Oligonucleotides with 1,4-dioxane-based nucleotide monomers

Article abstract of DOI:10.1021/jo300222q

An epimeric mixture of H-phosphonates 5R and 5S has been synthesized in three steps from known secouridine 1. Separation of the epimers has been accomplished by RP-HPLC, allowing full characterization and incorporation of monomers X and Y into 9-mer oligonucleotides using H-phosphonates building blocks 5R and 5S, respectively. A single incorporation of either monomer X or monomer Y in the central position of a DNA 9-mer results in decreased thermal affinity toward both DNA and RNA complements (δT_m = -3.5 °C/-3.5 °C for monomer X and δT_m = -11.0 °C/-6.5 °C for monomer Y). CD measurements do not reveal major rearrangements of the duplexes formed, but molecular modeling suggests that local rearrangement of the sugar phosphate backbone and decreased base interactions with neighboring bases might be the origin of the decreased stability of duplexes.

Full text of DOI:10.1021/jo300222q

Article
pubs.acs.org/joc
Oligonucleotides with 1,4-Dioxane-Based Nucleotide Monomers
Andreas S. Madsen and Jesper Wengel*
Nucleic Acid Center, Department of Physics, Chemistry and Pharmacy, University of Southern Denmark, 5230 Odense M, Denmark
S
* Supporting Information
ABSTRACT: An epimeric mixture of H-phosphonates 5R
and 5S has been synthesized in three steps from known
secouridine 1. Separation of the epimers has been accom-
plished by RP-HPLC, allowing full characterization and
incorporation of monomers X and Y into 9-mer oligonucleo-
tides using H-phosphonates building blocks 5R and 5S,
respectively. A single incorporation of either monomer X or
monomer Y in the central position of a DNA 9-mer results in
decreased thermal affinity toward both DNA and RNA
complements (ΔTm = −3.5 °C/−3.5 °C for monomer X
and ΔTm = −11.0 °C/−6.5 °C for monomer Y). CD
measurements do not reveal major rearrangements of the
duplexes formed, but molecular modeling suggests that local rearrangement of the sugar phosphate backbone and decreased base
interactions with neighboring bases might be the origin of the decreased stability of duplexes.
demonstrated potential as antisense drug by directing transla-
tional arrest of a targeted gene.22
INTRODUCTION
■
Formal expansion of the furanose ring of nucleosides to give
nucleoside analogues with six-membered rings as sugar moiety
has been a significant research field during the past few
decades.1,2 Simple cyclohexane derivatives3−5 and nucleoside
analogues with 1,4-heteroatom cyclohexanes including 1,4-
dioxane, 1,4-oxathiane, 1,4-dithiane, and morpholine as the
sugar moiety have also been reported,6−14 and formal 1,4-
dioxane derivatives formed by hydration of 2′,3′-dialdehyde-
2′,3′-seco-nucleosides have been reported as a stable form of
such dialdehydes.15,16 None of the above-mentioned nucleoside
analogues have been incorporated into oligonucleotides (ONs)
as they all, but the hydrates, lack either one or two hydroxy
groups to enable phosphodiester oligomerization.
Eschenmoser initiated the field of ONs with six-membered
sugar rings by the synthesis of homo-DNA and its allo-, altro-
and gluco-pyranosyl analogues, where the furanose ring is
formally expanded by a methylene (or hydroxymethylene)
group between C1′ and C2′.17 These studies demonstrated that
homo-DNA:homo-DNA exhibit very strong Watson−Crick
base-pairing but also reverse Hoogsteen adenine and guanine
self-pairings, whereas it does not form duplexes with DNA or
RNA. In a recent example, the synthesis and oligomerization of
a 2′-enopyranose thymidine was realized, and interestingly,
pairing with both poly(dA) and poly(rA) was demonstrated.18
Hexitol nucleic acid (HNA) and its hydroxylated (mannitol
and altrohexitol) derivatives, formally derived from pentofura-
noses by inserting a methylene group between C1′ and O4′,
represents another ON system which is able to form duplexes
not only with itself but also with RNA and DNA.19−21 The
hexitol skeleton mimics the C3′-endo-conformation adopted by
RNA in RNA:RNA duplexes (stability of duplexes: HNA:HNA
> HNA:RNA > HNA:DNA).19 Furthermore, HNA has
Incorporation of nucleoside analogues with saturated and
unsaturated cyclohexane rings into ONs have been realized by
the formal substitution of O4′ of DNA with an ethyl or ethenyl
moiety, resulting in cyclohexane nucleic acid (CNA)23 and
cyclohexene nucleic acid (CeNA),24,25 respectively. Both form
stable duplexes with RNA and DNA, and CeNA has been
shown to elicit RNase H activity, although at a lower rate than
phosphodiester and phosphorothioate DNA.25 Another mod-
ification having a 6-membered ribose substitute is the so-called
morpholino-NA26 which forms neutral oligomers by linking the
monomers via phosphordiamidate linkages.
Both homo-DNA and HNA are derived from RNA by formal
addition of a methylene group. As hydration of duplexes plays a
crucial role in duplex stability, introduction of a hydrophobic
methylene moiety might not be optimal. Alternatively, the ring
could be expanded by introduction of an oxygen atom.
However, 1,2- and 1,3-dioxanes would be expected to be very
labile, leaving only 1,4-dioxanes as potential nucleotide mimics.
1′,2′,4′- and 1′,2′,3′-substituted dioxanes (see Figure 1 for
numbering)27 bear closer resemblance to O2′-linked ribofur-
anosides, and 2′,3′,4′-substituted dioxanes are expected to be
very unstable resulting in cleavage to the dialdehyde with the
nucleobase acting as leaving group. However, 1′,3′,4′-substituted
dioxanes satisfy both substitution pattern and formal
resemblance to O3′-linked ribofuranosides, establishing nucleo-
tide monomers X and Y as potential nucleotide mimics (Figure
1).
Received: January 31, 2012
Published: March 22, 2012
© 2012 American Chemical Society
3878
dx.doi.org/10.1021/jo300222q | J. Org. Chem. 2012, 77, 3878−3886

The Journal of Organic Chemistry
Article
of the 2′,3′-diol or 2′-selective reduction of the corresponding
2′,3′-dialdehyde.
Of the many possible oxidation methods known, acidic
methods such as Swern or chromium-based oxidations were
unfeasible (leading to rapid cleavage of the trityl ether, giving
two primary hydroxy groups, distinguishable only by the
prochirality of the β-position (i.e., C4′)). Nonacidic DMSO-
based oxidations as Pfitzner−Moffatt and Parikh−Doering
̈
Figure 1. Structure of target monomers X and Y, with the used
numbering indicated at monomer X.
oxidation proved unsuccessful, as did the otherwise widespread
Dess−Martin periodinane oxidation. However, oxidation was
readily effected by IBX (o-iodoxybenzoic acid) in pyridine/
EtOAc at near-reflux.35,36 By NMR spectroscopy, oxidation to
give aldehyde 2 was confirmed by the disappearance of signals
corresponding to H3′ and 3′-OH and the appearance of a peak
at 9.75 ppm (1H NMR), and by a downfield shift to 200.3 ppm
of the peak corresponding to C3′ (13C NMR). Despite the
extremely easy workup for this reaction, varying amounts of the
corresponding hydrate could be detected in all reaction
mixtures isolated and analyzed. As a result, oxidation followed
by debenzoylation by sodium hydroxide in methanol was
carried out as a two step procedure without purification of the
aldehyde intermediate, affording hemiacetal 3R/S in an
Besides analogy with homo-DNA and HNA, monomers X
and Y are appealing as they have an oxygen atom in a position
expected to be close to the position of the 2′-OH group of
RNA. This oxygen could potentially act as hydrogen bond
acceptor, allowing preservation of structural integrity of a
duplex. The corresponding free nucleotides or nucleosides
could also potentially act as antiviral or anticarcinogenic drugs
by offering both hydrogen bond donor and acceptor sites which
could mimic the 2′- and 3′-hydroxy groups of natural
nucleotides.
From a structural point of view, six-membered rings are
furthermore interesting as the possible conformations and the
energies needed for their interconversions are entirely different
from those of the furanose ring of natural nucleotides. Thus,
whereas the furanose ring of natural nucleotides has a relatively
low energy barrier for the pseudorotation between N-type and
1
approximate 1:1 ratio (according to H NMR) in 67% yield
over two steps (Scheme 1). Disappearance of the signal from
the aldehyde moiety in 1H NMR upon cleavage of the benzoyl
ester indicates ring closure to the hemiacetal, with no detectable
presence of the hydroxyaldehyde.
S-type conformers (ΔG⧧ = 8−20 kJ/mol),28,29 inversion
Phosphoramidite Strategy. The epimeric mixture of
hemiacetal 3R/S reacted readily with 2-cyanoethyl N,N-
diisopropylphosphoramidochlorodite in CH2Cl2 resulting in
four diastereoisomers in 73% combined yield (Scheme 1). Even
though C3′ is the most relevant stereocenteras the
phosphorus atom of the resulting phosphodiesters after
oligonucleotide synthesis is nonchiralfull separation and
assignment of configuration at the two stereocenters was
desired for all four diastereomers, especially as we wanted to
obtain oligonucleotides containing monomers X and Y with
defined C3′ configuration. However, only partial separation of
the four diastereoisomers could be accomplished by silica gel
column chromatography. The mixture could be separated into
two fractions, each containing two phosphoramidites (31P
NMR (121 MHz, CDCl3) δ 149.5 ppm and 150.1 ppm (∼1:4
ratio, termed 4A/B) and 152.0 ppm and 152.4 ppm (∼6:1
ratio, termed 4C/D), respectively). Attempts to separate the
binary mixtures by RP-HPLC were also unsuccessful, giving
only partial separation even on analytical (1.0 mg) scale. As a
result of the poor resolution obtained, in combination with the
risk of degradation of the phosphoramidites during workup,
separation of phosphoramidites 4 was not pursued further.
c
between the two chair conformations of both unsubstituted
cyclohexane and 1,4-dioxane is found to proceed through
several higher energy transition states and metastable
intermediates (ΔG⧧ = 40 kJ/mol).30−33 As a result, 1,4-
c
dioxane nucleotides are expected to show a smaller conforma-
tional freedom than natural furanose nucleotides.
With the above-mentioned considerations in mind, we
therefore set to synthesize oligonucleotides containing
monomers X and Y.
RESULTS AND DISCUSSION
■
Synthesis of Hemiacetal 3. Hemiacetal 3, expected to
exist as an equilibrium between its two epimeric forms, was
identified as key intermediate from which both monomers X
and Y could conceivably be obtained. As selective O2′-
benzoylation of the corresponding 2′,3′-diol to give benzoate
ester 1 has recently been described en route to UNA (unlocked
nucleic acid),34 a synthetic scheme of oxidation of the free 3′-
hydroxy group of alcohol 1 followed by O2′-debenzoylation
was preferred over routes including either 3′-selective oxidation
Scheme 1. Synthesis of Phosphoramidites 4 and H-Phosphonates 5
3879
dx.doi.org/10.1021/jo300222q | J. Org. Chem. 2012, 77, 3878−3886

The Journal of Organic Chemistry
Article
Figure 2. 31P NMR spectra after coupling of phosphoramidite mixtures 4A/B and 4C/D (top and bottom, respectively) with 2-cyanoethanol,
demonstrating that both binary mixtures contain both C3′-epimers. Spectra of amidites and activated amidites are recorded at 81 MHz. Spectra of
final phosphites are recorded at 121 MHz.
Direct assignment of stereochemistry at C3′ (and at the chiral
employed.43 H-Phosphonate building blocks for natural
nucleosides are commercially available, and the corresponding
solid phase ON synthetic protocols are well established with
coupling yields comparable to those obtained with phosphor-
amidite building blocks.44 An appealing feature of the H-
phosphonate approach is the nonstereogenic nature of the H-
phosphonate moiety of the monoester building blocks, whereby
the major obstacle in the phosphoramidite strategy could be
circumvented. Furthermore, several examples of phosphoryla-
tion via H-phosphonates on epimeric positions have been
reported in the literature.45−48
phosphorus atom) of the two components of the binary
1
mixtures of phosphoramidites 4 was not feasible by H NMR
due to the complexity of the spectra and the low intensity of the
minor isomer. As an alternative to full separation and
configurational assignment, the relative C3′ stereochemistry of
the binary mixtures was determined. Each of the separated
binary mixtures was coupled with 3-hydroxypropionitrile (2-
cyanoethanol) in dry acetonitrile, using 1H-tetrazole as
activator, and the progress of the reactions was monitored by
31P NMR (see Figure 2). The phosphorus atom of the resulting
phosphite triesters is not stereogenic, thereby allowing relative
assignment of C3′ stereochemistry of the four phosphorami-
dites. After the phosphoramidites were stirred with 1H-
tetrazole for 60 min, a 31P NMR spectrum showed a single
peak at 125.1 ppm. The single peak observed is in agreement
with previous reports on activation of phosphoramidites,
probably as a result of epimerization of phosphorus via
exchange with excess tetrazole.37 The very long activation
time was needed to convert all starting material to the activated
tetrazolide as indicated by 31P NMR. While the need for shorter
reaction times has been reported for simultaneous activation
and coupling,38 i.e., mixing phosphoramidite, tetrazole, and
alcohol, the long reaction time needed in our investigation is
likely a consequence of the sequential approach used, leading to
slower kinetics in the activation step. Subsequently, excess
anhydrous 2-cyanoethanol was added, and the reaction mixture
was stirred until 31P NMR indicated full conversion of the
tetrazolide to the corresponding phosphite triesters. Both
binary mixtures, however, give rise to spectra having two peaks,
an intensive peak at 139.8 ppm and a less intensive peak at
140.8 ppm. Thus, both binary mixtures consist of two
diastereomers having alternate C3′ stereochemistry. Direct
incorporation of the two binary mixtures into oligonucleotides
would therefore yield an undesirable mixture of two
diastereomeric ONs (containing either monomer X or Y),
and as a result ON synthesis using this strategy was not
pursued.
Gratifyingly, H-phosphonates 5R/S were easily obtained by
treating hemiacetal 3 with PCl3/1,2,4-triazole followed by
hydrolysis, resulting in an epimeric mixture of H-phosphonates
1
5R/S in an approximate 1:1 ratio (according to H NMR) in
78% yield. Attempts to separate epimers 5R/S by silica gel
column chromatography were unsuccessful; however, separa-
tion could be realized by RP-HPLC on both small scale (∼50
μg) and a more practical scale (6−10 mg). A representative RP-
HPLC profile with retention times of 43.9 min (faster eluding
compound) and 45.8 min (slower eluding compound) is shown
in Figure 3. Subsequently, appropriate fractions were pooled,
Figure 3. Representative RP-HPLC profile for the separation of H-
phosphonate epimers 5R/S.
concentrated under reduced pressure, and then lyophilized to
afford the two separated epimers as white foams. This
procedure allowed isolation of enough material to realize full
characterization and individual incorporation of epimers 5R and
5S into ONs.
Assignment of 3′-Stereochemistry of H-Phospho-
nates 5. Several factors, including intramolecular steric and
electronic factors as well as solvation, contribute to the free
energy of the conformations of H-phosphonates 5R and 5S.
However, the dioxane ring would be expected to adopt a chair
conformation where two conformations of the dioxane ring are
possible for each of the epimers. This is in agreement with the
observed conformation of closely related (2R,6R)-6-hydrox-
H-Phosphonate Strategy. One of the drawbacks of the
phosphoramidite strategy is, as described above, the formation
of two stereogenic centers upon phosphitylation of hemiacetal
3, thereby adding complexity to the following separation step.
Phosphoramidite chemistry is usually the method of choice in
ON synthesis, due to commercial availability of standard
monomer building blocks and well-established automated solid-
phase synthetic protocols.39−42 However, phosphotriester and
H-phosphonate methodologies have also been successfully
3880
dx.doi.org/10.1021/jo300222q | J. Org. Chem. 2012, 77, 3878−3886

The Journal of Organic Chemistry
Article
Figure 4. Two possible conformations of H-phosphonate 5R and 5S, respectively, including the corresponding double-barreled Lewis structures, and
observed and estimated coupling constants.
a
Table 1. Thermal Denaturation Studies of Matched and Mismatched Duplexes
Tm [ΔTm/mod] (°C)
DNA: 3′-CAC TBT ACG
RNA: 3′-CAC UBU ACG
G
B =
A (ON4)
29.0
C (ON5)
G (ON6)
T (ON7)
A (ON8)
C (ON9)
<10
[<−17.0]
<10
[<−13.5]
<10
[<−10.5]
(ON10)
U (ON11)
5′-GTG ATA TGC (ON1)
5′-GTG AXA TGC (ON2)
5′-GTG AYA TGC (ON3)
13.0
[−16.0]
14.0
[−11.5]
<10
[<−8.0]
20.0
[−9.0]
20.5
[−5.0]
<10
[<−8.0]
14.0
[−15.0]
18.5 [−7.0]
27.0
24.0
[−3.0]
<10
[<−17.0]
13.5
[−10.0]
<10
[<−10.5]
25.5 [−3.5]
23.5 [−3.5]
20.5 [−6.5]
b
18.0 [−11.0]
17.0 [−1.0]
12.0
[−8.5]
a
Thermal denaturation temperatures of ON2 and ON3 toward DNA or RNA complements (ON4 and ON8, respectively) [Tm values/°C (ΔTm =
change in Tm value, for matched duplexes calculated relative to ON1:ON4 or ON1:ON8, for mismatched duplexes calculated relative to matched
duplex)] measured as the maximum of the first derivative of the melting curve (A260 vs temperature) recorded in medium salt buffer ([Na+] = 110
mM, [Cl−] = 100 mM, pH 7.0 (NaH2PO4/Na2HPO4)), using 1.0 μM concentrations of the two ONs. Tm values are averages of at least two
b
measurements. Broad transition.
ymethyl-2-(uracil-1-yl)-1,4-dioxane,13 where the 1,4-dioxane
ring adopts a chair conformation with uracil and hydroxymethyl
in equatorial positions, despite the potential stabilization of the
nucleobase adopting an axial conformation by the anomeric
effect. The size of the nucleobase and hydroxymethyl moieties
most probably lead to significant 1,3-diaxial interactions,
thereby destabilizing the diaxial conformation. These observa-
tions strongly suggest that the N1−C1′−O4′−C4′−C5′ part of
H-phosphonates 5R/S will induce a similar conformation with
the nucleobase and the 4′-substituent in equatorial positions,
whereas the anomeric effect of O3′ would be expected to
stabilize an axial position of the O3′-phosphonate functionality.
To determine the C3′-configuration of the separated H-
phosphonates, the coupling constants (J values) observed in 1H
NMR of the separated H-phosphonates were investigated and
ON Synthesis Using H-Phosphonates 5R and 5S. For
homo-DNA, data solely exist for fully modified ONs, whereas a
few HNA−DNA hybrids have been synthesized.19,22 Evaluation
of fully or multiply modified ONs of either of the dioxane
monomers introduced herein would be very interesting as they
could further improve the knowledge about six-membered
nucleotide mimics and elucidate if a continuous stretch of
dioxane monomers could lead to favorable backbone geometry.
However, such evaluation requires future development of
improved synthetic and separation procedures, and since a
preliminary evaluation of binding affinity can be accomplished
with only one incorporation, we decided not to pursue multiple
incorporations of monomers X and Y. Singly modified polyT
oligomers (e.g., dT6XdT6), as reported for HNA,19 could
possibly be synthesized, thereby allowing direct comparison,
but analysis of such duplexes is hampered by the possibility of
both parallel and antiparallel binding to a complement ON, and
we therefore chose to evaluate monomers X and Y as single
incorporations in a mixed sequence.
ON2 and ON3 containing a single central incorporation of
monomer X or Y, respectively, were synthesized by manual
solid support synthesis using the synthetic cycle outlined in the
Experimental Section. Coupling yields of H-phosphonates 5R
and 5S were 69% and 76%, respectively, according to the
absorbance measured after detritylation, whereas standard
nucleoside H-phosphonates coupled in >95% stepwise yields.
The synthesized oligonucleotides were purified by RP-HPLC,
́
compared to values predicted by the Dıez−Altona−Donders
equation49 as incorporated in the MestRe-J program50 (see
Figure 4). J1′,2′A, J1′,2′B, and J3′,4′ are all correlated to dioxane
conformation, and naturally, J3′,4′ is correlated with C3′-
configuration. Importantly, only the N1,C5′-equatorial/O3′-
axial conformation of H-phosphonate 5S is predicted to have
J1′,2′B and J3′,4′ matching the 10.2 and 1.3 Hz, respectively,
observed for the slower eluting epimer. As a consequence, the
faster eluting epimer must be H-phosphonate 5R and
gratifyingly the N1,O3′,C5′ equatorial conformation of this is
the only conformation predicted to have a J3′,4′ matching the 7.9
Hz observed for the fast eluting epimer.
3881
dx.doi.org/10.1021/jo300222q | J. Org. Chem. 2012, 77, 3878−3886

The Journal of Organic Chemistry
Article
and their constitution and purity (>80%) was verified by
MALDI-MS analysis and ion-exchange-HPLC, respectively.
ON2 (monomer X) and ON3 (monomer Y) were both
obtained in ∼15% overall yield as estimated from UV
measurements and loading of the synthesis support used. The
aqueous stability of the synthesized ONs was tested by a sample
stored at rt for two weeks. Both unmodified (ON1) and
modified ONs (ON2 and ON3) proved stable under these
conditions with no observable cleavage as evaluated by IC-
HPLC (see Figure S1, Supporting Information).
Thermal Denaturation and Circular Dichroism Experi-
ments. The influence on thermal stability of DNA:DNA and
DNA:RNA duplexes of a single incorporation of either
monomers X and Y was evaluated by UV−thermal duplex
denaturation studies (Table 1). Unless otherwise noted, the
UV−thermal denaturation curves displayed smooth sigmoidal
monophasic transitions similar to the unmodified reference
duplexes.
Incorporation of a single monomer X resulted in a slight
decrease in thermal affinity toward DNA and RNA comple-
ments when compared to unmodified DNA (ΔTm = −3.5 °C
against both DNA and RNA), whereas single incorporation of
monomer Y resulted in a dramatic decrease in thermal affinity
toward a DNA complement (ΔTm = −11.0 °C) and a
substantial but less dramatic decrease in thermal affinity toward
an RNA complement (ΔTm = −6.5 °C). To investigate if the
uracil moiety of monomers X and Y forms hydrogen bonding
with an opposing complementary adenosine monomer, thermal
denaturation studies were carried out toward DNA and RNA
strands with an opposing mismatched nucleotide. For
monomer X, mismatch discrimination was observed for all
mismatches both against DNA and RNA, except for U:rG,
where a very broad transition did not allow determination of
Tm. For monomer Y good to excellent mismatch discrimination
was observed, except for the U:T mismatch.
Although the thermal stabilities are significantly different, CD
curves of ON1, ON2, and ON3 in duplexes with comple-
mentary DNA (ON4) are nearly identical (see Figure 5a). All
three curves display distinctive B-type DNA signatures,
including a positive peak at 280 nm and a negative peak at
250 nm. For duplexes formed with complementary RNA
(ON8), the overall features of the curves are identical, even
though the negative peaks observed in the spectrum of
ON1:ON8 at 210 and 240 nm are less prominent for
ON2:ON8 and even positive (though still distinctive local
minima) for ON3:ON8 (see Figure 5a). Consequently, duplex
ON2:ON8 must be very similar to the A/B-duplex type of
DNA:RNA, whereas ON3:ON8 attains a slightly different
duplex type. However, the CD spectra do not hint at the origin
of the reduced thermal affinity of ON2 and ON3 toward
complements.
Figure 5. Circular dichroism spectra: (a) ON1:ON4 (green),
ON2:ON4 (blue), ON3:ON4 (orange); (b) ON1:ON8 (green),
ON2:ON8 (blue), ON3:ON8 (orange). The spectra were recorded at
5 °C otherwise using the same experimental conditions as for thermal
denaturation experiments.
solvation model54 in MacroModel V9.1.51 However, the
optimized parameters do not take modified nucleotides into
account, and the original amber94 parameters were therefore
used for monomers X and Y and the 3′-neighboring nucleotide.
During simulation, structures were collected at evenly
distributed intervals to allow conformational analysis during
the simulation trajectory.
The altered sugar moiety of dioxane nucleotides X and Y
precludes direct conformational comparison with unmodified
DNA; however, some qualitative considerations can be made
by comparison of the two modified monomers. Furthermore,
overall nucleotide parameters (e.g., intrastrand phosphate
distance (Pn to Pn+1)) can evaluate the fit of the modified
nucleotides in a duplex context. The duplex conformations
obtained from the modeling trajectory demonstrate that
monomer Y is present in an almost ideal chair conformation
(C3′CC1′), whereas monomer X needs to adopt a boat
conformation (BO2′,O4′) to allow accommodation in the duplex
structure (see Figure 6).
As might be expected, the chair conformation of monomer Y
explores a much smaller fraction of the conformational space
than the energetically less favored twist boat conformation of
monomer X. However, neither of the monomers displays
radical conformational changes (e.g., shifts from C3′CC1′ to
Molecular Modeling of Duplexes ON2:DNA and
ON3:DNA. To rationalize the observed decreased thermal
affinities of ON2 and ON3 toward their DNA complement, the
behavior of monomers X and Y in DNA:DNA duplexes was
evaluated by molecular modeling of duplexes ON1:ON4,
ON2:ON4, and ON3:ON4. As CD curves indicated that the
overall duplex geometry was not changed upon incorporation
of the modified monomers, a standard B-type DNA duplex was
modified in MacroModel v9.151 to give the duplexes of interest.
The model structures were subjected to a 5 ns stochastic
dynamics simulation, using the amber94 force field52 with the
improved parambsc0 parameters53 incorporated and GB/SA
Figure 6. Close up of monomers X (left) or Y (right) as incorporated
in energy-minimized duplexes, displaying BO2′,O4′ and C3′CC1′
conformations, respectively.
3882
dx.doi.org/10.1021/jo300222q | J. Org. Chem. 2012, 77, 3878−3886

The Journal of Organic Chemistry
Article
Figure 7. Truncated structures of energy-minimized duplexes with incorporation of monomers X (left) or Y (right) demonstrating a shift of the
nucleobase into the major groove. Overlay in gray is the unmodified DNA:DNA duplex.
C1′CC3′). Furthermore, both dioxane monomers display a longer
1. Crucially, separation of the epimeric mixture by HPLC
allowed full characterization of the individual H-phosphonates
and individual incorporation of monomers X and Y into
oligonucleotides. These are the first examples of the use of
dioxane as sugar moiety in oligonucleotides. Thermal
denaturation and circular dichroism measurements demon-
strate that deoxyoligonucleotides containing a single incorpo-
ration of either monomer are capable of duplex formation with
both DNA and RNA complements. Furthermore, Watson
Crick base pairing capability is retained, despite a decrease of
Tm of the formed duplex. Overall the results indicate that the
1,4-dioxane skeleton of monomers X and Y are nonideal
mimics of the natural furanose moiety of DNA and RNA.
and much narrower distribution of the Pn−Pn+1 distance (6.8 Å,
0.2 Å for both monomers) than found in the unmodified
DNA:DNA duplex (5.0 Å, 0.6 Å). For monomer X, the longer
P−P distance is accompanied by a shift of the 3′-phosphate
group into the minor groove indicating the need for major
alterations of the backbone to accommodate the dioxane
nucleotide. Furthermore, the nucleobase is shifted approx-
imately 1.1 Å into the major groove compared to the
unmodified reference duplex, thereby significantly decreasing
stacking with A4 (the 5′-neighbor, see Figure 7).
Almost the same shift of the nucleobase is seen for monomer
Y (∼0.9 Å), and rearrangement of the phosphate backbone is
also observed, indicating a major rearrangement of the
backbone to accommodate the dioxane moiety. As a result of
the rearrangement, the 3′-phosphate group is perturbed into the
major groove compared to the unmodified reference duplex.
Neighboring nucleotides display an increased S-type con-
formation for both the 5′- and the 3′-neighbor of monomer X
(N/S(A4) = 1%/96% and N/S(A6) = 2%/92%, compared to
N/S(A4) = 4%/74% and N/S(A6) = 4%/81% for the
unmodified duplex). On the contrary, no indication of
conformational tuning for monomer Y is found. Conforma-
tional tuning of neighboring nucleotides by LNA monomers is
well established55−57 and likely to contribute to the observed
increased thermal stability of LNA duplexes. Likewise, the
conformational tuning effect of monomer X may influence the
thermal stability of the duplexes, and is possibly a contributing
factor to the difference between the destabilizing effects of
monomers X and Y.
EXPERIMENTAL SECTION
■
(2R/S,3R,5R)-3-(4,4′-Dimethoxytrityloxymethyl)-2-hydroxy-
5-(uracil-1-yl)-1,4-dioxane (3R/S). IBX (o-iodoxybenzoic acid, 3.04
g, 10.9 mmol) was suspended in anhydrous EtOAc (150 mL) and
pyridine (1.10 mL, 13.6 mmol), and the resulting suspension was
heated under stirring to 80 °C. Alcohol 1 (3.54 g, 5.43 mmol) was
dissolved in anhydrous EtOAc (50 mL) and added to the heated
suspension. After being stirred at 80 °C for 5 h, the suspension was
cooled to 0 °C, and the precipitate was filtered off and washed with
cold EtOAc (2 × 25 mL). The combined organic phase was
evaporated to dryness under reduced pressure to afford a white foam,
tentatively assigned as aldehyde 2 [TLC (5% EtOH in CH2Cl2): Rf(1)
= 0.3, Rf(2) = 0.35; 1H NMR (DMSO-d6) δ 11.43 (d, JNH,5 = 2.2, 1H,
NH), 9.56 (s, 1H, CHO), 7.91−7.98 (m, 2H, H2Bz,H6Bz), 7.75 (d, J6,5
= 8.0, 1H, H6), 7.65−7.72 (m, 1H, H4Bz), 7.50−7.57 (m, 2H,
H3Bz,H5Bz), 7.15−7.34 (m, 9H, H2DMT, H6DMT, H2′DMT, H6′DMT
2″DMT, H3″DMT, H4″DMT, H5″DMT, H6″DMT), 6.83−6.88 (m, 4H,
H3DMT, H5DMT, H3′DMT, H5′DMT), 6.13 (ddABX,X, J1′,2′B = 6.1, J1′,2′A
,
=
Displacement of phosphate groups into the grooves could
potentially disrupt the structural water surrounding the duplex,
leading to a destabilization of the structure. Likewise is the shift
of the nucleobase a likely origin of destabilization by decreased
stacking interactions and larger exposed hydrophobic surface.
For monomer X, the adopted boat conformation is expected to
be energetically unfavorable, which most likely contributes
significantly to the destabilization of the corresponding duplex.
However, the molecular modeling structures do not provide an
obvious explanation for the difference in thermal affinity
induced by the two modified duplexes, but the combination of
unfavorable conformations and low flexibility would be
expected to decrease the possibility of accommodating the
modified monomers in the duplex structure, leading in general
to the observed low stability structures.
5.4, 1H, H1′), 5.61 (dd, J5,6 = 8.0, J5,NH = 2.2, 1H, H5), 4.74−4.82
(mABX,A, J2′A,2′B = 11.7, J2′A,1′ = 5.4, 1H, H2′A), 4.60−4.68 (mABX,B, J2′B,2′A
= 11.7, J2′B,1′ = 6.1, 1H, H2′B), 4.42 (dd, J4′,5′A = 5.4, J4′,5′B = 3.0, 1H,
H4′), 3.72 (s, 6H, OCH3), 3.22−3.34 (m, 2H, H5′), traces of EtOAc
was detected; 13C NMR (DMSO-d6) δ 200.3 (CHO), 164.9 (COPh),
162.9 (C4), 158.0 (C4DMT,C4′DMT), 151.0 (C2), 144.4 (C1″DMT),
140.2 (C6), 135.1 (C1DMT/C1′DMT), 135.0 (C1DMT/C1′DMT), 133.7
(C4Bz), 129.6 (C2DMT, C6DMT, C2′DMT, C6′DMT), 129.2 (C2Bz, C6Bz/
C3Bz, C5Bz), 129.0 (C1Bz), 128.8 (C2Bz, C6Bz/C3Bz, C5Bz), 127.8
(C2″DMT, C6″DMT/C3″DMT, C5″DMT), 127.6 (C2″DMT, C6″DMT/C3″DMT
,
C5″DMT), 126.7 (C4″DMT), 113.2 (C3DMT, C5DMT, C3′DMT, C5′DMT),
102.3 (C5), 85.6 (CAr3), 82.1 (C1′/C4′), 81.2 (C1′/C4′), 63.2 (C2′/
C5′), 62.1 (C2′/C5′), 54.9 (OCH3) ppm; ESI-HiRes m/z 673.2132
([M + Na]+, C37H34N2O9Na+ calcd 673.2157)]. The resulting crude
aldehyde 2 was dissolved in anhydrous MeOH (10 mL) and added to
a solution of NaOH (1.30 g, 32.5 mmol) in anhydrous MeOH (200
mL). The resulting mixture was stirred under Ar at 0 °C. After the
mixture was stirred for 70 min, satd aq NH4Cl (50 mL) was added,
and the resulting mixture was stirred for 10 min. Brine (50 mL) was
added, and extraction with CH2Cl2 (3 × 200 mL) was performed. The
combined organic phase was evaporated to dryness under reduced
CONCLUSION
■
H-Phosphonates 5R and 5S were synthesized as an epimeric
mixture in 53% yield over three steps from known secouridine
3883
dx.doi.org/10.1021/jo300222q | J. Org. Chem. 2012, 77, 3878−3886

The Journal of Organic Chemistry
Article
pressure, and the resulting residue was purified by silica gel column
chromatography (30−60% EtOAc in toluene) to afford an epimeric
mixture of hemiacetal 3 (2.03 g, 68% from 1) as a white solid: TLC
(10% MeOH in CH2Cl2) Rf = 0.3; H NMR (DMSO-d6) δ 11.48 (d,
JNH,5 = 2.0, 1H, ex, NHR/NHS), 11.46 (d, JNH,5 = 2.1, 1H, ex, NHR/
was purified by silica gel column chromatography (0−10% MeOH in
CH2Cl2, containing 1 vol % Et3N) to afford an epimeric mixture of the
desired H-phosphonates 5R/S as a white foam (310 mg, 60%).
HPLC Separation of H-Phosphonates 5R and 5S. RP-HPLC
separation of H-phosphonates 5R/S was performed on 6−10 mg scale
using a Waters Prep LC 4000 system equipped with an XBridge Prep
C18, 5 μm, ODB (⌀ 19 mm × 100 mm). The gradient protocol used
was an isocratic hold of 100% buffer A for 2 min, followed by a linear
gradient to 30% buffer B over 5 min, and a linear gradient to 47%
buffer B over 45 min at a flow rate of 8.0 mL/min (buffer A: 0.05 M
aqueous triethylammonium acetate, pH = 7.4; buffer B: 75%
acetonitrile in buffer A (v/v)). The protocol gave sufficient separation
(tR(5R) = 43.9 min, tR(5S) = 45.8 min), and appropriate fractions
were pooled, concentrated under reduced pressure, and then
lyophilized affording the two epimers as white foams.
1
NHS), 7.74 (d, J6,5 = 8.1, 1H, H6S), 7.64 (d, J6,5 = 8.1, 1H, H6R), 7.36−
7.45 (m, 4H, H2″DMT, H6″DMT), 7.18−7.33 (m, 14H, H4″DMT, H2DMT
,
H6DMT, H2′DMT, H6′DMT, H2″DMT, H6″DMT), 6.99 (d, JOH,3′ = 6.8, 1H,
ex, 3′-OHS), 6.84−6.90 (m, 8H, H3DMT, H5DMT, H3′DMT, H5′DMT),
6.69 (d, JOH,3′ = 6.5, 1H, ex, 3′-OHR), 5.68−5.78 (m, 4H, H1′S, H1′R,
H5S, H5R), 4.85 (dd, J3′,OH = 6.5, J3′,4′ = 1.9, 1H, H3′R), 4.77 (dd, J3′,4′
=
8.0, J3′,OH = 6.7, 1H, H3′S), 4.10−4.16 (m, 1H, H4′R), 3.65−3.93 (m,
16H, 3 × H2′S/H2′R, H4′S, OCH3,S, OCH3,R), 3.48−3.57 (m, 1H,
H2′S/H2′R), 3.05−3.19 (m, 3H, 2 × H5′S, H5′R), 2.96−3.03 (m, 1H,
H5′R) ppm;56 13C NMR (DMSO-d6) δ 163.0 (C4R/C4S), 162.9 (C4R/
C4S), 158.08 (C4DMT,R, C4′DMT,R/C4DMT,S, C4′DMT,S), 158.05
(C4DMT,R, C4′DMT,R/C4DMT,S, C4′DMT,S), 150.1 (C2R/C2S), 150.0
(2R,3R,5R)-3-(4,4′-Dimethoxytrityloxymethyl)-5-(uracil-1-yl)-
1
1,4-dioxan-2-yl hydrogen phosphonate triethylamine (5R): H
NMR (CDCl3) δ 7.62 (d with 6.02, 0.5H, PH, JH,P = 642), 7.61 (d, J6,5
= 8.2, 1H, H6), 7.42−7.45 (m, 2H, H2″DMT, H6″DMT), 7.29−7.33 (m,
(C2R/C2S), 144.84 (C1″DMT,R/C1″DMT,S), 144.76 (C1″DMT,R
/
C1″DMT,S), 140.5 (C6R/C6S), 140.4 (C6R/C6S), 135.6 (C1DMT,R
C1′DMT,R/C1DMT,S, C1′DMT,S), 135.4 (C1DMT,R, C1′DMT,R/C1DMT,S
C1′DMT,S), 129.8 (C2DMT, C6DMT, C2′DMT, C6′DMT), 127.8 (C2″DMT
C3″DMT, C5″DMT, C6″DMT), 126.7 (C4″DMT), 113.1 (C3DMT, C5DMT
,
,
,
,
4H, H2DMT, H6DMT, H2′DMT, H6′DMT), 7.23−7.29 (m, 2H, H3″DMT
,
,
H5″DMT), 7.18−7.22 (m, 1H, H4″DMT), 6.78−6.82 (m, 4H, H3DMT
H5DMT, H3′DMT, H5′DMT), 6.02 (d with 7.62, 0.5H, PH, JH,P = 642),
5.92 (dd, J1′,2′B = 6.9, J1′,2′A = 3.1, 1H, H1′), 5.74 (d, J5,6 = 8.2, 1H, H5),
C3′DMT, C5′DMT), 102.1 (C5S/C5R), 102.0 (C5S/C5R), 91.8 (C3′S),
87.1 (C3′R), 85.5 (CAr3,R/CAr3,S), 85.3 (CAr3,R/CAr3,S), 78.7 (C4′S),
77.9 (C1′S/C1′R), 77.50 (C4′R/C1′S/C1′R), 77.45 (C4′R/C1′S/C1′R),
64.8 (C2′S/C2′R), 63.0 (C5′S/C5′R), 62.9 (C5′S/C5′R), 58.6 (C2′S/
C2′R), 55.0 (OCH3) ppm; ESI-HiRes m/z 569.1898 ([M + Na]+,
C30H30N2O8Na+ calcd 569.1894).
5.49 (dd, J3′,P = 8.4, J3′,4′ = 7.9, 1H, H3′), 4.13 (dd, J2′A,2′B = 12.1, J2′A,1′
=
3.1, 1H, H2′A), 3.81−3.86 (m, 1H, H4′), 3.80 (s, 1H, HNEt3), 3.78 (s,
6H, OCH3), 3.64 (dd, J2′B,2′A = 12.1, J2′B,1′ = 6.9, 1H, H2′B), 3.40
(mABX,A, J5′A,5′B = 10.6, J5′A,4′ = 2.2, 1H, H5′A), 3.33 (mABX,B, J5′B,5′A
=
10.6, J5′B,4′ = 4.6, 1H, H5′B), 3.01 (q, JCH2,CH3 = 7.3, 6H, NCH2), 1.28
(t, JCH3,CH2 = 7.3, 9H, CH2CH3) ppm;59 13C NMR (CDCl3) δ 162.6
(C4), 158.5 (C4DMT, C4′DMT), 149.8 (C2), 144.6 (C1″DMT), 139.9
(2R/S,3R,5R,PR/S)-2-Cyanoethoxy(diisopropylamino)-
phosphinoxy-3-(4,4′-dimethoxytrityloxymethyl)-5-(uracil-1-
yl)-1,4-dioxane (4RR/RS/SR/SS). An epimeric mixture of hemiacetal
3 (510 mg, 0.933 mmol) was coevaporated with 1,2-dichloroethane (2
× 3 mL) and redissolved in anhydrous CH2Cl2 (5.0 mL), and
anhydrous DIPEA (1.0 mL) was added at rt, followed by dropwise
addition of 2-cyanoethyl N,N′-(diisopropyl)phosphoramidochloridite
(0.31 μL, 1.37 mmol). The reaction mixture was stirred at rt for 65
min, EtOH (2 mL) was added, and stirring was continued at rt for 20
min. The reaction mixture was poured into CH2Cl2 (20 mL) and
washed with satd aq NaHCO3 (10 mL) and brine (10 mL). The
combined aqueous phase was back-extracted with CH2Cl2 (20 mL),
and the combined organic phase was evaporated to dryness under
reduced pressure. The resulting residue was purified by silica gel
column chromatography (0−35% EtOAc in toluene) to afford the
desired phosphoramidites 4A−D (512 mg, 73%) as white foams in
three fractions containing phosphoramidites [4A and 4B, 228 mg,
33%], [4C and 4D, 130 mg, 17%], and [4A−D, 154 mg, 22%],
respectively. Phosphoramidites 4A and 4B (1:4 ratio, based on 31P
NMR): TLC (50% EtOAc in PE) Rf = 0.55; 31P NMR (CDCl3) δ
149.5 (PA), 150.1 (PB) ppm; ESI-HiRes m/z 769.3008 ([M + Na]+,
C39H47N4O9PNa+ calcd 769.2973). Phosphoramidites 4C and 4D (6:1
ratio, based on 31P NMR): TLC (50% EtOAc in PE) Rf = 0.45; 31P
NMR (CDCl3) δ 152.0 (PC), 152.4 (PD) ppm; ESI-HiRes m/z
769.2953 ([M + Na]+, C39H47N4O9PNa+ calcd 769.2973).
(2R/S,3R,5R)-3-(4,4′-Dimethoxytrityloxymethyl)-5-(uracil-1-
yl)-1,4-dioxan-2-yl Hydrogen Phosphonate Triethylamine (5R/
S). Anhydrous Et3N (2.55 mL, 18.3 mmol) and imidazole (1.11 g, 16.4
mmol) were dissolved in anhydrous CH3CN (25 mL), and the
resulting mixture was cooled to 0 °C. PCl3 (0.45 mL, 5.14 mmol) was
added dropwise to the solution resulting in formation of a white
precipitate. Hemiacetal 3 (399 mg, 0.729 mmol) was coevaporated
with CH3CN (5 mL) and redissolved in anhydrous CH3CN (5 mL),
and the resulting mixture was added dropwise to the above suspension
under stirring at 0 °C. The reaction mixture was stirred for 22 h while
the temperature was allowed to rise to rt, and then aq Et3NH·HCO3
(2 M, 10 mL) was added. After being stirred for an additional 20 min,
the reaction mixture was partitioned between CH2Cl2 (100 mL) and
water (50 mL), and the organic phase was washed with satd aq
NaHCO3 (100 mL). The combined water phase was back-extracted
with CH2Cl2 (100 mL), and the combined organic phase was
evaporated to dryness under reduced pressure. The resulting residue
(C6), 135.8 (C1DMT/C1′DMT), 135.7 (C1DMT/C1′DMT), 130.2 (C2DMT
C6DMT, C2′DMT, C6′DMT), 128.3 (C2″DMT, C6″DMT), 127.8 (C3″DMT
,
,
C5″DMT), 126.9 (C4″DMT), 113.2 (C3DMT, C5DMT/C3′DMT, C5′DMT),
113.1 (C3DMT, C5DMT/C3′DMT, C5′DMT), 102.4 (C5), 92.4 (C3′), 86.2
(CAr3), 78.6 (C1′), 77.2 (C4′), 64.7 (C2′), 62.1 (C5′), 55.2 (OCH3),
45.5 (NCH2), 8.6 (NCH2CH3) ppm; 31P NMR (CDCl3) δ 1.95 ppm;
ESI-HiRes m/z 734.2783 ([M + Na]+, C36H46N3O10PNa+ calcd
734.2813).
(2S,3R,5R)-3-(4,4′-Dimethoxytrityloxymethyl)-5-(uracil-1-yl)-
1
1,4-dioxan-2-yl hydrogen phosphonate triethylamine (5S): H
NMR (CDCl3) δ 7.63 (d with 6.04, 0.5H, PH, JH,P = 636), 7.57 (d, J6,5
= 8.2, 1H, H6), 7.40−7.44 (m, 2H, H2″DMT, H6″DMT), 7.28−7.32 (m,
4H, H2DMT, H6DMT, H2′DMT, H6′DMT), 7.23−7.28 (m, 2H, H3″DMT
,
,
H5″DMT), 7.17−7.21 (m, 1H, H4″DMT), 6.77−6.82 (m, 4H, H3DMT
H5DMT, H3′DMT, H5′DMT), 6.04 (d with 7.63, 0.5H, PH, JH,P = 636),
5.89 (ddABX,X, J1′,2′A = 10.2, J1′,2′B = 3.2, 1H, H1′), 5.74 (d, J5,6 = 8.2, 1H,
H5), 5.39 (dd, J3′,P = 8.2, J3′,4′ = 1.3, 1H, H3′), 4.06−4.13 (m, 1H, H4′),
3.93 (ddABX,A, J2′A,2′B = 11.2, J2′A,1′ = 10.2, 1H, H2′A), 3.80 (s, 1H,
HNEt3), 3.782 (s, 3H, OCH3), 3.779 (s, 3H, OCH3), 3.66 (ddABX,B
,
J2′B,2′A = 11.2, J2′B,1′ = 3.2, 1H, H2′B), 3.38 (ddABX,A, J5′A,5′B = 10.1, J5′A,4′
=
6.8, 1H, H5′A), 3.18 (ddABX,B, J5′B,5′A = 10.1, J5′B,4′ = 4.8, 1H, H5′B), 3.00
(q, JCH2,CH3 = 7.3, 6H, NCH2), 1.27 (t, JCH3,CH2 = 7.3, 9H, CH2CH3)
ppm;60 13C NMR (CDCl3) δ 162.5 (C4), 158.5 (C4DMT, C4′DMT),
149.6 (C2), 144.7 (C1″DMT), 140.2 (C6), 136.0 (C1DMT/C1′DMT),
135.7 (C1DMT/C1′DMT), 130.2 (C2DMT, C6DMT/C2′DMT, C6′DMT),
130.1 (C2DMT, C6DMT/C2′DMT, C6′DMT), 128.2 (C2″DMT, C6″DMT),
127.8 (C3″DMT, C5″DMT), 126.8 (C4″DMT), 113.2 (C3DMT, C5DMT
/
C3′DMT, C5′DMT), 113.1 (C3DMT, C5DMT/C3′DMT, C5′DMT), 102.5
(C5), 89.1 (C3′), 86.3 (CAr3), 78.3 (C1′), 77.2 (C4′), 62.7 (C5′), 60.3
(C2′), 55.2 (OCH3), 45.5 (NCH2), 8.6 (CH2CH3) ppm; 31P NMR
(CDCl3) δ 1.85 ppm; ESI-HiRes m/z 734.2786 ([M + Na]+,
C36H46N3O10PNa+ calcd 734.2813).
Protocol for ON Synthesis. ONs were synthesized on 0.2 μmol
or 1 μmol scale using succinyl-linked LCAA-CPG (long-chain
alkylamine controlled pore glass, pore size 500 Å) DNA C columns.
Coupling cycle: deblocking (3% trichloroacetic acid in dichloro-
methane, 1.0 mL over 1 min); wash A (CH3CN, 2.0 mL over 1 min);
neutralizer (CH3CN/pyridine, 50:50, v/v, 1.0 mL over 1 min);
coupling (premixing H-phosphonate (200 μL, 50 mM, in CH3CN/
pyridine, 50:50, v/v) and adamantane carbonyl chloride (200 μL, 200
3884
dx.doi.org/10.1021/jo300222q | J. Org. Chem. 2012, 77, 3878−3886

The Journal of Organic Chemistry
Article
mM, in CH3CN/pyridine, 95:5, v/v), then adding over 2 min); wash B
(CH3CN/pyridine, 75:25, v/v, 1.0 mL over 1 min); coupling (as
above); wash B (as above); wash A (as above). Oxidation after full-
length ON synthesis: I2, (78 mM in pyridine/water, 98:2, v/v, 20 mL
over 30 min); wash A (as above).
(14) Maeda, M.; Kajimoto, N.; Yamaizumi, Z.; Okamoto, Y.;
Nagahara, K.; Takayanagi, H. Tetrahedron Lett. 1997, 38, 6841−6844.
(15) Jones, A. S.; Markham, A. F.; Walker, R. T. J. Chem. Soc., Perkin
Trans. 1 1976, 1567−1570.
(16) Jones, A. S.; McClean, M. J.; Tanaka, H.; Walker, R. T.;
Balzarini, J.; De Clercq, E. Tetrahedron 1985, 41, 5965−5972.
(17) Eschenmoser, A. Science 1999, 284, 2118−2124.
(18) Sabatino, D.; Damha, M. J. J. Am. Chem. Soc. 2007, 129, 8259−
8270.
(19) Van Aerschot, A.; Verheggen, I.; Hendrix, C.; Herdewijn, P.
Angew. Chem., Int. Ed. 1995, 34, 1338−1339.
(20) Hossain, N.; Wroblowski, B.; Van Aerschot, A.; Rozenski, J.; De
Bruyn, A.; Herdewijn, P. J. Org. Chem. 1998, 63, 1574−1582.
(21) Allart, B.; Khan, K.; Rosemeyer, H.; Schepers, G.; Hendrix, C.;
Rothenbacher, K.; Seela, F.; Van Aerschot, A.; Herdewijn, P. Chem.
Eur. J. 1999, 5, 2424−2431.
(22) Vandermeeren, M.; Preveral, S.; Janssens, S.; Geysen, J.; Saison-
́
Behmoaras, E.; Van Aerschot, A.; Herdewijn, P. Biochem. Pharmacol.
2000, 59, 655−563.
(23) Maurinsh, Y.; Rosemeyer, H.; Esnouf, R.; Medvedovici, A.;
Wang, J.; Ceulemans, G.; Lescrinier, E.; Hendrix, C.; Busson, R.;
Sandra, P.; Seela, F.; Van Aerschot, A.; Herdewijn, P. Chem.Eur. J.
1999, 5, 2139−2150.
(24) Wang, J.; Herdewijn, P. J. Org. Chem. 1999, 64, 7820−7827.
(25) Wang, J.; Verbeure, B.; Luyten, I.; Lescrinier, E.; Froeyen, M.;
Hendrix, C.; Rosemeyer, H.; Seela, F.; Van Aerschot, A.; Herdewijn, P.
J. Am. Chem. Soc. 2000, 122, 8595−8602.
(26) Summerton, J. Biochim. Biophys. Acta, Gene Struct. Expression
1999, 1489, 141−158.
(27) Since the synthetic strategy initiates from a ribofuranose the
numeration throughout the manuscript follows standard nucleoside
nomenclature (i.e., the nucleobase is attached to C1′, etc.)
(28) Altona, C.; Sundaralingam, M. J. Am. Chem. Soc. 1972, 94,
8205−8212.
After deprotection and cleavage from solid support (32% aq NH3,
12 h, 55 °C), ONs were purified by RP-HPLC using a Waters 600
system equipped with an Xterra MS C18 (10 μm, 7.8 × 10 mm)
precolumn and an Xterra MS C18 (10 μm, 7.8 × 150 mm) column
using the representative gradient protocol depicted in Table S1
(Supporting Information), followed by detritylation (80% aq AcOH,
20 min, rt), precipitation (abs EtOH, −18 °C, 12−16 h) and washing
(abs EtOH (2 × 300 μL)). The composition of the synthesized ONs
was verified by MALDI-MS analysis (ON2 (5′-GTG AXA TGC): [M
+ H]+ m/z 2756 calcd 2755; ON3 (5′-GTG AYA TGC) [M + H]+ m/
z 2753 calcd 2755) recorded in positive ion mode on a PerSeptive
Voyager STR using 3-hydroxypicolinic acid as a matrix. The purity
(>80%) was verified by ion-exchange HPLC using an a LaChrom L-
7000 system equipped with a Dionex PA100 column (4 × 250 mm) at
80 °C and pH 8 using the representative protocol shown in Table S2
(Supporting Information).
ASSOCIATED CONTENT
* Supporting Information
■
S
General experimental section, RP-HPLC profiles of phosphor-
amidites 4A/B, RP-HPLC ON purification protocol (Table
S1), IE-HPLC ON analytical protocol (Table S2), IE-HPLC
profiles of ONs after 2 weeks, and copies of NMR spectra of
compounds 2−5. This material is available free of charge via the
Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: jwe@sdu.dk.
■
(29) Olson, W. K.; Sussman, J. L. J. Am. Chem. Soc. 1982, 104, 270−
278.
(30) Anet, F. A. L.; Sandstrom, J. Chem. Commun. 1971, 1558−1559.
(31) Jensen, F. R.; Neese, R. A. J. Am. Chem. Soc. 1971, 93, 6329−
6330.
(32) Chapman, D. M.; Hester, R. E. J. Phys. Chem. A 1997, 101,
3382−3387.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We greatly appreciate funding from The Danish National
Research Foundation. We thank Ribotask ApS for assistance
with HPLC purification and Dr. Michael Wamberg for
assistance during oligonucleotide synthesis.
(33) Durig, J. R.; Zheng, C.; El Defrawy, A. M.; Ward, R. M.;
Gounev, T. K.; Ravindranath, K.; Rajeswara Rao, N. J. Raman Spectrosc.
2009, 40, 197−204.
(34) Langkjær, N.; Pasternak, A.; Wengel, J. Bioorg. Med. Chem. 2009,
17, 5420−5425.
REFERENCES
■
(35) More, J. D.; Finney, N. S. Org. Lett. 2002, 4, 3001−3003.
(36) Rohr, K.; Vogel, S. ChemBioChem 2006, 7, 463−470.
(1) Leumann, C. J. Bioorg. Med. Chem. 2002, 10, 841−854.
(2) Herdewijn, P. Chem. Biodivers. 2010, 7, 1−59.
(3) Robin, A.; Julienne, K.; Meslin, J.-C.; Deniaud, D. Eur. J. Org.
Chem. 2006, 3, 634−643.
(4) Barral, K.; Halfon, P.; Pep
2002, 43, 81−84.
(5) Barral, K.; Courcambeck, J.; Pep
(37) Berner, S.; Muhlegger, K.; Seliger, H. Nucleic Acids Res. 1989,
̈
17, 853−864.
(38) (a) Coppola, C.; Simeone, L.; De Napoli, L.; Montesarchio, D.
Eur. J. Org. Chem. 2011, 6, 1155−1165. (b) Fujita, S.; Oka, N.; Wada,
T. J. Org. Chem. 2011, 76, 2648.
̀
e, G.; Camplo, M. Tetrahedron Lett.
̀
e, G.; Balzarini, J.; Neyts, J.; De
(39) Sinha, N. D.; Biernat, J.; Koster, H. Tetrahedron Lett. 1983, 24,
̈
Clercq, E.; Camplo, M. J. Med. Chem. 2005, 48, 450−456.
(6) Vyas, D. M.; Szarek, W. A. Carbohydr. Res. 1973, 30, 225−227.
(7) Szarek, W. A.; Vyas, D. M.; Achmatowicz, B. J. Heterocycl. Chem.
1975, 12, 123−127.
(8) Hronowski, L. J. J.; Szarek, W. A. J. Med. Chem. 1982, 25, 522−
526.
5843−5846.
(40) Sinha, N. D.; Biernat, J.; McManus, J.; Koster, H. Nucleic Acids
̈
Res. 1984, 12, 4539−4557.
(41) Beaucage, S. L.; Iyer, R. P. Tetrahedron 1992, 48, 2223−2311.
(42) Beaucage, S. L.; Caruthers, M. H. In Current Protocols in Nucleic
Acid Chemistry; Wiley: New York, 2000, p 3.3.1.
(9) Iwasaki, T.; Nishitani, T.; Horikawa, H.; Inoue, I. Tetrahedron
Lett. 1981, 22, 1029−1032.
(43) (a) Reese, C. B. Tetrahedron 1978, 34, 3143−3179. (b) Itakura,
K.; Bahl, C. P.; Katagiri, N.; Michniewicz, J.; Wightman, R. H.; Narang,
S. A. Can. J. Chem. 1973, 51, 3469−3471. (c) Katagiri, N.; Itakura, K.;
Narang, S. A. J. Am. Chem. Soc. 1975, 97, 7332−7337. (d) Garegg, P.
(10) Jones, A. S.; Walker, R. T.; Wyatt, P. G.; Balzarini, J.; De Clercq,
E. J. Chem. Res., Synop. 1985, 336−337.
(11) Prisbe, E. J. J. Med. Chem. 1986, 29, 2445−2450.
(12) Van Aerschot, A.; Janssen, G.; Herdewijn, P. Bull. Soc. Chim.
Belg. 1990, 99, 769−777.
J.; Lindh, I.; Regberg, T.; Stawinski, J.; Stromberg, R.; Henrichson, C.
̈
Tetrahedron Lett. 1986, 27, 4051−4054. (e) Garegg, P. J.; Lindh, I.;
Regberg, T.; Stawinski, J.; Stromberg, R.; Henrichson, C. Tetrahedron
(13) Das, A. K.; Datta, M.; Mazumdar, S. K.; Das, N.; Van Aerschot,
A.; Low, J. N.; Howie, R. A. J. Cryst. Spectrosc. Res. 1992, 22, 439−442.
̈
Lett. 1986, 27, 4055−4058. (f) Froehler, B. C.; Matteucci, M. D.
3885
dx.doi.org/10.1021/jo300222q | J. Org. Chem. 2012, 77, 3878−3886

The Journal of Organic Chemistry
Article
Tetrahedron Lett. 1986, 27, 469−472. (g) Froehler, B. C.; Ng, P. G.;
Matteucci, M. D. Nucleic Acids Res. 1986, 14, 5399−5407.
(44) Stawinski, J.; Kraszewski, A. Acc. Chem. Res. 2002, 35, 952−960.
(45) Berkin, A.; Coxon, B.; Pozsgay, V. Chem.Eur. J. 2002, 8,
4424−4433.
(46) Ruhela, D.; Vishwakarma, R. A. J. Org. Chem. 2003, 68, 4446−
4456.
(47) Slattegard, R.; Teodorovic, P.; Kinfe, H. H.; Ravenscroft, N.;
̈
Gammon, D. W.; Oscarson, S. Org. Biomol. Chem. 2005, 3, 3782−
3787.
(48) Donahue, M. G.; Johnston, J. N. Bioorg. Med. Chem. Lett. 2006,
16, 5602−5604.
(49) Díez, E.; San-Fabian
Mol. Phys. 1989, 68, 49−63.
(50) Navarro-Vazquez, A.; Cobas, J. C.; Sardina, F. J.; Casanueva, J.;
́
, J.; Guilleme, J.; Altona, C.; Donders, L. A.
́
Díez, E. J. Chem. Inf. Comput. Sci. 2004, 44, 1680−1685.
(51) MacroModel, version 9.8, Schrodinger, LLC, New York, 2010.
̈
(52) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K.
M.; Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.;
Kollman, P. A. J. Am. Chem. Soc. 1995, 117, 5179−5197.
́ ́
(53) Perez, A.; Marchan, I.; Svozil, D.; Sponer, J.; Cheatham, T. E.;
Laughton, C. A.; Orozco, M. Biophys. J. 2007, 92, 3817−3829.
(54) Still, W. C.; Tempczyk, A.; Hawley, R. C.; Hendrickson, T. J.
Am. Chem. Soc. 1990, 112, 6127−6129.
(55) Petersen, M.; Nielsen, C. B.; Nielsen, K. E.; Jensen, G. A.;
Bondensgaard, K.; Singh, S. K.; Rajwanshi, V. K.; Koshkin, A. A.; Dahl,
B. M.; Wengel, J.; Jacobsen, J. P. J. Mol. Recognit. 2000, 13, 44−53.
(56) Nielsen, C. B.; Singh, S. K.; Wengel, J.; Jacobsen, J. P. J. Biomol.
Struct. Dyn. 1999, 17, 175−191.
(57) Nielsen, K. E.; Singh, S. K.; Wengel, J.; Jacobsen, J. P.
Bioconjugate Chem. 2000, 11, 228−238.
(58) As integrals of the hemiacetals were in a 1:1 ratio, the integrals
of resolved signals (corresponding to one proton) have been set to 1H
and accordingly overlapping signals (corresponding to one proton)
have integrals of 2H.
(59) The HP signal is at 6.82 (d, JH,P = 642, 1H, HP) but is given
here as separate peaks because of the large coupling constant.
(60) The HP signal is at 6.84 (d, JH,P = 636, 1H, HP) but is given
here as separate peaks because of the large coupling constant.
3886
dx.doi.org/10.1021/jo300222q | J. Org. Chem. 2012, 77, 3878−3886