4′-Alkoxy oligodeoxynucleotides: A novel class of RNA mimics

Article abstract of DOI:10.1039/c1ob06148h

4′-Alkoxy-oligothymidylates were prepared as model compounds to study the influence of a C4′-alkoxy group on hybridisation. The phosphodiester homooligomers (15 units long) containing either a 4′-methoxy or 4′-(2-methoxyethoxy) group were found to display increased hybridisation with both dA₁₅ and rA₁₅ complementary counterparts compared to the natural oligothymidylate. In addition, we found their hybridisation behaviour to be similar to that of the regioisomeric 2′-O-methyl-oligothymidylate. The formed complexes (duplexes and triplexes) were studied using UV spectroscopy and polyacrylamide gel electrophoresis (PAGE). Structural background of the hybridization behaviour was examined using NMR and MDS. The favourable hybridisation properties of the 4′-alkoxyoligothymidylates indicated that 4′-alkoxy modified nucleotides are promising compounds for the assembly of chimeric oligonucleotides with tunable properties. The Royal Society of Chemistry 2011.

Full text of DOI:10.1039/c1ob06148h

View Article Online / Journal Homepage / Table of Contents for this issue
Organic &
Biomolecular
Chemistry
Dynamic Article Links
Cite this: Org. Biomol. Chem., 2011, 9, 8261
www.rsc.org/obc
PAPER
4¢-Alkoxy oligodeoxynucleotides: a novel class of RNA mimics†
Radek Liboska,a Jan Sna´sˇel,a Ivan Barv´ık,b Milosˇ Budeˇsˇ´ınsky´,a Radek Pohl,a Zdeneˇk Tocˇ´ık,a Ondrˇej Pa´v,a
Dominik Rejman,a Pavel Nova´ka and Ivan Rosenberg*a
Received 12th July 2011, Accepted 2nd September 2011
DOI: 10.1039/c1ob06148h
4¢-Alkoxy-oligothymidylates were prepared as model compounds to study the influence of a C4¢-alkoxy
group on hybridisation. The phosphodiester homooligomers (15 units long) containing either a
4¢-methoxy or 4¢-(2-methoxyethoxy) group were found to display increased hybridisation with both
dA15 and rA15 complementary counterparts compared to the natural oligothymidylate. In addition, we
found their hybridisation behaviour to be similar to that of the regioisomeric 2¢-O-methyl-oligothymidy-
late. The formed complexes (duplexes and triplexes) were studied using UV spectroscopy and polyacryl-
amide gel electrophoresis (PAGE). Structural background of the hybridization behaviour was examined
using NMR and MDS. The favourable hybridisation properties of the 4¢-alkoxyoligothymidylates
indicated that 4¢-alkoxy modified nucleotides are promising compounds for the assembly of chimeric
oligonucleotides with tunable properties.
apeutic targets for antisense oligonucleotides revived chemists’
Introduction
interest in these compounds.26 Undoubtedly, the complexity
Attempts to use synthetic oligonucleotides (ONs) as tools for
of problems associated with specifying suitable candidates for
the regulation of gene expression, starting with the original
practical use of oligonucleotides as efficient therapeutics still
antisense concept by Zamecnik,1 has necessitated diverse chemical
warrants further research on novel modifications of the sugar–
modifications of the natural oligomeric chains to improve their
phosphate backbone to achieve optimal oligonucleotide proper-
hybridisation capabilities, resistance toward nucleases, uptake
ties.
into cells, and RNase H activation to convert them into ther-
We have devoted considerable attention to the nuclease-stable,
apeutic candidates.2–5 After exploitation of phosphorothioates6
as the first-generation of antisense compounds, the second
generation (gapmers) combined 2¢-O-alkyl nucleotides7–9 and
phosphorothioate/phosphodiester units. This generation seems
to be providing promising ON constructs for further therapeutic
development. So far, hundreds of synthetic alternatives have
been introduced, and their properties assessed.5,10 Currently, there
are numerous modifications available for in vivo experiments,
including methylphosphonates,11 boranophosphates12 (recruiting
RNase H activity13), cyclohexenyl oligonucleotides,14 morpholino
ONs,15,16 N3¢→P5¢ phosphoramidates,17,18 LNA,19–22 and PNA.23
The recently published screening study on various siRNA
oligonucleotides constructed with a number of currently existing
chemical modifications formulated certain guidelines for efforts
to enhance siRNA activities through chemical modifications,
which further highlighted the role of the often subtle structural
alterations.24 In addition, the discovery of miRNAs25 as ther-
phosphonate-based internucleotide linkages.27–31 Recently, we re-
ported a study on the conformation of a novel phosphonate C3¢-
O-P-CH2-O-C4¢ internucleotide linkage isosteric to that of the
natural phosphodiester C3¢-O-P-O-CH2-C4¢ bond. In contrast to
the natural unit 2, we found that the sugar moiety of the nucleoside
phosphonate units31 1a and 1b occurred predominantly in the C2¢-
endo (~95%, S-type) and C3¢-endo (~98%, N-type) conformations,
respectively (Fig. 1).
This dramatic difference in the sugar ring conformation seems
to be related to the anomeric effect32 elicited by the presence
of an exocyclic oxygen atom attached to the C4¢ carbon.
Considering these findings, we concluded that also 4¢-alkoxy-2¢-
deoxynucleosides 3a and 3b could exhibit a similar conformational
behaviour,33–35 and that, in all likelihood, the incorporation of 3a
with preferential C3¢-endo conformation35 into the phosphodiester
oligodeoxynucleotides would give rise to new interesting chimeras
that hybridise well to a complementary RNA. We started with
thymine as the nucleobase because of its natural relation to the
2¢-deoxysugar moiety, but uracil would have also been justified
with regard to the prevailing C3¢-endo (RNA-type) conformation
of the 4¢-alkoxy-2¢-deoxyribose sugar moiety.
aInstitute of Organic Chemistry and Biochemistry, Academy of Sciences of
the Czech Republic, v. v. i., Flemingovo 2, 166 10 Prague 6, Czech Republic.
E-mail: rosenberg@uochb.cas.cz
bDivision of Biomolecular Physics, Institute of Physics, Faculty of Math-
ematics and Physics, Charles University, Ke Karlovu 5, 12116 Prague 2,
Czech Republic
Thus, in this paper, we describe two types of 4¢-alkoxy olig-
othymidylates and compare their hybridisation properties towards
dA15 and rA15 with those of the natural oligothymidylate and 2¢-
O-methyl-oligoribothymidylate.
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c1ob06148h
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 8261–8267 | 8261
©

View Article Online
Fig. 1 Sugar pucker of the 4¢-phosphonoalkoxy and 4¢-alkoxynucleoside units.
Scheme 1 Synthesis of protected 4¢-alkoxypentofuranosylthymine-3¢-cyanophosphoramidites 9a,9b, and 12.
with 2-cyanoethoxy-N,N-diisopropylaminochlorophosphine af-
forded the phosphoramidites 9a and 9b as the units for oligonu-
cleotide assembly.
Results and discussion
Chemistry
4¢-(2-Methoxyethoxy)thymidine-3¢-phosphoramidite 12 (Scheme
1). The 4¢,5¢-didehydro compound 5 was treated with anhydrous
2-methoxyethanol and m-chloroperoxybenzoic acid to yield the
epimeric 4¢-(2-methoxythoxy) derivatives 10a and 10b, which were
separated by silica gel chromatography. Only the former, resem-
bling the structure of dT, was further derivatized. Dimethoxytrity-
lation afforded derivative 11 which was phosphitylated with 2-
cyanoethoxy-N,N-diisopropylaminochlorophosphine to yield the
desired monomer 12.
Synthesis of monomers for oligonucleotide synthesis. 4¢-Epi-
meric 4¢-methoxypentofuranosylthymine-3¢-phosphoramidites (9a
and 9b in Scheme 1). The 5¢-iodo derivative 436 prepared from
2¢-deoxythymidine and a triphenylphosphine–iodine complex was
dehydrohalogenated with sodium methoxide, providing a good
yield of the 4¢,5¢-didehydro compound 5.37 Its treatment with
tert-butyldimethylsilyl chloride afforded the silyl derivative 6.38
The electrophilic addition of methanol in the presence of m-
chloroperbenzoic acid provided the desired 4¢-methoxy epimers,
7a and 7b, which were easily separable on silica gel in good yields
of 45% and 31%, respectively. The assignment of C4¢ configuration
in 7a and 7b and their desilylated derivatives, 3a and 3b (see below
and in the ESI† p. S8–S12), was made by NMR spectroscopy.
This was in agreement with previously published data.39 Both
epimers, 7a (a-L-threo) and 7b (b-D-erythro), were protected by
a dimethoxytrityl group in the 5¢ position, and the appropriate
5¢-dimethoxytrityl derivatives were treated, after extraction into
DCM, with TBAF in THF to afford compounds 8a and 8b, respec-
tively. Subsequent phosphitylation of the free 3¢-hydroxyl groups
Conformational analysis of 4¢-methoxynucleosides 3a and 3b.
The key compounds that inspired this work, the nucleosides 3a
and 3b, have been previously reported,37,39 and the configuration
of these 4¢-epimers was assigned by Chattopadhyaya.39 Because we
worked with the 3¢-O-silylated starting compound 6 (not 3¢-OH or
3¢-OAc, as in references37,39), to prepare the respective 4¢ epimers in
good yields we needed assurance of the reaction outcome, isomeric
purity, and unambiguous configuration assignments. In addition,
we wanted to learn more about the conformational preferences of
these individual isomers.
8262 | Org. Biomol. Chem., 2011, 9, 8261–8267
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
The configuration at carbon C4¢ and preferred orientation
of thymine in compounds 3a and 3b was determined from the
observed NOE contacts in the 2D-H,H-ROESY spectra in DMSO
(Fig. S1†). The absence of hydrogen on the C4¢ reduces the
number of proton vicinal couplings and limits the common
NMR conformation analysis using the PSEUROT40,41 program.
Therefore we used an approximate method for the estimation of the
population of “south”-type conformers (usually C2¢-endo) based
on eqn ([1]), as described for deoxyribonucleosides by Rinkel and
Altona.42
contrast to the PAGE we found in most cases negative profiles of
melting curves suggesting the presence of a triplex structure (for
the A295-temperature plots see ESI† p. S39–S41). These results,
quite different from those obtained by PAGE are difficult to
explain. One can conclude that the measurement at 295 nm could
be more sensitive to the presence of a triplex structure even at its
very low equilibrium concentration. Therefore, in the following
text, the terms duplex and triplex will refer to results obtained
from PAGE.
To gain further insight into the structure of possible complexes
formed, we measured the thermal difference spectra (TDS) ob-
tained by the subtraction of the complexes’ UV spectra measured
(%) south C2¢-endo conformer = [31.5 - J(1¢,2¢¢) - J(2¢,2¢¢)
(1)
- J(2¢¢,3¢)]/10.9] ¥ 100
◦
above and below their Tm (60 and 15 C, respectively). Normal-
The observed J-values in 3a (Fig. S1†) then lead to a high
preference for the “north”-type C3¢-endo conformation (~84%).
On the other hand, the observed J-values in 3b (Fig. S1†) indicate
a high population of “south”-type C2¢-endo conformers (~83%).
isation of the TDS provided two types of patterns (see ESI† p.
S37–S38). The first type is common for both dA15 and rA15 in the
presence of Mg2+ ions (Mg-type), and the second type is common
for dA15 and rA15 in the presence of Na+ ions (Na-type). In contrast
to the Mg-type with one negative local minimum, the Na-type
exhibited two negative local minima. The shapes of the Na-type
individual complexes were very complicated to interpret and did
not seem to correspond to any typical pattern reported. In the light
of these findings, we have evaluated the complexes based on the
deconvolution of Raman or UV spectra of their complementary
strand mixtures using the factor analysis.46 This powerful method
provides information on the number of statistically significant
complexes, the equilibrium concentration of all components in
a mixture (i.e., duplex, triplex, and single strands), and enables the
calculation of thermodynamic parameters. However, these results
will be published elsewhere.
As obvious from Table 1, the complexes of (4¢-MeOEtOT)14T
and (4¢-MeOT)14T oligomers (entries 4, 5) exhibited significantly
increased thermal stabilities in comparison to the natural dT15
and (4¢-MeT)14T,30 but very similar to that of (T2¢-OMe)14T (entry 3)
prepared as a reference compound.
To examine the mutual conformational compatibility of the
4¢-MeOEtOT, 4¢-MeOT, and T2¢-OMe units, we verified the thermal stabilities
of complexes of the mixed oligothymidylates, (4¢-MeOEtOT-4¢-MeOT)7T,
1
To support the results obtained from H NMR conformational
analysis using eqn ([1]), we explored the conformational behaviour
of compounds 3a and 3b by molecular modelling. We used
the well-known concept of pseudorotation43 for the sugar ring
conformation of nucleosides and calculated the endocyclic torsion
angles. The results of this analysis showed a 98% preference for
an N-pucker in 3a and an 84% population of an S-pucker in 3b
(Table S2†). This is in good agreement with experimental NMR
data and confirms our discussions on the MDS results and overall
structural relationships.
Hybridisation study (Table 1)
All oligothymidylates were synthesized by a phosphoramidite
method on solid phase according to the standard protocol. Mea-
surements of the thermal characteristics of the oligonucleotide
complexes were performed at 260 and 295 nm on a CARY 100
Bio UV Spectrophotometer equipped with a Peltier temperature
controller and thermal analysis software. In all cases, we observed
a single transition profile. Tm values were determined from the first
derivative plots (dA260/dT versus temperature) as the temperatures
corresponding to the local maxima of dA260/dT (see Table 1; for
profiles see ESI† p. S22–S36).
(
4¢-MeOEtOTT2¢-OMe)7T, and (4¢-MeOTT2¢-OMe)7T (entries 9–11), with both
dA15 and rA15. Compared to dT15, all three types of oligomers
exhibited a decreased affinity towards dA15-Na+, a retained or
slightly higher affinity towards dA15-Mg2+, and an increased
affinity towards rA15-Na+ and rA15-Mg2+. All three modifications
clearly exerted a similar and beneficial effect on hybridisation,
which obviously arose from specific structural factors (discussed
below in the MD study).
Characterization of the type of the complexes was performed by
a native PAGE at 15 ◦C in the presence of sodium or magnesium
ions (see ESI† p. S39–S41). PAGE revealed formation of triplexes
of dT15 (entry 1) and (4¢-MeOEtOT)14T (entry 4) with rA15, and also of
(
4¢-MeOT)14T (entry 5) with dA15 only in the presence of magnesium
ions, which are known to promote the triple helical structure.44
Interestingly, the (T2¢-OMe)14T oligomer (entry 2) used as a reference
compound formed under these conditions the duplex as did
also the remaining studied oligomers. Surprisingly, the complex
In contrast to these findings, the oligothymidylates (T-T2¢-OMe)7T
and (T-4¢-MeOT)7T (entries 7, 8) composed of natural thymidine and
modified units exhibited a strongly reduced affinity towards both
dA15 and rA15 in the presence of Na+ ions compared to dT15 (Table
1). Both these types of oligomers had reduced Tm values (below
the Tm of the natural dT15) when in duplexes with dA15-Mg2+,
but they retained their affinity towards rA15-Mg2+, similar to dT15.
This behaviour could be seen not only in the light of a different
population of the individual sugar conformers in the T unit (~60%
C2¢-endo) on one hand and in 4¢-MeOT (~80% C3¢-endo) or 2¢-MeOT
units (~80% C3¢-endo)47 on the other, but also in the different
dynamics of the individual conformational changes.
(
4¢-MeOEtOT)14T-dA15-Mg2+ (Entry 4) exhibiting a high Tm value
showed only a weak band on the PAGE corresponding, by the
mobility shift, to the triplex (see ESI†, p. 40, Fig. S4, panel C, lane
4). This result was reproducible, and we do not have any plausible
explanation for this phenomenon.
Since in several cases the complexes were not detected on the
PAGE, we measured their thermal characteristics at 295 nm that
should distinguish between the duplex and triplex structures.45
Duplexes do not typically exhibit absorbance changes, whereas
triplexes produce a negative melting curve (decreasing profile). In
A disturbance of the binding of Na+ or Mg2+ ions due to a
lack of structural and conformational uniformity in the chain is
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 8261–8267 | 8263
©

View Article Online
Table 1 Hybridisation properties of modified oligothymidylates with dA15 and rA15
also likely to contribute to the destabilisation of these complexes.
Concerning the role of the monovalent and divalent cations used
in the present study in thermal stabilities, there was a significant
difference in the Tm of (4¢-MeOT)14T and (4¢-MeOEtOT)14T complexed
to dA15 in the presence of Na+ ions. Compared to (4¢-MeOT)14T,
with a Tm value identical to that of natural dT15, the (4¢-MeOEtOT)14T
had a significantly lower Tm (~10 ◦C). A similar behaviour was
observed when the mixed oligomers (4¢-MeOEtOT-4¢-MeOT)7T, (4¢-MeOEtOT-
T2¢-OMe)7T, and (4¢-MeOT-T2¢-MeO)7T were examined in a complex with
dA15 and Na+ ions, with Tm values far below that of dT15 and the
uniformly modified (T2¢-OMe)14T and (4¢-MeOT)14T. The reasons for
these observations are suggested below in the section devoted to
MDS.
We also prepared the oligothymidylate (4¢-MeOT)14T (entry 6) ana-
logue, which also contained the C4¢-methoxy units but epimeric
to those in (4¢-MeOT)14T (entry 5). As we expected, this oligomer
did not hybridise with any of the respective counterparts (dA15 or
rA15) due to the backbone with a different organisation.
8264 | Org. Biomol. Chem., 2011, 9, 8261–8267
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
MD studies
In contrast to the natural dT15, which hybridises more efficiently
with dA15 than with rA15, the oligomers (4¢-MeOT)14T, (4¢-MeOEtOT)14T,
and (T2¢-OMe)14T hybridised more tightly with rA15. This correlates
with the ab initio calculations and NMR findings showing the
preferential C3¢-endo conformation of the sugar moieties in the
corresponding nucleosides.
To gain further knowledge on how the measured thermal
stabilities (Tm) are related to the structures of the novel
oligomers, we performed molecular dynamics simulations (MDS)
with model oligonucleotide triple helical structures carrying
chemical modifications on their either C4¢ or C2¢ atoms:
T(4¢-MeOT)8T·rA10*T(4¢-MeOT)8T, T(4¢-MeOEtOT)8T·rA10*T(4¢-MeOEtOT)8T,
and T(T2¢-OMe)8T·rA10*T(T2¢-OMe)8T (· denotes Watson–Crick base
pairing, and * indicates Hoogsteen base pairing). Method details
are provided in the ESI.†
Fig. 3 Triple helical structure of T(4¢-MeOEtOT)8T·rA10*T(4¢-MeOEtOT)8T used
as a model in the MD simulations. Enlarged versions of these pictures
(either with or without hydrogen atoms) are provided in the ESI† (Fig.
S9, S10), including those produced for the reference triple helical structure
T(T2¢-OMe)8T·rA10*T(T2¢-OMe)8T (Fig. S5–S6).
Intra-strand effects – The C2¢-OCH3 moieties in the T2¢-OMe
-
modified chain were buried within the nucleic acid grooves (Fig.
S5–S6†), and the CH3 groups interacted tightly with the C5¢
atoms of the neighbour nucleotide residues through hydrophobic
effects. Moreover, the CH3 groups partially shielded the mutual
repulsion between the polarised O2¢, O2, and O4¢ atoms. There-
fore, the intra-strand hydrophobic and electrostatic complemen-
tarity effects could boost-stimulate helical preorganization and
provide additional stabilisation. In contrast, the surprisingly low
melting temperature (20 ◦C) observed for the (4¢-MeOEtOT)14T-dA15-
Na+ complex (Table 1, entry 4) could have been caused by a
crowding together of the central hydrophobic portions of the
C4¢-OCH2CH2OCH3 moieties in the 4¢-MeOEtOT chain, affecting the
ability of (4¢-MeOEtOT)14T to organise into a helix-like structure.
Direct inter-strand effects – In the triple helical structures,
the (4¢-MeOT)8T and (4¢-MeOEtOT)8T Hoogsteen strands (light blue
in Fig. 2, 3, S7–S10†) divided the major groove, thus creating
the minor-major and major-major grooves. The former was very
narrow, and the alkoxy groups in 4¢-MeOT and 4¢-MeOEtOT units were
oriented towards the phosphate groups of the central dA15/rA15.
Nevertheless, no apparent steric conflict was found, even in the
case of the relatively bulky C4¢-OCH2CH2OCH3 moiety in the
4¢-MeOEtOT unit bridging the minor-major groove. On the other hand,
there were no suitable hydrogen-bond donors in the dA15/rA15
chains that could provide direct inter-strand stabilisation through
interactions with the oxygen atoms of the alkoxy groups in the
4¢-MeOT-, 4¢-MeOEtOT-, and T2¢-OMe-modified strands.
Solvent mediated inter-strand effects – Nevertheless, additional
stabilisation could have been provided by water molecules and
cations. The thermal stabilities of the (4¢-MeOT)14T complexes with
dA15/rA15, expressed as Tm values, were slightly higher than that of
(T2¢-OMe)14T. Interestingly, these differences were more pronounced
with Na+ rather than Mg2+ ions (Table 1). Na+ ions were found
to reside in the proximity of C4¢-OCH3 groups if partial atomic
charges on oxygen atoms were calculated by either the CHARMM
force field or the ab initio Mulliken equations instead of the original
AMBER/RESP method. In this case, the oxygen atoms of the 4¢-
OCH3 and O3¢-phosphoester groups (originating from the same
chain) were directly involved in Na+-ion binding (Fig. S14†).
Consequently, the Na+ ions residing in the nucleic-acid grooves
diminished the interstrand repulsion between the nucleic acids
chains. The Na+ ions are only weakly bound to the ligands and
interchange many binding sites within the MD runs. In contrast,
the Mg2+ ions are usually firmly anchored towards the ligands, with
residence times that typically exceeded the duration of the MD
runs. Surprisingly, the Mg2+ ions placed in the Na+ binding sites of
the nearby of C4¢-OCH3 groups escaped quickly towards the non-
bridging oxygen atoms of the phosphate groups (O1P) located in
the same oligonucleotide chain. Aside from the strong coulombic
attraction of the highly charged Mg2+-O1P atoms, the substantially
smaller van der Waals radius of the Mg2+ ions (compared to
Na+) played a role in locking the C4¢-alkoxy sugars in a labile
intermediate conformer position between the C2¢- and C3¢-endo
energy minima. Such a position did not last long, and the sugars
typically underwent a conformational transition accompanied by
the Mg2+ ions escaping towards the O1P atoms within tens or
hundreds of picoseconds.
Fig. 2 Triple helical structure of T(4¢-MeOT)8T·rA10*T(4¢-MeOT)8T used as a
model in the MD simulations. Enlarged versions of these pictures (either
with or without hydrogen atoms) are provided in the ESI† (Fig. S7,
S8), including those produced for the reference triple helical structure
T(T2¢-OMe)8T·rA10*T(T2¢-OMe)8T (Fig. S5–S6).
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 8261–8267 | 8265
©

View Article Online
Interestingly, the Hoogsteen (4¢-MeOEtOT)14T chains carrying
the C4¢-OCH2CH2OCH3 groups could share the Mg2+ ions in
triple helical structures with phosphate groups from the central
adenosine strand (Fig. 4, S13†). This could partly explain the
highest measured melting temperatures (57 ◦C) determined for
the [(4¢-MeOEtOT)14T]-rA15-Mg2+ and [(4¢-MeOEtOT-4¢-MeOT)7T]-rA15-Mg2+
systems (Table 1, entries 4 and 9). Therefore, the Mg2+ ions were
placed in the proximity of non-bridging oxygen atoms from the
phosphate groups of the rA15 strand, as is found in many X-ray
structures. The C4¢-OCH2CH2OCH3 groups were also adjusted to
interact with the Mg2+ ions (Fig. S11–S12†). Two contacts between
the Mg2+ ion and the C4¢-OCH2CH2OCH3 moiety involving both
or only the terminal oxygen atoms were tested. The former
configuration (Mg2+- C4¢OCH2CH2OCH3) resulted in more stable
contacts, so that the Hoogsteen (4¢-MeOEtOT)14T strand and rA15 were
tightly attached together (Fig. 4, S13†). However, the resulting
triple helical structure was slightly disturbed in the centre. On the
other hand, the latter configuration (Mg2+- C4¢OCH2CH2OCH3)
resulted in less stable contacts, but the resulting triple helical
structure was more similar to the natural triplex.
comparably as phosphoramidate oligodeoxynucleotides.48 The
observed Tm enhancement was of similar magnitude to that of
the respective 2¢-O-methyl group-containing (T2¢-OMe)14T used as a
reference oligonucleotide, thus confirming the original assumption
that these two types of modifications sharing the presence of
an N-type sugar pucker in the parent nucleoside could provide
oligonucleotides with similar hybridisation properties. We are
aware of the importance of creating the mixed purine–pyrimidine
4¢-alkoxyoligonucleotides (i.e., containing all of the U, C, A,
and G nucleobases) for their direct use in biological experi-
ments. Therefore, we are currently working on a straightforward
synthesis of 4¢-alkoxy-2¢deoxynucleosides in the purine series to
make these compounds practically available on a preparative
scale. We believe that the 4¢-alkoxy-2¢-deoxynucleosides will be
interesting building units for assembly of the phosphodiester 4¢-
alkoxyoligodeoxynucleotides as new interesting RNA chimeras.49
Experimental
Synthesis of monomers
Preparation of the final nucleoside phosphoramidites for oligonu-
cleotide assembly and characterisation of compounds is described
in detail in the ESI.†
Synthesis of oligonucleotides
All oligonucleotides (Table 1) were synthesised from the appro-
priate monomers on a ~0.5 mmol scale by a standard phos-
phoramidite method using the LCAA CPG with attached 5¢-
O-dimethoxytritylthymidine-3¢-O-hemisuccinate as the first nu-
cleoside. Deprotection and release of each oligonucleotide from
CPG was achieved with gaseous ammonia (0.7 MPa) at r.t. for
8 h. Oligonucleotides were purified on Luna C18 columns (5 mm,
10 ¥ 250 mm) at a flow rate 3 ml min-1 using a linear gradient
of acetonitrile (0→50%, 40 min) in 0.1 M TEAA (pH 7.2 at
r.t.) and finally freeze-dried. All prepared oligonucleotides were
characterised by MALDI TOF (Table S1†).
Fig. 4 Two modes of contact between Mg2+ ions and C4¢-OCH2CH2-
OCH3 groups in the (4¢-MeOEtOT)28, (4¢-MeOEtOT)29 and rA6-rA9 portions of
the Hoogsteen strand of the [T(4¢-MeOEtOT)8T]W-C ·rA10*[T(4¢-MeOEtOT)8T]Hgstn
triplex, where rA10 consists of residues 1–10, Watson–Crick pyrimidine
strand – residues 11–20, and Hoogsteen strand – residues 21–30. The
enlarged version is available in the ESI† (Fig. S13), where binding of Mg2+
ions in the context of the whole triple helical structure can be found as well
(Fig. S11, S12†).
Stability of the modified oligothymidylates towards nucleases
The stability of the homooligomers (4¢-MeOT)14T and (4¢-MeOEtOT)14T
was verified under conditions described earlier.50 Complete sta-
bility in the L1210 cell-free extract and in the presence of
endonuclease P1 (EC 3.1.3.16) was observed. Experiments carried
out under conditions in which halftime cleavage of the natural
dT15 was less than 1 min showed that no cleavage of the modified
oligomers occurred within 2 h (determined by HPLC). However,
no resistance was found against snake venom exonuclease (EC
3.1.30.1).
Conclusion
In summary, we have presented a new and promising structural
alternative to the existing analogues of the natural phosphodiester
oligonucleotides. Modifications involving the presence of either a
4¢-methoxy or 4¢-(2-methoxyethoxy) group in oligothymidylates,
represented by the (4¢-MeOT)14T and (4¢-MeOEtOT)14T oligonucleotides,
were found to display markedly better hybridisation properties
with both the dA15 and rA15 counterparts compared to unmodified
dT15. The novel oligomers containing the 4¢-MeOEtOT and 4¢-MeOT
units are DNA constructs that mimicked the RNA behaviour
Hybridisation study (Table 1)
Thermal experiments with oligonucleotide complexes were per-
formed at 260 and 295 nm on a CARY 100 Bio UV Spec-
trophotometer (Varian Inc.) equipped with a Peltier temperature
controller and thermal analysis software. An equimolar mixture
of modified T15 and natural dA15 (or rA15) strands was prepared
to give a final concentration of 4 mM in 50 mM Tris-HCl (pH
7.2) and 1 mM EDTA containing either 100 mM Na+ or 10 mM
8266 | Org. Biomol. Chem., 2011, 9, 8261–8267
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Mg2+ ions. A heating–cooling cycle over a range of 18–90 ◦C with
a gradient of 0.5 ◦C min-1 was applied. Tm value of each complex
was determined from the first derivative plots (dA260/dT versus
temperature) as the temperature at a local maximum of dA260/dT.
For thermal stability curves at 260 and 295 nm and 1st derivation
dA260/dT versus T plots see the ESI† (p. S32–S36).
18 B. Shea-Herbert, K. Pongracz, J. W. Shay and S. M. Gryaznov,
Oncogene, 2002, 21, 638–642.
19 S. K. Singh, P. Nielsen, A. A. Koshkin and J. Wengel, Chem. Commun.,
1998, 455–456.
20 M. Petersen and J. Wengel, Trends Biotechnol., 2003, 21, 74–81.
21 S. Obika, D. Nanbu, Y. Hari, K. Morio, Y. In, T. Ishida and T. Imanishi,
Tetrahedron Lett., 1997, 38, 8735–8738.
22 D. Honcharenko, J. Barman, O. P. Varghese and J. Chattopadhyaya,
Biochemistry, 2007, 46, 5635–5646.
23 M. Egholm, O. Buchardt, L. Christensen, C. Behrens, S. M. Freier, D.
Driver, R. H. Berg, S. K. Kim, B. Norde´n and P. E. Nielsen, Nature,
1993, 365, 566–568.
24 J. B. Bramsen, M. B. Laursen, A. F. Nielsen, T. B. Hansen, C. Bus,
N. Langkjær, B. R. Babu, T. Højland, M. Abramov, A. Van Aerschot,
D. Odadzic, R. Smicius, J. Haas, C. Andree, J. Barman, M. Wenska, P.
Srivastava, C. Zhou, D. Honcharenko, S. Hess, E. Mu¨ller, G. V. Bobkov,
S. N. Mikhailov, E. Fava, T. F. Meyer, J. Chattopadhyaya, M. Zerial,
J. W. Engels, P. Herdewijn, J. Wengel and J. Kjems, Nucleic Acids Res.,
2009, 37, 2867–2881.
25 R. C. Lee, R. L. Feinbaum and V. Ambros, Cell, 1993, 75, 843–854.
26 C. Davis, B. Lollo, S. Freier and C. Esau, Nucleic Acids Res., 2006, 34,
2294–2304.
27 I. Rosenberg, in Frontiers in Nucleosides and Nucleic Acids; ed. R. F.
Schinazi and D. C. Liotta, IHL Press, Tucker, GA, 2004, pp. 519–
548.
Determination of the type of complex by PAGE
Modified oligonucleotides (mT15) were mixed with (32P) 5¢-end
labelled dA15 or rA15 in 10 mM Tris (pH 7.4) and 150 mM NaCl (or
10 mM MgCl2) at 1 : 1 molar ratios to obtain final concentrations
of 100 nM. Individual annealings were performed by heating the
mixtures to 80 ◦C and cooling them slowly to room temperature.
The samples were then mixed with 10 mM Tris (pH 7.4) and 20%
glycerol in a 1 : 1 ratio and loaded onto 20% native polyacrylamide
gels. Electrophoresis was run in 1xTBE (or 1xTB with 10 mM
MgCl2) at 12.5 V cm-1 for 10 h at 15 ◦C. The gels were dried and
visualised by autoradiography (Fig. S4,† p. S39–S40).
28 V. Vaneˇk, M. Budeˇsˇ´ınsky´, M. Rinnova´ and I. Rosenberg, Tetrahedron,
Acknowledgements
2009, 65, 862–876.
29 D. Rejman, P. Kocˇalka, R. Pohl, Z. Tocˇ´ık and I. Rosenberg, Collect.
Czech. Chem. Commun., 2009, 74, 935–955.
30 Z. Tocˇ´ık, I. Barv´ık Jr., M. Budeˇsˇ´ınsky´ and I. Rosenberg, Biopolymers,
2006, 83, 400–413.
31 Z. Tocˇ´ık, I. Barv´ık Jr., M. Budeˇsˇ´ınsky´ and I. Rosenberg, Biopolymers,
2009, 91, 514–529.
32 C. Thibodeau, P. Acharya and J. Chattopadhyaya, Stereoelectronic
effects in nucleosides and nucleotides and their structural implications,
Uppsala University Press, ISBN 91-506-1351-0, 2nd Ed., Uppsala,
2003.
Support from the grant 202/09/0193 (Czech Science Foundation)
and the Research Centres KAN200520801 (Acad. Sci. CR) and
LC060061 (Ministry of Education, CR), under the Institute
research project Z40550506, are gratefully acknowledged, in
addition to support by project No. MSM 0021620835 (Ministry
of Education, CR). The authors thank the staff of the Mass
Spectrometry Group of the IOCB AS, v.v.i. (Dr Josef Cvacˇka,
Head) for the HR-MS spectra.
33 J. P. H. Verheyden and J. G. Moffat, J. Am. Chem. Soc., 1975, 97,
4386–4395.
34 H. Maag, R. M. Rydzewski, M. J. McRoberts, D. Crawford-Ruth, J. P.
H. Verheyden and E. J. Prisbe, J. Med. Chem., 1992, 35, 1440–1451.
35 W. Tong, P. Agback and J. Chattopadhyaya, Acta Chem. Scand., 1993,
47, 145–156.
References
1 P. C. Zamecnik and M. L. Stephenson, Proc. Natl. Acad. Sci. U. S. A.,
1978, 75, 280–284.
2 E. Uhlmann and A. Peyman, Chem. Rev., 1990, 90, 543–584.
3 S. T. Crooke, Annu. Rev. Med., 2004, 55, 61–95.
4 I. Lebedeva and C. A. Stein, Annu. Rev. Pharmacol. Toxicol., 2001, 41,
403–419.
36 P. J. Garegg and B. Samuelsson, J. Chem. Soc., Perkin Trans. 1, 1980,
2866–2869.
37 H. Maag, R. M. Rydzewski, M. J. McRoberts, D. Crawford-Ruth, J. P.
H. Verheyden and E. J. Prisbe, J. Med. Chem., 1992, 35, 1440–1451.
38 Z. Tocˇ´ık, I. Dvorˇa´kova´, R. Liboska, M. Budeˇsˇ´ınsky´, M. Masoj´ıdkova´
and I. Rosenberg, Tetrahedron, 2007, 63, 4516–4534.
39 W. Tong, P. Agback and J. Chattopadhyaya, Acta Chem. Scand., 1993,
47, 145–156.
40 J. van Wijk, C. A. G. Haasnoot, F. A. A. M. de Leeuw, B. D. Huckriede,
A. Westra Hoekzema and C. Altona, PSEUROT 6.2 1993, PSEUROT
6.3 1999; Leiden Institute of Chemistry, Leiden University.
41 F. A. A. M. de Leeuw and C. Altona, J. Comput. Chem., 1983, 4,
428–437.
5 J. Kurreck, in TherapeuticOligonucleotides, ed. J. Kurreck, RSC Pub-
lishing, Cambridge (UK), 2008, pp. 1–22.
6 F. Eckstein, Antisense Nucleic Acid Drug Dev., 2000, 10, 117–121.
7 A. M. Iribarren, B. S. Sproat, P. Neuner, I. Sulston, U. Ryder and A. I.
Lamond, Proc. Natl. Acad. Sci. U. S. A., 1990, 87, 7747–7751.
8 C. Boiziau, B. Larrouy, B. S. Sproat and J. J. Toulme´, Nucleic Acids
Res., 1995, 23, 64–71.
9 M. Manoharan, Biochim. Biophys. Acta, 1999, 1489, 117–130.
10 S. M. Freier and K. H. Altmann, Nucleic Acids Res., 1997, 25, 4429–
4443.
11 M. A. Reynolds, R. I. Hogrefe, J. A. Jaeger, D. A. Schwartz, T. A. Riley,
W. B. Marvin, W. J. Daily, M. M. Vaghefi, T. A. Beck, S. K. Knowles, R.
E. Klem and L. J. Arnold, Jr., Nucleic Acids Res., 1996, 24, 4584–4591.
12 D. Sergueev, A. Hasan, M. Ramaswamy, B. R. Shaw and P. M. Gross,
Nucleosides Nucleotides Nucl. Acids, 1997, 16, 1533–1538.
13 X. Wang, M. Dobrikov, D. Sergueev and B. R. Shaw, Nucleosides
Nucleotides Nucl. Acids, 2003, 22, 1151–1153.
14 K. Nauwelaerts, E. Lescrinier, G. Sclep and P. Herdewijn, Nucleic Acids
Res., 2005, 33, 2452–2463.
15 J. Summerton and D. Weller, Antisense Nucleic Acid Drug Dev., 1997,
7, 187–195.
16 J. Heasman, Dev. Biol., 2002, 243, 209–214.
17 M. Faria, D. G. Spiller, C. Dubertret, J. S. Nelson, M. R. H. White,
D. Scherman, C. He´le`ne and C. Giovannangeli, Nat. Biotechnol., 2001,
19, 40–44.
42 L. J. Rinkel and C. Altona, J. Biomed. Struct. Dyn., 1987, 4, 621–649 .
43 C. Altona and M. Sundaralingam, J. Am. Chem. Soc., 1972, 94, 8205–
8212.
44 D. Coman and I. M. Russu, Nucleic Acids Res., 2004, 32, 878–883.
45 J.-L. Mergny, J. Li, L. Lacroix, S. Amrane and J. B. Chaires, Nucleic
Acids Res., 2005, 33, e138.
ˇ
46 J. Hanusˇ, I. Barv´ık, K. Ruszova´-Chmelova´, J. Steˇpa´nek, P.-Y. Turpin,
I. Rosenberg and M. Petrova´-Endova´, Nucleic Acids Res., 2001, 29,
5182–5194.
47 B. Bhat, E. E. Swayze, P. Wheeler, S. Dimock, M. Perbost and Y. S.
Sanghvi, J. Org. Chem., 1996, 61, 8186–8199.
48 M. Egli and S. M. Gryaznov, Cell. Mol. Life Sci., 2000, 57, 1440–1456.
49 M. Koizumi, Curr. Topics Med. Chem., 2007, 7, 661–665.
ˇ
50 D. Rejman, J. Sna´sˇel, R. Liboska, Z. Tocˇ´ık, O. Pacˇes, S. Kra´l´ıkova´,
M. Rinnova´, P. Koisˇ and I. Rosenberg, Nucleosides Nucleotides Nucl.
Acids, 2001, 20, 819–823.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 8261–8267 | 8267
©