ethylene glycol) linker is attached to the terminal phenol residue. Conjugation of this derivative to DNA sequences is accomplished by a reversed coupling protocol, one in which the 5'-terminal nucleoside residue of a fully protected DNA sequence is converted to a terminal phosphoramidite. In the presence of the Hoechst derivative and tetrazole the final coupling reaction is achieved to generate the conjugated nucleic acid. After deprotection and cleavage of the conjugate from the support, HPLC analysis indicates that the conjugation reaction proceeds with yields as high as 75%. The presence of the conjugated Hoechst derivative increases the stability of DNA duplexes typically by 10-16°C. A variety of sequence variants indicate that the tether length is sufficient to reach beyond the terminus of the DNA duplex and bind to internal A-T rich target sequences as far away as four base pairs from the site of attachment. A four base pair binding site appears to be necessary for effective helix stabilization by the conjugate, but in some cases can include a G-C base pair, which is consistent with a previous X-ray diffraction study regarding the binding of Hoechst 33258 to duplex DNA. When A-T base pairs alternate with G-C base pairs, a small but discernible increase is T(m) is observed (3.6°C), indicating that binding to this sequence still occurs, but not in the same manner as to A-T rich sequences. Upon formation of the conjugated duplex, an enhanced quantum yield for the fluorescence emission spectrum of the tethered Hoechst derivative is observed. When an A-T rich binding site is present, the enhanced quantum yield increases by at least 16- and in some cases to nearly 30-fold relative to the value obtained for the single-stranded DNA-Hoechst conjugate.

Full text of DOI:10.1021/jo9618536

J . Org. Chem. 1997, 62, 523-529
523
Hoech st 33258 Teth er ed by a Hexa (eth ylen e glycol) Lin k er to th e
5′-Ter m in i of Oligod eoxyn u cleotid e 15-Mer s: Du p lex Sta biliza tion
a n d F lu or escen ce P r op er ties
Sharanabasava B. Rajur, J ordi Robles, Kristin Wiederholt, Robert G. Kuimelis, and
Larry W. McLaughlin*
Department of Chemistry, Merkert Chemistry Center, Boston College, Chestnut Hill, Massachusetts 02167
Received September 30, 1996X
A fluorescent Hoechst 33258 derivative has been prepared in which a hexa(ethylene glycol) linker
is attached to the terminal phenol residue. Conjugation of this derivative to DNA sequences is
accomplished by a reversed coupling protocol, one in which the 5′-terminal nucleoside residue of a
fully protected DNA sequence is converted to a terminal phosphoramidite. In the presence of the
Hoechst derivative and tetrazole the final coupling reaction is achieved to generate the conjugated
nucleic acid. After deprotection and cleavage of the conjugate from the support, HPLC analysis
indicates that the conjugation reaction proceeds with yields as high as 75%. The presence of the
conjugated Hoechst derivative increases the stability of DNA duplexes typically by 10-16 °C. A
variety of sequence variants indicate that the tether length is sufficient to reach beyond the terminus
of the DNA duplex and bind to internal A-T rich target sequences as far away as four base pairs
from the site of attachment. A four base pair binding site appears to be necessary for effective
helix stabilization by the conjugate, but in some cases can include a G-C base pair, which is
consistent with a previous X-ray diffraction study regarding the binding of Hoechst 33258 to duplex
DNA. When A-T base pairs alternate with G-C base pairs, a small but discernible increase is Tm
is observed (3.6 °C), indicating that binding to this sequence still occurs, but not in the same manner
as to A-T rich sequences. Upon formation of the conjugated duplex, an enhanced quantum yield
for the fluorescence emission spectrum of the tethered Hoechst derivative is observed. When an
A-T rich binding site is present, the enhanced quantum yield increases by at least 16- and in some
cases to nearly 30-fold relative to the value obtained for the single-stranded DNA-Hoechst
conjugate.
In tr od u ction
effects for these dyes are sensitive enough to permit the
detection of one target cell per million in mixed cell
populations.13 Such highly sensitive applications suggest
that an oligonucleotide tethering one of the Hoechst dyes
could be valuable in the development of a sensitive
nucleic acid hybridization diagnostic.
A number of well-characterized ligands are known to
interact with double-stranded DNA by binding in the
minor groove of the duplex structure. Among these
agents are the fluorescent dyes Hoechst 33258 and
Hoechst 33342 as well as the related benzindole DAPI.
Crystal structures of DNA complexes containing Hoechst
33258 and DAPI have been reported,1-4 confirming the
minor groove binding nature of the fluorophores. Binding
in the minor groove by these dyes results in helix
stabilization and a dramatic enhancement in the ob-
served fluorescence emission quantum yield.5,6 The
enhanced emission characteristics are largely due to the
protection of the dye in the excited state when bound in
the minor groove from nonradiative relaxation effects,
presumably collisional processes involving water.5 The
fluorescence properties of the Hoechst dyes have previ-
ously been employed in a quantitative manner to auto-
mate DNA content assays,7-9 to determine cell num-
bers,10,11 and to sort chromosomes.12 The quantum yield
The ability to easily and rapidly prepare nucleic acid
sequences tethered to fluorescent agents, or those pos-
sessing other unique properties, provides a vehicle for
the development of a series of new hybridization probes,
nucleic acid-based diagnostics, and ultimately nucleic
acid-based therapeutics. The observation that Hoechst
dyes can cross the cell membrane and nuclear membrane
in order to stain chromosomes12 also suggests that a
DNA-Hoechst conjugate might be better able to cross
cell membranes than the unconjugated naked DNA
sequence. To prepare nucleic acid conjugates, one of the
most common techniques involves the incorporation of a
linker to the 5′-terminus of the oligonucleotide. When
the linker contains an appropriate nucleophilic site, such
as a sulfhydryl14 or aliphatic amino group,15 such linkers
then permit postsynthetic conjugation by a variety of
X Abstract published in Advance ACS Abstracts, J anuary 15, 1997.
(1) Pjura, P. E.; Grzeskowiak, K.; Dickerson, R. E. J . Mol. Biol. 1987,
197, 257-271.
(7) Sterzel, W.; Bedford, P.; Eisenbrand, G. Anal. Biochem. 1985,
147, 462-470.
(2) de C. T. Carrondo, M. A. A. F.; Coll, M.; Aymami, J .; Wang, A.
H.-J .; van der Marel, G. A.; van Boom, J . H.; Rich, A. Biochemistry
1989, 28, 7849-7859.
(8) Araki, T.; Yamanoto, A.; Yamada, M. Histochemistry 1987, 87,
331-338.
(9) Karawajew, L.; Rudchenko, S.; Wlasik, T.; Trakht, I. J . Immunol.
Methods 1990, 129, 277-282.
(3) Teng, M.-K.; Usman, N.; Frederick, C. A.; Wang, A. H.-J . Nucleic
Acids Res. 1988, 16, 2671-2675.
(10) Downs, T. R.; Wilfinger, W. W. Anal. Biochem. 1983, 131, 538-
(4) Quintana, J . R.; Lipanov, A. A.; Dickerson, R. E. Biochemistry
1991, 30, 10294-10306.
547.
(11) Adams, C. J .; Storrie, B. Histochem. Cytochem. 1981, 29, 326-
334.
(12) Arndt-J ovin, D. J .; J ovin, T. M. Cytometry 1990, 11, 80-89.
(13) Lee, B. R.; Haseman, D. B.; Reynolds, C. P. Cytometry 1989,
10, 256-266.
(5) Zimmer, C.; Wa¨hnert, U. Prog. Biophys. Mol. Biol. 1986, 47, 31-
112.
(6) Lootiens, F. G.; Regenfuss, P.; Zechel, A.; Dumortier, L.; Clegg,
R. M. Biochemistry 1990, 29, 9029-9039.
S0022-3263(96)01853-1 CCC: $14.00 © 1997 American Chemical Society

524 J . Org. Chem., Vol. 62, No. 3, 1997
Rajur et al.
7.53 (d, 2H). UV (H2O): λmax ) 233 nm. IR 3465, 2873, 2219,
1747, 1602, 1507, 1451, 1350, 1300, 1262 cm-1. HRMS: Calcd
for C19H3NO7 (M + H+) 384.20222, found 384.20270.
agents. An alternative approach simply converts the
conjugation agent into a phosphoramidite building block
to be used directly in the final coupling step during the
assembly of the oligonucleotide conjugate.16,17
2-{2-[4-[2-[P en t a (et h ylen e glycol)]et h oxy]p h en yl]-6-
b e n zim id a zolyl}-6-(1-m e t h yl-4-p ip e r a zin yl)b e n zim id -
a zole (4). To prepare the necessary imino methyl ester 2, 0.5
g (1.27 mmol) of 1 was dissolved in 15 mL of dichloromethane,
0.14 g (4.2 mmol) of anhydrous methanol was added, and the
reaction was cooled to 0 °C. The reaction mixture was
saturated with HCl and maintained at 0 °C overnight. The
resulting white precipitate was collected, washed thoroughly
with diethyl ether, and dried to yield 0.52 g (1.21 mmol, 95%)
of a white solid which was used in the following step without
purification.
A variety of DNA conjugates have been prepared in
order to enhance hybridization properties to complemen-
tary DNA or RNA target sequences. Intercalators,18
polyamines,19 and peptidelike agents related to dis-
tamycin20 have been attached to DNA sequences and
their hybridization properties subsequently examined. In
previous work we described the preparation of DNA
sequences tethering the Hoechst fluorescent dye 33258
to an internal site within the target binding site of the
A-T rich sequence.21 The preparation of these materials
was complicated by aggregation of the Hoechst dye
during the postsynthetic labeling procedures, and only
small quantities of materials could be obtained. The
helix-stabilizing and fluorescent properties of the double-
stranded complexes prepared from such conjugates were
also reported in that initial study. In the present work
we describe the synthesis and properties of conjugates
in which a Hoechst dye is tethered to the terminus of
the DNA sequence through a hexa(ethylene glycol) linker,
and conjugation of the oligonucleotide by the dye is
accomplished during the assembly of the DNA sequence.
To 0.2 g (0.62 mmol) of 2-(3,4-diaminophenyl)-6-(1-methyl-
4-piperazinyl)benzimidazole22 (3) dissolved in anhydrous metha-
nol (3 mL) was added 0.496 g (11.09 mmol) of imidate 2
followed by 0.66 mL (1.1 mmol) of glacial acetic acid. The
mixture was stirred at 65-70 °C for 3 h. TLC analysis
(dichloromethane:methanol, 7:3) containing trace of tri-n-
butylamine indicated the completion of the reaction. Solvents
were evaporated, and the residue was purified on a small
column of silica gel using ethyl acetate with a trace of tri-n-
butylamine and gradient of methanol to yield 0.198 g (0.287
mmol, 46%) of pure product. Mp ) 115-120 °C dec. Rf:
(dichloromethane:methanol 7:3 + trace of tri-n-butylamine) )
0.48. 1H NMR (300 MHz, CD3OD): δ ) 2.57 (3H, s), 2.93 (4H,
m), 3.30 (4H, m), 3.57 (2H, m), 3.87 (2H, t, J ) 4.3 Hz), 4.20
(2H, t, J ) 4.3 Hz), 7.06 (1H, dd, J ) 2.1, 8.7 Hz), 7.10 (2H, d,
J ) 9.0 Hz), 7.15 (1H, d, J ) 2.1 Hz), 7.52 (1H, d, J ) 8.7 Hz),
7.69 (1H, d, J ) 8.4 Hz), 7.94 (1H, dd, J ) 8.4 Hz, 1.8 Hz),
8.05 (2H, d, J ) 9.0 Hz), 8.25 (1H, d, J ) 1.8 Hz). UV (H2O):
λmax ) 212, 255, 334 nm, λmin ) 238, 288 nm. IR: 3172, 2866,
Exp er im en ta l Section
Ma ter ia ls. HPLC grade solvents were either obtained from
Aldrich (Milwaukee, WI) or from Fisher Scientific (Fair Lawn,
NJ ). 5′-Dimethoxytrityl nucleoside phosphoramidite mono-
mers as well as all ancillary reagents for nucleic acid synthesis
were obtained from BioGenex (San Ramon, CA) or from
Applied Biosystems, Inc. (Foster City, CA). Oligonucleotides
were synthesized using nucleoside phosphoramidite deriva-
tives and an Applied Biosystems 381A DNA synthesizer.
High-performance liquid chromatography (HPLC) was carried
out on ODS-Hypersil column (0.94 × 25cm or 0.46 × 25 cm,
Shandon Southern, England). Mass spectra were obtained
from the Mass Spectrometry Laboratory, School of Chemical
Sciences, University of Illinois, Urbana, IL.
Meth od s. 1-(p-Cya n op h en oxy)-2-p en ta (eth ylen e gly-
col)eth a n e (1). To an ice cooled solution of 0.105 g (0.88
mmol) of 4-cyanophenol, 0.5 g (1.77 mmol) of hexa(ethylene
glycol), 0.464 g (1.77 mmol) of triphenylphosphine in anhy-
drous dioxane 15 mL was added dropwise 0.28 mL (1.77 mmol)
of DEAD over a period of 30 min. The reaction mixture was
left for stirring at 0 °C for 2 h. TLC analysis (ethyl acetate:
methanol, 9:1) at this point indicated the completion of the
reaction. Solvents were evaporated, and the residue was
dissolved in ethyl acetate and washed with 5% sodium
bicarbonate, washed with water, and dried over anhydrous
sodium sulfate. The solvent was evaporated, and the residue
was purified by flash chromatography over silica gel using
ethyl acetate and a gradient of methanol (1-10%) to yield
0.250 g (0.950 mmol, 54%) of an yellow oil. Rf in ethyl acetate:
methanol 9:1 ) 0.39. 1H-NMR (CDCl3) δ ) 3.02 (bs, 1H),
3.52-3.73 (m, 20H), 3.84 (t, 2H), 4.14 (t, 2H), 6.91 (d, 2H),
23357, 1636, 1610, 1485, 1454, 1288, 1257, 1184, 1112 cm-1
.
HRMS: Calcd for C37H49N6O7 (M + H+) 689.36627, found
689.36660.
DNA Syn th esis. The 15-mers were synthesized using
standard phosphoramidite protocols.23 To attach the Hoechst
derivative to the 15-mers through the hexa(ethylene glycol)
linker, the following procedure was employed: After the final
nucleoside coupling on the solid support, the DMT group was
removed under standard conditions using trichloroacetic acid.
After washing the support with anhydrous acetonitrile, equal
volumes of 1 M (2-cyanoethoxy)bis(diisopropylamino)phosphine
and 0.5 M tetrazole, both dissolved in anhydrous acetonitrile,
were added to the support. After a reaction period of 30 min,
the support was washed with anhydrous acetonitrile, and this
phosphitylation step was repeated a second time. Alterna-
tively, a solution of 1.0 M of (2-cyanoethoxy)(diisopropylamino)-
chlorophosphine and 1.2 M diisopropylethylamine in aceto-
nitrile was employed. To complete the coupling, equal volumes
of 0.5 M tetrazole/acetonitrile and a 0.1 M solution of the
Hoechst derivative 4 dissolved in anhydrous DMF were added
to the support, and the reaction was continued for 16 h. After
washing the support with acetonitrile, the acetylation and
oxidation steps were performed in the usual manner. The
conjugate was deprotected and removed from the support by
treatment with concentrated aqueous ammonia at 50 °C for
16 h.
HPLC using a 9.4 × 250 mm column of ODS-Hypersil was
used to purify the conjugated oligonucleotides using a linear
gradient of 10-40% buffer B over 30 min (buffer A: 50 mM
triethylammonium acetate, pH 7.0, buffer B: 50 mM triethyl-
ammonium acetate, pH 7.0 containing 70% acetonitrile). An
estimation of the efficiency of the final coupling step could be
determined by comparing the ratio of the free 15-mer (failed
conjugation product) to the conjugated 15-mer obtained in the
HPLC isolation procedures. This final coupling step varied
somewhat in yield between 50 and 75%.
(14) Connolly, B. A.; Rider, P. Nucleic Acids Res. 1985, 13, 4485-
4502.
(15) Connolly, B. A. Nucleic Acids Res. 1987, 15, 3131-3139.
(16) Cocuzza, A. J . Tetrahedron Lett. 1989, 30, 6287-6290.
(17) Asseline, U.; Bonfils, E.; Kurfu¨rst, R.; Chassignol, M.; Roig, V.;
Thuong, N. T. Tetrahedron 1992, 48, 1233-1254.
(18) Thuong, N. T.; Asseline, U.; Moneney-Garestier, T. In Oligo-
deoxynucleotides. Anti-sense Inhibitors of Gene Expression; Cohen, J .
S., Ed.; CRC Press, Inc.: Boca Raton, FL, 1989; pp 25-52.
(19) Lemaitre, M.; Bayard, B.; Lebleu, B. Proc. Natl. Acad. Sci.
U.S.A. 1987, 84, 648-652.
(20) Sinyakov, A. N.; Lokhov, S. G.; Kutyavin, I. V.; Gamper, H. B.;
Meyer, R. B. J . Am. Chem. Soc. 1995, 117, 4995-4996.
(21) Wiederholt, K.; Rajur, S. B.; Giuliano, J . J .; O’Donnell, M. J .;
McLaughlin, L. W. J . Am. Chem. Soc. 1996, 118, 7055-7062.
Nu cleosid e An a lyses. To 0.5 A260 units of oligonucleotide
in 100 µL of 50 mM Tris‚HCl, pH 8.0, containing 100 mM
(22) Loewe, H.; Urbanietz, J . Arzneimettelforsch. 1974, 27, 1927-
1933.
(23) Matteucci, M. D.; Caruthers, M. H. J . Am. Chem. Soc. 1981,
103, 3185-3191.

Hoechst 33258 Tethered to Oligodeoxynucleotide 15-Mers
J . Org. Chem., Vol. 62, No. 3, 1997 525
MgCl2 were added 0.003 units of snake venom phoshodi-
esterase and 1 unit of alkaline phosphatase and the mixtures
incubated 2-6 h at 37 °C. The digestion products (ap-
proximately 0.2 A260 units) were analyzed by reversed phase
HPLC using a 4.6 × 250 mm column of ODS-Hypersil and a
gradient of 5% buffer B for 10 min, followed by a linear
gradient from 5-100% buffer B over 10 min (buffer A and
buffer B were as described above). Under these conditions,
the following retention times were observed (min): dC 3.1,
dm5C, 5.2, dG 7.5, T 8.0, dA 9.2, and Hoechst derivative 4 (see
Scheme 1) 17.5.
Sch em e 1
Th er m a l Den a tu r a tion Stu d ies. Thermal denaturation
studies were performed in 21 mM HEPES pH 7.5, 100 mM
NaCl and 20 mM MgCl2 at duplex concentrations of ∼1 µM.
Absorbance (260 nm) and temperature values were measured
with an AVIV 14DS UV/visible spectrophotometer equipped
with digital temperature control. The temperature of the cell
compartment was increased in 1.0 °C steps (from 0 to 95 °C),
and when thermal equilibrium was reached, temperature and
absorbance data were collected. Tm values were determined
both from first-order derivatives and by graphical analysis of
the absorbance vs temperature plots.
F lu or escen ce Stu d ies. Fluorescence emission spectra
were collected on a Shimadzu RF5000U fluorescence spectro-
photometer containing a Shimadzu DR-15 microprocessor and
graphics display terminal. Spectra were obtained in 21 mM
HEPES pH 7.5, 100 mM NaCl and 20 mM MgCl2 at duplex
concentrations of ∼1 µM. All measurements were performed
with the following list of parameters: slit width Ex/Em ) 10
nm/10 nm, low sensitivity, medium speed. Samples were
introduced into a 1.25 mL cell thermally isolated with a water
jacket. Temperature was controlled with a recirculating water
bath. Excitation wavelength ) 342 nm, emission range )
380-600 nm.
Resu lts a n d Discu ssion
(ethylene glycol) linker. With this design, relatively few
new steps were required; p-cyanophenol could be alky-
lated by hexa(ethylene glycol) using a Mitsunobu proto-
col29 to generate 1 (Scheme 1). The cyano group was then
converted to the imidate 2, and without purification, the
imidate was reacted with the diamine 3 obtained using
the described synthetic procedures for Hoechst 33258.22
The product of this cyclization (4) was simply the Hoechst
33258 fluorophore with a covalently attached hexa-
(ethylene glycol) linker.
The final step in this process required that the bis-
benzimidazole derivative with the hexa(ethylene glycol)
linker be converted into a phosphoramidite. This reac-
tion proved to be problematic. Regardless of the condi-
tions used, we were unable to isolate any significant
quantities of the phosphoramidite derivative 5 (Scheme
1). The most likely explanation for this difficulty is that
the relatively nucleophilic imidazole nitrogens competed
for the reactive chlorophosphine derivative. After stand-
ard aqueous workup, the phosphorus imidazole bond
would hydrolyze regenerating the starting materialsother
side reactions are of course also possible. Although we
considered the possibility of introducing protecting groups
at these sites, we opted for a more pragmatic approach.
The alternative strategy for the preparation of the DNA-
Hoechst conjugate involved coupling the linker to the 5′-
terminus of the oligonucleotide using what might be
termed a “reverse” protocol. That is, the phosphoramid-
ite would be incorporated into the support-bound oligo-
nucleotide, and the coupling would be facilitated by
adding 4 in the presence of tetrazole. Similar approaches
to coupling at the 5′-terminus have been described in
earlier studies in which phosphotriester chemistry was
In our initial attempts at preparing a DNA-Hoechst
33258 conjugate, we wanted to synthesize a bis-benz-
imidazole derivative tethering an appropriate linker and
then functionalize the linker as a phosphoramidite. We
chose hexa(ethylene glycol) for the linker. The 18 atom
length of this ethylene glycol oligomer should provide
adequate flexibility in order for the Hoechst derivative
to reach an appropriate A-T rich minor-groove binding
site, while the polyether character results in potentially
advantageous hydrophilic characteristics for complexes
formed in aqueous solutions. This linker has also
functioned well for conjugates used to study DNA tri-
plexes.24 Hexa(ethylene glycol) and similar glycol-based
linkers have been employed in other studies involving
both DNA25 and RNA26-28 complexes.
Syn th esis. Our attempts to attach the hexa(ethylene
glycol) linker directly to the parent Hoechst 33258
fluorophore failed under a variety of conditions. The
alternative strategy involved incorporating the linker into
the bis-benzimidazole chromophore during a complete
synthesis of a Hoechst 33258 derivative.22 In this latter
approach, we could take advantage of the described
synthetic procedures for the Hoechst fluorophore and
simply alter the components of the final cyclization step
to provide a derivative containing the desired hexa-
(24) Robles, J .; Rajur, S. B.; McLaughlin, L. W. J . Am. Chem. Soc.
1996, 118, 5820-5821.
(25) Durand, M.; Chevrie, K.; Chassignol, M.; Thuong, N. T.;
Mauizot, J . C. Nucleic Acids Res. 1990, 18.
(26) Benseler, F.; Fu, D. J .; Ludwig, J .; McLaughlin, L. W. J . Am.
Chem. Soc. 1993, 115, 8483-8484.
(27) Ma, M. Y.-X.; McCallum, K.; Climie, S. C.; Kuperman, R.; Lin,
W. C.; Sumner-Smith, M.; Barnett, R. W. Nucleic Acids Res. 1993, 21,
2585-2589.
(28) Thomson, J . B.; Tuschl, T.; Eckstein, F. Nucleic Acids Res. 1993,
21, 5600-5603.
(29) Mitsunobu, O. Synthesis 1981, 1-28.

526 J . Org. Chem., Vol. 62, No. 3, 1997
Rajur et al.
Sch em e 2
employed.30 Using this approach, a fully protected sup-
port-bound 15-mer was initially assembled using stand-
ard protocols. The terminal DMT protecting group was
removed using trichloroacetic acid to generate the sup-
port-bound oligomer such as 6, containing a 5′-terminal
hydroxyl (Scheme 2). Conversion of 6 to the correspond-
ing phosphoramidite 7 was accomplished by two rounds
of phosphitylation using (2-cyanoethoxy)bis(diisopropyl-
amino)phosphine and 0.5 equiv of tetrazole (or using the
corresponding chlorophosphine derivative in the presence
of base). Two treatments may not be required to generate
7, but since we could not confirm the extent of the
reaction, two cycles of phosphitylation seemed a prudent
course of action. The coupling reaction between 4 and 7
proceeded by adding sufficient solution (0.1 M of 4 and
0.5 M of tetrazole) to immerse the solid support. We
could detect significant, but less than optimal quantities
of conjugate after a 1-2 h reaction period (<50%), but
since the extent of conjugation can only be judged after
complete workup of the crude product followed by HPLC
analysis (see below), it was generally more efficient to
simply allow the reaction to proceed for an extended
period of time, typically overnight. After a reaction
period of approximately 16 h, the support was washed
and oxidized in the normal fashion to complete the
coupling protocol and generate conjugates of varying
sequence composition such as 8 (Scheme 2).
F igu r e 1. (a) HPLC analysis of the crude reaction mixture
obtained after ammonia deprotection of the conjugate 8. a: 15-
Mer polypyrimidine DNA sequence, b: Hoechst-15-mer con-
jugate, c: Hoechst-hexa(ethylene glycol) derivative 4. (b)
HPLC analysis of the snake venom phosphodiesterase/bacte-
rial alkaline phosphatase digest of a DNA 15-mer conjugated
to the Hoechst derivative 4. * marks unidentified contaminants
in the buffers.
of the crude material indicated the presence of three
species (Figure 1a). The earliest eluting peak corre-
sponds to the unconjugated 15-mer. The peak multiplic-
ity for this first product likely reflects some variation in
the end group character (with or without a terminal
phosphorus species). The latest eluting peak corresponds
to 4. We believe that its presence in the crude material
is mostly likely the result of nonspecific binding by 4 to
the solid support during the final coupling reaction.
Traces of 4 may remain nonspecifically bound to the
support after the various washings, and only under
extended treatment with hot ammonia are they removed.
Alternatively, the presence of 4 might represent some
minor cleavage of the conjugate 8 during ammonia
deprotection. The major peak of the analysis in Figure
1a corresponds to the fully deprotected derivative corre-
sponding to 8. Its identity could be confirmed by diges-
tion with snake venom phosphodiesterase and alkaline
phosphatase. Analysis of the digestion mixture indicated
The extent of the conjugation reaction could only be
judged after treatment of the coupling product with
aqueous ammonia to effect deprotection of the conjugate
and its removal from the solid support. HPLC analysis
(30) Kempe, T.; Sundquist, W. I.; Chow, F.; Hu, S.-L. Nucleic Acids
Res. 1985, 13, 45-57.

Hoechst 33258 Tethered to Oligodeoxynucleotide 15-Mers
J . Org. Chem., Vol. 62, No. 3, 1997 527
Ta ble 1. Th er m a l a n d F lu or escen ce P r op er ties of Dou ble-Str a n d ed DNA Com p lexes Teth er in g a Hoech st 33258
Der iva tive
Tm valuesa
duplexd
X ) C/M, °C
conjugatee
X ) M, °C
entry
1
sequencec
∆Tm,a °C
∆Fmax
b
(H)-eg6-TTTTTTTTXTXTXTT
AAAAAAAAGAGAGAA
(H)-eg6-XTTTTTTTXTXTXTT
GAAAAAAAGAGAGAA
(H)-eg6-XXTTTTTTXTXTXTT
GGAAAAAAGAGAGAA
(H)-eg6-XTXTTTTTXTXTXTT
GAGAAAAAGAGAGAA
(H)-eg6-TTGGTTTTXTXTXTT
AACCAAAAGAGAGAA
(H)-eg3-TTGGTTTTXTXTXTT
AACCAAAAGAGAGAA
(H)-eg6-TTTGGTTTXTXTXTT
AAACCAAAGAGAGAA
(H)-eg6-XTXTXTXTXTXTXTT
GAGAGAGAGAGAGAA
49.2/57.5
53.0/56.5
52.8/57.6
53.1/58.1
57.0/59.4
53.8/59.4
52.2/57.6
56.5/64.1
62.1/70.2f
12.9/12.7
29
2
3
4
5
6
7
8
72.4
5.9
13.8
13.0
13.4
13.1
10.5
3.6
19
25
20
22
16
16
5
71.4
71.1
72.8
72.5
68.1
67.7
a
Values are the average of at least two determinations (+0.5 °C). ∆Tm values represent the change in Tm when the conjugate was
b
incorporated into the M-containing duplex. Fluorescence enhancement measured as the ratio of the emission maxima for the double-
stranded conjugate relative to the single-stranded conjugate. c (H)-eg6- represents Hoechst 33258 tethered to the 5′-terminus of the DNA
sequence through a hexa(ethylene glycol) linker. (H)-eg3- is a similar conjugate, but the tether is a tri(ethylene glycol) linker. X ) C
(2′-deoxycytidine), or M (2′-deoxy-5-methylcytidine). Tm values for the parent 15-mer duplex lacking the Hoechst 33258 conjugate. e Tm
d
values for the conjugated 15-mer duplex. f Tm values for this conjugate are reported for both the C- and M-containing complexes (C/M).
the correct ratio of nucleosides and one equivalent of 4
(Figure 1b). A comparison of the early eluting peak(s),
representing the unconjugated 15-mer, with the major
peak, representing the conjugate (see Figure 1a), indi-
cated that the conjugation reaction (both phosphitylation
and coupling) proceeded with yields varying from 50 to
75%.
nonconjugated duplexes containing the methylated C-
derivative exhibited higher Tm values than those ob-
served for the C-containing sequences (Table 1). This
observation is consistent with those described previ-
ously31 in which DNA duplexes containing the C-5
methylated pyrimidine T always exhibited higher Tm
values than those containing the corresponding non-
methylated 2′-deoxy derivative U.
These yields do not compare with the high yield (∼98%)
phosphoramidite couplings typically observed with stand-
ard DNA assembly protocols, but they resulted in mate-
rial sufficient for the subsequent studies. In fact this
procedure produced much more material than could be
obtained using the previously described approach with
an internally placed linker conjugated by a postsynthetic
alkylation reaction.21 It should be noted that with
standard DNA assembly protocols,23 the reactive phos-
phoramidite species is present in solution at a concentra-
tion far in excess of the support-bound oligonucleotide.
With such protocols, trace amounts of water only mar-
ginally affect the quantity of the reactive species present
and do not impact the coupling efficiency to any great
extent. When this process is reversed, that is the reactive
phosphoramidite species is now attached to the support
bound oligonucleotide, there is no excess in the reactive
species present. In these cases, even minute quantities
of water will have a significant and detrimental effect
on the coupling efficiency by reacting with the support-
bound oligonucleotide. It is therefore necessary to ensure
that these conjugation reactions are performed under
anhydrous conditions.
Tm Va lu es for Hoech st-DNA Con ju ga tes. We have
examined the thermal characteristics of double-stranded
DNA complexes of varying sequence composition tether-
ing the Hoechst 33258 fluorophore (Table 1). We began
with a 15-mer sequence (entry 1, Table 1) in which the
terminal portion of the duplex (that nearest the conju-
gated Hoechst derivative) contained eight A-T base pairs.
Initially we examined a homopurine-homopyrimidine
duplex in which the C-residues of the pyrimidine strand
were either 2′-deoxycytidine (C) or the corresponding
5-methyl derivative (M). In all cases examined, the
Tethering the Hoechst derivative through the hexa-
(ethylene glycol) linker increased the Tm value for both
duplexes (e.g., C-containing and M-containing) duplex by
approximately 13 °C. This observation suggests that the
tethered groove-binder can fold back into the minor
groove as designed and provide additional binding inter-
actions to stabilize the DNA duplex. We cannot at this
time exclude the possibility that some intermolecular
ligand-helix binding occurs. That is, the ligand from one
conjugated duplex binds to the minor groove of a second
duplex. But at the concentrations employed, this process
seems unlikely. The combination of the tethered ligand
and the use of 2′-deoxy-5-methylcytidine (M) residues
results in Tm values that are increased by as much as 20
°C relative to the simple non-conjugated C-containing
hybridization complex (see entry 1, Table 1). With the
exception of this initial sequence, the remaining Hoechst
conjugates have been prepared from the more stable
M-containing sequences to take advantage of the en-
hanced stability provided by the C-5 methyl group in
these complexes. In order to judge the available flex-
ibility for minor-groove binding by the tethered Hoechst
ligand, we examined a series of duplexes formed from
various sequences in which the extended (A-T)8 minor
groove binding site was disrupted by the presence of
various G-C (C-G) base pairs. For example, the introduc-
tion of a terminal G-C base pair into the duplex (entry
2, Table 1) at the terminus of the duplex resulted in only
a minor increase in the Tm value for the native duplex.
The Tm for the corresponding Hoechst conjugate in-
creased by nearly 16 °C. Although the Hoechst deriva-
(31) Wang, S.; Kool, E. T. Biochemistry 1995, 34, 4125-4132.

528 J . Org. Chem., Vol. 62, No. 3, 1997
Rajur et al.
tives do not bind well to sites containing GC base pairs,
the length of the glycol linker presumably permits the
groove-binding agent to “reach” beyond the G-C base pair
and bind within that portion of the minor groove contain-
ing seven A-T base pairs.
“reach” a binding site four base pairs removed from the
end of the duplex suggests that the longer linker will be
effective in permitting binding to sites even further
removed from the terminus of the hybridization complex.
F lu or escen ce Ch a r a cter istics. Upon hybridization
to the target sequence, the fluorescence signal from the
tethered Hoechst derivative is enhanced, as the result
of binding in the minor groove. The enhanced quantum
yield observed for these tethered Hoechst conjugates
varies somewhat with target sites, but as long as a four
base pair binding site (with a minimum of three A-T base
pairs) is present, the emission quantum yield is observed
to increase by at least 16-fold and in some cases nearly
30-fold. These values are generally in the same range
as those reported previously for the free ligand6 and also
similar to previous work in which the ligand was tethered
to an internal site within the DNA duplex.21 The strong
fluorescence signals observed for all the complexes
containing an appropriate minor groove binding site
suggests that the tethered ligand is capable of penetrat-
ing deeply within the minor groove even when the target
binding site is removed from the site of the linker by four
or even five base pairs. In this respect, the tethered
ligand functions well to report upon the success of the
hybridization event. Recognition of the target sequence
by the single-stranded conjugate results in the formation
of the minor groove which triggers a ligand binding event
which results in a complex with a strong fluorescence
emission signal.
Six of the duplexes containing a four base pair A-T rich
binding site all result in Tm values of at least 71 + 1 °C;
the enhancement in quantum yield varies by as much
as a factor of nearly twosfrom 16-fold (entry 6) to nearly
thirty-fold (entry 1). These variations in fluorescence
characteristics likely reflect some polymorphism in ligand
binding to the sequences. With a four base pair A-T
binding site and the hexa(ethylene glycol) linker, fluo-
rescence enhancements of at least 20-fold are observed
and suggest similar modes of binding. Reducing the size
of the binding site from four to three A-T base pairs, or
reducing the length of the linker to tri(ethylene glycol)
(with the binding site removed by four base pairs) results
in a moderate decrease in fluorescent signal, presumably
as a result of some modulation in the nature of the ligand
binding. It is noteworthy that even with the alternating
A-T/G-C sequence, a 5-fold enhancement in quantum
yield was measured. This signal suggests that some
groove binding occurs even to sequences high in G-C
content. However, the presence of the guanine N2-amino
groups likely prevents deep penetration into the groove
structure, and in such locations, relaxation of the excited
state by radiative mechanisms is not likely to predomi-
nate. This latter complex (entry 8) results in only a 5-fold
enhancement in quantum yield.
The addition of a second adjacent G-C base pair to the
end of the duplex has only a moderate effect on the Tm
value for the conjugated complex. Its presence both
reduces the size of the A-T rich binding site, and shifts
it to a position that is now two base pairs from the end
of the duplex. The minimum required binding site for
Hoechst 33258 is generally considered to be four A-T (T-
A) base pairs, and the six A-T base pairs present in the
conjugate of entry 3 provide an effective target for the
tethered ligand. Replacement of positions 1 and 3 by G
residues (entry 4, Table 1) also results in a conjugated
complex with essentially the same Tm value as observed
for the previous entries (Table 1). The minor groove
binding site in this duplex is reduced to five A-T base
pairs, but is still sufficient for effective binding the
Hoechst dye. Placing two C-G base pairs at positions 3
and 4 of the duplex still results in a binding site of four
A-T base pairs (entry 5, Table 1), but the binding site is
shifted four residues from the end of the helix. The Tm
value for this complex is virtually indistinguishable from
those complexes containing longer runs of A-T base pairs
(entries 1-4, Table 1).
By comparison, placement of two G-C base pairs into
the center of the (A-T)8 double-stranded sequence results
in two (A-T)3 binding sites. In this case, the stability of
the double-stranded conjugate is reduced somewhat from
those containing longer A-T rich binding sites, but a 10
°C increase in Tm suggests that the Hoechst ligand is still
able to bind reasonably effectively to this target sequence
(entry 7, Table 1). One of the crystal structure analyses1
of a DNA duplex containing a bound Hoechst ligand
indicates that the bis-benzimidazole binds across three
of the A-T base pairs with the N-methylpiperazine ring
located within the minor groove of the adjacent G-C base
pair. In this case, the four-base pair binding site contains
three A-T base pairs and one G-C base pair. A similar
sequence target is present with the complex illustrated
by entry 7. We have not determined whether the
tethered ligand binds to the terminal or internal (A-T)3
target site. However the noted crystal structure analysis
indicates that the Hoechst ligand binds to a GAAA
sequence (internal site) rather than a CAAA sequence
(terminal site) (see entry 7). The final duplex of this
series is that represented by entry 8 (Table 1) which is
composed of alternating A-T and G-C base pairs. In
principle the Hoechst ligand binds poorly to such a
sequence, but the 3.6 °C increase in Tm for this conjugate
suggests that some form of binding to duplex is present
when the Hoechst ligand is conjugated to one of the
strands of the duplex, and this binding provides at least
a measurable increase is duplex stability.
Con clu sion s
The hexa(ethylene glycol) linker provides sufficient
flexibility that the ligand can reach a binding site at least
four to five base pairs from the end of the helix (entry 5,
Table 1). We took the conjugate represented by entry 5
of Table 1 and reduced the length of the linker from the
18 atoms present in hexa(ethylene glycol) to only 9 atoms
by using tri(ethylene glycol) as the linker. The subse-
quent conjugate was capable of stabilizing the duplex to
the same extent as did the conjugate containing the
longer linker (compare entries 5 and 6 of Table 1). That
the shorter 9-atom linker permits the tethered ligand to
Tethering a Hoechst chromophore to the terminus of
a DNA sequence through a polyethylene glycol linker can
be accomplished during the solid-phase based assembly
of DNA sequences. The presence of the ligand enhances
duplex stabilization, providing that a four base pair
binding site is present (containing at least three A-T base
pairs). Owing to the length of the tether, it is not
necessary that this binding site be at the terminus of the
hybridization complex, that is, near the site at which the
linker is attached. The binding site can be at least four

Hoechst 33258 Tethered to Oligodeoxynucleotide 15-Mers
J . Org. Chem., Vol. 62, No. 3, 1997 529
to five base pairs removed from the end of the duplex
and the minor groove binding ligand is still able “reach”
the binding site. The use of the tethered Hoechst
derivative, and the substitution of 5-methylcytosine in
place of C, results in Tm values that are increased by as
much as 20 °C relative to the nonconjugated C-containing
duplexes. Upon binding, a dramatic enhancement in the
fluorescence emission signal results, and these enhance-
ments, as much as 30-fold over the value observed for
the single-stranded conjugate, can be used to report upon
the success of the hybridization event.
Ack n ow led gm en t. J .R. is a fellow of the Ministerio
de Educacio´n y Ciencia of the Spanish Government.
This work was supported by a grant from the NIH (GM
37065).
J O9618536