Synthesis and structural studies of S-type/N-type-locked/frozen nucleoside analogues and their incorporation in RNA-selective, nuclease resistant 2′-5′ linked oligonucleotides

Article abstract of DOI:10.1039/c2ob26762d

2′-endo locked or frozen (S-type)/3′-endo locked or frozen (N-type) nucleoside analogues were synthesized. Conformational analysis based on ³J_HH and NOE measurements is presented which is further confirmed by X-ray crystal structural

Full text of DOI:10.1039/c2ob26762d

Organic &
Biomolecular
Chemistry
View Article Online
View Journal | View Issue
PAPER
Synthesis and structural studies of S-type/N-type-
locked/frozen nucleoside analogues and their
incorporation in RNA-selective, nuclease resistant 2’–5’
linked oligonucleotides†
Cite this: Org. Biomol. Chem., 2013, 11,
746
Namrata Erande,a Anita D. Gunjal,a Moneesha Fernandes,a Rajesh Gonnadeb and
Vaijayanti A. Kumar*a
2’-endo locked or frozen (S-type)/3’-endo locked or frozen (N-type) nucleoside analogues were syn-
3
thesized. Conformational analysis based on JHH and NOE measurements is presented which is further
Received 4th May 2012,
confirmed by X-ray crystal structural studies. 2’–5’ isoDNA oligonucleotides (ON) were synthesized using
these modified nucleoside analogues and UV-Tm studies of the resultant 2’–5’ isoDNA : RNA duplexes
reflect the site- and sequence-dependent effects and confirm that the S-type sugar conformations were
preferred over the N-type sugar geometry in such duplexes.
Accepted 7th November 2012
DOI: 10.1039/c2ob26762d
www.rsc.org/obc
in a variety of analogues such as locked nucleic acids (LNA/
BNA),6 hexitol nucleic acid (HNA)7 and tricyclic DNA (tcDNA),8
Introduction
The antisense therapeutic approach discovered more than all of which show increased affinity towards complementary
three decades ago by Zamecnik and Stephenson1,2 continues RNA with very high fidelity and without compromising specific
to attract chemists towards the design and synthesis of modi- base-pairing properties. The 2′–5′ linked isoDNA/isoRNA are a
fied antisense oligonucleotides (ASOs) till today, now with a unique class of oligonucleotides that bind exclusively only to
renewed interest, especially with the potential applications in RNA targets and not to their DNA counterparts.9 This interest-
siRNA3 and miRNA4 therapeutic targeting.
ing aspect of 2′–5′ linked oligonucleotides has not been
Preorganization of the structure of an oligonucleotide exploited for their applications in antisense research probably
single strand so as to match the structure of the duplex form, because of the relatively poor stability of the isoDNA/isoRNA :
can lead to increased duplex stability due to entropic advan- RNA duplexes. Damha et al. synthesized 2′–5′ RNA9 (isoRNA)
tage. The concept of conformational restriction5 has come to which occurs naturally10 but does not encode genetic infor-
the forefront for application in nucleic acid chemistry as seen mation. This synthesized isoRNA exhibited self-pairing as well
as pairing with RNA,9 but did not complex with complemen-
tary DNA. The CD patterns of isoRNA : RNA duplexes were
found to be very similar to those of the natural 3′-5′DNA : RNA
aOrganic Chemistry Division, CSIR-National Chemical Laboratory, Homi Bhabha
Road, Pune 411008, India. E-mail: va.kumar@ncl.res.in; Fax: +91 20 2590 2629;
duplexes, indicating comparable overall structures i.e.,
Tel: +91 20 2590 2340
bCMC Division, CSIR-National Chemical Laboratory, Homi Bhabha Road, Pune
compact A-form duplex structure.11 Solution-phase NMR
studies indicated that the ribose sugar pucker in the self-
411008, India
pairing isoRNA duplexes12 was C2′-endo (S-type). From these
CD and NMR studies, it appeared that the isoRNA strand in
the isoRNA : RNA duplex may be adopting the structural fea-
tures of a natural DNA and thus the sugar residues would be
existing in the C2′-endo conformation. In molecular modeling
studies by Yathindra and Lalitha,13 a critical examination of
the stereochemistry of 2′–5′- and 3′–5′-linkages revealed an
inverse relationship between the nucleotide geometry and
involvement of either the 2′- or 3′-hydroxyl group of ribose
sugar in forming the phosphodiester linkage. A 2′–5′-linkage
with a C2′-endo sugar produces a compact nucleotide geometry
†Electronic supplementary information (ESI) available: 1H & 13C NMR spectra of
compounds 2–7, 10a & 10b; mass spectra of compounds 2–7, 10a & 10b;
31P NMR and mass spectra of compounds 8, 11a & 11b; HPLC chromatograms of
purified modified and unmodified 2′ DNA oligomers; MALDI-TOF mass spectra
of 2′ DNA oligomers; MALDI-TOF mass spectra of 2′ DNA-1, 2′ DNAxUF-1d, 2′
DNArUF-1d, 2′ DNA-UN-1s and 2′ DNA-US-1s with deletion of the 2′ terminal ade-
nosine-5′-phosphate; UV-melting plots of 2′ DNA-1 : RNA1/DNA1 and 2′ DNA-2 :
RNA2/DNA2; UV-melting plots of duplexes of oligomers containing US, UN, XUF
and rUF with RNA1/RNA2; representative UV-melting plot of ss2′ DNA-US-1s,
ssRNA1 and 2′ DNA-US-1s : RNA1; representative RP-HPLC chromatograms for
SVPDE study with 3′DNA-1, 2′ DNA-1 and 2′ DNA-3. CCDC 900136 and 900137.
For ESI and crystallographic data in CIF or other electronic format see DOI:
10.1039/c2ob26762d
746 | Org. Biomol. Chem., 2013, 11, 746–757
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
conformational locks that would render the nucleosides to
adopt either the S-type (Fig. 2, I and III) or N-type (Fig. 2, II
and IV) sugar conformations. This work was inspired by the
successes met especially with LNA and HNA and 2′-ribo/arabino-
fluoro-substituted DNA where highly stable DNA : RNA
duplexes were formed by applying appropriate conformational
constraints with entropic advantages. We reported earlier20
our preliminary results summarising the effects of incorpor-
ation of these monomer units (Fig. 2, I–IV) with constrained or
frozen S-type/N-type sugar conformations in the isoDNA strand
of isoDNA : RNA duplexes. In this paper, we further give the
detailed synthesis and characterization of the constrained and
frozen monomer building blocks, the effect of the constrained
unit in the sequence context, site of modifications, effect of
the neighbouring nucleoside using UV analysis and also the
stability of the 2′–5′ linked oligomers to exonuclease degra-
dation by SVPDE in comparison with the unmodified DNA
sequences. In addition, conformational analysis of modified
nucleosides is presented based on X-ray crystallographic analy-
sis in our laboratory for 3′-deoxy-3′-ribofluorouridine, 3′-deoxy-
3′-xylofluorouridine, 3′-O,4′-C-methylene ribonucleoside21d in
Fig. 1 Sugar conformations of 2’–5’ nucleoside in single-stranded isoDNA/
isoRNA and double-stranded isoDNA/isoRNA : RNA duplex.
3
conjunction with JHH couplings and two-dimensional NOE
experiments in solution.
Fig. 2 S-type and N-type locked and frozen nucleosides.
Results and discussion
Synthesis of monomers
whereas the 3′–5′-linkage with the same C2′-endo sugar gives
rise to an extended nucleotide geometry. Thus, to have a The synthesis of constrained nucleoside units (Fig. 2, II–IV)
compact A-type overall structure of isoRNA : RNA duplex, the and their protected phosphoramidite building blocks amen-
S-type or C2′-endo sugar pucker in the repeating nucleotides is able for 2′–5′ linked oligonucleotide synthesis was accom-
stereochemically favoured in the isoRNA strand of the duplex. plished starting from D-glucose. The S-type locked nucleoside
Similar structural requirements were reported for 2′–5′-linked (Fig. 2, I) phosphoramidite was synthesized according to litera-
isoDNA : RNA duplexes by both Breslow and Sheppard,14 and ture procedure21 using uridine as starting material. The syn-
Switzer et al.15 CD spectroscopy15a and modelling studies16 thesis of N-type (Fig. 2, II) building block is outlined in
predicted that the isoDNA strand in isoDNA : RNA duplexes Scheme 1. Compound 1 was synthesized from the glucose di-
would predominantly have S-type sugar conformations. On the acetonide by following the earlier reported multistep synthesis
contrary, an extended structure with N-type sugar confor- procedure.22 The free hydroxyl group in 1 (Scheme 1) was pro-
mations was reported when NMR studies were carried out tected as its Alloc derivative to give compound 2 in 99% yield
using single stranded isoRNA17a and 2′–5′-d(G4C4).17b Also, the (purity >90% by NMR). The acetonide group in 2 was removed,
crystal structure18 obtained for a single strand 2′–5′-RNA and the compound was converted to the diacetate 3 in 75%
showed a C3′-endo or N-type sugar pucker. Hence the sugar yield by treatment with AcOH and Ac2O in the presence of a
conformations (N-type) observed in an isoDNA/isoRNA single catalytic amount of H2SO4. The reaction of 3 with bis(tri-
strand are not the same as that predicted for the isoDNA/ methylsilyl)uracil (U·2TMS) under Vorbrüggen’s conditions
isoRNA strand in an isoRNA/isoDNA : RNA duplex (S-type). In a afforded only the β-anomer of uridine derivative 4 (93%). The
single-stranded 2′–5′-DNA oligomer, the gauche and anomeric Alloc protecting group in 4 could be selectively cleaved in pres-
effects of the 2′-hydroxyl group would dominate the preference ence of the 3′-acetate using Pd(0) to get 5 (70%). The free
for 3′-endo (N-type) sugar conformation in the absence of a secondary hydroxyl group in 5 was protected as its 4,4′-
3′-hydroxyl group. Therefore in the isoDNA : RNA duplex it was dimethoxytrityl (DMT) group using excess of DMT chloride
assumed that the RNA strand would impose its structure19 on and a catalytic amount of DMAP in dry pyridine to get 6 (85%).
the isoDNA strand causing the individual nucleosides in the Compound 6 was then subjected to ammonolysis to give free
isoDNA to adopt C2′-endo conformations so as to form a stable 2′-hydroxy compound
7 (93%). Phosphitylation at the
duplex with RNA, leading to a conformationally and topologi- 2′-hydroxy group with N,N-diisopropylamino-2-cyanoethylpho-
cally constrained isoDNA in the isoDNA : RNA duplex (Fig. 1).16 sphino-chloridite afforded the desired phosphoramidite build-
In view of the above discussion, we planned to study the ing block 8 (60%) as N-type locked monomer. All the
duplex stability of isoDNA : RNA duplexes by applying compounds were purified by column chromatography. All the
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 746–757 | 747

View Article Online
Paper
Organic & Biomolecular Chemistry
Table 1 Conformational analysis of the nucleosides using H1’–H2’ coupling
constants from 1H NMR spectra
H1′–H2′ coupling
J in Hz
Nucleoside derivative
%S
Reference
3′-Deoxyuridine
1.2
7.5
3.5
3
94
36
97
21
21
21c
24
S-Locked uridine (US, I)
N-Locked uridine (UN, II)
3′-Deoxy-3′-ribofluoro-uridine 7.7
(rUF, III)
3′-Deoxy-3′-xylofluoro-uridine 1.0
(XUF, IV)
0.3 23
Scheme 1 Synthesis of N-type locked monomer. Reagents and conditions: (i)
Dry CH2Cl2–dry pyridine, Alloc-Cl,
0 °C 1 h, 99%. (ii) AcOH–Ac2O–H2SO4,
10 : 1.0 : 0.1, RT, 12 h, 75%. (iii) Uracil, CH3CN, BSA, 70 °C, TMS-OTf, 0 °C, reflux,
3 h, 93%. (iv) CH2Cl2, PPh3, piperidine, Pd(dba)2, RT, 10 min, 70%. (v) DMTCl, dry
pyridine, DMAP, 24 h, 85%. (vi) MeOH–aq.NH3, RT, 1 h, 93%. (vii) dry CH2Cl2,
EtN(iPr)2, N,N-diisopropylamino-2-cyanoethylphosphino-chloridite, RT, 3 h, 60%.
Fig. 3 Percent NOE cross peak and possible conformers for all four monomers.
Scheme 2 Synthesis of frozen S-type and N-type fluoro monomers. Reagents
and conditions: (i) DMTCl, DMAP, dry pyridine, RT, 62%. (ii) Dry CH2Cl2, EtN(iPr)2,
N,N-diisopropylamino-2-cyanoethylphosphino-chloridite, RT (11a, 58%; 11b,
43%).
new compounds were appropriately characterised using 1H,
13C & 31P NMR and high resolution mass spectrometric analy-
sis. The synthesis of 3′-deoxy-3′-xylofluoro- and 3′-deoxy-3′-ribo-
fluoro-phosphoramidite building blocks is outlined in
Scheme 2. The 3′-deoxy-3′-xylofluorouridine 9a and 3′-deoxy-3′-
ribofluorouridine 9b were synthesized from D-glucose23 and
D-xylose24 respectively following reported procedures. Com-
pounds 9a and 9b were crystallized from CH2Cl2–CH3OH and
analysed by X-ray crystallography. The free primary hydroxyl
group in 9a/b was protected with DMT group using excess of
DMT chloride and a catalytic amount of DMAP in dry pyridine
to give 10a/b respectively with 62% yield. Phosphitylation of
the 2′-hydroxy group in 10a/b with N,N-diisopropylamino-2-cya-
noethylphosphino-chloridite afforded the desired phosphora-
midite building blocks 11a (58%) and 11b (43%). All
compounds were purified by column chromatography. All the
new compounds were appropriately characterised using 1H,
13C & 31P NMR and high resolution mass spectrometric
analysis.
Fig. 4 Two possible conformers for N-type locked monomer due to flexible
5-membered locked ring in II.
Conformational analysis of nucleoside analogues
We compared the sugar conformations of 3′-deoxyuridine, the
S-type/N-type locked monomer units (I and II) and also those
of the 3′-ribofluoro/xylofluoro nucleoside analogues (III, IV)
(Table 1, Fig. 3–5). The NOE cross peaks between H2′–H6
protons in compounds I–IV suggest the anti conformation of
uracil base. The anti conformations of the nucleobases were
also supported by χ angle obtained from the X-ray crystal struc-
ture data of III and IV.
3
The homonuclear JH1′–H2′ coupling constants are very sen-
sitive to the dihedral angle and therefore directly reflect the
sugar pucker in five-membered ribonucleosides. %S confor-
mation for each chosen unit was calculated from the H1′–H2′
NMR coupling constants as earlier reported.21 A smaller
748 | Org. Biomol. Chem., 2013, 11, 746–757
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
J value of 0–3 Hz is typical of N-type sugar whereas the J value the N-type geometry is preferred, the observed NOE (4%) is
between 6–9 Hz is indicative of S-type sugar pucker.16 Thus, relatively weaker (due to increased distance between the two
the 3′-deoxyuridine was shown to be in 97% N-type,21 i.e. C3′- interacting protons).25 Further, the preferred O5′–C5′–C4–C3′
endo sugar pucker, which is consistent with the O4′–C1′–C2′– geometry in solution would be +sc for the ribofluoro derivative
O2′ gauche effect as well as the anomeric effect which brings and is in equilibrium with −sc and t for the xylofluoro deriva-
the nucleobase in pseudoaxial orientation.19 The 3′-O-4′-C- tives based on the 3JH4′−H5′ and 3JH4′−H5′′26 (ESI,† page S-3).
methylene-linked uridine (US, I, Fig. 3 and 5) and the xylo-3′-O-
The conformational features based on coupling constants
5′-C-methylene- linked uridine (UN, II, Fig. 3 and 4) would lock and NOE measurements are confirmed by X-ray crystallo-
the monomer in S-type and N-type conformations, respectively. graphic data for compounds IV,23 reported earlier. We col-
The crystal structural data reported earlier21d for thyminyl lected this data for compound III for the first time in the
monomer of compound I clearly shows the structure locked in present study.‡ We also collected crystal structure data for
S-type sugar with a pseudorotation phase angle (P) of 136.2°, compound IV for comparison. The results from these studies
maximum out-of-plane pucker (νm) of 32.3° and the O5′–C5′– are depicted in Fig. 6 in the form of an ORTEP picture for com-
C4′–C3′dihedral angle γ in +sc range (Table 2). Our two-dimen- pound III. Three crystal forms were observed and Form I is
sional NOE NMR experiments revealed a strong NOE cross shown in Fig. 6. The superimposition of the three forms for
peak (I, 9.7%, Fig. 3) for the H2′ and nucleobase H6, thus con- the III is shown in ESI† (page S 48). The 3′-ribo-configured
firming the 2′-endo sugar pucker for compound I, in solution fluorosugar in III was found to be C2′-endo, S-type, as against
also.25 The limiting values for J4′5′ and J4′5′′ in the staggered the C3′-endo, N-type sugar conformation that was found for
O5′–C5′–C4–C3′ rotamers were calculated to be as follows:
Rotamer γ+ (+sc) J4′5′ = 2.4 Hz, J4′5′′ = 1.3 Hz, Rotamer γt (t) J4′5′
2.6 Hz, J4′5′′ = 10.5 Hz and Rotamer γ- (−sc) J4′5′ = 10.6 Hz, J4′5′′
=
=
‡X-ray intensity data measurements of compound III and IV were carried out on
a Bruker SMART APEX II CCD diffractometer with graphite-monochromatized
(MoKα = 0.71073Å) radiation at 100(2) K. The X-ray generator was operated at 50
kV and 30 mA. A preliminary set of cell constants and an orientation matrix were
calculated from 694 (for III) harvested from three sets of 90 frames for URF and
36 frames for UXF. Data were collected with ω scan width of 0.5° at different set-
3.8 Hz.26 In the absence of the H4′ proton in I, it is not poss-
ible to comment on the solution rotameric population indi-
cated by γ but the crystal data indicated that this falls in +sc
range when the base was thymine (Table 2).
In compound II, the flexible 5-membered ring allowed tings of φ and 2θ with a frame time of 10 and 5 s respectively keeping the
sample-to-detector distance fixed at 5.00 cm. The X-ray data collection was moni-
∼36% S-type character to the sugar ring, (conformational
tored by APEX2 program (Bruker 2006, APEX2, SAINT and SADABS. Bruker AXS
analysis based on the 3JH1′–H2′). Two forms of the sugar confor-
Inc., Madison, Wisconsin, USA.).
mations for II could be possible, in which the C5′-carbon is
fixed in R configuration as depicted in Fig. 4. The O4′–C4′–C3′–
Crystal data of III. C9H11FN2O5, M = 246.20, colorless plate, 0.59 × 0.17 ×
0.10 mm3, monoclinic, space group P21, a = 15.4056(4), b = 6.8529(2), c = 15.8005(4)
O3′ gauche effect operative in N-type xylo-sugar would stabilize Å, β = 116.0190(10)°, V = 1499.04(7) Å3, Z = 6, T = 100(2) K, 2θmax = 56.00°,
Dcalc (g cm−3) = 1.636, F(000) = 768, μ (mm−1) = 0.146, 24 993 reflections col-
the N-type sugar ring conformation shifting the equilibrium to
N-type. An additional proof for the sugar conformation in II
lected, 7180 unique reflections (Rint = 0.0295), 6731 observed (I > 2σ(I)) reflec-
tions, multi-scan absorption correction, Tmin = 0.919, Tmax = 0.986, 484 refined
being predominantly N-type, comes from a relatively less
parameters, S = 1.067, R1 = 0.0342, wR2 = 0.0861 (all data R = 0.0376, wR2
=
strong NOE cross peak (5%) between H2′ and H6.25 An
0.0882), maximum and minimum residual electron densities; Δρmax = 0.30,
additional positive NOE cross peak (1.8%) between H6 and H6′ Δρmin = −0.22 (e Å−3). The asymmetric unit contains three symmetry independent
molecules. Structure overlay of all the three symmetry independent molecules
is observed which would indeed not be observed if S-type con-
shows major conformational difference at the orientation of uracil moiety. The
formation was predominant in II (Fig. 3 and 4). The preferred
OH group at C2′ and CH2OH group at C4′ also showed noticeable difference.
O5′–C5′–C4–C3′ geometry would be then expected to be (γt)
Crystal data of IV C9H11FN2O5, M = 246.20, colorless plate, 0.58 × 0.30 ×
0.22 mm3, monoclinic, space group P21, a = 6.3007(3), b = 7.3372(3), c = 10.8989(4)
due to the fixed R configuration at C5′ in the preferred N-type
3
conformer of II. The observed JH4′–H5′ coupling (J = 4.1 Hz, Å, β = 92.277(2)°, V = 503.45(4) Å3, Z = 2, T = 100(2) K, 2θmax = 56.00°, Dcalc (g
cm−3) = 1.624, F(000) = 256, μ (mm−1) = 0.144, 10 593 reflections collected, 2440
ESI,† page S-3) would not be able to differentiate between the
two possible structures for compound II, in the absence of
H4′–H5′′ proton for which expected J would be ≈10.5 Hz in
North conformation, and ≈4.5 Hz for South conformation.
unique reflections (Rint = 0.0175), 2398 observed (I > 2σ(I)) reflections, multi-scan
absorption correction, Tmin = 0.921, Tmax = 0.969, 162 refined parameters, S =
1.054, R1 = 0.0239, wR2 = 0.0643 (all data R = 0.0242, wR2 = 0.0645), maximum and
minimum residual electron densities; Δρmax = 0.27, Δρmin = −0.17 (e Å−3). All the
The monomeric conformations of 3′-deoxy-3′-ribofluorouri- data were corrected for Lorentzian, polarization and absorption effects using
dine (rUF, III), and 3′-deoxy-3′-xylofluorouridine (XUF, IV) were
SAINT and SADABS programs (Bruker, 2006). SHELX-97 was used for structure
solution and full matrix least-squares refinement on F2 (G. M. Sheldrick, Acta
found to be almost frozen in S- and N-type conformations
Crystallogr., 2008, A64, 112).
3
respectively, based on the vicinal coupling JH1′–H2′. The con-
Hydroxyl H-atoms compounds III were located in difference Fourier map and
refined isotropically. Other H-atoms in both structures were placed in geometri-
formational preferences are consistent with the strong O4′–
C4′–C3′–F3′ gauche effect of fluoro-substituted nucleoside ana- cally idealized position and constrained to ride on their parent atoms.
NOESY experiments were performed on a Bruker AV 400 MHz NMR spec-
logues over the anomeric effect of C-1′ substitution. The proof
for the solution structure further comes from the observed
strong NOE cross peaks as shown in Fig. 3. The NOE cross
trometer, and data was analyzed using the Bruker TOPSpin 1.3 software. The
spectra were collected at 25 °C with 1 s mixing time. The total acquisition time of
each two-dimensional experiment was 2–4 hours and 8–16 scans were taken.
peak between the beta oriented H2′ and nucleobase H6 is
Percent nOe were calculated from the sub spectra of the row passing through
12.6% in III (the S-type sugar), whereas in compound IV where nosey, and the diagonal peak area has been assigned as 100.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 746–757 | 749

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 6 ORTEP Drawing for compound III.
Fig. 5 Preferred conformers for I, III, and IV.
Bioautomation DNA synthesizer and commercially available
5′-O-DMT-3′-deoxy-2′-phosphoramidites. The modified units
were incorporated at pre-determined positions using an
extended coupling time to yield the modified oligomers
efficiently. The site of the modified units in the sequences was
chosen so that these were either separated by 4–5 nucleosides
or were consecutive. Also, the 3′-end neighbour could be a
purine or pyrimidine. The synthesized oligomers (Table 3,
entry no. 1–18) were purified by HPLC subsequent to
ammonia treatment and characterized by MALDI-TOF–TOF
mass spectrometric analysis. Their purity was also re-checked
by analytical HPLC and found to be >95% prior to their use in
experiments. The thermal stability of the isoDNA duplexes
with complementary DNA/RNA was evaluated by UV-melting
studies. The melting temperatures, Tms, of the complexes
containing the S-type and N-type locked (US & UN) and frozen
the 3′-xylo-configured fluorosugar in compound IV (ESI,† page
S48), also reported earlier. Using the information available
from the X-ray studies, it can be deduced that the sugar pucker
pseudorotation phase angle (P) for form I crystals of com-
pound III is 171.2° and the maximum out of the plane pucker
(νmax) is 38.6°, that is characteristic of C2′-endo, S-type sugar
pucker (Table 2). The geometric parameters for compound IV
were found to be concurring with C3′-endo, N-type sugar
pucker (pseudorotation phase angle (P) for crystals of com-
pound IV is 0.88° and the maximum out of the plane pucker
(νmax) is 37.6°, that is characteristic of C3′-endo, N-type sugar
pucker). The preferred O5′–C5′–C4–C3′ geometry is consistent
(γ +sc range) with the NMR solution studies for compound III,
whereas in the case of compound IV, the O5′–C5′–C4–C3′ geo-
metry was found to be γt range in the crystal structure
(Table 2).
(rUF
&
XUF) units were determined and compared with the
unmodified duplexes formed using 2′ DNA-1 (Table 3, entry 1)
and 2′-DNA-2 (Table 3, entry 12). The results obtained for the
Synthesis of oligomers and UV-melting studies
The two sequences of the current study are of biological rel- isoDNA : RNA complexes are summarized in Table 3. Similar to
evance: DNA-1 is used for miRNA down-regulation27 and the isoRNA sequences reported by Damha et al.,9 the isoDNA
DNA-2 is used in the splice-correction assay developed by Kole sequences studied here (Table 3, entries 1–18) were also
et al.,28 The oligomers were synthesized using an automated found to bind only to complementary RNA and not to
Table 2 Geometrical parameters obtained from X-ray crystal structure for compounds I, III and IV
3′-Deoxy-3′-ribofluoro-uridine (rUF, III)
Mol A
Mol B
Mol C
3′-Deoxy-3′-xylofluoro-uridine (XUF, IV)
S-Locked thymine (TS)
Torsion angles (°)
ν0 = [C4′–O4′–C1′–C2′]
ν1 = [O4′–C1′–C2′–C3′]
ν2 = [ C1′–C2′–C3′–C4′]
ν3 = [ C2′–C3′–C4′–O4′]
ν4 = [ C3′–C4′–O4′–C1′]
χ = [O4′–C1′–N1–C2]
γ = [O5′–C5′–C4′–C3′]
Pseudorotation parameters (°)
Phase angle (P)
−18.12(18)
34.95(17)
−38.14 (16)
28.46(17)
−6.57(18)
−128.73(16)
50.5(2)
−7.84(17)
−35.54(16)
35.24(17)
−21.94(18)
2.00(18)
21.35(17)
−178.67(14)
49.9(2)
11.58(10)
−30.98(11)
37.52(10)
−31.59(10)
12.76(10)
−164.34(8)
−172.13(9)
−29.6
32.42
−23.5
7.12
14.22
−107.95
74.0
26.27(16)
−34.03(16)
30.32(17)
−14.32(17)
−149.70(14)
42.0(2)
171.2
38.60
185.7
34.16
126.8
36.62
0.88
37.59
136.31
32.50
Puckering amplitude(Vm)
750 | Org. Biomol. Chem., 2013, 11, 746–757
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Table 3 Oligonucleotide sequences, characterization using MALDI-TOF mass analysis and comparative UV-Tm values
Entry no.
Sequence code
Sequence (5′ → 2′)
Masscalcd./Massobsd.
Tm °C RNA
ΔTm (°C)
1
2
3
4
5
6
7
8
2′ DNA-1
CACCATTGTCACACTCCA
5364/5362
5377/5376
5391/5391
5377/5380
5391/5393
5391/5390
5368/5365
5372/5374
5368/5366
5372/5372
5372/5373
5369/5368
5383/5386
5397/5396
5383/5385
5397/5395
5373/5376
5377/5379
50.5
49.4
42.8
51.0
52.8
48.5
45.7
44.5
50.5
51.7
49.2
46.0
46.4
41.9
47.0
46.7
48.3
49.4
0.0
−1.1
−7.7
+0.5
+2.3
−2.0
−4.8
−6.0
0.0
+1.2
−1.3
0.0
+0.4
−4.1
+1.0
+0.7
+2.3
+3.4
2′ DNA-UN-1s
2′ DNA-UN-1d
2′ DNA-US-1s
2′ DNA-US-1d
2′ DNA-US-1d2
2′ DNA-XUF1s
2′ DNA-XUF-1d
2′ DNA-rUF-1s
2′ DNA-rUF-1d
2′ DNA-rUF-1d2
2′ DNA-2
CACCATTGTCACACUN
̲
CCA
̲ ̲ CCA
̲
̲
̲
CACCATTGUN
CACACUN
CACCATTGTCACACUS
̲ CCA
̲
CACCATTGUS
̲
CACACUS
̲
CCA
US
̲ ̲ GTCACACTCCA
̲
̲
CACCAUS
CACCATTGTCACACXUF
̲ CCA
̲
̲
̲
̲
̲
CACCATTGXUF
̲
CACACXUF
̲
CCA
̲ CCA
̲
̲
̲
̲
̲
̲
9
CACCATTGTCACACrUF
̲
̲
̲
10
11
12
13
14
15
16
17
18
CACCATTGrUF
̲
CACACrUF
̲
CCA
GTCACACTCCA
CCTCTTACCTCAGTTACA
CACCArUFr
̲ ̲
UF
̲
̲
̲
2′ DNA-UN-2s
2′ DNA-UN-2d
2′ DNA-US-2s
2′ DNA-US-2d
2′ DNA-rUF-2s
2′ DNA-rUF-2d
CCTCTTACCTCAGTUN
̲
ACA
̲
̲
̲
CCTCTTACCUN
̲ ̲ ACA
CAGTUN
CCTCTTACCTCAGTUS
̲
ACA
̲
CCTCTTACCUS
̲
CAGTUS
̲
ACA
̲
̲
CCTCTTACCTCAGTrUF
̲
ACA
̲
̲
CCTCTTACCrUF
̲
CAGTrUF
̲
ACA
̲
̲
̲
̲
Melting temperatures (Tms) were obtained from the maxima of the first derivatives of the melting curves (A260 nm versus temperature), measured
in buffer containing 10 mM sodium phosphate, 150 mM sodium chloride, pH 7.4, using 1 μM concentration of each of the two complementary
strands. Each experiment was repeated at least thrice and the values are accurate to 0.5 °C. Complementary RNA sequences for DNA-1 and
DNA-2 are RNA-1 and RNA-2 respectively. RNA1 = 5′-UGGAGUGUGACAAUGGUG-3′. RNA2 = 5′-UGUAACUGAGGUAAGAGG-3′. Tm of 2′
DNA-1 : cDNA-1 = 59.0 °C and 2′ DNA-2 : cDNA-2 = 54.0 °C. ΔTm = Tm − Tm(control).
complementary DNA (ESI,† page S-41–46). Sharp, well-defined the destabilizating effect caused by the N-type frozen units
sigmoid-shaped melting curves were observed for the com- (XUF). Thus, the complexes containing XUF units were destabi-
plexes with complementary RNA with a 10–20% hyperchromi- lized to a much greater extent (ΔTm = −4.8 & −6.0 °C respect-
city at 260 nm, while, for the complexes with complementary ively, for complexes containing one and two modified units
DNA, there was no sigmoid transition, indicating no complexa- respectively; Table 3, entries 7 & 8 respectively) in comparison
tion in the latter case. As reported in our earlier communi- to the stabilization observed for complexes containing rUF
cation,20 for complexes containing modified/unmodified 2′ units (ΔTm = 0.0 & +1.2 °C respectively, for complexes contain-
DNA-1, the S-type locked units were found to favour stable ing one and two modified units; Table 3, entries 9 & 10
complex formation with the complementary RNA, while the respectively). This could be due to the inability of the S-type
N-type locked units caused a destabilization of the resulting frozen conformation to further strengthen the stacking and
complexes with complementary RNA. This stabilizing/destabi- hydrogen-bonding interactions when the nucleobase is
lizing effect was found to increase with an increase in the oriented pseudoequatorially. The O4′–C4′–C3′–F3′ gauche effect
number of modified units. Thus, the inclusion of two modi- is probably responsible for the resistance to the N- to S-type
fied units (US or UN) caused a larger stabilization/destabiliza- conformational change that must take place in adopting the
tion of the resulting complexes (ΔTm = +2.3 & −7.7 °C duplex state, considering comparable steric interactions
respectively; Table 3, entries 5 & 3 respectively) compared to between the fluorine and hydrogen atoms.23,24,29 This domi-
the oligomers containing only a single modified unit (ΔTm
=
nating O4′–C4′–C3′–F3′ gauche effect is again responsible for
+0.5 & −1.1 °C respectively; Table 3, entries 4 & 2 respectively). the preferred sugar pucker in the UF/xUFderivative. To study
The stabilization caused by the S-type locked monomer (US) the additive stabilization effect of the conformational con-
and the destabilization caused by the N-type locked monomer straint, we modified consecutive sites in DNA-1 as in 2′
(UN) was in accordance with the prediction of S-type confor- DNA-US-1d2 and 2′ DNA-rUF-1d2 (entries 6, 11, Table 3). In
mation for the isoDNA strand in an isoDNA : RNA duplex.16 either case, the consecutive modified sites were not able to
However, the orientation of the nucleobase in the US unit is cause additive stabilization. In fact, the duplexes formed were
pseudoequatorial, which effects weaker stacking and hydro- marginally destabilized (ΔTm ≈ −1.5 °C). This result is also in
gen-bonding interactions18 in comparison to the instance contrast to the 3′-5′ LNA : RNA duplexes where consecutive
when the nucleobase is pseudoaxial (as in the UN unit and modified units show much better stabilization due to
LNA). This could be the reason why the stability imparted by increased base-stacking interactions when the nucleobases are
the US was not as large as that by LNA (ΔTm ∼ +4 °C/modifi- axially oriented.6
r
cation)6 in 3′-5′-LNA : RNA duplexes. A similar effect was
To further study the effect of modifications in the sequence
X
observed for the S- and N-type frozen units (rUF & UF respect- context and the site of modifications, we synthesized another
ively). In this case too, the S-type frozen units (rUF) were found sequence 2′ DNA-2 (entry 12, Table 3), and oligomers having
to effect a much more modest stabilization in comparison to one or two S-type locked (US), N-type locked (UN) and
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 746–757 | 751

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 7 Duplex CD spectra for complexes derived from I–IV.
Fig. 8 Stability assay of the ONs to degradation by SVPDE. #In the case of 2’
DNA oligomers, the 2’-end adenine nucleoside-5’-phosphate is cleaved (ESI,†
page S-39–40).
ribofluoro (rUF) monomeric units (2′ DNA-UN-2s, 2′ DNA-UN-
2d, 2′ DNA-US-2s, 2′ DNA-US-2d, 2′ DNA-rUF-2s, 2′ DNA-rUF-2d
(Table 3, entries 13–18 respectively) and compared the stability
of the duplexes in terms of Tm-studies with the unmodified
control duplex (2′ DNA-2 : RNA-2). The modified units were
spaced by four unmodified units. The results were qualitatively
similar to those observed for DNA-1. The N-type locked
monomer again caused destabilization of the resulting duplex
whereas the S-type locked/frozen units stabilized the duplex
with complementary RNA. In fact, the 3′-deoxy-3′-ribofluorouri-
dine-derived sequences stabilized the duplex much better
(ΔTm ∼ +2 °C/modification, entries 17, 18, Table 3). In the
sequence context, the difference is that one of the modified
units is flanked by a purine instead of pyrimidine at the 2′-
end. The result presented in this paper thus confirms that in
isoDNA : RNA duplexes, the DNA strand would prefer to
assume S-type geometry that is complemented by S-type
locked and S-type frozen nucleoside units and the stability
would be improved further, depending on the flanking base
sequence.
As an additional support for duplex formation in all the
modifications used in the present studies, CD spectroscopic
studies of complexes of modified oligomers with RNA1
(entries 2, 4, 7 and 9) were also carried out. In all these cases,
it was observed that the duplex CD spectra were similar for all
the complexes studied (Fig. 7) and resemble A-type DNA : RNA
duplex CD spectra.30 The additive CD spectrum of the two
strands (RNA1 + 2′ DNA-US-1s) was different from that for the
duplex (RNA: 2′ DNA-US-1s). The overall adopted structures
due to the incorporation of modified units (I–IV) were not
found to alter the overall structural features of the complexes
formed, although thermal stability was affected.
natural ssDNA was completely digested within 30 min. The
unmodified as well as modified 2′–5′ linked 3′-deoxy-oligo-
nucleotides were found to be considerably stable under these
conditions, although the 3′-deoxy-riboadenosine-5′phosphate
at the 2′-end of the oligomers was cleaved under these con-
ditions. SVPDE is known to cleave the tetrameric 2′–5′-adenine
sequences31 and in this particular study, the sequences used
happened to contain an adenine nucleoside at the 2′-end. The
HPLC peak arising after this cleavage (ESI,† page S-47) was
identified by MALDI-TOF mass spectrometry (ESI,† page
S-39–40) as the oligomer in which one adenosine-5′-phosphate
unit is deleted. To confirm this finding, we further synthesized
an unmodified 2′–5′ linked 3′-deoxy oligomer (ESI,† page S-47;
2′ DNA-3, 5′-GAAGGGCTTCTTCCTTAT-2′), that has a thymine
nucleoside at the 2′-end. This oligomer was found to be stable
towards nucleolytic cleavage. Thus, the 2′–5′ oligonucleotides
were found to be resistant to digestion by the 3′-exonuclease
except for the hydrolytic cleavage of the 2′-terminal adenosine-
5′-phosphate.
Experimental
General
All the reagents were purchased from Sigma-Aldrich and used
without further purification. SVPDE was purchased from
Sigma. DMF, pyridine were dried over KOH and 4 Å molecular
sieves. TLCs were run on pre-coated silica gel GF254 sheets
(Merck 5554). All reactions were monitored by TLC and usual
workup implies sequential washing of the organic extract with
water and brine followed by drying over anhydrous sodium
Stability of oligonucleotide to SVPDE
We examined the stability of unmodified and modified 2′-5′- sulfate and evaporation under vacuum. Column chromato-
linked oligonucleotides towards 3′-exonuclease [snake venom graphy was performed for purification of compounds on
phosphodiesterase (SVPDE)] degradation and compared them silica gel (60–120 mesh or 100–200 mesh, Merck). TLCs were
with natural DNA. After incubation of each ON solution at performed using dichloromethane–methanol or petroleum
37 °C in the presence of SVPDE, the percentage of intact ONs ether–ethyl acetate solvent systems for most compounds. Com-
was analyzed at several time-points by RP-HPLC (Fig. 8). The pounds were visualized with UV light and/or by spraying with
752 | Org. Biomol. Chem., 2013, 11, 746–757
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
30% perchloric acid–EtOH solution and heating. 1H (200 MHz) reaction mixture was stirred overnight at room temperature.
and 13C (50 MHz) NMR spectra were recorded on a Bruker ACF TLC indicated complete product formation. The reaction was
200 spectrometer fitted with an Aspect 3000 computer and quenched with ice and 5% aqueous NaHCO3, then extracted
31P NMR spectra were recorded on a 400 MHz Bruker ACF with dichloromethane, followed by water wash and drying over
instrument. All the chemical shifts (δ/ppm) are referred to sodium sulfate. After solvent removal the crude product was
1
internal TMS for H and chloroform-d/DMSO-d6 for 13C NMR. purified by silica gel column chromatography using petroleum
1H NMR data are reported in the order of chemical shift, mul- ether and ethyl acetate as eluants. The compound 3 was eluted
tiplicity (s, singlet; d, doublet; t, triplet; m, multiplet and/or in 30% ethyl acetate in petroleum ether. Yield: 1.32 g, 76.3%
multiple resonance), number of protons. Mass spectra were 1H NMR (CDCl3, 200 MHz): δ 2.07 (s, 3H, CH3), 2.09 (s, 3H,
recorded on a APQSTAR spectrometer, LC-MS on a Finnigan- CH3), 3.84–3.92 (m, 1H, H6), 4.05–4.11 (m, 1H, H6′), 4.64–4.67
Matt instrument. DNA oligomers were synthesized on CPG (m, 2H, H3, H4), 4.75–4.80 (m, 1H, H2), 4.95–5.0 (m, 2H, allyl-
solid support using Bioautomation Mer-Made 4 synthesizer. H), 5.26–5.41 (m, 2H, allyl-H), 5.84–6.23 (m, 1H, allyl-H), 6.52
The RNA oligonucleotides were obtained commercially (Sigma- (d, 1H, H1, J = 4.2 Hz). 13C NMR (CDCl3, 50 MHz): δ 20.3,
Aldrich). RP-HPLC was carried out on a C18 column using 20.75, 68.4, 68.5, 74.9, 80.3, 83.0, 85.2, 100.2, 118.7, 131.0,
either a Varian system (Analytical semi-preparative system con- 153.8, 169.0. IR (CHCl3): νmax (cm−1) 3024, 1749, 1650,
sisting of Varian Prostar 210 binary solvent delivery system and 1428.16, 1371.84 1266.85 1114.45, 1061.37, 1024.79, 957.51,
Dynamax UV-D2 variable wavelength detector and Star chrom- 879.82, 755.49, 667.5, 599.59. HRMS(ESI) calcd for
atography software) or a Waters system (Waters Delta 600 e C14H18O9Na (M + Na+) 353.0843, found: 353.080.
quaternary solvent delivery system and 2998 photodiode array
2′-O-Acetyl-3′,6′-anhydro-5′-O-allyloxy-uridine (4)
detector and Empower2 chromatography software). MALDI-
TOF spectra were recorded on a Voyager-De-STR (Applied Bio- Compound 3 (1.11 g, 3.36 mmol), obtained from the previous
systems) MALDI-TOF instrument or a AB Sciex TOF/TOFTM step was dissolved in anhydrous acetonitrile (35 ml). The reac-
Series ExplorerTM 72 085 instrument and the matrix used for tion flask was flushed with nitrogen and uracil (0.45 g,
analysis was THAP (2′,4′,6′-trihydroxyacetophenone). UV experi- 4.03 mmol) was added. N,O-Bis(trimethylsilyl)acetamide (BSA)
ments were performed on a Varian Cary 300 UV-VIS spectro- (0.27 ml, 1.1 mmol) was added to the reaction flask under
photometer fitted with
a Peltier-controlled temperature nitrogen atmosphere. Then the reaction mixture was refluxed
programmer. CD spectra were recorded on a Jasco J-715 Spec- at 70 °C for one hour, followed by cooling in an ice bath.
tropolarimeter, with a ThermoHaake K20 programmable water TMSOTf (0.23 ml, 1.27 mmol) was added slowly with a syringe
circulator for temperature control of the sample.
and the reaction mixture was refluxed for three hours. TLC
showed disappearance of starting material and appearance of
a lower moving UV-positive spot which charred on acid spray-
5-O-Allyloxy-3,6-anhydro-1,2-isopropylidene-glucofuranose (2)
3,6-Anhydro-1,2-isopropylidene-5-hydroxy-glucofuranose (0.46 g, ing and heating. The reaction mixture was cooled to room
2.28 mmol) was dissolved in dry dichloromethane (9 ml). temperature, diluted with dichloromethane, washed with
Anhydrous pyridine (0.38 ml) was added and the reaction NaHCO3 and water, dried over sodium sulfate followed by
mixture was cooled to 0 °C in an ice-bath. Allyloxycarbonyl solvent removal. The crude product was purified by silica gel
chloride (0.3 ml, 2.91 mmol) was added dropwise and then the column chromatography using dichloromethane and metha-
reaction was stirred at room temperature for two hours, when nol as eluants. Compound 4 eluted in 2.5% methanol in
TLC showed absence of starting compound. The reaction dichloromethane. Yield: 1.19 g, 92.96%. 1H NMR (CDCl3,
mixture was extracted with dichloromethane, followed by water 200 MHz): δ 2.13 (s, 3H, CH3), 4.09–4.12 (m, 2H, H6′, H6′′),
wash and drying over sodium sulfate. Removal of solvent 4.54–4.57 (m, 1H, H4′), 4.62–4.66 (m, 2H, H3′, H2′), 4.93–4.98
yielded a sticky gum of compound 2, which was used with- (m, 1H, H5′), 5.15–5.41 (bm, 4H, allyl-H), 5.78–5.83 (dd, 1H, J =
out any further purification. Yield: 0.650 g, >90% by NMR. 10.53, 2.14 Hz, H5) 5.84–6.01 (m, 1H, allyl-H), 6.24 (d, 1H, H1′,
1H NMR (CDCl3, 200 MHz): δ 1.34 (s, 3H, CH3), 1.49 (s, 3H, J = 3.88 Hz), 7.62 (d, 1H, J = 8.21 Hz, H6), 8.84 (bs, 1H, NH).
CH3), 3.78–3.86 (m, 1H, H6), 4.00–4.08 (m, 1H, H6′), 4.54–4.56 13C NMR (CDCl3, 50 MHz): δ 20.4, 69.0, 70.9, 76.0, 79.2, 81.2,
(d, 1H, J = , H5), 4.61–4.67 (m, 3H, H2, H3, H4), 4.99–5.09 85.5, 90.5, 103.4, 119.4, 130.8, 139.7, 150.3, 153.9, 163.1, 169.4.
(m, 2H, allyl-H), 5.25–5.43 (m, 2H, allyl-H), 5.84–6.03 (bm, 1H, IR (CHCl3): νmax (cm−1) 3391.85, 3020.46, 2924.68, 2853.76,
allyl-H) 6.00 (d, 1H, H1, J = 3.55 Hz). 13C NMR (CDCl3, 1751.23, 1694.89, 1458.89, 1381.57, 1265.47, 1216.15, 1090.08,
50 MHz): δ 26.7, 27.3, 66.7, 68.8 76.0, 80.7, 84.8, 85.2 107.1, 1060.10, 758.86, 668.59. HRMS(ESI) calcd for C16H19O9N2
112.7, 119.1, 131.1, 154.1. HRMS(ESI) calcd for C13H19O7 (M + (M + H+) 383.1085, found: 383.1086.
H+) 287.1125, found: 287.1123.
5′-Hydroxy-2′-O-acetyl-3′,6′-anhydro-uridine (5)
1,2-Di-O-acetyl-3,6-anhydro-5-O-allyloxy-α,β-glucofuranose (3)
Compound 4 (1.19 g, 3.14 mmol) obtained in the previous
Compound 2 was desiccated (1.50 g, 0.52 mmol) and dissolved step, was dissolved in dichloromethane (45 ml). PPh3 (0.54 g,
in acetic acid (16 ml). Acetic anhydride (1.6 ml) was added, 2.06 mmol) was added, followed by piperidine (2.0 ml,
after cooling the reaction flask to 10 °C, followed by dropwise 0.02 mmol) and tris(dibenzylideneacetone)dipalladium
and slow addition of concentrated sulfuric acid (0.16 ml). The [Pd2(dba)3] (0.15 g, 0.16 mmol). The reaction mixture was
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 746–757 | 753

View Article Online
Paper
Organic & Biomolecular Chemistry
stirred for 10 minutes. TLC showed absence of starting com- 135.8, 136.0, 136.0, 140.6, 144.9, 149.6, 150.8 158.8, 164.2.
pound. Solvent was removed and the crude product was given HRMS(ESI) calcd for C31H30O8N2Na (M + Na+) 581.1894,
a wash with solvent ether. Column purification done on a found: 581.1896.
silica gel column yielded the pure compound 5, which eluted
5′-O-Dimethoxytrityl-3′-O,5′-C-methylene(uridine)
xylonucleoside-2′-O-phosphoramidite (8)
in methanol (3.5%) in dichloromethane. Yield: 0.65 g, 70%.
1H NMR (CDCl3, 500 MHz): δ 2.13 (s, 3H, CH3), 2.75 (d, 1H,
J = 7.98 Hz, OH), 3.85–3.88 (m, 1H, H6′′), 4.04–4.08 (m, 1H, Compound 7 (0.20 g, 0.36 mmol) was co-evaporated with dry
H6′), 4.51–4.52 (m, 2H, H4′, H5′), 4.71–4.72 (m, 1H, H3′), dichloromethane, and then dissolved in dry dichloromethane
5.28–5.29 (m, 1H, H2′) 5.82 (d, 1H, H5, J = 8.19 Hz), 6.23 (d, (3.0 ml). Diisopropylethylamine (DIPEA) (0.19 ml, 1.07 mmol)
1H, H1′, J = 3.6 Hz), 7.59 (d, 1H, H6, J = 8.32 Hz), 8.45 (s, 1H, was added, followed by chloro(2-cyanoethoxy)-N,N-diisopropyl-
NH). 13C NMR (CDCl3, 50 MHz): δ 20.5, 72.3, 73.8, 79.8, 82.7, amino)-phosphine (0.16 ml, 0.72 mmol) at 0 °C. The reaction
85.5, 91.3, 103.8, 140.4, 150.0, 162.4, 169.5. HRMS(ESI) calcd mixture was stirred under argon atmosphere at room tempera-
for C12H15O7 N2 (M + H+) 299.0874, found: 299.0873.
ture for 3 hours, when TLC indicated the absence of starting
material. The reaction mass was diluted with dichloro-
methane, washed with NaHCO3 and water, dried over sodium
5′-O-Dimethoxytrityl-2′-O-acetyl-3′,6′-anhydro-uridine (6)
The substrate 5 (0.43 g, 1.44 mmol) was co-evaporated with sulfate, followed by solvent removal. The crude product was
anhydrous pyridine twice, and then dissolved in anhydrous purified by silica gel column chromatography using
a
pyridine (10 ml). 4,4′-Dimethoxytrityl chloride (1.46 g, 1 : 1 mixture of dichloromethane : ethyl acetate and 1% tri-
4.32 mmol) was added in one lot. The reaction mixture was ethylamine. Yield: 0.16 g, 58.9%.31P NMR (CDCl3, 50 MHz):
stirred overnight at room temperature, when TLC showed a δ 150.7, 152.0. HRMS(ESI) calcd for C40H49O9N4(M + H+)
faster moving trityl-positive spot which charred on acid spray- 759.3153 found: 759.3155, C40H48O9N4Na (M + Na+) 781.2973
ing and heating. The reaction was quenched with methanol, found: 781.2974, and C40H49O9N4K (M + K+) 797.2712 found:
extracted with dichloromethane, washed with NaHCO3 and 797.2712.
water, dried over sodium sulfate, followed by solvent removal.
The crude product was purified by silica gel column chromato-
graphy using dichloromethane, methanol and pyridine (0.5%)
5′-O-Dimethoxytrityl-3′-deoxy-3′-fluoro-xylofuranosyluridine
(10a)23
as eluants. Compound 6 is eluted in methanol (1.5%) in A mixture of compound 9a (0.50 g, 2.03 mmol), 4,4′-dimethoxy-
dichloromethane. Yield: 0.73 g, 84.3%. 1H NMR (CDCl3, trityl chloride (0.67 g, 2.03 mmol) and 4-dimethylaminopyri-
200 MHz): δ 2.09 (s, 3H, CH3), 3.22–3.44 (m, 2H, H6′, H6′′), dine (0.03 g, cat.) were dissolved in pyridine (5.0 ml). The
3.80 (s, 6H, OCH3), 4.13–4.30 (m, 3H, H4′, H3′, H2′), 5.14–5.16 reaction mixture was stirred at room temperature for 2 h. The
(m, 1H, H5′), 5.79 (d, 1H, H5, J = 8.10 Hz), 6.05 (d, 1H, H1′, J = pyridine was removed under vacuum. The residue was dis-
3.4 Hz), 6.82–6.87 (m, 4H, DMT), 7.31–7.5 (m, 9H, DMT), 7.81 solved in ethyl acetate (100 ml), washed with saturated
(d, 1H, H6, J = 8.17 Hz), 8.65 (s, 1H, NH). 13C NMR (CDCl3, NaHCO3 (2 × 50 ml) and saturated aqueous NaCl (2 × 30 mL).
50 MHz): δ 20.6, 55.3, 70.9, 74.3, 80.5, 82.5, 84.7, 87.6, 90.5, The ethyl acetate layer was dried over Na2SO4, filtered and
103.0, 113.4, 172.2, 172.9, 128.1, 130.0, 135.8, 136.0, 139.8, evaporated to dryness. The crude was purified by silica gel
144.8, 149.7, 150.3, 158.8, 163.1, 169.4. HRMS(ESI) calcd for (neutralized with Et3N) column chromatography using ethyl
C33H32O9N2Na (M + Na+) 623.2000, found: 623.2000.
acetate : petroleum ether (9 : 1) to offer the title compound as a
1
white solid. Yield: 0.70 g, 62.8%. H NMR (CDCl3, 200 MHz):
5′-O-Dimethoxytrityl-2′-hydroxy-3′,6′-anhydro-uridine (7)
1H NMR (CDCl3): δ: 3.53 (m, 2H), 3.80 (s, 6H), 4.40 (d, 1H),
The substrate 6 (0.70 g, 1.16 mmol) was dissolved in AR grade 4.51–4.73 (m, 1H), 4.86–5.10 (dd, 1H, J3′,F = 50.7 and J3′,4′
=
methanol (50 ml). Aqueous ammonia (15 ml) was added and 3.1 Hz), 5.63 (d, 1H, J65 = 8.2 Hz), 6.87 (m, 4H) 5.81(s, 1H),
the pinkish slightly turbid reaction mixture was stirred for one 7.26–7.35 (m, 9H), 7.46 (d, 1H, J5,6 = 8.2 Hz). 13C NMR(CDCl3,
hour at room temperature. TLC showed the absence of starting 50 MHz): δ 55.1, 60.3, 78.4, 82.8, 86.5, 93.0, 96.4, 101.7, 113.1,
compound. The solvents were removed to get a yellowish solid. 126.9, 127.8, 128.0, 130.0, 135.5, 139.8, 144.3, 150.9, 158.5,
The solid was redissolved in dichloromethane and given a 164.1. HRMS(ESI) calcd for C30H29O7N2FNa (M + Na+) 571.1851
water wash. The organic layer was dried over sodium sulfate found: 571.1852.
and concentrated to get a pale yellow solid foam. Purification
was done by silica gel column chromatography using dichloro-
methane, methanol and pyridine (0.5%) as eluants. Com-
5′-O-Dimethoxytrityl-3′-deoxy-3′-fluoro-ribofuranosyluridine
(10b)
pound 7 eluted in methanol (2.5%) in dichloromethane.
A mixture of 3′-deoxy-3′-fluoro-5-hydroxy-ribofuranosyluridine
Yield: 0.62 g, 93.2%. 1H NMR (CDCl3, 200 MHz): δ 3.18–3.34 9b (0.50 g, 2.03 mmol), 4,4′-dimethoxytrityl chloride (0.67 g,
(m, 2H), 3.79 (s, 6H), 4.16–4.17 (d, 1H, J = 3.52 Hz), 4.33–4.44 2.03 mmol) and a catalytic amount of 4-dimethylaminopyri-
(m, 2H), 5.71 (d, 1H, J = 8.1 Hz) 5.77 (s, 1H), 6.83–6.86 (m, 4H), dine were dissolved in pyridine (5 ml). The reaction mixture
7.27–7.51 (m, 9H) 8.30 (d, 1H, J = 8.1 Hz), 10.16 (bs, 1H). 13C was stirred at room temperature for 2 h. Pyridine was removed
NMR (CDCl3, 50 MHz): 55.2, 72.2, 75.1, 80.1, 85.3, 86.7, 88.0, under vacuum. The residue was dissolved in ethyl acetate
96.0, 101.3, 113.3–113.4, 123.8, 127.1, 127.8, 128.1, 129.9, (100 ml), washed with saturated NaHCO3 (2 × 50 ml) and
754 | Org. Biomol. Chem., 2013, 11, 746–757
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
saturated aqueous NaCl (2 × 30 ml). The ethyl acetate layer was was performed. After synthesis, the sequences were cleaved
dried over Na2SO4, filtered and evaporated to dryness. The from the support by ammonia treatment to cleave the ester
crude product was purified by silica gel (neutralized with Et3N) functionality that joins support to the 2′-terminus of the oligo-
column chromatography using EtOAc–petroleum ether (9 : 1) mers and deprotects the exocyclic amino protecting groups
1
to get compound 10b as white foam. Yield: 0.70 g, 62.8%. H used during the synthesis. RP-HPLC was carried out on a C18
NMR (CDCl3, 200 MHz): δ 3.39–3.54 (m, 2H). 3.78 (s, 6H), column using either a Varian system (analytical semi-prepara-
4.36–4.45 (m, 2H), 4.94–5.23 (dd, 1H, J3′,F = 54.3 and J3′,4′
=
tive system consisting of Varian Prostar 210 binary solvent
2.0 Hz), 5.45 (d, 1H, J6,5 = 8.0 Hz), 6.15 (d, 1H, J1,2 = 6.5 Hz), delivery system and Dynamax UV-D2 variable wavelength detec-
6.84 (m, 4H), 7.22–7.33 (m, 9H),7.71 (d, 1H, J5,6 = 8.0 Hz). 13C tor and Star chromatography software) or a Waters system
NMR (CDCl3, 50 MHz):δ 55.2, 62.7, 74.8, 82.2, 87.4, 90.0, 93.6, (Waters Delta 600e quaternary solvent delivery system and
103.0, 113.3, 127.2, 128.0, 129.9, 134.8, 139.7, 143.8, 151.2, 2998 photodiode array detector and Empower2 chromato-
158.7, 163.4. HRMS(ESI) calcd for C30H29O7N2FNa (M + Na+) graphy software), using an increasing gradient of acetonitrile
571.1851 found: 571.1848.
in 0.1 N triethylammonium acetate, pH 7.0, and characterized
by MALDI-TOF mass spectrometry. The MALDI-TOF spectra
were recorded on Voyager-De-STR (Applied Biosystems) MALDI-
TOF instrument and the matrix used for analysis was THAP
5′-O-(Dimethoxytrityl)-β-D-xylofuranosyl 3′-deoxy-3′-fluoro-
uridinyl-2-O-phosphoramidite (11a)
Compound 10a (0.20 g, 0.36 mmol) was co-evaporated with dry (2′,4′,6′-trihydroxyacetophenone). The purity was re-checked
dichloromethane, and then dissolved in dry dichloromethane and found to be >95%.
(3.0 ml). Diisopropylethylamine (DIPEA) (0.19 ml, 1.09 mmol)
UV-Tm measurements
was added, followed by chloro(2-cyanoethoxy)-N,N-diisopropyl-
amino)-phosphine (0.16 ml, 0.73 mmol) at 0 °C. The reaction The concentration was calculated on the basis of absorbance
mixture was stirred under argon atmosphere at room tempera- from molar extinction coefficients of the corresponding
ture for 3 hours, when TLC indicated the absence of starting nucleobases of DNA/RNA.32 The experiments were performed
material. The reaction mass was diluted with dichloro- at 1 μM concentrations. The complexes were prepared in
methane, washed with NaHCO3 and water, dried over sodium 10 mM sodium phosphate buffer, pH 7.4 containing NaCl
sulfate, followed by solvent removal. The crude product was (150 mM) and were annealed by keeping the samples at 90 °C
purified on
a silica gel column using 1 : 1 mixture of for 2 min followed by slow cooling to room temperature and
dichloromethane : ethylacetate and 1% triethylamine. Yield: refrigeration for at least two hours prior to running the exper-
0.16 g, 58.6%. 31P NMR (50 MHz, CDCl3): δ 151.4, 153.5. iments. Absorbance versus temperature profiles were obtained
HRMS(ESI) calcd for C39H48O8N4FP (M + H+)749.3110 found: by monitoring the absorbance at 260 nm from 10–85 °C at a
749.3110 and for (M + Na+) C39H47O8N4FPNa 771.2930 found: ramp rate of 0.5 °C per minute. The data were processed using
771.2935.
Microcal Origin 6.1 and Tm (°C) values were derived from the
maxima of the first derivative plots. All values are an average of
at least 3 experiments and accurate to within 0.5 °C.
5′-O-(Dimethoxytrityl)-β-D-ribofuranosyl 3′-deoxy-3′-fluoro-
uridinyl-2-O-phosphoramidite (11b)
CD experiments
Compound 10b (0.20 mg) obtained from previous step was dis-
solved in dry DCM (3 ml) followed by the addition of N,N-diiso- The samples for CD experiments were prepared as for the UV-
propylethylamine (0.18 ml, 1.09 mmol) and chloro(2- Tm experiments. Thus, 1 μM concentration of each strand was
cyanoethoxy)-(N,N-diisopropylamino)-phosphine (0.15 ml, used. The complexes were prepared in 10 mM sodium phos-
0.72 mmol) and the reaction mixture was stirred at room temp- phate buffer, pH 7.4 containing NaCl (150 mM) and were
erature for 2 h. The contents were then diluted with dry DCM annealed by keeping the samples at 90 °C for 2 min followed
and washed with 5% NaHCO3 solution. The organic phase was by slow cooling to room temperature and refrigeration for at
dried over anhydrous Na2SO4 and concentrated to get a foamy least two hours prior to running the experiments. CD spectra
solid. The residue was dissolved in DCM and precipitated with were recorded in a 1 cm pathlength cuvette at 10 °C, as an
hexane to obtain compound 11b (0.12 g) Yield: 43%. 31P NMR accumulation of 3 scans, and using a resolution of 1 nm,
(CDCl3): δ 151.4, 152.5. HRMS (ESI) calcd for C39H48O8N4FP bandwidth of 1 nm, sensitivity of 20 mdeg, response of 1 s and
(M + H+) 749.3110 found: 749.3112 and for (M + Na+) a scan speed of 200 nm min−1
.
C39H47O8N4FPNa 771.2930 found: 771.2929.
Nuclease resistance study
Synthesis of oligonucleotides
Enzymatic hydrolysis of the ONs (7.5 μM) was carried out at
DNA oligomers were synthesized on a CPG solid support using 37 °C in buffer (100 μl) containing 100 mM Tris-HCl (pH 8.5),
Bioautomation Mer-Made 4 synthesizer using standard β-cyano- 15 mM MgCl2, 100 mM NaCl and SVPD (2 μg, 1.2 × 10−4 U).
ethyl phosphoramidite chemistry. The 2′–5′-linked oligo- Aliquots were removed at several time-points; a portion of each
mers were synthesized in 2′–5′ direction using universal reaction mixture was removed and heated to 90 °C for 2 min to
columns as solid support. 5-Ethyl-thiotetrazole was used as the inactivate the nuclease. The amount of intact ONs was ana-
activator. For the modified units, double coupling (300 s × 2) lyzed at several time points by RP-HPLC. The percentage of
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 746–757 | 755

View Article Online
Paper
Organic & Biomolecular Chemistry
intact ON was then plotted against the exposure time to obtain
the ON degradation curve with time.
7 (a) C. Hendrix, H. Rosemeyer, I. Verheggen, F. Seela, A. Van
Aerschot and P. Herdewijn, Chem.–Eur. J., 1997, 3, 110;
(b) A. Van Aerschot, I. Verheggen, C. Hendrix and
P. Herdewijn, Angew. Chem., Int. Ed. Engl., 1995, 34, 1338.
8 (a) R. Steffens and C. J. Leumann, J. Am. Chem. Soc., 1997,
119, 11548; (b) R. Steffens and C. J. Leumann, J. Am. Chem.
Soc., 1999, 121, 3249; (c) D. Renneberg and C. J. Leumann,
J. Am. Chem. Soc., 2002, 124, 5993.
9 (a) M. J. Damha, P. A. Giannaris, P. Marfey and L. S. Reid,
Tetrahedron Lett., 1991, 32, 2573; (b) P. A. Giannaris and
M. J. Damha, Nucleic Acids Res., 1993, 21, 4742;
(c) M. J. Damha, B. Meng, D. Wang, C. G. Yannopoulos and
G. Just, Nucleic Acids Res., 1995, 23, 3967.
10 (a) J. C. Wallace and M. Edmons, Proc. Natl. Acad.
Sci. U. S. A., 1983, 80, 950; (b) I. M. Kerr and R. E. Brown,
Proc. Natl. Acad. Sci. U. S. A., 1978, 75, 256.
11 M. Wasner, D. Arion, G. Borkow, A. Noronha, A. H. Uddin,
M. A. Parniak and M. J. Damha, Biochemistry, 1998, 37,
7478.
Conclusions
A detailed study comprising the 2′–5′ oligonucleotides with
N-type and S-type locked/frozen nucleoside analogues was
undertaken. The results indicate that in the 2′–5′ linked oligo-
mers, the preferred geometry of nucleosides is S-type. More-
over, the S-type frozen ribofluoro uridine nucleoside was
found to exhibit higher stability when flanked by a purine
rather than a pyrimidine at its 2′-end. The stability of these oli-
gomers towards SVPDE is much better compared to the
natural 3′-5′ phosphodiester-linked oligomers, except for a
2′-terminal adenosine-5′-phosphate. Further work with other
nucleosides is currently being carried out in our laboratory.
Acknowledgements
12 B. J. Premraj, P. K. Patel, E. R. Kandimalla, S. Agarwal,
R. V. Hosur and N. Yathindra, Biochem. Biophys. Res.
Commun., 2001, 283, 537.
13 V. Lalitha and N. Yathindra, Curr. Sci., 1995, 68, 68.
14 T. L. Sheppard and R. C. Breslow, J. Am. Chem. Soc., 1996,
118, 9810.
N. E. thanks University Grants Commission, New Delhi for
senior Research Fellowship and V. A. K. thanks Wellcome
Trust UK, and Council of Scientific and Industrial Research,
New Delhi for research grants.
15 (a) T. P. Prakash, K.-E. Jung and C. Switzer, Chem.
Commun., 1996, 1793; (b) K.-E. Jung and C. Switzer, J. Am.
Chem. Soc., 1994, 116, 6059.
Notes and references
1 M. L. Stephenson and P. C. Zamecnik, Proc. Natl. Acad. 16 B. J. Premraj, S. Raja and N. Yathindra, Biophys. Chem.,
Sci. U. S. A., 1978, 75, 285. 2002, 95, 253.
2 P. C. Zamecnik and M. L. Stephenson, Proc. Natl. Acad. 17 (a) M. Polak, M. Manoharan, G. B. Inamati and J. Plavec,
Sci. U. S. A., 1978, 75, 280.
3 (a) S. Dutta, N. Bhaduri, N. Rastogi, S. G. Chandel,
J. K. Vandavasi, R. S. Upadhaya and J. Chattopadhyaya,
Nucleic Acids Res., 2003, 31, 2066; (b) B. J. Premraj, S. Raja,
N. S. Bhavesh, K. Shi, R. V. Hosur, M. Sundaralingam and
N. Yathindra, Eur. J. Biochem., 2004, 271, 2956.
Med. Chem. Commun., 2011, 2, 206; (b) Y.-L. Chiu and 18 S. Y. Wodak, M. Y. Liu and H. W. Wyckoff, J. Mol. Biol.,
T. M. Rana, RNA, 2003, 9, 1034; (c) C. F. Bennet and 1977, 116, 855.
E. E. Swayze, Annu. Rev. Pharmacol., 2010, 50, 259; 19 W. Saenger, in Principles of Nucleic Acid Structure, Springer-
(d) F. Lima, T. P. Prakash, H. M. Murray, G. A. Kinberger, Verlag, New York, 1984.
W. Li, A. E. Chappell, C. S. Li, S. F. Murray, H. Gaus, 20 N. D. Erande, A. D. Gunjal, M. Fernandes and V. A. Kumar,
P. P. Seth, E. E. Swayze and S. T. Crooke, Cell, 2012, 150, 883. Chem. Commun., 2011, 47, 4007.
4 (a) Z. Wang, Methods Mol. Biol., 2011, 676, 51; (b) S. Davis, 21 (a) S. Obika, K.-i. Morio, D. Nanbu and T. Imanishi, Chem.
B. Lollo, S. Freier and C. Esau, Nucleic Acids Res., 2006, 34,
2294.
Commun., 1997, 1643–1644; (b) S. Obika, K.-i. Morio,
D. Nanbu, Y. Hari, I. Hiromi and T. Imanishi, Tetrahedron,
2002, 58, 3039; (c) S. Obika, K.-i. Morio, Y. Hari and
T. Imanishi, Bioorg. Med. Chem. Lett., 1999, 9, 515;
(d) S. Obika, K.-i. Morio, D. Nanbu, Y. Hari and
T. Imanishi, Chem. Commun., 1999, 2423–2424.
5 (a) T. Imanishi and S. Obika, Chem. Commun., 2002, 1653;
(b) C. J. Leumann, Bioorg. Med. Chem., 2002, 10, 841;
(c) P. Herdewijn, Biochim. Biophys. Acta, Gene Struct.
Expression, 1999, 1489, 167; (d) C. Mathé and C. Périgaud,
Eur. J. Org. Chem., 2008, 1489.
22 R. Zou and M. D. Matteucci, Tetrahedron Lett., 1996, 37,
941.
6 (a) A. Koshkin, S. K. Singh, P. Nielsen, V. K. Rajwanshi,
R. Kumar, M. Meldgaard, C. E. Olsen and J. Wengel, Tetra- 23 H. Hayakawa, F. Takai, H. Tanaka, T. Miyasaka and
hedron, 1998, 54, 3607; (b) S. Obika, D. Nanbu, Y. Hari, K. Yamaguchi, Chem. Pharm. Bull., 1990, 38, 1136.
J.-i. Andoh, K.-i. Morio, T. Doi and T. Imanishi, Tetrahedron 24 I. A. Mikhailopulo, N. E. Poopeiko, T. I. Pricota,
Lett., 1998, 39, 5401; (c) D. A. Braasch and D. R. Corey,
Chem. Biol., 2001, 8, 1; (d) B. Vester and J. Wengel, Biochem-
G. G. Sivets, E. I. Kvasyuk, J. Balzarini and E. De Clercq,
J. Med. Chem., 1991, 34, 2195.
istry, 2004, 43, 13233; (e) M. Koizumi, Biol. Pharm. Bull., 25 N. Zhang, S. Zhang and J. W. Szostak, J. Am. Chem. Soc.,
2004, 27, 453.
2012, 134, 3691–3694.
756 | Org. Biomol. Chem., 2013, 11, 746–757
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
26 J. Plavec, L. H. Coole, A. Sandstrom and J. Chattopadhyaya, 30 H. T. Steely Jr., D. M. Gray and R. L. Ratliff, Nucleic Acids
Tetrahedron, 1991, 47, 7363–7376. Res., 1986, 14, 10071.
27 J. Krutzfeldt, N. Rajewsky, R. Braich, K. G. Rajeev, 31 (a) K. Morita, S. Obika, T. Imanishi, Y. Kitade and
T. Tuschl, M. Manoharan and M. Stoffel, Nature, 2005
438, 685.
28 S.-H. Kang, M.-J. Cho and R. Kole, Biochemistry, 1998
37, 6235.
M. Koizumi, Nucleic Acids Symp. Ser., 2008, 52, 637;
(b) K. Nagaoka, S. Kito, Y. Kitamura, Y. Ueno and Y. Kitade,
Nucleic Acids Symp. Ser., 2009, 53, 123; (c) K. Nagaoka,
Y. Kitamura, Y. Ueno and Y. Kitade, Bioorg. Med. Chem.
Lett., 2010, 20, 1186.
29 (a) A. M. Kawasaki, M. D. Casper, S. M. Freier, E. A. Lesnik,
M. C. Zounes, L. L. Cummins, C. Gonsalez and P. D. Cook, 32 Oligonucleotide Synthesis,
A
Practical Approach, ed.
J. Med. Chem., 1993, 36, 831.
M. J. Gait, IRL Press, Oxford, 1984.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 746–757 | 757