Solid phase synthesis of oligonucleotides tethered to oligo-glucose phosphate tails

Article abstract of DOI:10.1016/S0040-4020(02)00684-1

Oligonucleotides conjugated at both 3′ and 5′-ends with glucose residues, 4,6-linked through a phosphodiester bridge, have been synthesized by sequential addition of a 6-O-DMT-glucose-4-phosphoramidite building block following a standard automated ODN ass

Full text of DOI:10.1016/S0040-4020(02)00684-1

Tetrahedron 58 (2002) 6697–6704
Solid phase synthesis of oligonucleotides tethered to oligo-glucose
phosphate tails
a
a
Matteo Adinolfi,a Lorenzo De Napoli,a Giovanni Di Fabio, Alfonso Iadonisi,
,
*
Daniela Montesarchioa and Gennaro Picciallib
a
`
Dipartimento di Chimica Organica e Biochimica, Universita degli Studi di Napoli ‘Federico II’, via Cynthia 4, 80126 Napoli, Italy
`
Dipartimento di Chimica delle Sostanze Naturali, Universita degli Studi di Napoli ‘Federico II’, via D. Montesano 49, 80131 Napoli, Italy
b
Received 8 March 2002; revised 31 May 2002; accepted 20 June 2002
Abstract—Oligonucleotides conjugated at both 30 and 50-ends with glucose residues, 4,6-linked through a phosphodiester bridge, have been
synthesized by sequential addition of a 6-O-DMT-glucose-4-phosphoramidite building block following a standard automated ODN assembly
procedure. Two 30,50-bis-glycoconjugated 18-mers, designed for antisense experiments, have been prepared and their hybridization
properties with a complementary DNA fragment evaluated by UV thermal analysis. q 2002 Elsevier Science Ltd. All rights reserved.
1. Introduction
natural ribose moieties in ODNs with suitable mimics in
many cases led to enhancements in terms of bioavailability
and nuclease resistance. As a matter of fact, a large number
of differently modified ODNs are in clinical trials7,8 and
an antisense 21-mer phosphorothioate ODN (ISIS-2922,
Vitravene) has been recently approved as an antiviral
agent.9 However, severe modifications, involving dedicated,
sometimes cumbersome synthetic protocols, in most cases
limit the accessibility of modified ODNs to very specialized
laboratories.
Since the pioneering work of Zamecnick and Stephenson in
the early 80’s,1 the scientific community recognized the
wide potential of synthetic oligodeoxyribonucleotides
(ODNs) as new therapeutic agents, able to specifically and
efficiently block exogenous genetic messages without
damaging the normal functions of the hosting cell. Gene
expression inhibition by ODNs of specific sequence can
be achieved targeting mRNA (antisense strategy)2 by
formation of stable hybrid DNA–RNA duplexes, where
the RNA strand is rapidly degraded by endogenous RNase
H. Alternatively, synthetic ODNs of specific sequences can
hybridize double stranded DNA tracts (antigene strategy)3
through formation of triple helical complexes, thus directly
interfering with transcription processes. Moreover, ODNs
can show highly selective recognition properties for specific
proteins as well, acting as aptamers,4 thus offering an
extremely efficient and versatile approach to intervene at
any stage of a pathogenic genetic information transfer, as,
for example, of a virus or of an oncogene.
As a valid alternative to a major chemical modification of
the native structure of ODNs, conjugation is currently
largely exploited for in vivo applications of ODNs: this
strategy involves the preservation of the original ODN
backbone (and therefore of the affinity towards the natural
target) to which, at one or both ends of the chain, are
covalently attached molecules imparting peculiar functions,
such as peptides, polyethylene glycol, intercalators, fluor-
escent labels, or hydrophobic molecules as steroids, fatty
acids, long chain alcohols, fullerene derivatives, etc.10–12
By masking one or both ends of the ODN with any sterically
demanding residue, a high degree of in vivo stabilization is
definitively achieved, since exonucleases—by far the most
abundant nucleases in cells—require free 30 or 50-OH ends
to show nucleolytic activity. On the other hand, in order to
reach an effective ODN concentration in cells, a specific
delivery strategy has to be addressed, for example by linking
specific ‘carriers’, as RGD or nuclear localization signal
(NLS) peptide sequences, cationic tails as polylysine resi-
dues, PEG units, nanoparticles or hydrophobic molecules,
to the ODN chain.
The dramatically low cellular uptake of ODNs associated
with their very short half-life in cells, due to rapid
degradation by nucleases, generally renders ‘natural’
ODNs completely inactive in in vivo systems. To improve
the pharmacological profile of oligonucleotides, various
chemical modifications, either at the level of the hetero-
cyclic bases or of the sugar-phosphate backbone of the ODN
chain, have been extensively investigated.5,6 The replace-
ment of the normal 30,50-phosphodiester linkages or of the
Keywords: nucleic acids analogues; phosphoramidites; solid phase
synthesis; carbohydrate mimetics; antitumor compounds.
Oligosaccharides can also be useful carriers, relying on
specific recognition mechanisms based on sugar-binding
*
Corresponding author. Fax: þ39-081-674393; e-mail: denapoli@unina.it
0040–4020/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(02)00684-1

6698
M. Adinolfi et al. / Tetrahedron 58 (2002) 6697–6704
Scheme 1. (a) Ac2O, pyridine, quantitative yield; (b) I2, Et3SiH cat., CH3OH 70%; (c) DMTCl, TEA, DMAP cat., pyridine, 90%; (d) ClP(OCE)N(i Pr)2, DIEA,
DCM, 90%; (e) succinic anhydride, DMAP, pyridine, 92%; (f) CPG-NH2, DCCI, pyridine.
membrane receptors (lectins).13 In this context, several
protocols have been recently proposed to link a carbo-
hydrate moiety to ODNs.14–16 We recently reported17,18 the
first examples of direct, on-line glycosidation of the 50-end
of ODNs while 0still anchored to the CPG solid matrix and
synthesized 30,5 -bis-glucosylated oligomers starting from a
new glucose functionalized support. However, this strategy
is strictly limited to the insertion of a single sugar residue at
the end of the ODN chains and its extension can hardly be
envisaged to the synthesis of oligosaccharide–oligonucleo-
tide conjugates. In fact, despite the notable efforts devoted
to the solid phase synthesis of oligosaccharides,19,20 the
available methodologies are still far from being optimized
so as to be successfully applied to a repetitive, high yielding
oligomerization process. Further constraints are also present
in this case, where the synthetic strategy for the assembly of
the oligosaccharide portion must also be completely
chemically compatible with the presynthesized DNA tract.
In order to circumvent these limitations, mainly due to the
glycosidation step, suitable oligosaccharide mimics can
therefore be fruitfully devised.
growing oligonucleotide chain by standard, automated
ODN synthesis protocols.27
With the aim of extending this methodology to the synthesis
of stable and easy-to-handle oligosaccharide surrogates as
new conjugating molecules for ODNs, we describe herein
the synthesis and utilization of new glucose phosphorami-
dite building block 5. This molecule, bearing the DMT as a
transient protecting group for the 6-OH function, and the
2-cyanoethyl, N,N-diisopropylamino-phosphoramidite
group as the reactive centre linked to the 4-position, can
be sequentially coupled on an insoluble matrix. The usage of
5 in connection with glucose-functionalized support 717
has allowed us to successfully prepare 30,50-bis-glucose-
phosphate ODNs conjugates.
2. Results and discussions
Glucose building block 5 contains all the typical functional
groups required in nucleotide monomers for the standard
solid phase ODN assembly. This allows the efficient,
sequential addition of a defined number of glucose-
phosphate residues tethered to an ODN sequence. Its
synthesis (see Scheme 1) has been carried out in four
steps and 57% overall yields, starting from methyl 4,6-O-
benzylidene-a-D-glucopyranoside 1, which was converted
into peracetylated 2 in quantitative yields by treatment with
Ac2O/pyridine. Then neat removal of the benzylidene
protecting group was achieved by reaction with I2 and cat.
Et3SiH in CH3OH, furnishing 3.28 Selective dimethoxy-
tritylation of the primary hydroxy function, performed by
reaction with DMTCl and cat. DMAP in pyridine/TEA,
gave desired 4. This was next phosphitylated by addition of
2-cyanoethyl, N,N-diisopropylamino-chlorophosphorami-
dite and DIEA in anhydrous DCM, leading to 5. Support
7 was synthesized starting from intermediate 4, which, after
A very common procedure to conjugate a precious molecule
to an ODN chain involves its conversion into the corre-
sponding phosphoramidite derivative.10 Indeed this strategy
has also been exploited in the case of sugars14 or of base-
glycosylated nucleosides,21–25 which have been incorpo-
rated on the solid phase in high yields by following the same
efficient chemistry adopted for ODN assembly,26 resulting
in stable phosphodiester linkages between the conjugating
molecule and the 50 or the 30-OH terminus of the ODN.
A similar approach has also been recently proposed by
Joyce and co-workers, who designed a new class of
DNA–carbohydrate conjugates, called nucleo-glyco-
conjugates, based on the synthesis of three new carbo-
hydrate phosphoramidite derivatives, inserted at the
extremities or at defined, internal positions of the

M. Adinolfi et al. / Tetrahedron 58 (2002) 6697–6704
6699
amidite 5 was exploited in the solid phase synthesis of small
hybrids 8–10, 12 and 13 which, after purification, have been
fully characterized by 1H, 31P NMR and ESI MS techniques.
As described in Scheme 2, support 7 was first exploited in a
standard coupling with a nucleoside phosphoramidite
affording, after deprotection and detachment, 8 as a model
compound for the 30-conjugation.
As a test to study and optimize the extension of the glucose-
phosphate tail, consisting of glucose residues 4,6-linked
through phosphodiester bonds, 5 was then coupled to
glucosylated support 7, affording 9. The combined use of 5
and support 7 allowed the preparation of 30-diglucosylated
dinucleotide 10. In order to check the efficiency of coupling
with the 50-OH function of a nucleoside, 5 was then reacted
with CPG-bound deoxycytidi0ne 11, leading to0 12. As a
model compound for the 30,5 -conjugation, 30,5 -bisgluco-
sylated trinucleotide 13 was also synthesized. Remarkably,
in all the studied cases, phosphoramidite 5, coupled
following a standard, automated protocol to the 50-end of
a CPG-bound nucleoside (as for 12), of a growing ODN
chain (as for 13) or of a 6-OH function of a solid phase-
linked glucose (as for 9 and 10) showed absolutely similar
efficiencies as classical nucleoside phosphoramidites
(.98% yields, as evaluated by spectroscopic DMT test).
On
this
basis,
two
18-mers
of sequence
0
0
d(5 GCGTGCCTCCTCACTGGC3 ), carrying at both the 30
and 50-ends one or two glucose-phosphate residues (14 and
15, Fig. 1), have been synthesized starting from support 7,
following standard automated phosphoramidite procedures.
This ODN sequence is complementary to a tract of the
mRNA of protein kinase A type 1 (PKA1), subunit RIa,
which is overexpressed in the majority of human cancers
and well recognized as a relevant target for therapeutic
intervention against neoplasia; therefore several modifi-
cations of this sequence have been designed for antisense
experiments and are currently under clinical trial.31
Scheme 2. I: (a) 1% DCA in DCM; (b) coupling with T 30-phosphoramidite
building block; (c) 1% DCA in DCM; (d) NH4OH, 558C, overnight. II:
(a); (b) coupling with 5; (c); (d). III: (a); (b) three coupling cycles with 5,
dC and dA 30-phosphoramidite building blocks, respectively; (c); (d). IV:
(a); (b) four coupling cycles with dA, dC, T 30-phosphoramidite building
blocks and 5, respectively; (c); (d).
Detachment from the solid support and deprotection of the
conjugated oligomers were achieved by treatment with
concentrated aq. ammonia. Purification was carried out by
HPLC on an anionic exchange Nucleogel column; the
HPLC profile of crude 15 is showed in Fig. 2.
succinylation of the 4-OH function giving 6,30 was then
reacted with CPG-NH2 solid support as previously
reported,17 leading to an average functionalization of
0.050 mequiv./g, evaluated by DMT test. The identity and
purity of all the synthesized compounds have been
confirmed by NMR (1H, 13C and 31P) and mass spectral
data.
The collected peaks were then desalted by gel filtration
chromatography on a Sephadex G25 column eluted
with H2O. The isolated products were checked for purity
by HPLC on an analytical RP18 column, showing them
to be .98% pure. The synthesized compounds were
To test the feasibility of the proposed synthetic strategy for
the preparation of new ODN conjugates, glucose phosphor-
Figure 1. 30,50-Bis-glucoconjugates ODNs 14 and 15.

6700
M. Adinolfi et al. / Tetrahedron 58 (2002) 6697–6704
the number of coupling cycles with building block 5. This
strategy offers several advantages: (i) 5 can be obtained by
very straightforward and high yielding reactions; (ii) it can
be coupled in extremely high yields following standard
phosphoramidite protocols either to nucleoside-, oligo-
nucleotide-, or sugar-immobilized OH functions; (iii) it can
be exploited for several successive coupling cycles in an
automated synthesis, since no apparent reduction in the
efficiency of the coupling has been observed with increasing
the number of glucose-phosphate residues. Several small
hybrids have been synthesized and fully characterized
to test the feasibility of this synthetic protocol. Two
30,50-conjugated 18-mers, complementary to a region of
the mRNA of PKA1 (subunit RIa) which is a well studied
antisense target, have been synthesized and characterized by
MALDI TOF mass spectrometry. Experiments of duplex
formation of 14 or 15 with a complementary DNA 22-mer,
monitored by UV thermal denaturation analysis, showed
that the presence of the oligo-glucose phosphate tails at the
ends of the ODN chain did not impair its hybridization
properties. In addition, as demonstrated by the experiment
carried out on 14, this modification renders the ODNs
resistant to exonucleases.
Figure 2. HPLC profile of crude 15 on a Nucleogel SAX column (see
Section 4).
characterized by mass spectrometry; particularly, MALDI
TOF MS spectra of each ODN conjugate showed a single
peak in agreement with the calculated molecular weight.
Further studies are currently in progress to extend this
methodology to the on-line solid phase synthesis of ODNs
conjugated to phosphate-connected chains of other sugars or
preformed oligosaccharides, previously converted into
suitable phosphoramidite derivatives.
In order to assess the influence of the glucosyl-phosphate
tethers on the hybridization properties of the synthesized
ODNs, thermal denaturation studies were carried out by
mixing conjugates 14 or 15 with complementary 22-mer
0
0
d(5 TTGCCAGTGAGGAGGCACGCAT3 ) in a 1:1 ratio in
comparison with the same unmodified duplex. In both cases,
the affinity towards the complementary sequence, expressed
in terms of melting temperatures of the duplexes
(Tm¼608C), showed not to be affected by the presence of
the glucose-phosphate residues, thus demonstrating that this
kind of conjugation does not negatively interfere with the
recognition abilities of the ODN sequence. In renaturation
experiments, the absorbance vs temperature profile was
almost superimposable with the melting curve, showing the
process to be, as expected, quasi-reversible.
4. Experimental
4.1. Materials and methods
NMR spectra were recorded on Bruker WM-400 and Varian
Gemini 200 and 300 spectrometers. All chemical shifts are
expressed in ppm with respect to the residual solvent signal.
31P NMR spectra were recorded on a Bruker WM-400
spectrometer using 85% H3PO4 as external standard. The
solid support functionalizations were carried out in a short
glass column (5 cm length, 1 cm i.d.) equipped with a
sintered glass filter, a stopcock and a cap. The oligonucleo-
tides were assembled on a Millipore Cyclone Plus DNA
synthesizer, using commercially available 30-O-(2-cyano-
ethyl)-N,N-diisopropyl-phosphoramidite 20-deoxyribo-
nucleosides as building blocks.
The protecting effect of the glucose-phosphate residues
on the synthesized ODNs towards nuclease activities
was studied by incubating 14 with a 30-exonuclease in a
Tri0s–HCl buffer at 378C. The unmodified 18-mer
0
d(5 GCGTGCCTCCTCACTGGC3 ) was also incubated
with the enzyme in the same experimental conditions for
comparison. While the unmodified sequence was com-
pletely degraded within 2 h, glyco-conjugate 14 was totally
unaffected even after 72 h. Experiments aimed at evaluating
the antisense activity of the conjugated ODNs are currently
underway in collaboration with specialized laboratories.
The following abbreviations were used throughout the text:
4,40-dimethoxytrityl (DMT), dichloromethane (DCM), diiso-
propylethylamine (DIEA), N,N-dicyclohexylcarbodiimide
(DCCI), N,N-dimethylaminopyridine (DMAP), triethylamine
(TEA).
HPLC analyses and purifications were performed on a
Beckman System Gold instrument equipped with a UV
detector module 166 and a Shimadzu Chromatopac C-R6A
integrator. MALDI TOF mass spectrometric analysis was
performed on a PerSeptive Biosystems Voyager-De Pro
MALDI mass spectrometer using a picolinic/3-hydroxy-
picolinic acids mixture as the matrix. For the ESI MS
analysis a Waters Micromass ZQ instrument, equipped with
an Electrospray source, was used in the negative mode.
3. Conclusions
An efficient, on-line automated solid phase synthesis of a
new class of ODNs conjugated to oligosaccharide mimics,
easily obtained by sequential addition of novel DMT-
glucose phosphoramidite 5, has been described. The
proposed method allows the elongation of ODNs with
glucose-phosphate tails of the desired length by modulating

M. Adinolfi et al. / Tetrahedron 58 (2002) 6697–6704
6701
1
Thermal denaturation experiments were carried out on a
Jasco V-530 UV spectrophotometer with detection at l¼
260 nm and equipped with a Jasco ETC-505T temperature
controller unit.
2:3, v/v). H NMR (200 MHz, CDCl3) d 6.83–7.58 (13H,
m, aromatic protons); 5.30 (1H, t, J2,3¼J3,4¼9.7 Hz, H-3);
4.89 (1H, d, J1,2¼3.8 Hz, H-1); 4.83 (1H, dd, J1,2¼3.8 Hz,
J2,3¼9.7 Hz, H-2); 3.79 (6H, bs, –OCH3 DMT group);
3.62–3.76 (4H, overlapped signals, H-4, H-5, H2-6); 3.39
(3H, s, 1-OCH3); 2.75 (1H, d, J¼2.8 Hz, 4-OH); 2.08 (6H,
overlapped s’s, two –COCH3). 13C NMR (50 MHz, CDCl3)
d 171.2 and 170.4 (COCH3); 158.5, 135.7, 129.9, 128.0,
127.8, 126.7 and 113.1 (aromatic carbons DMT group); 96.5
(C-1); 86.4 (quaternary C DMT group); 72.8, 70.8, 70.8 and
69.8 (C-2, C-3, C-4 and C-5); 63.6 (C-6); 55.1 (–OCH3
DMT group); 54.9 (1-OCH3); 20.8 and 20.7 (two COC H3).
FAB MS (positive ions): m/z 581 (MþH)þ; 303 (DMT)þ.
4.2. Synthesis of phosphoramidite 5
Compound 1 (564 mg, 2 mmol) was dissolved in pyridine
(4 mL) and then acetic anhydride (2 mL) was added. After
2 h, CH3OH was added, the mixture was diluted with DCM
and the organic phase was washed three times with water.
Concentration of the organic phase afforded pure 2 (732 mg,
quantitative yield).
4.2.1. Compound 2. White solid, mp: 107–108.58C
(recrystallized from ethanol/water). Rf¼0.75 (hexane–
ethyl acetate 2:3, v/v). IR (KBr, cm21): 1748, 1460, 1379,
1235, 1200, 1146, 1126, 1093, 1061, 995, 963, 928, 775,
743, 703. 1H NMR (300 MHz, CDCl3) d 7.20–7.45 (5H, m,
aromatic protons); 5.56 (1H, t, J2,3¼J3,4¼9.8 Hz, H-3); 5.52
(1H, s, Ph-CH ); 4.91 (1H, d, J1,2¼3.8 Hz, H-1); 4.88 (1H,
To a solution of compound 4 (637 mg, 1.1 mmol) in
anhydrous DCM (7 mL) DIEA (580 mL, 3.3 mmol) and
2-cyanoethyl-N,N-diisopropylamino-chlorophosphorami-
dite (380 mL, 1.7 mmol) were added at room temperature
under argon. After 1 h the solution was diluted with ethyl
acetate and the organic phase was washed with brine and
then concentrated to afford a residue, which was then
purified by silica gel chromatography (eluent petroleum
ether–ethyl acetate 7:3 with few drops of TEA) to give 5 as
a mixture of diastereoisomers (778 mg, 90% yield).
dd, J1,2¼3.8 Hz, J2,3¼9.8 Hz, H-2); 4.30 (1H, dd, J5,6eq
¼
3.7 Hz, J6ax,6eq¼9.8 Hz, H-6eq); 3.92 (1H, m, H-5); 3.76
and 3.63 (1H each, two t’s, H-4 and H-6ax); 3.39 (3H, s,
–OCH3); 2.05 and 2.00 (3H each, s’s, two –COCH3). 13C
NMR (50 MHz, CDCl3) d 171.2 and 170.4 (COCH3); 136.6,
128.6, 127.8 and 125.8 (aromatic carbons); 101.1 (benzyl-
idene CH); 97.1 (C-1); 78.7, 71.1, 68.6 and 68.3 (C-2, C-3,
C-4 and C-5); 61.9 (C-6); 54.8 (1-OCH3); 20.3 (COCH3).
FAB MS (positive ions): m/z 367 (MþH)þ.
4.2.4. Compound 5. Glassy solid. Rf¼0.55 (hexane–ethyl
acetate 3:2, v/v). 1H NMR (400 MHz, CDCl3) d 6.85–7.61
(26H, m, aromatic protons); 5.48 (2H, overlapped t’s,
J2,3¼J3,4¼9.6 Hz, two H-3); 4.99 and 4.97 (2H, two d’s,
J1,2¼3.7 Hz, two H-1); 4.81 and 4.85 (2H, two dd’s, two
H-2); 3.67 (12H, overlapped s’s, –OCH3 of DMT group);
3.54 and 3.56 (3H each, two s’s, two 1-OCH3); 4.00–3.15
(16H, overlapped signals, H-4, H-5, H2-6, –OCH2CH2CN
and two CH(CH3)2); 2.30 and 2.54 (4H, overlapped t’s,
J¼9.8 and 9.8 Hz, two OCH2CH2CN); 2.01, 2.03, 2.06
and 2.09 (3H each, four s’s, four COCH3); 1.03, 1.01, 0.98,
0.92 (24H, four d’s, Jvic¼6.6 Hz, CH(CH3)2). 13C NMR
(50 MHz, CDCl3) d 170.0 and 169.6 (two COCH3); 158.1,
144.8, 136.0, 135.7, 129.9, 129.8, 127.9, 127.3, 126.3 and
112.6 (aromatic carbons DMT group); 117.2 and 116.9
(CN); 96.1 (C-1); 85.5 and 85.4 (quaternary C of DMT
group); 71.9, 71.1, 70.3 and 70.1 (C-2, C-3, C-4 and C-5);
63.2 and 62.5 (C-6); 58.4 and 58.0 (OC H2CH2CN); 54.7
and 54.6 (OCH3 of DMT group and 1-OCH3); 42.7 and 42.5
[C H(CH3)2]; 24.0 [CH(C H3)2]; 20.8 and 20.4 (two
COC H3); 20.0, 19.9, 19.6 and 19.4 (OCH2CH2CN). 31P
NMR (161.98 MHz, CDCl3) d 152.1 and 151.0.
To a solution of 2 (732 mg, 2.0 mmol) in CH3OH (11 mL)
iodine (220 mg, 0.87 mmol) and triethylsilane (16 mL,
0.10 mmol) were added. After 3 h at room temperature the
mixture was diluted with ethyl acetate and the organic phase
washed with an aqueous solution of sodium bicarbonate and
sodium thiosulfate. The organic phase was concentrated
under reduced pressure and the residue was purified by silica
gel chromatography (eluent DCM–CH3OH from 98:2 to
95:5) to afford diol 3 (385 mg, 70% yield).
4.2.2. Compound 3. Colourless oil. Rf¼0.45 (dichloro-
methane–methanol 9:1, v/v). 1H NMR (200 MHz, CDCl3) d
5.30 (1H, t, J2,3¼J3,4¼9.8 Hz, H-3); 4.91 (1H, d, J1,2
¼
3.8 Hz, H-1); 4.83 (1H, dd, J1,2¼3.8 Hz, J2,3¼9.8 Hz, H-2);
3.65–3.90 (4H, overlapped signals, H-4, H-5, H2-6);
3.40 (3H, s, 1-OCH3); 2.10 and 2.08 (3H each, s’s, two
–COCH3). 13C NMR (50 MHz, CDCl3) d 171.2 and 170.4
(COCH3); 96.6 (C-1); 72.6, 71.1, 70.8 and 68.6 (C-2, C-3,
C-4 and C-5); 61.2 (C-6); 55.0 (1-OCH3); 20.7 and 20.6
(COCH3). FAB MS (positive ions): m/z 279 (MþH)þ.
4.3. Functionalization of the resin. Support 7
Succinic anhydride (164 mg, 1.6 mmol) was added under
argon to a solution of 4 (630 mg, 1.1 mmol) and DMAP
(240 mg, 2.0 mmol) in anhydrous pyridine (5 mL). The
mixture was left under stirring for 48 h at room temperature
and then concentrated under reduced pressure. Silica gel
chromatography of the residue (eluent dichloromethane–
methanol 95:5, v/v) afforded pure 6 (680 mg, 92% yield).
To a solution of compound 3 (372 mg, 1.3 mmol) and
DMAP (9 mg, 0.07 mmol) in anhydrous pyridine (5 mL)
515 mg of 4,40-dimethoxytriphenylmethyl chloride
(1.5 mmol) and 207 mL of TEA (1.5 mmol) were added.
The mixture was left under stirring overnight and then
concentrated under reduced pressure. Silica gel chroma-
tography of the residue (eluent petroleum ether–ethyl
acetate 6:4 with few drops of TEA) afforded compound 4
(700 mg, 90% yield).
Compound 6:30 glassy solid. Rf¼0.19 (dichloromethane–
1
methanol 9:1, v/v). H NMR (200 MHz, CDCl3) d 6.82–
7.55 (13H, m, aromatic protons); 5.45 (1H, t, J2,3¼J3,4
¼
9.8 Hz, H-3); 5.09 (1H, dd, J3,4¼J4,5¼9.8 Hz, H-4); 5.00
(1H, d, J1,2¼3.6 Hz, H-1); 4.91 (1H, dd, J1,2¼3.6 Hz,
4.2.3. Compound 4. Oil. Rf¼0.80 (hexane–ethyl acetate

6702
M. Adinolfi et al. / Tetrahedron 58 (2002) 6697–6704
J2,3¼9.8 Hz, H-2); 3.91 (1H, m, H-5); 3.77 (6H, bs, –OCH3
DMT group); 3.46 (3H, s, 1-OCH3); 3.23 (2H, m, H2-6);
2.10–2.50 (4H, m, succinic CH2–CH2); 2.08, 1.97 (3H
each, s’s, two –COCH3). 13C NMR (50 MHz) 174.6
(–COOH); 170.6, 170.1 and 170.0 (two COCH3 and
succinic ester CO); 158.2, 144.3, 137.1, 135.6, 129.8,
128.0, 127.5 and 126.5 (aromatic carbons); 96.3 (C-1); 85.7
(quaternary C of DMT group); 70.9, 69.9, 68.9 and 68.4
(C-2, C-3, C-4 and C-5); 61.7 (C-6); 54.9 (1-OCH3 and
–OCH3 of DMT group); 28.7 and 28.6 (succinic CH2); 20.4
(COC H3). ESI MS (negative ions): m/z 679.28 (M2H)2.
3.47 (3H each, s’s, 2OCH3). 31P NMR (161.98 MHz, D2O),
d 2.01 ESI-MS: calculated mass: 449.11; found: m/z 448.61
(M2H)2.
1
4.4.3. Compound 10. H NMR (400 MHz, D2O), d 8.26
(1H, s, H-2 A); 8.18 (1H, s, H-8 A); 7.68 (1H, d, J¼8.0 Hz,
H-6 C); 6.36 and 6.23 (1H each, dd’s, J¼6.8, 6.8 Hz, 2H-10);
5.75 (1H, d, J¼8.0 Hz, H-5 C); 4.86 (2H, m’s, partially
submerged by the residual solvent signal, two H-30); H-4
Glu-phosphate is submerged by the residual solvent signal;
4.70 (2H, overlapped signals, two H-1 Glu); 4.30 (2H,
overlapped m’s, two H-40); 4.12–4.07 (6H, overlapped m’s,
H2-50 C and two H2-6 Glu adjacent to a phosphate residue);
3.98–3.57 (8H, overlapped signals, H2-50 A, two H-2, two
H-5 and two H-3); 3.46 (1H, dd, J¼9.2, 9.2 Hz, H-4); 2.76
and 2.36 (2H each, m’s, H2-20). 31P NMR (161.98 MHz,
D2O), d 1.53, 1.10, 0.68; ESI-MS: calculated mass:
1052.22; found: 1052.70.
Compound 6 (305 mg, 0.45 mmol) and DCCI (277 mg,
1.34 mmol) were dissolved in anhydrous pyridine (10 mL).
After 3 min the mixture was added under argon to a flask
containing the CPG resin (300 mg). After 3 days under
horizontal shaking, the resin was exhaustively washed with
DCM and methanol and successively dried under reduced
pressure. Spectroscopic DMT test on a weighed sample
of dried support 717 allowed the determination of the
incorporation of the glucose derivative, which was
0.050 mequiv./g.
1
4.4.4. Compound 12. H NMR (400 MHz, D2O), d 8.02
(1H, d, J¼7.6 Hz, H-6 C); 6.37 (1H, dd, J¼6.4, 6.4 Hz,
H-10); 6.17 (1H, d, J¼7.6 Hz, H-5 C); 4.88 (1H, d, J¼
4.0 Hz, H-1 Glu); H-4 Glu is submerged by the residual
solvent signal; 4.60 (1H, m, H-30); 4.30–4.18 (3H,
overlapped signals, H-40 and H2-50); 3.99 (1H, dd, J¼8.0,
8.0 Hz, H-3 Glu); 3.88 (2H, m, H2-6 Glu); 3.78 (1H, m, H-5
Glu); 3.66 (1H, dd, J¼4.0, 8.0 Hz, H-2 Glu); 3.51 (3H, s,
OCH3); 2.45 (2H, m, H2-20). 31P NMR (161.98 MHz, D2O),
d 1.86 ESI-MS: calculated mass: 482.12; found: m/z 481.31
(M2H)2.
4.4. Synthesis of hybrids 8–10, 12 and 13. General
procedure
50 mg of the starting support (11 or 7) were used for each
synthesis. The coupling with 5 was performed on an
automated DNA synthesizer following standard phosphor-
amidite chemistry,26 using a 40 mg/mL solution of the
amidite unit in anhydrous CH3CN. Incorporation of the
glucose phosphate residue within the chain always occurred
with the same efficiency as for the usual nucleoside
phosphoramidites, i.e. was always .98% yields (by DMT
test). After completion of the desired ODN sequence and
final DMT removal, the support was treated overnight with
conc. aq. ammonia at 558C. The filtered solution and the
washings were concentrated under reduced pressure and
purified by HPLC on a Nucleosil 100-5 C18 column
(4.6£250 mm, 7 mm) eluted with a linear gradient of
CH3CN in TEAB buffer (0.1 M, pH 7.0). The final products
were lyophilized and characterized by NMR and MS data.
For compounds 10 and 13, the reported mass is the
calculated value on the basis of the combination of the
multiply charged ions found.
1
4.4.5. Compound 13. H NMR (400 MHz, D2O), d 8.54
(1H, s, H-2 A); 8.26 (1H, s, H-8 A); 7.72 (1H, s, H-6 T); 7.62
(1H, d, J¼7.6 Hz, H-6 C); 6.53 (1H, dd, J¼6.8 and 6.8 Hz,
H-10); 6.31 (1H, dd, J¼7.2, 6.4 Hz, H-10); 6.15 (1H, dd,
J¼7.2, 6.0 Hz, H-10); 6.06 (1H, d, J¼7.6 Hz, H-5 C); 5.15
and 4.91 (1H each, m’s, two H-30); 4.88 (1H, d, J¼3.6 Hz,
H-1 Glu); one H-30, H-4 Glu phosphate and one H-1 Glu are
submerged by the residual solvent signal; 4.55 and 4.39 (1H
each, m’s, two H-40); 4.21 (9H, overlapped signals, one
H-40, three H2-50 and H2-6 Glu adjacent to a phosphate
residue); 4.06–3.56 (8H, overlapped signals, two H-2, two
H-5, one H2-6 Glu, two H-3 and one H-4); 2.96, 2.44
and 1.90 (2H each, m’s, three H2-20); 1.99 (3H, s, CH3 T).
31P NMR (161.98 MHz, D2O), d 1.48, 1.32, 0.83, 0.58;
ESI-MS: calculated mass: 1356.26; found 1356.23.
4.4.1. Compound 8. 1H NMR (400 MHz, D2O), d 7.65 (1H,
s, H-6 T); 6.30 (1H, dd, J¼6.0, 6.0 Hz, H-10); 4.82 (1H, m,
partially submerged by the residual solvent signal, H-30);
4.76 (1H, d, J¼4.0 Hz, H-1 Glu); 4.19 (1H, m, H-40); 4.09
(2H, m, H2-6 Glu adjacent to a phosphate residue); 3.80
(1H, dd, J¼4.0, 10.0 Hz, H-2 Glu); 3.75 (1H, m, H-5 Glu);
3.67 (2H, m, H2-50); 3.63 (1H, d, J¼10 Hz, H-3 Glu); 3.46
(1H, dd, J¼10, 10 Hz, H-4 Glu); 3.39 (3H, s, OCH3); 2.50
(2H, m, H2-20); 1.87 (3H, s, CH3 T). 31P NMR (161.98 MHz,
D2O), d 1.42. ESI-MS: calculated mass: 498.13; found: m/z
497.14 (M2H)2.
4.5. Synthesis of conjugates 14 and 15
ODN chain assembly was performed on 50 mg (25 mmol) of
support 7 on an automated DNA synthesizer following a
standard phosphoramidite protocol with final DMT
removal.26 Two sequences were assembled (14 and 15)
inserting at the extremities of the growing sequence two
or three coupling cycles with phosphoramidite 5. The
oligomers were detached from the support and deprotected
by conc. ammonia treatment as described for hybrids 8–10,
12 and 13. The supernatant was filtered and the support
washed with H2O. The combined filtrates and washings
were concentrated under reduced pressure, redissolved in
H2O and analyzed and purified by HPLC on a Nucleogel
SAX column (Macherey-Nagel, 1000-8/46); buffer A:
20 mM KH2PO4 aq. solution, pH 7.0, containing 20%
4.4.2. Compound 9. 1H NMR (400 MHz, D2O), d 4.90 (2H,
overlapped signals, 2H-1); H-4 Glu-phosphate is submerged
by the residual solvent signal; 4.20 (2H, m, H2-6 adjacent to
a phosphate residue); 4.07–3.45 (8H, overlapped signals,
two H-2, two H-3, two H-4, two H-5 and H2-6OH); 3.48 and

M. Adinolfi et al. / Tetrahedron 58 (2002) 6697–6704
6703
(v/v) CH3CN; buffer B 1 M KCl, 20 mM KH2PO4 aq.
References
solution, pH 7.0, containing 20% (v/v) CH3CN; a linear
gradient from 0 to 100% B in 20 min, flow rate 0.8 mL/min,
was used. The isolated oligomers, having the following
retention times: 14¼15.17 min; 15¼16.85 min, were col-
lected and successively desalted by gel filtration on a
Sephadex G25 column eluted with H2O. By HPLC analysis
on a Nucleosil 100-5 C18 analytical column (4.6£250 mm,
7 mm), the isolated oligomers were .98% pure. Compound
14 MALDI TOF-MS: calculated mass: 6193; found: m/z
6192.10 (M2H)2. Compound 15 MALDI TOF-MS:
calculated mass: 6449; found: m/z 6447.22 (M2H)2.
1. Zamecnik, P. C.; Stephenson, L. M. Proc. Natl Acad. Sci. USA
1978, 75, 280–284.
2. Uhlmann, E.; Peyman, A. Chem. Rev. 1990, 90, 543–584.
´ `
3. Thuong, N. T.; Helene, C. Angew. Chem., Int. Ed. Engl. 1993,
32, 666–690.
4. Famulok, M.; Mayer, G.; Blind, M. Acc. Chem. Res. 2000, 33,
591–599.
5. Beaucage, S. L.; Iyer, R. P. Tetrahedron 1993, 49,
6123–6194.
6. Carbohydrate Modifications in Antisense Research, ACS
Symposium Series; Sanghvi, Y. S., Dan Cook, P., Eds.;
ACS: Washigton, 1994.
4.6. Thermal denaturation experiments
7. Hogrefe, R. I. Antisense Nucleic Acid Drug Dev. 1999, 9,
351–357.
The concentration of the synthesized ODNs was determined
spectrophotometrically at l¼260 nm and at 858C, using the
molar extinction coefficient calculated for the unstacked
oligonucleotide using the following extinction coefficients:
8. Grillone, L. R. Antisense Drug Technol. 2001, 725–738.
9. Orr, R. M. Curr. Opin. Mol. Ther. 2001, 3, 288–294.
10. Beaucage, S. L.; Iyer, R. P. Tetrahedron 1993, 49,
1925–1963.
15,400 (A); 11,700 (G); 7300 (C); 8800 (T) cm21 M21 32
.
For bis-conjugated oligomers 14 and 15, the contribution of
the glucose residues to the absorbance at l¼260 nm was
considered negligible. A 100 mM NaCl, 10 mM NaH2PO4,
aq solution at pH 7.0 was used for the melting experiments.
Melting curves were recorded using a concentration of
approximately 1 mM for each strand in 1 mL of the tested
solution in Teflon stoppered quartz cuvettes of 1 cm optical
path length. The resulting solutions were then allowed to
heat at 808C for 15 min, then slowly cooled and kept at 58C
for 20 min. After thermal equilibration at 108C, UV
absorption at l¼260 nm was monitored as a function of
the temperature, increased at a rate of 0.58C/min, typically
in the range 20–808C. The melting temperatures, deter-
mined as the maxima of the first derivative of absorbance vs.
temperature plots, are the average values of three inde-
pendent melting experiments. For all the studied cases, both
the modified duplexes and the unmodified one, within the
experimental error (^0.58C), exhibited a melting tempera-
ture value of 608C.
11. Markiewicz, W. T.; Groger, G.; Rosch, R.; Zebrowska, A.;
Markiewicz, M.; Klotz, M.; Hinz, M.; Godzina, P.; Seliger, H.
Nucleic Acids Res. 1997, 25, 3672–3680.
12. Tung, C.-H.; Stein, S. Bioconjugate Chem. 2000, 11,
605–618.
13. Monsigny, M.; Quetard, C.; Bourgerie, S.; Delay, D.; Pichon,
C.; Midoux, P.; Mayer, R.; Roche, A. C. Biochimie 1998, 80,
99–108.
14. Akhtar, S.; Routledge, A.; Patel, R.; Gardiner, J. M.
Tetrahedron Lett. 1995, 36, 7333–7336.
15. Ren, T.; Liu, D. Tetrahedron Lett. 1999, 40, 7621–7625.
16. Forget, D.; Renaudet, R.; Defrancq, E.; Dumy, P. Tetrahedron
Lett. 2001, 42, 7829–7832.
17. Adinolfi, M.; Barone, G.; De Napoli, L.; Guariniello, L.;
Iadonisi, A.; Piccialli, G. Tetrahedron Lett. 1999, 40,
2607–2610.
18. Adinolfi, M.; De Napoli, L.; Di Fabio, G.; Guariniello, L.;
Iadonisi, A.; Messere, A.; Montesarchio, D.; Piccialli, G.
Synlett 2001, 6, 745–748.
19. Davis, B. J. J. Chem. Soc., Perkin Trans. 1 2000, 2137–2160.
20. Seeberger, P. H.; Haase, W. Chem. Rev. 2000, 100,
4349–4394.
4.7. Enzymatic experiments
3.0 OD of glycoconjugate 14 were incubated at 378C with
1 unit of purified Exonuclease I (Amersham Biosciences,
UK) in a 10 mM Tris–HCl, 50 mM KCl and 1.5 mM MgCl2
buffer at pH 8.3. Aliquots of the mixture were taken at
defined time intervals, kept at 808C for 15 min, then left
at room temperature and analyzed by HPLC on an
analytical anion exchange column. A parallel experiment,
21. Hunziker, J. Bio. Med. Chem. Lett. 1999, 9, 201–204.
22. de Kort, M.; Ebrahimi, E.; Wijsman, E. R.; van der Marel,
G. A.; van Boom, J. H. Eur. J. Org. Chem. 1999,
2337–2344.
23. Matsuura, K.; Hibino, M.; Kataoka, M.; Hayakawa, Y.;
Kobayashi, K. Tetrahedron Lett. 2000, 41, 7529–7533.
24. de Kort, M.; de Visser, P. C.; Kurzeck, J.; Meeuwenoord, N. J.;
van der Marel, G. A.; Ruger, W.; van Boom, J. H. Eur. J. Org.
Chem. 2001, 2075–2082.
car0ried out on 3.0 OD of unmodified ODN sequence
0
d(5 GCGTGCCTCCTCACTGGC3 ), showed that the oligo-
nucleotide was completely digested within 2 h. On the
contrary, glycoconjugate 14 was totally stable even after
72 h.
25. Matsuura, K.; Hibino, M.; Yamada, Y.; Kobayashi, K. J. Am.
Chem. Soc. 2001, 123, 357–358.
26. Oligonucleotides and Analogues: A Practical Approach;
Eckstein, F., Ed.; IRL: Oxford, 1991.
27. Sheppard, T. L.; Wong, C.-H.; Joyce, G. F. Angew. Chem., Int.
Ed. Engl. 2000, 39, 3660–3663.
Acknowledgments
28. This unprecedented procedure was found after having
observed that the known methodology (based on a I2/
CH3OH system)29 was uneffective on 2.
We thank MURST (PRIN 2001), CNR and Regione
Campania (Legge 41) for grants in support of this
investigation and Centro di Metodologie Chimico-Fisiche
`
(CIMCF), Universita di Napoli ‘Federico II’, for the NMR
facilities.
29. Szarek, W. A.; Zamojski, A.; Tiwari, K. N.; Ison, E. R.
Tetrahedron Lett. 1986, 27, 3827–3830.

6704
M. Adinolfi et al. / Tetrahedron 58 (2002) 6697–6704
30. Adinolfi, M.; Barone, G.; De Napoli, L.; Iadonisi, A.; Piccialli,
G. Tetrahedron Lett. 1998, 39, 1953–1956.
32. Nucleic Acids. 3rd ed. Handbook of Biochemistry and
Molecular Biology; Fasman, G. D., Ed.; CRC, 1975; Vol. 1,
p 589.
31. Tortora, G.; Bianco, R.; Damiano, V.; Fontanini, G.; De
Placido, S.; Bianco, R.; Ciardiello, F. Clin. Cancer Res. 2000,
6, 2506–2512.