Alkynyl phosphonate DNA: A versatile click able backbone for DNA-based biological applications

Article abstract of DOI:10.1021/ja3026714

Major hurdles associated with DNA-based biological applications include, among others, targeted cell delivery, undesirable nonspecific effects, toxicity associated with various analogues or the reagents used to deliver oligonucleotides to cells, and stability toward intracellular enzymes. Although a plethora of diverse analogues have been investigated, a versatile methodology that can systematically address these challenges has not been developed. In this contribution, we present a new, Clickable, and versatile chemistry that can be used to rapidly introduce diverse functionality for studying these various problems. As a demonstration of the approach, we synthesized the core analogue, which is useful for introducing additional functionality, the triazolylphosphonate, and present preliminary data on its biological properties. We have developed a new phosphoramidite synthon-the alkynyl phosphinoamidite, which is compatible with conventional solid-phase oligonucleotide synthesis. Postsynthesis, the alkynylphosphonate can be functionalized via Click chemistry to generate the 1,2,3-triazolyl or substituted 1,2,3-triazolyl phosphonate-2′-deoxyribonucleotide internucleotide linkage. This manuscript describes the automated, solid-phase synthesis of mixed backbone oligodeoxyribonucleotides (ODNs) having 1,2,3-triazolylphosphonate (TP) as well as phosphate or thiophosphate internucleotide linkages and also 2′-OMe ribonucleotides and locked nucleic acids (LNAs) at selected sites. Nuclease stability assays demonstrate that the TP linkage is highly resistant toward 5′- and 3′-exonucleases, whereas melting studies indicate a slight destabilization when a TP-modified ODN is hybridized to its complementary RNA. A fluorescently labeled 16-mer ODN modified with two TP linkages shows efficient cellular uptake during passive transfection. Of particular interest, the subcellular distribution of TP-modified ODNs is highly dependent on cell type; a significant nuclear uptake is observed in HeLa cells, whereas diffuse cytoplasmic fluorescence is found in the WM-239A cell line. Cytoplasmic distribution is also present in human neuroblastoma cells (SK-N-F1), but Jurkat cells show both diffuse and punctate cytoplasmic uptake. Our results demonstrate that triazolylphosphonate ODNs are versatile additions to the oligonucleotide chemist's toolbox relative to designing new biological research reagents.

Full text of DOI:10.1021/ja3026714

Article
pubs.acs.org/JACS
Alkynyl Phosphonate DNA: A Versatile “Click”able Backbone for
DNA-Based Biological Applications
Heera Krishna and Marvin H. Caruthers*
Department of Chemistry and Biochemistry, University of Colorado at Boulder, Boulder, Colorado 80309, United States
S
* Supporting Information
ABSTRACT: Major hurdles associated with DNA-based biological
applications include, among others, targeted cell delivery, undesirable
nonspecific effects, toxicity associated with various analogues or the
reagents used to deliver oligonucleotides to cells, and stability toward
intracellular enzymes. Although a plethora of diverse analogues have
been investigated, a versatile methodology that can systematically
address these challenges has not been developed. In this contribution, we
present a new, Clickable, and versatile chemistry that can be used to
rapidly introduce diverse functionality for studying these various
problems. As a demonstration of the approach, we synthesized the
core analogue, which is useful for introducing additional functionality,
the triazolylphosphonate, and present preliminary data on its biological
properties. We have developed a new phosphoramidite synthonthe
alkynyl phosphinoamidite, which is compatible with conventional solid-phase oligonucleotide synthesis. Postsynthesis, the
alkynylphosphonate can be functionalized via “Click” chemistry to generate the 1,2,3-triazolyl or substituted 1,2,3-triazolyl
phosphonate-2′-deoxyribonucleotide internucleotide linkage. This manuscript describes the automated, solid-phase synthesis of
mixed backbone oligodeoxyribonucleotides (ODNs) having 1,2,3-triazolylphosphonate (TP) as well as phosphate or
thiophosphate internucleotide linkages and also 2′-OMe ribonucleotides and locked nucleic acids (LNAs) at selected sites.
Nuclease stability assays demonstrate that the TP linkage is highly resistant toward 5′- and 3′-exonucleases, whereas melting
studies indicate a slight destabilization when a TP-modified ODN is hybridized to its complementary RNA. A fluorescently
labeled 16-mer ODN modified with two TP linkages shows efficient cellular uptake during passive transfection. Of particular
interest, the subcellular distribution of TP-modified ODNs is highly dependent on cell type; a significant nuclear uptake is
observed in HeLa cells, whereas diffuse cytoplasmic fluorescence is found in the WM-239A cell line. Cytoplasmic distribution is
also present in human neuroblastoma cells (SK-N-F1), but Jurkat cells show both diffuse and punctate cytoplasmic uptake. Our
results demonstrate that triazolylphosphonate ODNs are versatile additions to the oligonucleotide chemist’s toolbox relative to
designing new biological research reagents.
synthesis of highly modified ODNs is often cumbersome
(synthons having difficult to prepare modifications are needed),
involves numerous additional protection/deprotection steps,
and occasionally suffers from low coupling efficiency which
leads to poor overall yields. Postsynthetic modifications in
solution phase are possible, especially when a biologically
relevant analogue is unable to survive the harsh conditions
employed during synthesis. Nevertheless, such postsynthetic
functionalization needs to be highly efficient, specific, and result
in the quantitative conversion of the ODN to the final, fully
modified product.
The copper(I) catalyzed azide−alkyne [3 + 2] cycloaddition
(CuAAC), developed by Meldal et al.10 and Sharpless et al.11
has emerged as the method of choice for postsynthetic
modification of numerous biologically relevant molecules.12
This is because the reaction is highly efficient, regioselective,
INTRODUCTION
■
Chemically modified, synthetic oligodeoxyribonucleotides
(ODNs) 15−25 2′-deoxyribonucleotides in length act in a
sequence-specific manner to modulate gene expression and
thus have gained popularity in diverse areas of genomic
research. These include, among others, their use as aptamers,1
ribozymes,2 enhancer DNA decoys,3 CpG immunostimulatory
oligomers,4 diagnostic DNA chips,5 and DNA-based functional
nanomaterials.6 ODNs having potential therapeutic drug
applications are also being tested as antisense reagents,1 triple
helix-forming oligodeoxyribonucleotides (TFOs),2 siRNAs,3
and as antagomirs to microRNAs.4,5 In order to be effective
for various applications, these ODNs contain many modifica-
tions such as 2′-fluoro7 to enhance nuclease resistance, a 2′-4′
methylene bridge8 (LNAs) which enhances duplex melting
temperatures, and cholesterol or other bioactive molecules9 for
increasing cellular uptake. However, given the repetitive nature
of the reagents used during solid-phase synthesis and the
considerable chemical functionality on these ODNs, the
Received: March 19, 2012
Published: May 21, 2012
© 2012 American Chemical Society
11618
dx.doi.org/10.1021/ja3026714 | J. Am. Chem. Soc. 2012, 134, 11618−11631

Journal of the American Chemical Society
Article
has a high pH tolerance range, retains facile synthetic
accessibility of starting materials, and the alkynes and azides
are inert to most biological and organic reaction conditions.13
The reaction can be performed in water under ambient
conditions14 and generates a 1,4-disubstituted 1,2,3-triazolyl
linkage that functions as a chemical linker for conjugating two
desirable chemical or biological entities.
toward nucleases and possess favorable hydrogen-bonding and
solubility properties. An added advantage would be the
inherent modularity of the Click approach as it permits further
modifications of the chemical and/or biological properties of
the ODN by appending cellular targeting ligands or other
desirable functionalities to 1,2,3-triazoles.
In order to test our hypothesis, we developed a novel
backbone modificationthe 1,2,3-triazolylphosphonate inter-
nucleotide (TP) linkage wherein one of the nonbridging
oxygen atoms of phosphate is replaced by a 1,2,3-triazole
moiety linked to phosphorus via C4 of the triazole ring. TP
linkages are generated as a two-step process. Initially the
ethynylphosphonate internucleotide linkage is introduced at
designated sites with conventional phosphoramidite chemistry
and automated solid-phase synthesis. The next step is to use
CuAAC to introduce the appropriate azide before deprotection
and cleavage from the solid support.39 By coupling these two
steps, we have synthesized chimeric ODNs (16−23mers) that
contain up to six TP modifications as well as other
functionalities such as LNA nucleotides, 2′-OMe ribonucleo-
tides, and thiophosphate. We have studied the thermal stability
and nuclease resistance of these ODNs and explored their
utility as efficient self-delivery agents. The results of these
experiments and the unique subcellular localization patterns of
TP ODNs in different cell lines have encouraged us to continue
investigating their potential as antagomirs for targeting
microRNA expression.
Apart from being a stable, biocompatible moiety, 1,2,3-
triazoles have many appealing features. Due to high dipole
moments, they are capable of hydrogen-bonding interactions
through their N-atoms as well as dipole−dipole and π-stacking
interactions with neighboring entities.15 The aromatic character
of the 1,2,3-triazole ring imparts chemical inertness to the
linkage as well as stability toward reduction, oxidation and acid
or base-catalyzed hydrolysis. These interesting electronic
properties have led to their use as π-conjugated linkers in
intramolecular electron transfer processes.16 Owing to their
ability to bind both cations and anions, they have been explored
as chemical sensors,17 and the presence of several donor sites
for metal coordination (N3, N2, C5) has led to the
development of the ‘Click-to-chelate’ approach wherein
transition metal chelates are installed onto biomolecules in a
single step.18 It is noteworthy that 1,4-substituted 1,2,3-triazoles
have been extensively employed as peptidomimetics owing to
their structural and electronic similarity to a linear trans-peptide
linkage.19 They have also been explored as bioisosteres for
internucleotide phosphate linkages.20−23 Although 1,2,3-
triazoles do not occur naturally, they are known to be
biologically active, and their potency as pharmacophores has
been reviewed.24
Despite the many desirable attributes of 1,2,3-triazoles,
synthetic ODNs incorporating these heterocycles as backbone
modifications have not been reported. However a literature
search reveals that “Click” chemistry has been successfully
applied to ligation, cyclization, bioconjugation, fluorophore
labeling of DNA25 and RNA,26 synthesis of modified
nucleosides,27 PNA−DNA chimeras,28 DNA−carbohydrate
dendrimers,13 site-specific labeling of siRNA for targeted
delivery29 and coupling ODNs to monolayers.30 These
postsynthetic functionalizations of DNA by CuAAC involves
the use of alkyne-modified dU phosphoramidites,31 phosphor-
amidites derived from alkyne-labeled 7-deazapurines and
pyrimidines,32 and 2′-O-alkynyl-modified sugars.33 Click
chemistry at the phosphorus center has not been explored
except for functionalization at the 5′-position of a DNA strand.
For example, an oligodeoxynucleotide containing a single 5′-
phosphate triester appended to a long spacer arm with a
terminal alkyne has been used for template-mediated chemical
ligation with a 3′-azide-modified oligodeoxynucleotide.25 Addi-
tionally recent reports describe the synthesis of phosphor-
amidites containing one, two, or three alkyne functions
appended to a phosphate triester via a side-chain, and
introduced at the 5′- end of oligodeoxynucleotides.34,35 The
only attempt at multiple functionalization of the phosphate
backbone with triazolyl linkages using the phosphate triester
methodology has been unsuccessful owing to the instability of
the triazolyl-containing triester side chain toward the basic
cleavage conditions required for chemical synthesis.36
In this contribution, we describe the chemical synthesis of
nucleoside 3′-O-alkynylphosphinoamidites, solid-phase syn-
thesis protocols, hybridization experiments (Tm), nuclease
susceptibilities, and preliminary in vivo studies on a series of
mixed-sequence ODNs having 1,2,3-triazolylphosphonate
internucleotide linkages.
RESULTS AND DISCUSSION
■
Copper-catalyzed azide−alkyne cycloadditions (CuAAC) have
been successfully employed using nucleosides that are
alkynylated at the 5′-, 2′-, or 3′-hydroxyls which generates the
corresponding O-propargylated derivatives.29 A major break-
through from the Carell group has enabled high-density
conversion of alkyne functions on ODNs containing up to six
consecutive alkyne-modified nucleobases.31 This was accom-
plished by introducing a long spacer arm between the
nucleobase and the alkyne functionality, thus making it more
accessible to Click reagents. Morvan et al. have synthesized
phosphoramidite building blocks having phosphorus appended
to the alkynyl function through a P−O linkage.34,35 However,
owing to the inherent lability of phosphotriesters toward basic
deprotection and cleavage conditions, this type of linkage
would be partially lost during synthesis.
Indeed, Tanabe et al. recently reported that aqueous
methylamine (20 h, 40 °C), which is used to remove ODNs
from a support, can effectively hydrolyze Click functionalized
side chains tethered by a P−O linkage to give unmodified
DNA.36 To overcome this limitation, we developed an alkynyl
functionalized phosphinoamidite that was joined to phosphorus
by a P−C as opposed to a P−O linkage. Using this synthon,
oligodeoxyribonucleotides having alkynylphosphonates at
selected sites were first synthesized on a solid support.
Triazolylphosphonate was then introduced using an appro-
priate azide via the CuAAC method. This approach has the
advantage that excess Click reagents can be used for better
product yields and then removed simply by washing the
We hypothesized that a more attractive synthetic target
would be a P−C linked triazolyl phosphonate because it would
be resistant to degradation during both synthesis and biological
studies. Given the metabolic inertness of triazoles,37,38 we also
hypothesized that such an ODN might have excellent stability
11619
dx.doi.org/10.1021/ja3026714 | J. Am. Chem. Soc. 2012, 134, 11618−11631

Journal of the American Chemical Society
Article
Scheme 1. Synthesis of 5′-O-Dimethoxytrityl-3′-O-ethynyl-2′-deoxyribonucleoside Phosphinoamidite and Generation of a
a
Triazolylphosphonate Internucleotide Linkage
a
Reaction conditions: (i) 1.1 equiv of HCCP[N(i-C3H7)2]2 (5), CH2Cl2, 1H-tetrazole, 60 min; (ii) 2′-deoxyribonucleoside linked to a solid
support through the 3′ hydroxyl, 0.25 M 5-ethylthio-1H-tetrazole, CH3CN, 600s; (iii) 0.02 M iodine in THF/water/pyridine; (iv) R1−N3 (6 equiv),
+
where R1 = SiMe3 or N(Me)3 CH2CH2, CuSO4·5H2O (0.8 equiv), sodium ascorbate (6 equiv), tris(1-benzyl-1H-1,2,3-triazol-4-yl)methylamine (7
equiv) in H2O/MeOH/THF (1.2 mL, 2:2:1 v/v/v).
support prior to deprotection of the bases and cleavage of the
ODN from the solid phase.
a 1.5 mL screw-cap glass reaction vial. The CuAAC reaction
was then carried out using the appropriate azide. A
heterogeneous mixture of the solid-support bound ODN (1.0
μM) and a solution of H2O/MeOH/THF (1.2 mL, 2:2:1 v/v/
v) containing azide (6 equiv based upon the alkyne),
CuSO4·5H2O (0.8 equiv), freshly prepared sodium ascorbate
(6 equiv), and tris[(1-benzyl-1H-1,2,3-triazol-4-yl)methyl]-
amine (TBTA,7 equiv) was allowed to react at room
temperature for 7 h to obtain the triazolyl phosphonate
ODN. The support was then washed successively with water (2
× 20 mL), methanol (2 × 10 mL), acetonitrile (2 × 10 mL)
and dichloromethane (2 × 10 mL). After drying, a 1:1 solution
(2 mL) of anhydrous ethylenediamine in toluene was added,
the support stirred by vortexing, and cleavage of the ODN was
allowed to proceed to completion over 2 h at room
temperature. The cleavage mixture was discarded42 and the
solid-support evaporated to dryness in a SpeedVac. The crude
ODN was washed from the support with 5% acetonitrile in
water (1 mL) and purified by RP-HPLC using a linear gradient
of 0 to 100% B over 40 min at a flow rate of 1.5 mL/min
(buffer A: 50 mM TEAB, pH 8.5; buffer B: acetonitrile). The
trityl-on fractions were combined, evaporated to dryness in a
SpeedVac, and treated with an 80% acetic acid solution for 2 h
in order to remove the trityl group. A second RP-HPLC was
then carried out to enhance the purification. A similar HPLC
profile was obtained for all ODNs, with the capped failure
sequences eluting at ∼20 min, whereas the trityl-on product
was observed at ∼35 min. Broad product peaks were
anticipated due to the presence of multiple chiral phosphorus
centers. The average yields (95% purity) for a 1.0 μM synthesis
ranged from 120 to 200 OD. An ODN that required a
fluorescein tag was synthesized as trityl-off and the last coupling
was completed with a commercially available fluorescein
phosphoramidite (Glen Research, VA). Aliquots of all ODNs
were characterized by analytical HPLC. Figures S7−S11 in the
SI show analytical RP-HPLC profiles for ODNs 1, 3, 5, 10, and
11 respectively. The mass data for selected ODNs having a
combination of TP, phosphate, or thiophosphate internucleo-
tide linkages and 2′-deoxyribonucleosides, 2′-OMe ribonucleo-
sides, and LNA nucleosides are listed in Table 1. Even after two
Chemical Synthesis. Our initial objective was to prepare
ethynylphosphinoamidite synthons of the four 2′- deoxyribo-
nucleosides (1−4). This strategy would allow us to introduce
the TP linkage at any selected site. The phosphitylating reagent,
bis(N,N-diisopropylamino)ethynyl phosphine (5), was synthe-
sized in high yield (90%) via a Grignard reaction from bis(N,N-
diisopropylamino)chlorophosphine and ethynylmagnesium
bromide.40 This phosphine was allowed to react with the
appropriate N-protected 5′-O-DMT-2′-deoxyribonucleoside (A,
G, C) or 5′-O-DMT-2′-deoxythymidine to yield 5′-O-DMT-2′-
deoxyribonucleoside-3′-O-ethynylphosphinoamidite (Scheme
1, Compounds 6−9). These synthons were obtained in high
purity after column chromatography with average yields ranging
from 40 − 66%. Characterization by NMR and mass spectral
analysis (see Experimental Section) confirmed the assigned
structures.
Solid-Phase Synthesis. Chimeric ODNs having TP,
phosphate, and thiophosphate internucleotide linkages as well
as 2′-deoxyribonucleoside, 2′-OMe ribonucleoside, or LNA
nucleosides were synthesized on solid supports using
ethynylphosphinoamidites 6−9, the commercially available 5′-
O-DMT-2′-deoxyribonucleoside 3′-O-phosphoramidites, and
the corresponding 3′-O-phosphoramidites of 2′-OMe ribonu-
cleosides and LNAs. Table S1(Supporting Information [SI])
summarizes the synthesis cycle. Briefly, the coupling wait time
was increased to 600 s for synthons 6−9, whereas the reaction
time for the remaining nucleoside phosphoramidites was
according to the manufacturers’ specifications. The activator
was 5-ethylthio-1H-tetrazole (0.25 M in anhydrous acetoni-
trile). After each condensation step, the support was washed
with acetonitrile for 40 s and failure sequences blocked using
Unicap phosphoramidite (Glen Research). Standard oxidation
with 0.02 M iodine in THF/water/pyridine generated
phosphate or alkynylphosphonate. The phosphorothioate
internucleotide linkage was prepared using Beaucage reagent41
as the oxidant. These synthesis steps were repeated until the
ODN of the desired sequence/length was prepared. Postsyn-
thesis, the support was removed from the column and placed in
11620
dx.doi.org/10.1021/ja3026714 | J. Am. Chem. Soc. 2012, 134, 11618−11631

Journal of the American Chemical Society
Article
Table 1. ODN Sequences used for NMR, Tm studies, and Molecular Weight Analysis
mol wt
†
no
ODNs
calc
obs
ODN 1
ODN 2
ODN 3
ODN 4
ODN 5
ODN 6
ODN 7
ODN 8
ODN 9
ODN 10
ODN 11
ODN 12
ODN 13
ODN 14
Aa CpCaAa TaGpAa TpGpTp GpCpTp GpCpT
5125.9
6741.2
7360.3
7753.4
7416.3
7161.2
5762.9
7751.3
7510.2
5130.9
7363.3
7108.2
7448.3
7193.2
5190.0
6805.6
7423.1
7817.2
7480.5
7163.5
5827.4
7813.4
7511.1
5195.4
7427.5
7109.3
7513.3
7194.4
TpGpTa ApAaAp CaCpAa TpGpAa TaGpTp GpCpTp GpCpT
TaGpTp ApAaAp CaCpAa TpGpAa TpGpTp GpCpTp GpCpTp AaT
UpMeGaUp ApMeAaAp CaCpMeAa UpMeGpMeAa UpMeGpMeUp GpMeCpMeUp GpMeCpMeUp ApMeUMe
Me
Me
Me
Me
Me
TaGpMeUp ApMeAaAp CaCpMeAa UpMeGpMeAa UpMeGpMeUp GpMeCpMeUp GpMeCpMeUp AMe
Me
Me
Me
Me
Me
TpGpMeUp ApMeApAp CpCpMeAp UpMeGpMeApUpMeGpMeUp GpMeCpMeUp GpMeCpMeUp AMe
Me
Me
Me
Me
Me
Me
Me
Me
Ap CaCpMeAa UpMeGpMeAa UpMeGsMeUp GsMeCpMeUp GsMeCpMeUsMe AaT
Me
Me
Me
Me
TaGpMeUp ApMeAaAp CaCpMeAa UpMeGpMeAa UpMeGsMeUp GsMeCpMeUp GsMeCpMeUsMe AaT
Me
Me
Me
Me
TpGpMeUp ApMeApAp CpCpMeApUpMeGpMeAp UpMeGsMeUp GsMeCpMeUp GsMeCpMeUsMe ApT
L
L
AaCpCa ApTp Gp AaTp Gp TpGpCp TpGpCp TpAaT
L
L
Tp Gp TpAaAp AaCpCa ApTp Gp AaTpGp TpGpCp TpGpCp TpAaT
L
L
Tp Gp TpApAp ApCpCpApTp Gp ApTpGp TpGpCp TpGpCp TpApT
L
L
L
L
L
Tp Gp TpAaAp AaCpCa ApTp Gp AaTpGp Tp GpCp Tp GpCp Tp AaT
L
L
L
L
L
Tp Gp TpApAp ApCpCpApTp Gp ApTpGp Tp GpCp Tp GpCp Tp ApT
†
Subscripts: “p”, phosphate; “s”, thiophosphate; “a”, 1,2,3-triazolylphosphonate. Superscripts: “Me” represents a 2′-OMe ribonucleoside; “L” denotes
a LNA nucleoside. All ODNs have 2′-deoxyribonucleosides except where designated as 2′-OMe ribonucleoside or LNA.
Figure 1. 31P NMR spectrum of ODN 4 (300 MHz, D2O, 13,064 scans) (UpMeGaUpMeApMeAaApMeCaCpMeAaUpMeGpMeAa
UpMeGpMeUpMeGpMeCpMeUpMeGpMeCpMeUpMeApMeUMe) (See Table 1 for an explanation of these abbreviations.).
RP-HPLC purification and desalting (illustra NAP-10 columns,
GE Healthcare), chelated copper ions were detected during the
mass spectral analysis. Washing the solid-support with an
EDTA solution prior to cleavage did not remove the copper
adducts. The observed peaks were broad, the m/z values were
higher than the expected mass and corresponded to one or
more copper adducts (see Table 1). This phenomenon has
been reported earlier29,43 and copper toxicity remains one of
the major potential problems with the copper-catalyzed “Click”
chemistry approach. This is because copper ions exhibit a high
affinity for thiol groups and may therefore severely hamper
many metabolic functions in the cell. Copper also catalyzes the
generation of highly reactive triplet oxygen species that can
degrade oligonucleotides44 and induce cytotoxicity in biological
systems.45,46 The interactions of copper with nucleotides and
nucleic acids have been studied extensively in the past.47 It is
well-known that there are two main sites of interaction of
copper with DNA, namely the phosphate backbone and the
11621
dx.doi.org/10.1021/ja3026714 | J. Am. Chem. Soc. 2012, 134, 11618−11631

Journal of the American Chemical Society
Article
Figure 2. Time-dependent enzymatic hydrolysis of ODN T4 in the presence of SVPDE.
electron-donor groups on the nucleoside bases. The ratios of
nucleoside base/phosphate binding affinities for the early
transition metals follow the order Mg(II) < Co(II), Ni (II) <
Mn (II) < Zn (II) < Cd(II) < Cu(II). Thus, copper ions are
known to exhibit the greatest affinity for nucleoside bases.48 As
indicated by several studies,49,50 Cu2+ may attach initially to
phosphate and later become bound to the bases since its
binding affinity (Ka) to the nucleosides is greater by an order of
102 M−1.51 Hence it is unclear whether the copper in the ODNs
prepared here is bound to phosphate or the nucleoside bases.
The 31P NMR analyses of ODNs display a signal at 9.7 ppm
(TP), and a sharp peak at −1.05 ppm (phosphate) (Figure 1).
31P NMR chemical shifts in the range 8.0−10.3 ppm have been
reported for bis(2-chloroethyl)-1H-1,2,3-triazolylphospho-
nates.52 Similar chemical shifts have also been reported for a
series of stable triazolylalanine phosphonate analogues.53
All ODNs used for biological studies were purified by PAGE
and desalted using illustra NAP-25 columns (GE Healthcare).
Polyacrylamide-gel electrophoresis of ODNs having variable
numbers of triazolyl phosphonate linkages was carried out using
an unmodified oligodeoxyribonucleotide as a marker in order to
determine the effect of this modification on electrophoretic
mobility. As anticipated, mobility decreased with an increase in
TP modifications due to charge neutralization of the phosphate
backbone. Using a series of polythymidylate 16-mer ODNs
(T1, T2, and T3), T3 having six TP modifications had reduced
mobility when compared to T2 with four modifications.
Similarly T1 with only two modifications had increased
mobility relative to T2 (Supporting Information, Table S2).
Neutralization of the DNA phosphate backbone is desirable for
many applications. Charge neutralized backbones have been
shown to improve DNA flexibility which enhances protein-
induced bending.54 It can also augment DNA-binding motifs in
DNA−protein recognition processes,55 improve hybridization
to a complementary strand,56 aid transport of ODNs across
biomembranes57 and prevent cleavage by RNase H.58,59 An
added advantage of the TP linkage is retention of aqueous
solubility despite the backbone charge neutralization since the
triazole ring can engage in hydrogen-bonding interactions with
neighboring entities. The P−Me or other alkyl or arylphosph-
onate hydrophobic backbone modifications do not possess
these characteristics and exhibit significant solubility con-
cerns.60
Enzymatic Studies. Since unmodified DNA is rapidly
degraded by cellular nucleases, one of the key requirements for
synthetic DNA analogues employed in biological research is
that they are nuclease resistant in subcellular environments.
Among chemically modified ODNs, the nuclease stability of the
nonionic methylphosphonate P−C bond as a substitute for P−
O has been thoroughly investigated.61,62 Because of these
studies, triazolylphosphonates were appealing since the P−C
bond was expected to be both nuclease resistant and have
enhanced aqueous solubility due to improved hydrogen
bonding.53
To evaluate the relative exonuclease susceptibility of
triazolylphosphonate modified ODNs, we synthesized two 23-
mer polythymidylate ODNs, T4 and T5, having one or two
triazolylphosphonate “caps” on both the 5′- and the 3′ ends
respectively. Unmodified polythymidylate DNA (23-mer) was
used as the control. The oligomers were tested for stability
against snake venom phosphodiesterase (SVPDE, 3′-exonu-
clease) and calf-spleen phosphodiesterase (CSPDE, 5′-
exonuclease).
For the SVPDE assay, a sample of the oligomer (0.7 OD)
was incubated at 37 °C with 5 μg of the enzyme in Tris-HCl
buffer (pH 9.0) containing 10 mM MgCl2. The CSPDE assay
was carried out using 0.7 OD of the oligomer and 0.1 U of the
11622
dx.doi.org/10.1021/ja3026714 | J. Am. Chem. Soc. 2012, 134, 11618−11631

Journal of the American Chemical Society
Article
Figure 3. Time-dependent enzymatic hydrolysis of ODN T4 in the presence of CSPDE.
Table 2. Tm Data for ODN/RNA Duplexes
number and type of modifications
b
c
d
e
a
name
duplex
TP linkages
2′-OMe
LNA
Thio
Tm (°C)
ΔTm (°C)
ΔTm′ (°C)
ODN 1
ODN 2
ODN 3
ODN 4
ODN 5
ODN 6
ODN 7
ODN 8
ODN 9
ODN 10
ODN 11
ODN 12
ODN 13
ODN 14
−
1/15
2/15
5
6
6
5
6
0
4
6
0
4
5
0
5
0
0
0
0
0
0
0
0
0
0
0
0
0
2
2
2
5
5
0
0
0
0
0
0
0
0
4
4
4
0
0
0
0
0
0
0
39.5
43.9
41.6
55.2
51.5
53.1
60
7
2.6
−
−
0.8
−0.4
−0.3
−
−
−0.5
−
−
−0.8
−
−0.9
−
−
3/15
0
4.9
4/15
17
16
16
12
16
16
0
10.7
14.4
12.8
5.9
5/15
6/15
7/15
8/15
59
6.9
9/15
62.1
50.4
47.6
51.6
69.5
74
3.8
10/15
11/15
12/15
13/15
14/15
15’/15
16/15
−3.9
−1.1
−5.1
−23
−27.5
−
0
0
0
0
0
46.5
65.9
−
0
0
−
−
a
b
ODNs 1−14 are defined in Table 1. 15 = 5′-UAG CAG CAC AUC AUG GUU UAC A-3′; 15′ = 5′-TGT AAA CCA TGA TGT GCT GCT A-3′;
16 = 5′- UGU AAA CCA UGA UGU GCU GCU A-3′(15′ is DNA, 15 and 16 are RNA). All Tm’s represent an average of at least three experiments;
ΔTm represents Tm unmodified control − Tm ODN. ΔTm′ represents the ΔTm per modification: (Tm modified control − Tm ODN)/number of
modifications.
c
d
e
enzyme in ammonium acetate buffer (pH 6.8) containing 2.5
mM EDTA at 37 °C. Aliquots from both reactions were
removed at various times, denatured and analyzed by RP-
HPLC as detailed in the SI. The half-lives were roughly based
upon the time for 50% degradation (area under the peak) of
the ODNs as observed by RP-HPLC. ODN T4 with one
11623
dx.doi.org/10.1021/ja3026714 | J. Am. Chem. Soc. 2012, 134, 11618−11631

Journal of the American Chemical Society
Article
triazolylphosphonate at the 5′ and the 3′ ends had a half-life of
∼2.5 h with SVPDE (Figure 2). A rapid degradation of internal
phosphodiester linkages was anticipated once the TP cap had
been removed by the enzyme. This was because the unmodified
DNA control degraded in less than 5 min under the same
experimental conditions. ODN T5 with two TP modifications
on each end was found to be more stable than ODN T4, with
an approximate half-life of 3.5 h. With CSPDE, even a single
modification at each end (ODN T4) prevented degradation of
the vicinal phosphodiester linkages. Less than 20% degradation
of ODN T4 was observed even after 12 h (Figure 3). The
unmodified DNA control was fully degraded by CSPDE in less
than 20 min under the same conditions.
Thermal Denaturation Studies. The biological potency of
an antisense ODN or an antagomir is influenced significantly by
its ability to bind target RNA with high specificity under
physiological conditions.63 Because 1,2,3- triazoles are hetero-
aromatic, they have the ability to engage in hydrophobic
interactions with nucleobases. For example, Kocalka et al. have
shown that stacking four consecutive 1,2,3-triazole modifica-
tions within the major groove of a DNA−RNA duplex
enhances its thermal stability by +4.3 °C per modification.64
Interestingly, a polythymidylate decamer of TLDNA (all
internucleotide phosphodiester linkages substituted by 1,2,3-
triazoles) was shown to have a Tm that was higher than the
unmodified DNA/DNA control by >30 °C. This effect was
attributed to the complete charge neutralization of the
phosphate backbone and retention of six-bond periodicity in
TLDNA.20
RNA duplex (depending on the modifications involved),
divided by the number of TP modifications.
The ΔTm′ was determined in order to provide a rough
estimate of the ΔTm for the TP linkage after approximately
discounting the effect of the other modifications on the duplex
stability. The ΔTm′ has been calculated only for those full
length ODNs that have corresponding controls. All ODN
duplexes showed an increase in A260 upon increasing temper-
ature and gave the characteristic sigmoidal melting curve. As
anticipated, the 16-mer ODN 1 (5 TP linkages) with a
truncation of 5 nucleotides from the 5′ end of the antagomir
showed the lowest Tm when hybridized with complementary
RNA (ΔTm = 7 °C). For ODNs 2 and 3 with six TP linkages,
the ΔTm was 0.6−0.8 °C per modification. The slightly lower
Tm observed in the case of ODN 3 (TP caps at 5′-and 3′ end)
when compared to ODN 2 (TP linkages toward the center of
the strand) suggests that the positioning of the TP linkages
along the strand could affect hybridization stability. The 2′-
OMe ribonucleoside containing ODNs (five or six TP linkages,
4, 5) showed only ∼0.3−0.4 °C depression per modification.
This modest reduction can be attributed to the well-known fact
that Tm depression can be partially offset by 2′-OMe sugar
residues in a molecule65(ODNs 4 and 5 have 17 and 16 2′-
OMe ribonucleosides each). Although the introduction of
thiophosphate linkages generally lower the melting temperature
of 2′-deoxyoligonucleotides by about 0.6 °C per linkage,66
a
depression of 0.5 °C per TP modification (ΔTm’) was observed
for ODN 8 which contained four thiophosphates in addition to
the sixteen 2′-OMe ribonucleosides. The hybridization affinity
improved dramatically when LNA nucleosides were added to
the construct.67 Thus, for an 18-mer ODN 10 having 2 LNA
nucleosides and 4 TP linkages, the Tm was higher than for the
unmodified DNA/RNA control duplex (∼4 °C). A further
increase in Tm (|ΔTm| = 27.5 °C) resulted when 5 LNA
nucleosides were present (ODN 13, five TP linkages).
Discounting the positive effect of the LNA, the Δ Tm’ value
of a TP linkage as calculated using ODN 13 is 0.9 °C. Thus, as
is the case with most phosphate backbone modifications, a
triazolylphosphonate linkage slightly destabilizes the duplex
when incorporated into a DNA and hybridized to comple-
mentary RNA. Since the triazolyl moiety is directly linked to
the phosphorus backbone, we presume that the heteroaromatic
triazole ring is not available for stacking interactions with the
nucleobases during duplex formation. Whether the incorpo-
ration of a flexible spacer arm between the phosphorus atom
and the 1,2,3-triazole would enhance duplex stabilization
remains to be investigated. Molecular modeling simulations
of twisted, intercalated nucleic acids (TINA) (the phosphorus
is tethered via an 8−10 carbon spacer to the N1-pyrenyl-1,2,3-
triazole moiety) showed that only the pyrenyl fragment was
engaged in stacking interactions with the nucleobases of a
DNA/DNA duplex.68
Biological Activity of 1,2,3-Triazolylphosphonates. In
a manner similar to peptidomimetics, the triazole moiety has
been widely studied as a pharmacophore,69 and several
molecules with the 1,2,3-triazole core are in the final stages
of clinical trials as anticancer, antibacterial, and antifungal
agents.24,70 Moreover, the pharmacokinetics of triazole based
drugs have been carefully studied.71 Given the vast literature on
the biocompatibility of these molecules and our own data on
the thermal and enzymatic stability of TP-modified ODNs, we
were motivated to explore the cellular uptake and subcellular
localization patterns of these ODNs.
In order to quantify contributions of TP linkages toward
duplex hybridization, we studied the thermal denaturation
characteristics of a series of triazolylphosphonate linked ODNs
1−14 (miR-15b antagomirs, Table 1) with the miR-15b RNA
comprising the sequence 5′-UAGCAGCACAUCAUG-
GUUUACA-3′. Each ODN was mixed with mir-15b RNA in
a 1:1 ratio at low salt conditions (10 mM sodium phosphate, 10
mM NaCl, pH 7.1) to obtain an overall concentration of 0.4
A260 units of the duplex. The samples were heat denatured at 85
°C for 10 min, cooled to 15 °C at a rate of 1 °C/min, and
maintained at this temperature for 10 min. Melting was
performed by heating the duplexes from 20 to 85 at 1 °C/min
with the absorbance (260 nm) being recorded at one min
intervals. Melting temperatures (Tm) were determined at the
maximum of first derivative plots.
For ODNs having TP/phosphate backbones (Table 1,
ODNs 1−3), and ODN 11 with TP/phosphate linkages and
two LNAs, the ΔTm (Table 2) was calculated as the difference
of melting temperatures of the ODN/RNA duplex and that of a
DNA/RNA control duplex 15′/15. For TP/phosphate/LNA
ODNs with five LNA (ODN 13) and all TP/2′-OMe
ribonucleoside ODNs (ODNs 4-9), the control was an
RNA/RNA duplex (16/15) as ODNs with many LNA and
2′-OMe ribonucleoside modifications usually generate A-form
helices.
In order to discount the effect of modifications other than
TP (LNA nucleosides, 2′-OMe ribonucleosides, thiophosphate
internucleotide linkages) on the Tm of an ODN, control ODNs
with the above modifications but devoid of TP linkages (6, 9,
12, and 14) were synthesized (Tables 1, 2). These controls
were used to determine the ΔTm′, which is defined as the ΔTm
per modification after discounting the effect of other
modifications. It was calculated as the difference of the Tm of
the antagomir/RNA duplex and the corresponding control/
11624
dx.doi.org/10.1021/ja3026714 | J. Am. Chem. Soc. 2012, 134, 11618−11631

Journal of the American Chemical Society
Article
It is well-known that oligonucleotide−cation conjugates
possess superior cellular uptake properties and facilitate
endosomal release via the proton sponge mechanism.72,73
Since trimethylaminoethyl side chains can be easily appended
to the triazolyl moiety using the Click approach,74 we
synthesized ODNs 17−20 carrying increasing numbers of
this positively charged side chain, and compared their uptake
properties with those of unsubstituted triazolylphosphonate
modified ODNs 21 and 22 (Table 3). A 5′-fluorescein tag was
Flow cytometry indicates that these ODNs are efficiently
taken up by cells without transfecting agents and that increased
uptake was dose dependent. However increasing the number of
modifications beyond four did not improve cellular uptake. At
0.5 μM, ODNs 18−20 had comparable (30−36%) uptake.
However, ODNs 21 and 22, without positively charged side
chains, had considerably higher (56%) uptake at 0.5 μM than
was observed for ODNs 18 and 19 which had the same number
and similar sites modified with the cationic trimethylaminoethyl
group. Even though the cellular uptake mechanism of the
triazolylphosphonate linkage is unknown, this preliminary study
suggests that the effect of positive charges on a TP-modified
backbone might be important only when the number of
positive charges far exceeds the negative charges on the
backbone. This conclusion is supported by recent literature
which shows that the selective introduction of cationic
trimethylammonium groups onto a polysaccharide (chitosan)
using Click chemistry resulted in improved cellular uptake of
the polymer only at nitrogen/phosphorus ratios (N/P) of 10/
1.75 Interestingly, for TP-modified ODNs, excellent cellular
uptake was observed even when the strand had only 13% TP
linkages (ODN 21).
a
Table 3. ODNs Used for Flow Cytometry
no.
ODN
17
18
19
20
21
22
5′-Fs-T†ApApCpApCpGpApTpApCpGpCpGpApT-3′
5′-Fs-T†ApApCpApCpGpApTpApCpGpCpGpA†T-3′
5′-Fs-T†A†ApCpApCpGpApTpApCpGpCpG†A†T-3′
5′-Fs-T†A†A†CpApCpGpApTpApCpGpC†G†A†T-3′
5′-Fs-TaApApCpApCpGpApTpApCpGpCpGpAaT-3′
5′-Fs-TaAaApCpApCpGpApTpApCpGpCpGaAaT-3′
a
Superscript “†” represents the 1-trimethylaminoethyl-1,2,3-triazolyl-
phosphonate linkage and subscript “a” represents the 1,2,3-
triazolylphosphonate linkage; Fs represents a fluorescein tag coupled
to the ODN via a thiophosphate linkage.
Further studies with ODNs carrying increasing number of
triazolylphosphonate linkages showed that the optimum cellular
uptake was observed for ODNs carrying four TP linkages (27%
modified). A further increase to 40% modification did not show
any enhancement in uptake efficiency. Increasing the ODN
concentration in the transfection media (ODN conc. > 6.0 μM)
did not result in a further shift of the fluorescence intensity
peak during flow cytometry. Thus the uptake pathway appeared
to be saturable at higher concentrations which suggest that
receptor-mediated endocytosis might be involved.76 Cells
treated with unmodified DNA exhibited very weak fluorescence
at all concentrations studied.
Fluorescence Microscopy. It is well-known that some
chemically modified ODNs such as methylphosphonates and
phosphorothioates adhere to cell membranes and that the
amount of cell-associated fluorescence observed by FACS
might not accurately reflect cell-internalized ODNs.76,77 Hence,
we investigated the subcellular distribution pattern of
fluorescently labeled ODN 22 on four different cell lines
using fluorescence microscopy.
introduced on each ODN in order to monitor cellular uptake.
This tag was linked through a thiophosphate to the ODN in
order to increase resistance toward degradation by nucleases.
Flow Cytometry. Preliminary studies on the cell uptake of
ODNs 17−22 were carried out on HeLa cells and the results
quantified by fluorescence assisted cell sorting (FACS).
Exponentially growing HeLa cells maintained as subconfluent
monolayers in Dulbecco’s Modified Eagle Medium (DMEM)
were seeded on 12-well culture plates 24 h before transfection.
For transfection, the medium was replaced with Opti-MEM
(reduced serum medium, Gibco) premixed with the ODNs at
various concentrations (0.5, 1.0, and 3.0 μM) and the cells were
incubated for 16 h. The cells were then rinsed three times with
D-PBS (Dulbecco’s phosphate buffer saline), trypsinized, and
resuspended in D-PBS. Fluorescence intensity was analyzed for
cells presenting higher fluorescence than the background. The
background was defined as the autofluorescence of cells
incubated only with Opti-MEM (no ODN) during the 16 h
transfection. Single-stranded 5′-fluorescein-labeled DNA having
the same sequence was used as control. The results are shown
in Chart 1.
HeLa, WM-239A, Jurkat and SK-N-F1 cells were incubated
with 6.0 μM fluorescein tagged ODN 22 for 20 h in the
absence of transfecting agents. The adherent cells (HeLa, WM-
Chart 1. Bar Diagram Showing the Percentage of HeLa Cells with Uptake, at Various Concentrations, of ODNs 17−22 after 16
h of Transfection in Opti-MEM (reduced serum medium)
11625
dx.doi.org/10.1021/ja3026714 | J. Am. Chem. Soc. 2012, 134, 11618−11631

Journal of the American Chemical Society
Article
239A, SK-N-F1) were then repeatedly washed with D-PBS and
fixed with formalin and prepared for analysis as described in the
Experimental Section. Jurkat cells were pelleted and resus-
pended three times with D-PBS and fixed with formalin before
analysis.
We observed that the subcellular localization pattern of ODN
22 was highly dependent on cell type. In HeLa cells, the ODNs
were localized predominantly in the nucleus (Figure 4A−D). In
Figure 5. Visualization of TP ODN uptake and intracellular trafficking
by fluorescence microscopy. Fluorescein labeled ODN 22 (6.0 μM in
Roswell Park Memorial Institute Medium, RPMI) was transfected into
WM-239A cells followed by 20 h of incubation at 37 °C. They were
then washed with PBS and fixed using buffered formalin solution.
Nuclei of the cells were counterstained with DAPI (blue). A. Phase
contrast image (40X) of WM-239A cells. B. Image showing the DAPI
nuclear localization of the same cells. C. Image showing fluorescein
localization. D. Image showing the merge of images B and C.
Figure 4. A. Visualization of TP ODN uptake and intracellular
trafficking by fluorescence microscopy. Fluorescein labeled ODN 22
(6.0 μM in reduced serum medium (Opti-MEM)) was transfected into
HeLa cells followed by 20 h of incubation at 37 °C after which the
cells were washed with PBS and fixed using buffered formalin solution.
Nuclei of the cells were counterstained with DAPI (blue). Phase
contrast image (40×) of HeLa cells. B. Image showing the DAPI
nuclear localization of the same cells. C. Image showing fluorescein
localization. D. Image showing the merge of images B and C.
WM-239A cells, localization was found to be highly diffuse in
the cytoplasm with little or no nuclear uptake. In addition to
being highly diffuse, the cytoplasmic fluorescence was not
punctuated or granular (Figure 5A−D). Since the ODN does
not appear to be sequestered in endosomal compartments, it is
anticipated that it might be readily available for biological
activity in the cytoplasm of these cells. In Jurkat cells, the
fluorescence was both punctate and diffuse while being located
primarily in the cytoplasm (Figure 6A−D). Given the
inherently poor transfection efficiency observed in neuro-
blastoma cell lines,78 we were encouraged by the finding that
ODN 22 was taken up by the SK-N-F1 cell line (human
neuroblastoma; Figure 7A−D). In these cells, the subcellular
distribution appeared to be predominantly cytoplasmic. A
literature search for similar findings revealed that nuclear
localization was observed when HeLa cells were transfected
with fluorescein labeled, polyimidazole conjugated deoxyoligo-
nucleotides (6 histamine residues, T12). The uptake was found
to proceed via active transport followed by membrane
destabilization.79 Imidazole containing polymers have also
been explored as effective gene delivery agents because of
their ability to release cargo into the cytoplasm via endo-
somolysis.80
Figure 6. Visualization of TP ODN uptake and intracellular trafficking
by fluorescence microscopy. Fluorescein labeled ODN 22 (6.0 μM in
RPMI) was transfected into Jurkat cells followed by 20 h of incubation
at 37 °C. They were then washed, pelleted and resuspended in PBS,
and fixed using buffered formalin solution. Nuclei of the cells were
counterstained with DAPI (blue). A. Phase contrast image (40X) of
Jurkat cells. B. Image showing the DAPI nuclear localization of the
same cells. C. Image showing fluorescein localization. D. Image
showing the merge of images B and C.
ization pathways adopted by the cell type as well as the
subsequent subcellular trafficking mechanism. Many different
internalization pathways for ODNs have been discovered. For
instance, phagocytosis takes place only in specialized cells such
as macrophages and granulocytes, while the other pathways
such as clathrin or caveolin-mediated uptake, macropinocytosis,
receptor mediated uptake (transferrin receptor, LDL receptor,
Toll-like receptors), are predominant in other cell types.81
Once the ODN enters the cell by one of the above
It is interesting that the localization of the synthetic ODN
was highly dependent on cell types. However, a simple
explanation to this phenomenon is currently unavailable. The
subcellular distribution might vary depending on the internal-
11626
dx.doi.org/10.1021/ja3026714 | J. Am. Chem. Soc. 2012, 134, 11618−11631

Journal of the American Chemical Society
Article
propidium iodide (PI), and then kept in the dark for 10 min
and sorted by flow cytometry. Each experiment was repeated
three times for statistical relevance. Cells that were not treated
with ODN 22 but stained with annexin-V FITC and PI were
used as the negative control. Cells stained with either annexin-V
FITC or PI were used to correct for compensation. Cells that
were not stained with both dyes served as controls for
optimizing the voltage settings on the instrument. From these
assays, we concluded that the cytotoxicity of the triazolyl-
phosphonate ODN was negligible in all cell lines studied. The
presence of copper adducts of ODNs did not show significant
toxicity in these experiments. In HeLa cells, the control showed
13.7% cell death and the ODN treated sample 14.1%. With
WM-239A cells, the control and ODN-treated samples showed
2.5% and 3% cell death respectively, whereas control Jurkat
cells showed 8.8% cell death as opposed to 9.7% for the sample.
Since no cytotoxicity owing to the presence of the copper
was detected in the cell lines tested, a literature search for
studies that evaluate the extent of toxicity of copper to
mammalian cells was carried out. It has been reported that in
the case of PBMCs (peripheral blood mononuclear cells), the
cell viability was virtually unaffected in the concentration range
0−10 μM, but a sharp decrease in viability was observed at
concentrations ranging from 50−150 μM.86 In human
hepatoma cells (HepG2), exposure to >64 μM of Cu2+ caused
up to 18% necrosis as opposed to <5% below this
concentration.87 In HeLa cells, copper ions at concentrations
ranging from 1−100 μM did not interfere with cell growth
although retarded cell growth was observed at 500 μM.88 Thus,
copper toxicity in mammalian cells is negligible at low
concentrations up to 10 μM, and hence remained undetected
in this study.
Figure 7. Visualization of TP ODN uptake and intracellular trafficking
by fluorescence microscopy. Fluorescein-labeled ODN 22 (6.0 μM in
reduced serum medium (Opti-MEM)) was transfected into SK-N-F1
cells followed by 20 h of incubation at 37 °C. They were then washed
with PBS and fixed using buffered formalin solution. Nuclei of the cells
were counterstained with DAPI (blue). (A) Phase contrast image 40×
SK-N-F1 cells. (B) Image showing the DAPI nuclear localization of the
same cells. (C) Image showing fluorescein localization. (D) Image
showing the merge of images B and C.
mechanisms, it encounters a complex maze of intracellular
pathways that ultimately determine its subcellular distribution
pattern.82 To the best of our knowledge, there have been no
reports on the differential uptake of the same chemically
modified ODN in different cell types. However, a search for
parallels in the literature reveals that the subcellular localization
patterns of the well-studied phosphorothioate oligonucleotides
(PS-ODNs) vary depending on cell type. Nuclear localization
was observed in bovine adrenal cortex cells,83 but in K562
human leukemia cells, these ODNs were localized within or at
the periphery of endosomes, and were also visualized free
within the cytoplasm and within the nucleus.84 In human
keratinocytes, PS-ODNs assumed a punctate distribution in the
cytoplasm with no nuclear uptake.85 Taken together, these
studies suggest that the cellular uptake and distribution pattern
of ODNs depend on a multitude of factors such as the nature of
the chemical modifications, the sequence and the length of the
ODN, the mechanism of uptake and the internalization
pathways adopted.
Currently, the internalization pathway (or pathways)
adopted by triazolylphosphonate modified ODNs are un-
known. Presumably, different pathways might be preferred by
each cell type. Alternatively, the ODN might leak from
endomembrane compartments during intracellular trafficking
or facilitate either endosomal rupture or membrane fusion.81
Cytotoxicity Assays. For cytotoxicity experiments, ∼1 ×
105 HeLa or WM-239A cells (or ∼2 × 105 Jurkat cells) were
seeded onto 12-well plates and incubated in appropriate
medium containing 10% fetal bovine serum, penicillin (100 U/
mL) and streptomycin (100 U/mL) for 24 h. All three cell lines
were treated with 6.0 μM ODN 22 and incubated for 20 h at 37
°C. After incubation, the cells were washed in PBS, (followed
by trypsinization in the case of HeLa and WM-239A), and
suspended in the Annexin-V binding buffer. Suspensions of
cells were counted and adjusted to a density of 1 x106cells/ml.
They were stained with annexin-V FITC followed by
SUMMARY
■
We have prepared a new 2′-deoxynucleoside 3′-phosphorami-
dite synthon, the alkynylphosphinoamidite, and demonstrated
that it is compatible with standard phosphoramidite chemistry.
To the best of our knowledge, alkynyl phosphinoamidites are
novel compounds that have not previously been synthesized. A
unique feature of this chemistry is that it facilitates introduction
of chemical modifications onto the DNA backbone via CuAAC
at any position along the DNA strand in a sequence
independent manner and allows site specific postsynthetic
modifications that are incompatible with conventional DNA
synthesis reagents. Postsynthetic conjugation on the solid-phase
also has the advantage that excess reagents can be used to drive
the reaction to completion. Moreover, this chemistry can be
tailored so as to develop a 5′-terminating ligand which can
“Click” two molecules of interest together, or to introduce
multiple Click reactions89 onto an ODN. Using this chemistry,
chimeric triazolylphosphonate modified ODNs having phos-
phates and thiophosphate internucleotide linkages, LNA
nucleosides, and 2′-OMe ribonucleosides (16−23 mers) have
been synthesized in good yields. A single TP modification at
the 3′- and 5′-ends generated an oligomer that was
approximately 30 times more stable than native DNA toward
SVPDE digestion. A 40-fold increase in nuclease resistance
toward CSPDE was also observed. This exceptional stability
could be partially attributed to the P−C bond. The absence of
triazoles in biological systems might also contribute to nuclease
stability as they may not be recognizable by nucleases. Although
this linkage is nonionic and can engage in hydrogen-bonding
interactions, TP-modified ODNs displayed a mild helix
11627
dx.doi.org/10.1021/ja3026714 | J. Am. Chem. Soc. 2012, 134, 11618−11631

Journal of the American Chemical Society
Article
mixture was allowed to attain room temperature, filtered under
nitrogen, and concentrated on a rotary evaporator. The resulting
viscous oil was extracted three times with anhydrous hexanes during
which the oil transformed into a solid. The solid was dissolved in a
minimum volume of anhydrous acetonitrile, and this solution was
extracted twice with anhydrous hexanes. All hexane fractions were
combined and concentrated in vacuo to give a translucent white oil,
which was found to be >99% pure by 31P NMR spectroscopy.Yield:
90%; 1H NMR (CDCl3): δ 3.53 (m, 4H), 3.02 (d, 1H), 1.19, 1.18 (d,
12H) and 1.17, 1.16 (d, 12H). 31P NMR (CDCl3): δ 28.0 (s). 13C
NMR (CDCl3): δ 68.1, 65.7, 25.9, 14.8.
destabilizing effect when hybridized to complementary RNA.
This effect might be due to the steric bulk of a rigid triazolyl
moiety which could pucker the sugar−phosphate backbone into
an energetically unfavorable conformation. Alternatively, one
diastereomer of the triazolylphosphonate could be destabilizing
toward double helix formation. Presumably, the ODN-bound
copper impurity might also exert a minor influence on Tm
values, depending on whether it is bound to the phosphate or
the nucleobase. Fluorescence microscopy revealed that TP-
modified ODNs are taken up, in the absence of cationic lipids,
by difficult-to-transfect cell lines (Jurkat, SK-N-F1) and that the
fluorescence intensities observed by FACS does not arise from
membrane-bound ODNs. The most intriguing observation
however was the differential uptake of the same ODN by
various cell types. While the fluorescence was predominantly
nuclear in HeLa cells, a diffuse cytoplasmic distribution was
observed in the WM-239A cell line. A mechanistic study might
provide valuable insights into the highly efficient uptake
pathway of these ODNs even when the oligonucleotide is
only 27% modified. It remains to be seen whether nuclear
uptake will be observed in other cell types and if this
observation can be exploited for designing efficient antisense
or antigene agents that require nuclear localization. Perhaps as
well, the observation that certain cell lines localize these ODNs
almost exclusively in the cytoplasm may facilitate the use of
antagomirs where cytoplasmic distribution is necessary. Future
research will address important challenges concerning the
effective design of potent siRNAs and antagomirs having the
TP linkage as well as efficient targeted delivery of these ODNs
in vivo.
The general procedure for the synthesis of 5′-O-dimethoxytrityl-2′-
deoxyribonucleoside 3′-O-ethynylphosphinoamidites (6−9, Scheme 1)
is as follows: Appropriately base-protected 5′-O-dimethoxytrityl-2′-
deoxyribonucleosides (1−4) (0.02 mols, 1.0 equiv), 1H-tetrazole (0.03
mol, 1.5 equiv) and anhydrous dichloromethane (100 mL) were taken
up in a 250 mL round-bottom flask and stirred under nitrogen until
the 2′-deoxyribonucleoside dissolves completely. Anhydrous dichloro-
methane (250 mL), bis(N,N-diisopropylamino)ethynylphosphine (5)
(0.02 mol, 1.0 equiv), and a magnetic stir bar were placed in a 500 mL
round-bottom flask fitted with a 150 mL addition funnel. The freshly
prepared solution of 2′-deoxyribonucleoside and 1H-tetrazole was
taken up in the addition funnel and added slowly over 30 min to the
500 mL round-bottom flask containing (5) under nitrogen. The
mixture was allowed to stir for an additional 30 min. Triethylamine
(approximately 0.2 mL) was added to neutralize the solution and the
solvent was removed in vacuo. The crude product was isolated by
chromatography using a 30−90% gradient of ethyl acetate in hexane
containing 1% triethylamine.
Compound 6. 5′-O-Dimethoxytrityl-N6-tert-butylphenoxyacetyl-
2′-deoxyriboadenosine 3′-O-ethynylphosphinoamidite: Yield: 66%;
1
31P NMR (CDCl3): δ 97.9, 96.4; H NMR (CDCl3): δ 8.74 (s, 1H),
8.24 (s, 1H), 7.40−7.18 (m, 10H), 7.02−6.78 (m, 7H), 5.32 (s, 1H),
4.84 (s, 1H), 4.34 (m, 1H), 3.79 (s, 6H), 3.42 (m, 2H), 3.07 (d, 1H),
2.71−2.61 (m, 1H), 2.06 (s, 1H), 1.32 (s, 9H), 1.22−1.19 (m, 12H).
13C NMR (CDCl3): δ 166.8, 158.5, 154.5, 152.4, 147.9, 135.6, 130.0,
CONCLUSION
■
In summary, we have developed new phosphoramidite
synthons of all four bases (A, G, C, and T) that can be used
in conjunction with CuAAC reaction for the site-specific
backbone modification of DNA. Our methodology enables the
multiple, high-density functionalization of the phosphorus
backbone of an oligodeoxynucleotide using Click chemistry.
Given the versatility of the CuAAC reaction, various bio-
logically relevant moieties, labeling dyes, diagnostic elements
etc. can be introduced at the phosphorus center as opposed to
the bases. This is advantageous because the base-pairing ability
of the DNA will not be significantly altered which makes this
the method of choice for adding desirable properties onto a
DNA strand without compromising their hybridization ability.
The ODNs generated by this strategy were found to be highly
stable to nucleases and displayed excellent cell uptake
properties without the aid of transfection agents. Subcellular
distribution patterns of the ODN were found to be highly
dependent on cell type. Taken together, these studies
demonstrate that triazolylphosphonate-modified ODNs are
ideal candidates for studying gene silencing and RNAi.
127.8, 126.6, 114.5, 113.1, 91.5, 84.4, 68.3, 55.2, 34.2, 31.5, 22.6.
Compound 7. 5′-O-Dimethoxytrityl-N2-tert-butylphenoxyacetyl-
2′-deoxyriboguanosine 3′-O-ethynylphosphinoamidite: Yield: 50%;
1
31P NMR (CDCl3): δ 97.3, 96.3; H NMR (CDCl3): δ 7.93 (s, 1H),
7.23−6.73 (m, 17H), 6.31 (t, 1H), 5.29 (s, 2H), 4.64 (s, 2H), 4.29 (m,
1H), 3.77 (s, 6H), 3.30 (m, 2H), 3.06 (d, 2H), 2.61 (m, 2H), 2.04 (s,
1H), 1.33 (s, 9H), 1.21−1.12 (m, 12H). 13C NMR (CDCl3): δ 168.3,
161.5, 158.2, 154.1, 145.8, 144.2, 130.4, 127.7, 126.8, 112.7, 92.4, 87.7,
87.4, 67.9, 65.4, 55.2, 31.4.
Compound 8. 5′-O-Dimethoxytrityl-N4-tert-butylphenoxyacetyl-
2′-deoxyribocytidine 3′-O-ethynylphosphinoamidite: Yield: 55%;
31P NMR (CDCl3): δ 100.2, 96.6; 1H NMR (CDCl3): δ 9.10 (s,
1H), 8.39 (d, 1H) 8.27 (d, 1H), 7.43−7.25 (m, 10H), 6.88−6.82 (m,
7H), 4.58 (s, 2H), 3.80 (s, 6H), 3.49 (s, 1H), 3.47 (s, 1H), 3.04 (d,
2H), 2.99 (d, 2H), 2.82 (m, 2H), 2.33 (m, 2H), 2.04 (s, 1H), 1.30 (s,
9H), 1.19−1.16 (m, 12H). 13C NMR (CDCl3): δ 169.6, 158.6, 155.3,
154.2, 147.7, 145.9, 144.4, 136.9, 130.3, 128.1, 127.9, 126.9, 114.4,
92.1, 86.5, 83.7, 67.1, 63.5, 55.2, 53.4, 34.3, 31.4.
Compound 9. 5′-O-Dimethoxytrityl-2′-deoxyribothymidine 3′-O-
ethynylphosphinoamidite: Yield: 45%; 31P NMR (CDCl3): δ 98.35,
96.20; 1H NMR (CDCl3): δ 7.63 (s, 1H), 7.60 (d, 2H), 7.30−7.23 (m,
10H), 6.83 (d, 2H), 6.47 (t, 1H), 6.38 (t, 1H), 4.66 (t, 1H), 4.16−4.10
(m, 1H), 3.90−3.87 (m, 1H), 3.78 (s, 6H), 3.33 (m, 3H), 3.04 (d,
2H), 2.60−2.46 (m, 1H), 2.33 (m, 2H), 2.04 (s, 3H), 1.25−1.06 (m,
18H). 13C NMR (CDCl3): δ 163.8, 158.5, 150.7, 143.9, 137.6, 136.2,
129.3, 128.4, 126.3, 114.6, 111.3, 78.8, 77.4, 63.6, 55.8, 45.9, 38.4, 22.4,
15.7.
Solid-Phase Synthesis. Solid-phase synthesis was carried out
using an ABI model 394 automated DNA synthesizer. The synthetic
protocol (SI, Table S1) was based on the conventional DMT-
phosphoramidite method. The 5′-O-dimethoxytrityl-2′-deoxyribonu-
cleoside 3′-O-ethynylphosphinoamidites (6−9) were used for coupling
in positions where triazolylphosphonate modification was necessary.
To generate 2′-deoxyribonucleoside phosphate linkages, commercially
EXPERIMENTAL SECTION
■
Chemical Synthesis. Synthesis of Bis(N,N-diisopropylamino)-
ethynylphosphine (5). To a flame-dried 500-mL Schlenk flask
equipped with a magnetic stir bar and septum under nitrogen was
added 0.02 mols of bis(N,N-diisopropylamino) chlorophosphine.
Anhydrous diethyl ether (350 mL) was added and the mixture was
cooled with stirring, to 0 °C. To this solution, ethynylmagnesium
bromide (0.5 M in THF, 1.1 equiv) was added dropwise via a syringe
over a period of 15 min, and the reaction mixture was allowed to stir at
0 °C for one hour. The 31P NMR spectrum after one hour indicated
complete conversion of the starting material to product. The reaction
11628
dx.doi.org/10.1021/ja3026714 | J. Am. Chem. Soc. 2012, 134, 11618−11631

Journal of the American Chemical Society
Article
available ultramild 5′-O-dimethoxytrityl-3′-O-cyanoethylphosphorami-
dites 2′-deoxyribonucleosides of A, G, C, and T were purchased from
ChemGenes (MA) and used without further purification. 5′-O-
Dimethoxytrityl-LNA nucleoside 3′-O-cyanoethylphosphoramidites
were obtained from Exiqon (UK). Commercially available ultramild
5′-O-dimethoxytrityl-2′-OMe ribonucleoside 3′-O-cyanoethylphophor-
amidites of A, G, C, and U were purchased from Glen Research (VA).
All appropriately protected deoxyribo and ribonucleoside 3′-O-
cyanoethylphosphoramidites were dissolved in anhydrous acetonitrile
at a concentration of 100 mM and placed on the appropriate ports of
the synthesizer. Prior to synthesis, the 5′-DMT group on the support-
bound 2′-deoxyribonucleoside was removed with 3% dichloroacetic
acid in dichloromethane. The coupling wait time was 600 s for the 5′-
O-dimethoxytrityl-2′-deoxyribonucleoside 3′-O-ethynylphosphinoami-
dites (6−9). All commercially available deoxyribo and ribonucleoside
phosphoramidites were used according to the manufacturer’s
recommendations. The activator was 5-ethylthio-1H-tetrazole (0.25
M in anhydrous acetonitrile, Glen Research, VA). After each
condensation step, the support was washed with acetonitrile for 40 s
and Unicap phosphoramidite (Glen Research, VA) was used according
to the manufacturer’s recommendations for capping failure sequences.
The standard oxidation reagent (0.02 M iodine in THF/water/
pyridine) was used for oxidation. Beaucage Reagent dissolved in
acetonitrile (1 g/100 mL) was used to generate thiophosphate
linkages. This cycle was repeated until the ODN of the desired
sequence and length was synthesized. Postsynthesis the CPG linked to
the ODN was removed from the column and placed in a 1.5 mL screw
cap, conical glass reaction vial. The Click reaction was carried out
using an appropriate azide (See SI). A heterogeneous mixture of the
solid-support bound ethynylphosphonate ODN, the azide,
CuSO4·5H2O, sodium ascorbate, and TBTA (1:6:0.8:6:7 molar
equiv. respectively) in water/methanol/THF (1.2 mL, 2:2:1 v/v/v)
was allowed to react at room temperature for 7 h in order to obtain the
respective triazolyl phosphonate ODN. The support was then washed
successively with water (2 × 20 mL) followed by methanol (2 × 10
mL), acetonitrile (2 × 10 mL) and dichloromethane (2 × 10 mL).
After drying the support, a 50% solution of anhydrous ethylenedi-
amine in toluene was added. The vial was vortexed in order to stir the
contents, and cleavage of the ODN from the support was allowed to
proceed for 2.5 h at room temperature. The cleavage mixture was
discarded and the solid-support was evaporated to dryness in a
SpeedVac. The ODN product adsorbed on the support was dissolved
in 5% acetonitrile−water and purified by RP-HPLC.
Nuclease Stability Experiments. Snake venom phosphodiester-
ase I and calf spleen phosphodiesterase II were purchased from Sigma
(St. Louis, MO). While performing enzymatic hydrolysis experiments,
10 μL of 1 M Tris buffer, 10 μL of 0.1 M MgCl2·6H2O, ODN T4 or
T5 (0.7 OD) and snake venom phosphodiesterase (5 μg) were mixed
and the total volume made up to 100 μL; 20 μL aliquots of the
reaction mixtures were removed at the indicated time points,
quenched by the addition of 1.0 M EDTA, and stored in dry ice
until analyzed by analytical RP-HPLC. For the CSPDE assay, 5 μL of 5
M ammonium acetate buffer (pH 6.8), 2.5 mM EDTA (38 μL), ODN
T4 or T5 (0.7 OD) and 0.07 U of CSPDE were mixed and made up to
a total volume of 100 μL. Twenty microliter aliquots of the reaction
mixtures were removed at the indicated time points, quenched by the
addition of 2.0 M urea and stored in dry ice until analyzed by analytical
RP-HPLC.
Preparation of Cells. HeLa cells, Jurkat cells, and SK-N-F1 cells
were obtained from American Type Culture Collection (Rockville,
MD) and serially maintained at monolayer cultures in a humidified
atmosphere of 5% carbon dioxide at 37 °C in Dulbecco’s Modified
Eagle’s Medium (HeLa, SK-N-F1) and Roswell Park Memorial
Institute Medium (WM-239A, Jurkat), containing 10% fetal bovine
serum, penicillin (100 U/mL) and streptomycin (100 U/mL). They
were counted and adjusted to the appropriate density, seeded onto 12-
well culture plates, and incubated for 24 h prior to ODN transfection.
HeLa cells were used at passages 12−18 and SK-N-F1 cells were used
at passages 3−7. Jurkat cells were used at passages 7−12. The WM-
239A cell line was a gift from Prof. Natalie Ahn, Department of
Chemistry and Biochemistry, University of Colorado at Boulder,
Boulder, CO, United States. These cells were routinely maintained as
monolayers in Roswell Park Memorial Institute Medium (RPMI)
containing 10% fetal bovine serum, penicillin (100 U/mL) and
streptomycin (100 U/mL) and used at passages 7−10.
Transfection Experiments. For transfection experiments, ∼1 ×
105 HeLa cells/well (12-well plates) were incubated in Dulbecco’s
Modified Eagle’s Medium (DMEM) containing 10% fetal bovine
serum, penicillin (100 U/mL) and streptomycin (100 U/mL) for 24 h.
The concentration (OD) of 5′-fluorescein-labeled ODN dissolved in
siRNA buffer 1× (Thermo Scientific) was measured by UV
spectroscopy. The media was removed and the cells were transfected
with the ODN in reduced serum medium (Opti-MEM) to give the
required final concentration. The cells were then incubated at 37 °C
for 16 h. After incubation, the media was removed from the wells and
cells were washed three times with 0.5 mL D-PBS. Cells were then
treated for 3 min at 37 °C with a prewarmed solution of trypsin−
EDTA (1×) until all cells became round and detached from the plates.
The cells from each plate were then taken up in 1 mL D-PBS and
pelleted by centrifugation at 1000 rpm for 5 min. The pellets were
washed and resuspended in D-PBS and kept at 0 °C in the dark until
analyzed by flow cytometry. Jurkat cells (∼2 × 105 cells) were grown
in suspension in 12-well plates and incubated in Roswell Park
Memorial Institute Medium (RPMI) containing 10% fetal bovine
serum, penicillin (100 U/mL) and streptomycin (100 U/mL) for 24 h.
The cells were transfected with the ODNs and incubated at 37 °C for
16 h. After incubation, the cells were pelleted by centrifugation at 1000
rpm for 5 min, resuspended in D-PBS, and washed twice with D-PBS.
The final pellet was resuspended in 100 μL of D-PBS and kept at 0 °C
in the dark until used for flow cytometric analysis.
Flow Cytometry. Flow cytometric data on at least 10,000 cells per
sample were acquired on a Moflow flow-cytometer (Beckman-
Coulter) equipped with a single 488 nm argon laser, 530/40 nm
emission filter (Fluorescein). Raw flow cytometry data were
manipulated and visualized using Summit 4.3 software (Beckman-
Coulter). Fluorescence intensity of the 5′-fluorescein tag was analyzed
for cells presenting higher fluorescence than the background. The
background was defined as the auto fluorescence of cells.
Fluorescence Microscopy. An inverted microscope (Olympus IX
81) equipped with a Hamamatsu C4742−95 CCD and CoolSNAP ES
digital camera (Photometrics) was used for fluorescence microscopy.
For microscopic analysis, ∼1 × 105 cells (HeLa, WM-239A) were
seeded onto sterile coverslips (18 mm diameter, VWR) placed in 12-
well plates and incubated in appropriate medium containing 10% fetal
bovine serum, penicillin (100 U/mL) and streptomycin (100 U/mL)
for 24 h. The cells were transfected with ODN 22 and incubated at 37
°C for 20 h. After incubation, the medium was removed from the wells
and cells washed four times with 0.5 mL PBS. The cells were covered
with 1 mL of 10% neutral, buffered formalin solution (Sigma-Aldrich)
for 15 min. The formalin solution was removed and the cells washed,
and covered with 2 mL of D-PBS for 10 min at RT. The coverslips
were carefully removed from the wells and mounted upside-down on
cover slides (25 × 75 mm, 1.0 mm thick) using Fluoromount-G with
DAPI (Southern Biotech) as the mounting medium, and observed
under the microscope.
Jurkat cells (∼2 × 105 cells) were seeded in suspension in 12-well
plates in the absence of coverslips and incubated in RPMI containing
10% fetal bovine serum, penicillin (100 U/mL) and streptomycin (100
U/mL) for 24 h. The SK-N-F1 cell line was purchased from American
Type Culture Collection (ATCC) and passaged in DMEM containing
10% fetal bovine serum, penicillin (100 U/mL) and streptomycin (100
U/mL). Prior to the fluorescence microscopy experiments, they were
seeded (∼2 × 105 cells) onto sterile coverslips (18 mm diameter,
VWR) placed in 12-well plates in appropriate culture media and
incubated for 24 h. Both cell lines were transfected with ODN 22 and
incubated at 37 °C for 20 h. After incubation, the cells were
trypsinized, pelleted by centrifugation at 1000 rpm for 5 min,
resuspended in D-PBS, and washed twice with D-PBS. The final pellet
was resuspended in 100 μL of D-PBS, fixed with formalin, washed with
D-PBS, pelleted, and resuspended in 100 μL D-PBS. Cells were then
11629
dx.doi.org/10.1021/ja3026714 | J. Am. Chem. Soc. 2012, 134, 11618−11631

Journal of the American Chemical Society
Article
mounted onto cover slides using DAPI-Fluoromount G and observed
under the microscope.
(22) El-Sagheer, A. H.; Sanzone, A. P.; Gao, R.; Tavassoli, A.; Brown,
T. Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 11338.
(23) El-Sagheer, A. H.; Brown, T. Chem. Commun. 2011, 47, 12057.
(24) Agalave, S. G.; Maujan, S. R.; Pore, V. S. Chem.Asian J. 2011,
6, 2696.
(25) Kumar, R.; El-Sagheer, A.; Tumpane, J.; Lincoln, P.;
Wilhelmsson, L. M.; Brown, T. J. Am. Chem. Soc. 2007, 129, 6859.
(26) Paredes, E.; Das, S. R. ChemBioChem 2011, 12, 125.
(27) Gutsmiedl, K.; Fazio, D.; Carell, T. Chem.Eur. J. 2010, 16,
6877.
ASSOCIATED CONTENT
■
S
* Supporting Information
General procedures for chemical synthesis, analytical RP-HPLC
profiles of ODNs, MALDI data on select ODNs, data on
PAGE, apoptosis assay protocol. This material is available free
of charge via the Internet at http://pubs.acs.org.
(28) Gogoi, K.; Mane, M. V.; Kunte, S. S.; Kumar, V. A. Nucleic Acids
Res. 2007, 35, e139.
AUTHOR INFORMATION
■
(29) Yamada, T.; Peng, C. G.; Matsuda, S.; Addepalli, H.;
Jayaprakash, K. N.; Alam, M. R.; Mills, K.; Maier, M. A.; Charisse,
K.; Sekine, M.; Manoharan, M.; Rajeev, K. G. J. Org. Chem. 2011, 76,
1198.
(30) Cutler, J. I.; Zheng, D.; Xu, X.; Giljohann, D. A.; Mirkin, C. A.
Nano Lett. 2010, 10, 1477.
Corresponding Author
Marvin.Caruthers@colorado.edu
Notes
The authors declare no competing financial interest.
(31) Gierlich, J.; Burley, G. A.; Gramlich, P. M. E.; Hammond, D. M.;
Carell, T. Org. Lett. 2006, 8, 3639.
ACKNOWLEDGMENTS
■
(32) Seela, F.; Sirivolu, V. R. Chem. Biodiversity 2006, 3, 509.
(33) Berndl, S.; Herzig, N.; Kele, P. t.; Lachmann, D.; Li, X.;
Wolfbeis, O. S.; Wagenknecht, H.-A. Bioconjugate Chem. 2009, 20, 558.
(34) Lietard, J.; Meyer, A.; Vasseur, J.-J.; Morvan, F. J. Org. Chem.
2007, 73, 191.
(35) Ligeour, C.; Meyer, A.; Vasseur, J.-J.; Morvan, F. Eur. J. Org.
Chem. 2012, 1851.
We thank Rich Shoemaker for assistance with the NMR facility,
Theresa Nahreini for FACS data collection, Prof. Natalie Ahn
for the fluorescence microscope, and the University of
Colorado Central Analytical Laboratories for the mass spectral
facility. This research was supported by the University of
Colorado at Boulder.
(36) Tanabe, K.; Ando, Y.; Nishimoto, S. Tetrahedron Lett. 2011, 52,
7135.
(37) Horne, W. S.; Yadav, M. K.; Stout, C. D.; Ghadiri, M. R. J. Am.
REFERENCES
■
(1) Keefe, A. D.; Pai, S.; Ellington, A. Nat. Rev. Drug Discovery 2010,
9, 537.
(2) El-Sagheer, A. H.; Brown, T. Proc. Natl. Acad. Sci. U.S.A. 2010,
Chem. Soc. 2004, 126, 15366.
(38) Dalvie, D. K.; Kalgutkar, A. S.; Khojasteh-Bakht, S. C.; Obach,
R. S.; O’Donnell, J. P. Chem. Res. Toxicol. 2002, 15, 269.
(39) Presumably, a stepwise use of the CuAAC method is also
possible which would introduce more than one substituted trizolyl
moiety into an ODN.
107, 15329.
(3) Cho, Y. S.; Kim, M.-K.; Cheadle, C.; Neary, C.; Park, Y. G.;
Becker, K. G.; Cho-Chung, Y. S. Proc. Natl. Acad. Sci. U.S.A. 2002, 99,
15626.
(40) It is noteworthy that a synthetic route involving bis(N,N-
diisopropylamino)chlorophosphine and sodium acetylide generated
the ethynylphosphinyldiamidite which was used as a precursor for
synthesizing mono and bicyclic phosphorus heterocycles. See:
Timmer, M. S. M.; Ovaa, H.; Filippov, D. V.; van der Marel, G. A.;
van Boom, J. H. Tetrahedron Lett. 2001, 42, 8231.
(41) Iyer, R. P.; Egan, W.; Regan, J. B.; Beaucage, S. L. J. Am. Chem.
Soc. 1990, 112, 1253.
(42) Oligonucleotides (greater than ∼15 bases) stick to the CPG-
solid support when anhydrous toluene−ethylene diamine mixture is
used for cleavage;
(43) Liu, X.-M.; Quan, L.-d.; Tian, J.; Laquer, F. C.; Ciborowski, P.;
Wang, D. Biomacromolecules 2010, 11, 2621.
(4) Kandimalla, E. R.; Bhagat, L.; Li, Y.; Yu, D.; Wang, D.; Cong, Y.-
P.; Song, S. S.; Tang, J. X.; Sullivan, T.; Agrawal, S. Proc. Natl. Acad.
Sci. U.S.A. 2005, 102, 6925.
(5) Miller, M. B.; Tang, Y.-W. Clin. Microbiol. Rev. 2009, 22, 611.
(6) Seeman, N. C. Annu. Rev. Biochem. 2010, 79, 65.
(7) Kawasaki, A. M.; Casper, M. D.; Freier, S. M.; Lesnik, E. A.;
Zounes, M. C.; Cummins, L. L.; Gonzalez, C.; Cook, P. D. J. Med.
Chem. 1993, 36, 831.
(8) Campbell, M. A.; Wengel, J. Chem. Soc. Rev. 2011, 40, 5680.
(9) Jeong, J. H.; Mok, H.; Oh, Y.-K.; Park, T. G. Bioconjugate Chem.
2008, 20, 5.
(10) Tornøe, C. W.; Christensen, C.; Meldal, M. J. Org. Chem. 2002,
67, 3057.
(44) Kanan, M. W.; Rozenman, M. M.; Sakurai, K.; Snyder, T. M.;
Liu, D. R. Nature 2004, 431, 545.
(11) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B.
Angew. Chem., Int. Ed. 2002, 41, 2596.
(45) Chen, Z.; Meng, H.; Xing, G.; Chen, C.; Zhao, Y.; Jia, G.; Wang,
T.; Yuan, H.; Ye, C.; Zhao, F.; Chai, Z.; Zhu, C.; Fang, X.; Ma, B.;
Wan, L. Toxicol. Lett. 2006, 163, 109.
(46) Hein, C.; Liu, X.-M.; Wang, D. Pharm. Res. 2008, 25, 2216.
(47) Izatt, R. M.; Christensen, J. J.; Rytting, J. H. Chem. Rev. 1971,
71, 439.
(48) Eichhorn, G. L.; Shin, Y. A. J. Am. Chem. Soc. 1968, 90, 7323.
(49) Eichhorn, G. L.; Clark, P. Proc. Natl. Acad. Sci. U.S.A. 1965, 53,
586.
(12) Best, M. D. Biochemistry 2009, 48, 6571.
(13) El-Sagheer, A. H.; Brown, T. Chem. Soc. Rev. 2010, 39, 1388.
(14) Seo, T. S.; Li, Z.; Ruparel, H.; Ju, J. J. Org. Chem. 2002, 68, 609.
(15) Whiting, M.; Muldoon, J.; Lin, Y.-C.; Silverman, S. M.;
Lindstrom, W.; Olson, A. J.; Kolb, H. C.; Finn, M. G.; Sharpless, K. B.;
Elder, J. H.; Fokin, V. V. Angew. Chem., Int. Ed. 2006, 45, 1435.
(16) de Miguel, G.; Wielopolski, M.; Schuster, D. I.; Fazio, M. A.;
Lee, O. P.; Haley, C. K.; Ortiz, A. L.; Echegoyen, L.; Clark, T.; Guldi,
D. M. J. Am. Chem. Soc. 2011, 133, 13036.
(50) Bach, D.; Miller, I. R. Biopolymers 1967, 5, 161.
(51) Sagripanti, J.-L.; Goering, P. L.; Lamanna, A. Toxicol. Appl.
Pharmacol. 1991, 110, 477.
(17) Lau, Y. H.; Rutledge, P. J.; Watkinson, M.; Todd, M. H. Chem.
Soc. Rev. 2011, 40, 2848.
(18) Struthers, H.; Mindt, T. L.; Schibli, R. J. Chem. Soc., Dalton
(52) Anisimova, N. A.; Kuzhaeva, A. A.; Berkova, G. A.; Deiko, L. I.;
Berestovitskaya, V. M. Russ. J. Gen. Chem. 2005, 75, 689.
(53) Mukai, S.; Flematti, G.; Byrne, L.; Besant, P.; Attwood, P.;
Piggott, M. Amino Acids 2011, DOI: 10.1007/s00726-011-1145-2.
(54) Hausheer, F. H.; Singh, U. C.; Palmer, T. C.; Saxe, J. D. J. Am.
Chem. Soc. 1990, 112, 9468.
Trans. 2010, 39, 675.
(19) Pedersen, D. S.; Abell, A. Eur. J. Org. Chem. 2011, 2399.
(20) Isobe, H.; Fujino, T.; Yamazaki, N.; Guillot-Nieckowski, M.;
Nakamura, E. Org. Lett. 2008, 10, 3729.
(21) Lucas, R.; Zerrouki, R.; Granet, R.; Krausz, P.; Champavier, Y.
Tetrahedron 2008, 64, 5467.
11630
dx.doi.org/10.1021/ja3026714 | J. Am. Chem. Soc. 2012, 134, 11618−11631

Journal of the American Chemical Society
Article
(55) Okonogi, T. M.; Alley, S. C.; Harwood, E. A.; Hopkins, P. B.;
Robinson, B. H. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 4156.
(56) Manoharan, M. Antisense Nucleic Acid Drug Dev. 2002, 12, 103.
(57) Levis, J. T.; Butler, W. O.; Tseng, B. Y.; Ts’O, P. O. P. Antisense
Res. Dev. 1995, 5, 251.
(58) Quartin, R. S.; Brakel, C. L.; Wetmur, G. Nucleic Acids Res. 1989,
17, 7253.
(59) Dash, C.; Scarth, B. J.; Badorrek, C.; Gotte, M.; Le Grice, S. F. J.
̈
Nucleic Acids Res. 2008, 36, 6363.
(60) Dias, N.; Stein, C. A. Mol. Cancer Ther. 2002, 1, 347.
(61) Miller, P. S.; McParland, K. B.; Jayaraman, K.; Tso, P. O. P.
Biochemistry 1981, 20, 1874.
(62) Thatcher, G. R. J.; Campbell, A. S. J. Org. Chem. 1993, 58, 2272.
(63) Monia, B. P.; Sasmor, H.; Johnston, J. F.; Freier, S. M.; Lesnik,
E. A.; Muller, M.; Geiger, T.; Altmann, K.-H.; Moser, H.; Fabbro, D.
Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 15481.
̌
(64) Kocalka, P.; Andersen, N. K.; Jensen, F.; Nielsen, P.
ChemBioChem 2007, 8, 2106.
(65) Freier, S. M.; Altmann, K.-H. Nucleic Acids Res. 1997, 25, 4429.
(66) Kibler-Herzog, L.; Zon, G.; Uznanski, B.; Whittier, G.; Wilson,
W. D. Nucleic Acids Res. 1991, 19, 2979.
(67) Kaur, H.; Wengel, J.; Maiti, S. Biochem. Biophys. Res. Commun.
2007, 352, 118.
(68) Geci, I.; Filichev, V. V.; Pedersen, E. B. Chem.Eur. J. 2007, 13,
́
6379.
(69) Maag, D.; Castro, C.; Hong, Z.; Cameron, C. E. J. Biol. Chem.
2001, 276, 46094.
(70) Meldal, M.; Tornøe, C. W. Chem. Rev. 2008, 108, 2952.
(71) Nnane, I. P.; Njar, V. C. O.; Brodie, A. A. J. Steroid Biochem. Mol.
Biol. 2001, 78, 241.
(72) Liu, X.; Wu, J.; Yammine, M.; Zhou, J.; Posocco, P.; Viel, S.; Liu,
C.; Ziarelli, F.; Fermeglia, M.; Pricl, S.; Victorero, G.; Nguyen, C.;
Erbacher, P.; Behr, J.-P.; Peng, L. Bioconjugate Chem. 2011, 22, 2461.
(73) Gagnon, K. T.; Watts, J. K.; Pendergraff, H. M.; Montaillier, C.;
Thai, D.; Potier, P.; Corey, D. R. J. Am. Chem. Soc. 2011, 133, 8404.
(74) Francavilla, C.; Low, E.; Nair, S.; Kim, B.; Shiau, T. P.; Debabov,
D.; Celeri, C.; Alvarez, N.; Houchin, A.; Xu, P.; Najafi, R.; Jain, R.
Bioorg. Med. Chem. Lett. 2009, 19, 2731.
(75) Gao, Y.; Zhang, Z.; Chen, L.; Gu, W.; Li, Y. Biomacromolecules
2009, 10, 2175.
(76) Shoji, Y.; Akhtar, S.; Periasamy, A.; Herman, B.; Juliano, R. L.
Nucleic Acids Res. 1991, 19, 5543.
(77) Shoji, Y.; Shimada, J.; Mizushima, Y.; Iwasawa, A.; Nakamura,
Y.; Inouye, K.; Azuma, T.; Sakurai, M.; Nishimura, T. Antimicrob.
Agents Chemother. 1996, 40, 1670.
(78) Jordan, E. T.; Collins, M.; Terefe, J.; Ugozzoli, L.; Rubio, T. J.
Biomol. Tech. 2008, 19, 328.
(79) Morvan, F.; Castex, C.; Vives
Nucleotides Nucleic Acids 2001, 20, 805.
(80) Pack, D. W.; Putnam, D.; Langer, R. Biotechnol. Bioeng. 2000, 67,
217.
̀
, E.; Imbach, J.-L. Nucleosides,
(81) Juliano, R. L.; Ming, X.; Nakagawa, O. Bioconjugate Chem. 2012,
23, 147.
(82) Angers, C. G.; Merz, A. J. Semin. Cell Dev. Biol. 2011, 22, 18.
(83) Li, B.; Hughes, J. A.; Phillips, M. I. Neurochem. Int. 1997, 31,
393.
(84) Beltinger, C.; Saragovi, H. U.; Smith, R. M.; LeSauteur, L.; Shah,
N.; DeDionisio, L.; Christensen, L.; Raible, A.; Jarett, L.; Gewirtz, A.
M. J. Clin. Invest. 1995, 95, 1814.
(85) Wingens, M.; van Hooijdonk, C. A. E. M.; de Jongh, G. J.;
Schalkwijk, J.; van Erp, P. E. J. Arch. Dermatol. Res. 1998, 290, 119.
(86) Singh, R.; Kumar, S.; Nada, R.; Prasad, R. Mol. Cell. Biochem.
2006, 282, 13.
(87) Aston, N. S.; Watt, N.; Morton, I. E.; Tanner, M. S.; Evans, G. S.
Hum. Exp. Toxicol. 2000, 19, 367.
(88) Hultberg, B.; Andersson, A.; Isaksson, A. Toxicology 1997, 117,
89.
(89) Gramlich, P. M. E.; Warncke, S.; Gierlich, J.; Carell, T. Angew.
Chem., Int. Ed. 2008, 47, 3442.
11631
dx.doi.org/10.1021/ja3026714 | J. Am. Chem. Soc. 2012, 134, 11618−11631