Structural modifications of antisense oligonucleotides

Article abstract of DOI:10.1016/S0014-827X(03)00022-3

Antisense oligonucleotides are efficient tools for the inhibition of gene expression in a sequence specific way. Natural oligonucleotides are decomposed rapidly in biological systems, which strongly restrict their application. In contrast, artificial oligonucleotides are designed to be more stable against degradation than the target mRNA, which results in a catalytic effect of the drug. Modification of the phosphate linkage has been the first successful strategy for antisense drug developments and Fomivirsene the first antisense drug in therapy. The launch of Fomivirsene has resulted in a revolutionary spin off to antisense research leading to a second generation of antisense oligonucleotides, which are stable against oligonucleotide cleaving enzymes. Among these, oligonucleotides bearing an alkoxy substituent in position 2′ were the most successful ones. The third generation of antisense oligonucleotides contains structure elements, which enhance the antisense action. Zwitterionic oligonucleotides show remarkable results, first, because the stability against ribozymes is largely increased, and secondly, because the electrostatic repulsion between the anionic sense and the zwitterionic antisense cords is minimized. Promising new target molecules in antisense reseach are oligonucleotide chimaeres, which enhance the antisense action (chimaeres with intercalators, chelators or polyamines) or enable an application as sequence specific detectors (chimaeres with biotin, fluorescein or radioligands).

Full text of DOI:10.1016/S0014-827X(03)00022-3

Il Farmaco 58 (2003) 243Á258
/
www.elsevier.com/locate/farmac
Structural modifications of antisense oligonucleotides
Ernst Urban, Christian R. Noe *
Institut fur Pharmazeutische Chemie, Universitat Wien, Pharmaziezentrum, Althanstraße 14, 1090 Wien, Austria
¨ ¨
Accepted 4 July 2002
Abstract
Antisense oligonucleotides are efficient tools for the inhibition of gene expression in a sequence specific way. Natural
oligonucleotides are decomposed rapidly in biological systems, which strongly restrict their application. In contrast, artificial
oligonucleotides are designed to be more stable against degradation than the target mRNA, which results in a catalytic effect of the
drug. Modification of the phosphate linkage has been the first successful strategy for antisense drug developments and Fomivirsene
the first antisense drug in therapy. The launch of Fomivirsene has resulted in a revolutionary spin off to antisense research leading to
a second generation of antisense oligonucleotides, which are stable against oligonucleotide cleaving enzymes. Among these,
oligonucleotides bearing an alkoxy substituent in position 2? were the most successful ones. The third generation of antisense
oligonucleotides contains structure elements, which enhance the antisense action. Zwitterionic oligonucleotides show remarkable
results, first, because the stability against ribozymes is largely increased, and secondly, because the electrostatic repulsion between
the anionic sense and the zwitterionic antisense cords is minimized. Promising new target molecules in antisense reseach are
oligonucleotide chimares, which enhance the antisense action (chimares with intercalators, chelators or polyamines) or enable an
¨
application as sequence specific detectors (chimares with biotin, fluorescein or radioligands).
¨
¨
´
# 2003 Editions scientifiques et me´dicales Elsevier SAS. All rights reserved.
Keywords: Artificial oligonucleotides; Backbone modified oligonucleotides; 2?-Modified oligonucleotides; Zwitterionic oligonucleotides; Peptide
nucleic acids
1. Introduction
for the inhibition of gene expression in a sequence
specific way [1,2].
The essential steps in rational drug design are the
identification of an appropriate target responsible for a
certain disease and the development of a drug with a
specific affinity to that target. One of the most general
approaches of drug targeting is the specific manipula-
tion of gene expression at the DNA or RNA stage of
protein synthesis. Recently, the elucidation of the
pathogenic role of individual target proteins for certain
diseases is rapidly progressing, most notably in cancer
research. This development causes a large demand on
specific tools for manipulating the synthesis of proteins
involved in pathogenic processes. Certainly, antisense
oligonucleotides are recognized to be very efficient tools
2. The antisense principle
The antisense principle is based on a specific recogni-
tion of certain DNA and RNA regions by an antisense
oligonucleotide, which inhibits the translation by a
selective pairing of the ‘sense’ with the complementary
‘antisense oligonucleotide strand. For construction of an
efficient antisense oligonucleotide a sequence of twelve
to twenty bases is sufficient to guarantee a specific
recognition of the target DNA region [3]. The effects of
an antisense oligonucleotide aims at a selective suppres-
sion of certain genes, which are responsible for the
synthesis of undesired proteins. When the affinity of the
antisense oligonucleotide towards the natural DNA or
RNA strand exceeds a certain level, a suppression of the
corresponding protein synthesis results [1,2] (Scheme 1).
* Corresponding author.
E-mail address: christian.noe@univie.ac.at (C.R. Noe).
´
0014-827X/03/$ - see front matter # 2003 Editions scientifiques et me´dicales Elsevier SAS. All rights reserved.
doi:10.1016/S0014-827X(03)00022-3

244
E. Urban, C.R. Noe / Il Farmaco 58 (2003) 243Á258
/
Scheme 1. Recognition of DNA and RNA regions by an antisense oligonucleotide.
3. Targets of antisense oligonucleotides
Special targets on the mRNA are the capping region,
the exonÁintron-transition and the polyadenylation site.
/
The first step of the protein synthesis may be inhibited
by triple helix formation and successively blockade of
transcription. Secondly, the antisense oligonucleotide
may interfere with the processing proteins, which are
responsible for the transformation of the primary DNA
transcript into the maturated mRNA. Formation of a
dual strand between the antisense oligonucleotide and
the mRNA may disable the transport of the mRNA
from the nucleus to the cytoplasm [3,4] (Scheme 2).
Dual strand formation in the cytoplasm results in a
blockade of protein synthesis in the ribosomes. Finally,
the aggregate formation between an antisense oligonu-
cleotide and the mRNA enhances the degradation by
ribozymes. If the antisense oligonucleotide is more
stable against degradation than the mRNA, a catalytic
effect results which multiplies the effect of the antisense
drug [3,4] (Scheme 3).
4. Artificial antisense oligonucleotides
Disadvantages of natural oligonucleotides are the
unsatisfactory binding affinity, the instability against
Scheme 2. The inhibition of the protein synthesis by antisense
oligonucleotides.
Scheme 3. Special targets on the mRNA for antisense oligonucleo-
sides.

E. Urban, C.R. Noe / Il Farmaco 58 (2003) 243Á
/258
245
cellular nucleases, the insufficient membrane penetra-
tion and the low bioavailability.
5. Backbone modified oligonucleotides (first generation)
Natural oligonucleotides are rapidly cleaved by nu-
cleases attacking at the phosphoric ester bonds. There-
fore, it is no surprise that modifications of the
phosphate linkage were the first targets of antisense
drug developments [1,11,12] (Scheme 6).
One strategy to overcome these problems is the
development of a suitable application system, which
protects the oligonucleotide, enhances the penetration
and enables a targeting to the compartment of action
[5Á8].
/
While synthesis of macromolecular nucleotides is
nowadays is exclusively a field of the polymerase chain
reaction, synthesis of natural (DNA and RNA) and of
all types of modified oligonucleotides has remained a
domain of chemical synthesis. Solid support based
techniques, by which an oligomer is built up adding
building blocks step by step, have been shown to be the
most efficient approach in oligonucleotide synthesis.
Out of a series of competitive methods, the so-called
phosphoramidite method is by far the most important
method for making the phosphorodiester bond [1,13]
(Scheme 7).
Although the phosphoramidite method has been
developed for the synthesis of natural oligonucleotides,
it can proceed for the synthesis of backbone modified
oligonucleotides as well, if modified protocols are used.
Scheme 8 shows a selection of sulfurization reagents
suitable for the synthesis of phosphorothioates. Re-
markable are the shorted sulfurization times of recently
Another strategy is the structural modification of
natural oligonucleotides leading to artificial structures.
But structural modifications may lead to a loss of the
base pairing ability. Thus, the most important test for a
new artificial oligonucleotide is the measurement of the
base pairing ability with natural DNA. The method of
choice for this purpose is the study of the duplex
stability by circular dichroism spectroscopy [9,10].
Certainly, the transition temperature (Tm) is the most
common parameter characterizing the transformation of
a double strand helix to a mixture of two single strand
oligonucleotides. For the determination of the transition
temperature, a series of circular dichroism spectra has to
be measured at different temperatures (Scheme 4).
A second demand for an artificial antisense oligonu-
cleotide is a sufficient stability against DNA and RNA
cleaving enzymes. Natural oligonucleotides are decom-
posed rapidly in biological systems, which strongly
restrict their application. In contrast, artificial oligonu-
cleotides reveal in many cases increased stability to-
wards the cleaving enzymes [1].
developed reagents [14Á16].
/
With respect to the stereochemistry, substitution of
one unbonded oxygen atom of the prochiral phosphorus
linkage in an oligonucleotide by sulfur, nitrogene or
borane creates a chiral center leading to diastereomeric
mixtures of RP and SP compounds. Since the number of
In principle structural modifications of antisense
oligonucleotides may concern the oligonucleotide back-
bone, the sugar moiety and the nucleic bases (Scheme 5).
Scheme 4. Observation of the transformation of a double strand helix to a mixture of two single strand oligonucleotides by circular dichroism
spectroscopy.

246
E. Urban, C.R. Noe / Il Farmaco 58 (2003) 243Á258
/
Scheme 5. Structural modifications of artificial antisense oligonucleotides.
therapy with other antiviral drugs like ganciclovir failed
due to the development of a drug resistance. Side effects
of Fomivirsene are limited to the treated eye(s), because
the drug is administered locally [2,21].
Although the field of application for Fomivirsene is
rather limited in therapy, the launch of the first
antisense drug resulted in a revolutionary spin off to
antisense research. A sometimes seriously questioned
innovative drug concept with gene therapeutic back-
ground has at once become an established drug therapy.
This situation has generated a unique chance to look at
this topic. Bearing in mind that antisense action may
prevent synthesis of any pharmacologically relevant
protein, this new type of drugs have been applicated to
several novel disease targets (Table 1).
Scheme 6. The structure of ‘classical’ backbone modified antisense
oligonucleotides.
diastereomers increases by 2n, this sums up to the
formation of about half a million of diastereomeric
compounds during the synthesis of a 20-mer. Although
methods have been developed for the enantioselective
One of the drug candidates with a more general field
of application is the antisense oligonucleotide Augmer-
osen.
synthesis of phosphorothioates [17Á20], the majority of
/
backbone modified oligonucleotides is prepared with
unselective methods. Nevertheless, antisense activity
remains intact in such diastereomeric mixtures [1,2]
(Scheme 9).
5.2. Augmerosen (Genasense, G3139)
(P-thio)-T-C-T-C-C-C-A-G-C-G-T-G-C-G-C-C-A-T-
desoxy-ribonucleic acid Augmerosen (Genasense,
G3139) is a phosphorothioate oligonucleotide which
has been designed to inhibit the mRNA of the BCL2
group proteins. The 18-mer phosphothioate oligonu-
cleotide is complementary to the first six codons of the
open reading frame of BCL2 gene. BCL2 is an apoptosis
inhibitor, which is implicated in the pathogenesis of
cancer. High concentrations of this protein are observed
in lymphatic and myelotic leukaemia, several types of
lymphomas, human melanoma cells and other malig-
nant tumors. Furtheron, the BCL2 group proteins are
implicated not only in the pathogenesis of cancer but
also in resistance to cancer treatment, because antic-
ancer drugs and radiation ultimately destroy cells by
induction of apoptosis [2,22] (Table 2).
5.1. Fomivirsene (Vitravene†)
(P-thio)-G-C-G-T-T-T-G-C-T-C-T-T-C-T-T-C-T-T-
G-C-G-desoxy-ribonucleic acid Fomivirsene (ISIS 2922)
is a phosphorothioate oligonucleotide which has been
constructed to inhibit the replication of the human
cytomegalo virus by an antisense mechanism. The
nucleotide sequence of Fomivirsene is complementary
to the major immediate early region 2 (IE2) of the
human cytomegalo virus. This part of the viral genome
is encoding a series of proteins, which are responsible for
the virus replication.
In 1998 Fomivirsene has been approved by the US
Food and Drug Administration (FDA) for the treat-
ment of cytomegalo virus induced retinitis in patients
with an acquired immune deficiency syndrome (AIDS).
The antisense drug inhibits the virus replication in the
Recently, several phase I and phase II studies have
been started investigating the therapeutical utility of
combinations of Augmerosen and classical anticancer
drugs in patients with different cancer types (Table 3).
target organ in submicromolar concentrations (EC50ꢀ
/
0.0390.02 mM) and proved to be active even when a
/

E. Urban, C.R. Noe / Il Farmaco 58 (2003) 243Á
/258
247
Scheme 7. The reaction cycle of the phosphoramidite method for the synthesis of DNA.

248
E. Urban, C.R. Noe / Il Farmaco 58 (2003) 243Á258
/
Scheme 8. Some sulfurization reagents suitable for the synthesis of phosphorothioates.
Generally, systemic treatment with phosphorothioate
oligonucleotides is well tolerated and side effects are
dose dependent [2].
Table 1
Currently running antisense programs in pharmaceutical companies
Companies
Program
6. 2?-Modified oligonucleotides (second generation)
AVI BioPharma
Calgene
restenosis
antisense technology with plants
restenosis
antisense screening
CV Therapeutics
Cytoclonal Pharmaceu-
tics
Nevertheless, some fundamental requirements for an
unrestricted use of antisense oligonucleotides in drug
therapy have not been resolved satisfactorily till now.
Responsible for this main insufficiency of the known
antisense oligonucleotides is the structural similarity of
phosphorothioate based antisense oligonucleotides to
the natural DNA and RNA. This has caused the
development of a second generation of structurally
modified antisense oligonucleotides, which are based
on more rigorous modifications of the oligonucleotide
structure. One of the most attractive targets for varia-
tion of oligonucleotides is the sugar moiety, especially
Enzo Biochem
antisense technology with plants, viru-
statics
Genta
Gilead Sciences
restenosis, cardiovascular, oncology
gene block technology
HoffmannÁ
/
La Roche
virustatics, HBV, HCV etc.
Human Genome Sciences antisense technology via genomics
Hybridon virustatics, HIV, HBV, HCV etc.
Immune Response Corp. TGF-b antisense, oncology
INEX Pharmaceuticals
Isis Pharmaceuticals
Japan Tobacco
liposomal technology for antisense
CMV, virustatics, oncology
antisense technology with plants
oncology
RhoneÁPoulenc
/
the 2?-position [23Á27] (Scheme 10).
/
Scheme 9. A method for the enantioselective synthesis of phosphorothioates.

E. Urban, C.R. Noe / Il Farmaco 58 (2003) 243Á
/258
249
Table 2
Clinical trials investigating the therapeutical utility of combinations of Augmerosen and classical anticancer drugs
Type of cancer
Trial location
Status
Partner drug
Melanoma
Prostate
Prostate
University of Vienna, Austria
Memorial Sloan-Kettering Cancer Center, NY, USA
University of Texas, TX, USA
phase II
Dacarbazine
Taxotere†
Taxotere†
Taxotere†
Camptosar†
Taxotere†
Fludara†
phase IÁ
phase II
/
II
Lung (small cell)
Colorectal
Breast
Ohio State University, OH, USA
University of Texas, TX, USA
phase IÁ
phase II
/II
Georgetown University, Washington, DC, USA
Ohio State University, OH, USA
phase IÁ
phase IÁ
/II
Acute leukaemia
/II
Table 3
Dosage dependence of side effects of phosphorothioates
Dosage (mg/ Side effects
kg)
100
80
immune stimulation and hepatotoxicity in mice
clotting and complement effects in monkeys proximal
tubular degradation
20
10
lymphoid hyperplasia in mice
Scheme 11. Sequence of the model nucleotides of 2?-O-alkyl- and 2?-
O-ethylenoxy-oligonucleotides.
inhibition of clotting times Complement activation
atrophic and regenerative changes in proximal tubules
slight inhibition of clotting times
3
2.5
highest dosage in humans
The base pairing ability of 2?-O-alkyl derivatives was
investigated using artificial nucleotides, which were
constructed with an increasing number of modified
adenosine derivatives. A comparison with the unmodi-
fied 12-mer of adenosine revealed the influence of the
substituents in position 2? of the oligonucleotides [25Á
/
27] (Scheme 12).
In principle, any modification of the oligonucleotide
backbone may result in a loss of the base pairing ability,
but 2?-modified oligonucleotides have proved to form
stable duplices. Stability of duplices with the comple-
mentary DNA or RNA strands decrease with the length
of alkyl chain and with an increasing number of
modified nucleotides in a given sequence [25Á
/27]
(Scheme 13).
One reason for the low Tm values of oligonucleotides
bearing long alkyl residues may be the high lipophility
Scheme 10. Structure of artificial 2?-O-alkyl-oligonucleotides [1].
Oligonucleotides bearing an alkoxy substituent in
position 2? have prove to be stable against DNA or
RNA cleaving enzymes. This increased stability can be
explained by a sterical hinderance due to the substituent
in position 2?, which inhibits an attack of the DNA
degrading enzymes at the phosphate groups [25Á
(Scheme 11).
/27]
Scheme 12. Impact of the alkyl substituents on the transition
temperatures (0.15 M NaCl, 0.01 M TrisÁHCl, pH 7.0, 9 mM).
/

250
E. Urban, C.R. Noe / Il Farmaco 58 (2003) 243Á258
/
Scheme 15. Zwitterionic 2?-O-(6-aminohexyl)-modified oligonucleo-
tides.
group, first, because the stability of the 6-aminohexyl-
modified oligonucleotides against the catalytic degrada-
tion effect of ribozymes is largely increased, and
secondly, because the electrostatic repulsion between
the anionic sense and the zwitterionic antisense cords is
minimized. The flexible alkyl group between the oxygen
of the ribose and the terminal amino function acts as a
spacer enabling an optimal arrangement of the amino
group for a linkage to the complementary DNA or
RNA strands [28,29] (Scheme 15).
Scheme 13. Structure of artificial 2?-O-ethylenoxy-oligonucleotides.
For the synthesis of 6-aminohexyl-modified oligonu-
cleotides suitably protected 6-aminohexyl-modified nu-
cleotides are necessary. A representative strategy for the
versatile choice of protecting groups can be deduced
from a recent report on the synthesis of a 6-aminohexyl-
modified adenosine derivative [30] (Scheme 16).
Scheme 14. Impact of the oxygen atoms on the transition tempera-
tures (0.15 M NaCl, 0.01 M TrisÁHCl, pH 7.0, 9 mM).
/
Thus, in the first step of the synthesis adenosine was
reacted to the 5?-dimethoxytrityl derivative 1. This
readily soluble adenosine derivative proved to be the
optimal intermediate for the introduction of a phthaloyl
protected 6-aminohexyl group in position 2?. Acetyla-
tion of 2 under mild conditions resulted in a selective
formation of the 3?-O-acetyl derivative 3, which was
protected at the amino function to yield the urethane
derivative 4. Finally, deprotection of 4 by methanolysis
gave the key intermediate 5, which is a versatile building
block for the synthesis of zwitterionic oligonucleotides.
Protective groups of the nucleoside 5 have been chosen
according to state of the art to enable a versatile solid
phase synthesis of artificial oligonucleotides by the
phosphor-amidite method [30] (Scheme 17).
of these alkyl substituents. Thus, the alkyl residues were
substituted by ethylenoxy groups with an increasing
number of ethylenglycol units [25Á27] (Scheme 14).
/
The base pairing ability of 2?-O-ethylenoxy deriva-
tives have been investigated by circular dichroism
spectroscopy. Stability of duplices of 2?-O-ethylenoxy
derivatives with complementary DNA strands proved to
be more stable than duplices of the corresponding 2?-O-
alkyl derivatives. Nevertheless, the duplex stability
decreases with the length of alkyl chain and with an
increasing number of modified nucleotides in a given
sequence, but replacement of alkyl groups by an 2?-O-
ethylenoxy moiety reduces this negative effect [25Á27].
/
Studies on the acidity of uridine showed, that the
7. Zwitterionic oligonucleotides (third generation)
imide NH can easier be deprotonated (pKaꢀ
the hydroxy groups at C-2? (pKaꢀ12.5) or C-3? and C-
5? (pKaꢀ14.5Á16.5). Thus, for the synthesis of 6-
aminohexyl modified uridine derivatives the NH group
/
9.38) than
/
One of the most successful strategies for the design of
artificial antisense oligonucleotides is the modification
at the sugar moiety by introduction of a 6-aminohexyl
/
/

E. Urban, C.R. Noe / Il Farmaco 58 (2003) 243Á
/258
251
Scheme 16. Synthesis of the protected 2?-O-(6-aminohexyl)-adenosine derivative 5.
must be protected before alkylation. Using strong bases
like sodium hydride for deprotonation, DMF as a
solvent and alkyl halide 7 as alkylating agent leads
mentary DNA strands have proved to be more stable
than duplices of the corresponding 2?-O-alkyl deriva-
tives. Nevertheless, the duplex stability decreases with
the length of the alkyl chain and with an increasing
number of modified nucleotides in a given sequence, but
replacement of alkyl groups by an 6-aminohexyl moiety
reduces this negative effect [29] (Scheme 19).
preferable (8:9ꢀ70:30) to the 2?-O-alkyl nucleoside 8,
/
which was isolated by chromatography.
Reacting the stanyl derivative 10 and the alkyl halide
11 at high temperatures (130 8C) leads to the formation
of equal amounts (12:13ꢀ55:45) of 2?- and 3?-O-
/
Zwitterionic 2?-O-[L-lysyl-(6-aminohexyl)] oligonu-
alkyluridines, which may be separated by medium
pressure chromatography after dimethoxy-tritylation
[29].
A comparison of the two alkylation methods revealed,
that the stanyl-process is significantly superior, because
no protection of the imido group is necessary, the yield
in the alkylation step is higher, the chromatographic
separation is easier and finally the protecting group at
the side chain is removable by standard protocols from
the oligonucleotides [29].
cleotides have been chosen as target molecules, because
we expected, that an additional amino function should
improve the linkage between the antisense and the
complementary sense strands [29] (Scheme 20).
Synthesis of the required nucleoside building block 17
started from the phthaloyl-protected (6-aminohexyl)-
uridine 14. Deprotection by hydrazinolysis and subse-
quent acylation by the activated lysine derivative 16
gave the desired nucleotide 17 [29] (Scheme 21).
The base pairing ability of 2?-O-(6-aminohexyl) and
The duplex stability of 2?-O-(6-aminohexyl) deriva-
tives was investigated using artificial nucleotides, which
were constructed with an increasing number of modified
uridine derivatives. A comparison with the unmodified
12-mer revealed the influence of the substituents in
position 2? of the oligonucleotides [29] (Scheme 18).
The base pairing ability of 6-aminohexyl modified
uridine derivatives have been investigated by circular
dichroism spectroscopy. Stability of duplices of 6-
aminohexyl modified uridine derivatives with comple-
2?-O-[L-lysyl-(6-aminohexyl)] oligonucleotides has been
investigated by circular dichroism spectroscopy. Stabi-
lity of duplices with complementary DNA strands have
been reported to decrease with an increasing number of
modified nucleotides in a given sequence.
Generally, the linkage of additional lysyl residues
does not affect the duplex stability of the model
oligonucleotides-with one exception-the oligonucleotide,
which contains six 2?-O-[L-lysyl-(6-aminohexyl)]-uridine

252
E. Urban, C.R. Noe / Il Farmaco 58 (2003) 243Á258
/
Scheme 17. A comparison of the two methods for the aminoalkyation of uridines the ‘direct alkylation approach’ vs. the ‘stanylene process’.
units, shows a significant higher duplex stability than
the other zwitterionic oligonucleotides [29] (Scheme 22).
8. Oligonucleotides with an amide type backbone
Replacement of natural structures by artificial bioi-
sosteric elements is a general strategy in drug synthesis
also valuable for nucleic acids. Peptide type nucleic acids
Scheme 18. Sequence of the model nucleotides.
have had a significant impact on antisense drug devel-

E. Urban, C.R. Noe / Il Farmaco 58 (2003) 243Á
/258
253
opment, because they are achiral and neutral molecules
[31,32]. Hybridization with DNA and RNA and forma-
tion of triple helices have been shown [33]. The
oligomers are resistant to endo- and exo-nuclease
mediated degradation as well as protease digestion
[34]. Nevertheless, all drug candidates presently in the
stage of clinical development exhibit only minor mod-
ifications of natural oligonucleotides [1,2]. One of the
main reasons for the lower importance of peptide type
nucleic acids is the lack of a well established method for
the oligomerization reaction (Scheme 23).
The peptide type oligonucleotide B has a backbone
containing both phosphate and amide linkages in
alternating sequence. Thus, hybrid B may exhibit both
the properties of an artificial peptide type nucleic acid
Scheme 19. Impact of the amino group on the transition temperatures
(0.15 M NaCl, 0.01 M TrisÁHCl, pH 7.0, 9 mM).
/
Scheme 20. Zwitterionic 2?-O-[
L
-lysyl-(6-aminohexyl)] oligonucleotides.
Scheme 21. Synthesis of the protected 2?-O-[L-lysyl-(6-aminohexyl)]-uridine 17.

254
E. Urban, C.R. Noe / Il Farmaco 58 (2003) 243Á258
/
tions on models of B indicate, that double helices with
natural DNA should be stable [35,36]. Nuclease resis-
tance may be anticipated due to the character of
substituents in position 2? (Scheme 24).
Dinucleotide 18 is the logical building block for the
synthesis of the oligonucleotide B. The preparation of
dinucleotides 18 requires the 2-amino-2,3-didesoxyri-
bose 19 and the 5-desoxy-allofuranose 20 as synthetic
precursors [37,38].
The unnatural 2-amino-ribose derivative 19 is not
directly accessible from the chiral pool, but has to be
prepared by a multi step reaction sequence from an
available chiral starting material. We decided to use
Scheme 22. Transition temperatures of zwitterionic oligonucleotides.
easily available D-glucosamine as adduct for the pre-
paration of 19 [37], since the configurations of all
important asymmetric carbon atoms were as desired
and the advantage of accessibility by the well established
phosphoramidite method. Molecular dynamics calcula-
Scheme 23. A comparison of DNA (A) to an artificial peptide oligonucleotide bearing both phosphate and amide linkages in alternating sequence
(B).
Scheme 24. Building blocks for the synthesis of the peptide type oligonucleotide B.

E. Urban, C.R. Noe / Il Farmaco 58 (2003) 243Á
/258
255
Scheme 25. Synthesis of the 2-amino-nucleoside 30, a precursor for dinucleotide 18.
and there was no need for introduction of the amino
function.
After protecting the amino function of D-glucosa-
by reaction with carbon disulfide and methyl iodide.
Radical reaction of 23 with tributyltin hydride [41],
activated by AIBN (a,a?-azoiso-butyronitril) yielded
deoxygenated product 24.
Treatment with toluenesulfonic acid opened the 5,6-
acetonide ring and the oxazoline moiety too. Oxidative
cleavage of 25 with sodium periodate gave ribose
derivative 26.
mine, the resulting benzoate 21 [39] was reacted with
acetone and Lewis acid to form 22 [40]. This oxazoline is
an important intermediate product because the gluco-
samine is fixed in the desired furanose form and all
reactive groups are protected with exception of the
hydroxy group in 3. The free hydroxy group was
transformed to the corresponding dithiocarbonate 23
Reaction of 26 with lithium aluminum hydride
resulted in a reduction of the aldehyde and benzylamide

256
E. Urban, C.R. Noe / Il Farmaco 58 (2003) 243Á258
/
Scheme 26. Synthesis of the artificial nucleoside 42, a precursor for dinucleotide 18.
functions to afford 27. In the next reaction step the
benzyl group was removed using catalytic hydrogena-
tion to yield 28. Finally, phenoxyacetylation and sub-
sequent nucleosidation gave the desired nucleoside 30
(Scheme 25).
as chiral precursor, because this sugar has the right
number of carbon atoms and the configuration at the
asymmetric carbons differs from
tion 3.
D-allose only in posi-
In the first reaction step D-glucuronolactone was
transformed to the acetonide 31 according to a known
procedure [40]. Acetonide 31 was recognized to be the
For the preparation of the 5-desoxy-allofuranose
derivative 20 [38] we have chosen D-glucuronolactone

E. Urban, C.R. Noe / Il Farmaco 58 (2003) 243Á
/258
257
optimal intermediate for the following steps, because the
desired furanose form is fixed by the acetal protecting
group and secondly because carbon 5 was the only
position bearing a free hydroxy function. The free
hydroxy group was transformed to the corresponding
dithiocarbonate 32 by reaction with carbon disulfide
and methyl iodide. Radical reaction with tributyltin
hydride [41] gave the desoxysugar 33. Reaction with
lithium aluminum hydride resulted in a reductive open-
ing of the lactone ring to give 34.
First, antisense oligonucleotides may be combined
with structure elements, which enhance the antisense
action. Promising candidates are intercalators, ion
chelating moieties and polyamine like structure ele-
ments, which may be bond covalently to antisense
oligonucleotides.
Secondly, antisense oligonucleotides may be com-
bined with structure elements, which modify the biolo-
gical effects of oligonucleotide. Possible targets are the
phosphorylation reaction and the enzymatic degrada-
tion of nucleotides.
Third, antisense oligonucleotides may be used as
sequence specific detectors for the localization of certain
oligonucleotides. For this purpose oligonucleotides
should be coupled with biotin, with fluorescein like
compounds or radioactive ligands.
In the next step the primary hydroxy function at C-6
of 34 was selectively benzoylated to give 35. Then the
configuration at position 3 was inverted via oxidation of
the
sponding ketone 36 and by selective reduction of ketone
36 with sodium borohydride to the -configurated allose
D-configurated glucose derivative 35 to the corre-
L
derivative 37. Acid catalyzed hydrolysis resulted in a
removal of the acetal protecting group to give derivative
38. After peracetylation we obtained the 5-desoxy-
allofuranose 39, which is the key intermediate for the
nucleosidation reaction. According to our expectations
the nucleosidation of the 2?-O-acetyl derivative 39 gave
high yields of the cytidine derivative 40, which was
finally protected at the amino group and selectively
deprotected at the hydroxy functions leading to the
desired artificial nucleotide 42 (Scheme 26).
In conclusion, we have prepared the 2-amino-nucleo-
side 30 [37] and the 5-desoxy-allo-furanose derivative 42
[38], which are interesting synthetic precursors for the
synthesis of the dinucleotide 18. Nevertheless, large
efforts are necessary in the synthetic field of artificial
nucleotides to realize an oligomerization reaction of
dinucleotide 18 to oligonucleotide B.
Certainly, a novel generation of antisense oligonu-
cleotides will rise up enabling an improved insight to the
processes of gene regulation and affording an important
contribution to modern drug development.
References
[1] C.R. Noe, L. Kaufhold, Chemistry of antisense oligonucleotides,
in: Methods and Principles in Medicinal Chemistry, New Trends
in Synthetic Medicinal Chemistry, 2000, pp. 261Á
/
347.
[2] I. Tamm, B. Dorken, G. Hartmann, Antisense therapy in
¨
oncology: new hope for an old idea, Lancet 358 (2001) 489Á497.
/
[3] C. Helene, J.J. Toulme, Specific regulation of gene expression by
antisense, sense and antigene nucleic acids, Biochim. Biophys.
Acta 1049 (1990) 99Á125.
/
[4] C.J. Marcus-Sekura, Techniques for using antisense oligodeoxyr-
ibonucleotides to study gene expression, Anal. Biochem. 172
(1988) 289Á295.
/
[5] A. Zimmer, S. Atmaca-Abdel Aziz, M. Gilbert, D. Werner, C.R.
Noe, Synthesis of cholesterol modified cationic lipids for liposo-
mal drug delivery of antisense oligo-nucleotides, Eur. J. Pharm.
9. A preview to future developments in antisense research
Biopharm. 47 (1999) 175Á178.
/
[6] H.-P. Zobel, A. Zimmer, S. Atmaca-Abdel Aziz, M. Gilbert, D.
Werner, C.R. Noe, J. Kreuter, F. Stieneker, Evaluation of
aminoalkylmethacrylate nanoparticles as colloidal drug carrier
systems. Part I: synthesis of monomers, dependence of the
physical properties on the polymerization methods, Eur. J.
Finally, the question arises, what will be the next
structural modifications of antisense oligonucleotides
and what will be the new targets in antisense research in
the near future (Table 4).
Pharm. Biopharm. 47 (1999) 203Á213.
/
[7] H.P. Zobel, F. Stieneker, S. Atmaca-Abdel Aziz, M. Gilbert, D.
Werner, C.R. Noe, J. Kreuter, A. Zimmer, Evaluation of
aminoalkylmethacrylate nanoparticles as colloidal drug carrier
systems. Part II: characterization of antisense oligonucleotides
loaded copolymer nanoparticles, Eur. J. Pharm. Biopharm. 48
Table 4
New trends in antisense research in the near future
Effectors (enhance antisense action)
Intercalators
Ion chelators
(1999) 1Á12.
/
[8] H.P. Zobel, D. Werner, M. Gilbert, C.R. Noe, F. Stieneker, J.
Kreuter, A. Zimmer, Effect of ultrasonication on the stability of
oligonucleotides absorbed on nanoparticles and liposomes, J.
Zwitterionic oligonucleotides
Modulators (modify biological effects)
Phosphorylation
Microencapsulation 16 (1999) 501Á509.
/
Inhibitors of enzymatic degradation
Detectors (allow localization of oligonucleotides)
Biotinylation
[9] V.I. Ivanov, L.E. Minchenkova, A.K. Schyolkina, A.I. Poletayev,
Different conformations of double-stranded nucleic acid in
solution as revealed by circular dichroism, Biopolymers 12
Fluorescein-linked oligonucleotides
Radioactive oligonucleotides
(1973) 89Á100.
/
[10] W. Saenger, Principles of Nucleic Acid Structure, Springer, New
York, 1984.

258
E. Urban, C.R. Noe / Il Farmaco 58 (2003) 243Á258
/
[11] C. Cazenave, C.A. Stein, N. Loreau, N.T. Thuong, L.M. Neckers,
C. Subasinghe, C. Helene, J.S. Cohen, J.J. Toulme, Comparative
inhibition of rabbit globin mRNA translation by modified
antisense oligodeoxynucleotides, Nucleic Acids Res. 17 (1989)
ethyl ribonucleoside-3?-O-phosphoramidites, Nucleic Acid Res.
19 (1991) 5965Á5971.
[25] H. Brunar, Herstellung, Paarungseigenschaften und antivirale
/
Wirksamkeit von funktionalisierten Oligonukleotiden, thesis,
4255Á
/
4273.
Technische Universitat Wien, 1993.
¨
[12] A. Storey, D. Oates, L. Banks, L. Crawford, T. Crook, Anti-sense
phosphorothioate oligonucleotides have both specific and non-
specific effects on cells containing human papillomavirus type 16,
[26] D. Werner, Synthese von 2?-O-Alkyl- und 2?-O-Oxyethylen-
Adenosin-Synthese von Oligonukleotiden und Untersuchung der
Paarungseigenschaften, thesis, Universitat Frankfurt am Main,
¨
Nucleic Acids Res. 19 (1991) 4109Á
/
4114.
1997.
[27] D. Wemer, H. Brunar, C.R. Noe, Investigations on the influence
[13] S. Scott, P. Hardy, R.C. Sheppard, M.J. McLean, Innovations
and perspectives in solid phase synthesis, Third International
Symposium, Mayflower Worldwide, 1994.
of 2?-O-alkyl modifications on the base pairing properties of
oligonucleotides, Pharm. Acta Helv. 73 (1998) 3Á10.
/
¨
[28] C.R. Noe, H. Brunar, Int. Appl. WO9516696 A2 950622, Osterr.
[14] H. Vu, B.L. Hirschbein, Internucleotide phosphite sulfurization
with tetraethylthiuram disulfide, phosphorothioate oligonucleo-
tide synthesis via phosphoramidite chemistry, Tetrahedron Lett.
Patent 2505/93.
[29] M. Gilbert, Optimierung von Antisense-Wirkstoffen-zwitterio-
32 (1991) 3005Á
[15] M.V. Rao, C.B. Reese, Z. Zhengyun, Dibenzoyl tetrasulphide*
rapid sulfur transfer agent in the synthesis of phosphorothioate
analogues of oligonucleotides, Tetrahedron Lett. 33 (1992) 4839Á
/
3008.
nische Nukleinsauren und Tragersystemvariationen, thesis, Uni-
¨ ¨
versitat Frankfurt am Main, 2000.
/
a
¨
[30] C. Novak, S. Wurst, E. Urban, C.R. Noe, Sixteenth Scientific
/
Congress of Austrian Pharmaceutical Society, Vienna, Austria,
4842.
[16] W.J. Stec, B. Uznanski, A. Wilk, B.L. Hirschbein, K.L. Fearon,
2001. Sci. Pharm. 69 (2001) 76Á77.
[31] P.E. Nielsen, M. Egholm, R.H. Berg, O. Buchardt, Sequence-
/
B.J. Bergot, Bis(O,O-diisopropoxy phosphinothioyl) disulfide*
/
a
highly efficient sulfurizing reagent for cost-effective synthesis of
selective recognition of DNA by strand displacement with a
thymine-substituted polyamide, Science 254 (1991) 1497Á1500.
/
oligonucleotide phosphorothioates, Tetrahedron Lett. 34 (1993)
[32] P.E. Nielsen, M. Egholm, O. Buchardt, Peptide nucleic acid
5317Á5320.
/
(PNA). A DNA mimic with a peptide backbone, Bioconjug.
[17] M. Fuji, K. Ozaki, A. Kume, M. Sekine, T. Hata, Acylpho-
sphonates. 5. A new method for stereospecific generation of
phosphorothioate via aroylphosphonate intermediate, Tetrahe-
Chem. 5 (1994) 3Á7.
/
[33] V. Demidov, M.D. Frank-Kamenetskii, M. Egholm, O. Buchardt,
P.E. Nielsen, Sequence selective double strand DNA cleavage by
peptide nucleic acid (PNA) targeting using nuclease S1, Nucleic
dron Lett. 26 (1986) 935Á938.
/
[18] R. Cosstick, D.M. Williams, An approach to the stereoselective
synthesis of Sp-dinucleoside phosphorothioates using phospho-
Acids Res. 21 (1993) 2103Á2107.
/
[34] V. Demidov, V.N. Potaman, M.D. Frank-Kamenetskii, M.
Egholm, O. Buchard, S.H. Sonnichsen, P.E. Nielsen, Stability of
peptide nucleic acids in human serum and cellular extracts,
triester chemistry, Nucleic Acids Res. 15 (1987) 9921Á9932.
/
[19] B. Uznanski, A. Grajkowski, B. Krzyzanowska, A. Kazmier-
kowska, W.J. Stec, M.W. Wieczorek, J. Blaszczyk, Stereochem-
Biochem. Pharmacol. 48 (1994) 1310Á1313.
/
istry
of
base-catalyzed
ring
opening
of
oxathiaphospholanes. Absolute configuration of 2-{N-[(Rc)-1-(-
1,3,2-
[35] C.R. Noe, N. Windhab, G. Haberhauer, Novel three-atom 2?-5?
linkages in antisense nucleotides: synthesis and pairing properties
of dinucleotides with a carboxylic ester linkage, Arch. Pharm.
naphthyl)ethyl]amino}-2-thiono-1,3,2-oxathiaphospholanes and
O,S-dimethyl
N-[(Rc)-1-(-naphthyl)ethyl]phosphoramidothio-
10202.
(Weinheim, Germany) 328 (1995) 743Á745.
[36] S. Jordanova, G. Haberhauer, G.F. Ecker, C.R. Noe, Sixteenth
Scientific Congress of Austrian Pharmaceutical Society, Vienna,
/
ates, J. Am. Chem. Soc. 114 (1992) 10197Á
[20] A. Okruszek, Methods of synthesis of modified oligonucleotides
and their analogs-potential inhibitors of gene expression by
/
Austria, 2001. Sci. Pharm. 69 (2001) 328Á329.
/
antisense strategy, Biotechnologia (1994) 16Á
[21] W. Jasek, Austria-Codex Fachinformation Osterr, Apotheker-
Verlagsgesellschaft, Wien, 2001.
/
39.
¨
[37] J. Winkler, E. Urban, C.R. Noe, Sixteenth Scientific Congress of
Austrian Pharmaceutical Society, Vienna, Austria, 2001. Sci.
Pharm. 69 (2001) 78Á79.
/
[22] B. Jansen, V. Wacheck, E. Heere-Ress, H. Schlagbauer-Wadl, C.
Hoeller, T. Lucas, M. Hoermann, U. Hollenstein, K. Wolff, H.
Pehamberger, Chemosensitisation of malignant melanoma by
[38] A. Bauchinger, E. Urban, C.R. Noe, Sixteenth Scientific Congress
of Austrian Pharmaceutical Society, Vienna, Austria, 2001. Sci.
Pharm. 69 (2001) 74Á
/
75.
BCL2 antisense therapy, Lancet 18 (2000) 1728Á1733.
/
[39] Y. Inouye, K. Onodera, S. Kitaoka, S. Hirano, Some fatty-acid
derivatives of
4724.
[23] M. Cotten, B. Oberhauser, H. Brunar, A. Holzner, G. Issakides,
C.R. Noe, G. Schaffner, E. Wagner, M.L. Birnstiel, 2?-O-Methyl,
2?-O-Ethyloligoribo-nucleotides and phosphothioate oligodeox-
yribonucleotides as inhibitors of the in vitro U7 snRNP-depen-
D
-glucosamine, J. Am. Chem. Soc. 78 (1956) 4722Á
/
[40] H. Mack, J.V. Basabe, R. Brossmer, 2-Acetamido-2-deoxy-
gluco- and -manno-furanose: simple preparation of 2-
acetamido-2-deoxy- -mannose, Carbohydr. Res. 175 (1988)
311Á316.
[41] D.H.R. Barton, S.W. McCombie, New method for the deox-
ygenation of secondary alcohols, Perkin Trans. I (1975) 1574Á
1585.
D-
-
D
a
dent mRNA processing event, Nucleic Acid Res. 19 (1991) 2629Á
2635.
/
D
/
[24] E. Wagner, B. Oberhauser, A. Holzner, H. Brunar, G. Issakides,
G. Schaffner, M. Cotten, M. Knollmuller, C.R. Noe, A simple
¨
procedure for TBC preparation of protected 2?-O-methyl or 2?-O-
/