Fragment-based solid-phase assembly of oligonucleotide conjugates with peptide and polyethylene glycol ligands

Article abstract of DOI:10.1016/j.ejmech.2016.05.001

Ligand conjugation to oligonucleotides is an attractive strategy for enhancing the therapeutic potential of antisense and siRNA agents by inferring properties such as improved cellular uptake or better pharmacokinetic properties. Disulfide linkages enable dissociation of ligands and oligonucleotides in reducing environments found in endosomal compartments after cellular uptake. Solution-phase fragment coupling procedures for producing oligonucleotide conjugates are often tedious, produce moderate yields and reaction byproducts are frequently difficult to remove. We have developed an improved method for solid-phase coupling of ligands to oligonucleotides via disulfides directly after solid-phase synthesis. A 2′-thiol introduced using a modified nucleotide building block was orthogonally deprotected on the controlled pore glass solid support with N-butylphosphine. Oligolysine peptides and a short monodisperse ethylene glycol chain were successfully coupled to the deprotected thiol. Cleavage from the resin and full removal of oligonucleotide protection groups were achieved using methanolic ammonia. After standard desalting, and without further purification, homogenous conjugates were obtained as demonstrated by HPLC, gel electrophoresis, and mass spectrometry. The attachment of both amphiphilic and cationic ligands proves the versatility of the conjugation procedure. An antisense oligonucleotide conjugate with hexalysine showed pronounced gene silencing in a cell culture tumor model in the absence of a transfection reagent and the corresponding ethylene glycol conjugate resulted in down regulation of the target gene to nearly 50% after naked application.

Full text of DOI:10.1016/j.ejmech.2016.05.001

European Journal of Medicinal Chemistry 121 (2016) 132e142
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Research paper
Fragment-based solid-phase assembly of oligonucleotide conjugates
with peptide and polyethylene glycol ligands
Mehrdad Dirin, Ernst Urban, Christian R. Noe, Johannes Winkler*
University of Vienna, Department of Pharmaceutical Chemistry, Althanstraße 14, 1090, Vienna, Austria
a r t i c l e i n f o
a b s t r a c t
Article history:
Ligand conjugation to oligonucleotides is an attractive strategy for enhancing the therapeutic potential of
antisense and siRNA agents by inferring properties such as improved cellular uptake or better phar-
macokinetic properties. Disulfide linkages enable dissociation of ligands and oligonucleotides in reducing
environments found in endosomal compartments after cellular uptake. Solution-phase fragment
coupling procedures for producing oligonucleotide conjugates are often tedious, produce moderate
yields and reaction byproducts are frequently difficult to remove. We have developed an improved
method for solid-phase coupling of ligands to oligonucleotides via disulfides directly after solid-phase
synthesis. A 20-thiol introduced using a modified nucleotide building block was orthogonally depro-
tected on the controlled pore glass solid support with N-butylphosphine. Oligolysine peptides and a
short monodisperse ethylene glycol chain were successfully coupled to the deprotected thiol. Cleavage
from the resin and full removal of oligonucleotide protection groups were achieved using methanolic
ammonia. After standard desalting, and without further purification, homogenous conjugates were
obtained as demonstrated by HPLC, gel electrophoresis, and mass spectrometry. The attachment of both
amphiphilic and cationic ligands proves the versatility of the conjugation procedure. An antisense
oligonucleotide conjugate with hexalysine showed pronounced gene silencing in a cell culture tumor
model in the absence of a transfection reagent and the corresponding ethylene glycol conjugate resulted
in down regulation of the target gene to nearly 50% after naked application.
Received 21 December 2015
Received in revised form
25 April 2016
Accepted 1 May 2016
Available online 3 May 2016
Keywords:
Antisense oligonucleotides
PEGylation
Disulfide bioconjugates
Solid phase synthesis
Gene silencing
© 2016 Elsevier Masson SAS. All rights reserved.
1. Introduction
uptake of the ONs is the development of bioconjugates with uptake
enhancing ligands [4,5]. By attaching specific ligands, oligonucle-
Oligonucleotide gene silencing and splice-switching technolo-
gies have grown into indispensable tools in biomedical research as
well as promising therapeutic modalities. However, their wide-
spread therapeutic application is still significantly challenged by
their poor pharmacokinetics, particularly insufficient cellular up-
take [1e3]. Considering the overall moderate clinical success of
oligonucleotides so far, it seems imperative to develop novel
oligonucleotide derivatizations that properly shield the active
agent up to its site of action without perturbing its pairing avidity.
One of the most promising approaches to improve the cellular
otide distribution and cellular uptake can be rationally modulated
to a certain extent. Depending on the ligand, efficient shielding
from degrading enzymes, increased circulation times, receptor-
specific cell binding and uptake, or enhanced membrane perme-
ation can be conferred to oligonucleotides. Revusiran is the prime
example of GalNAc-siRNA conjugates which make use of highly
efficient receptor-mediated uptake in hepatocytes through the
asialoglycoprotein receptor [6e8].
The generation of ligand-oligonucleotide conjugates requires
efficient fragment-(or stepwise-) based synthesis schemes, which
are ideally versatile to enable coupling of a variety of different li-
gands [9,10]. Precise and site-specific conjugation of small molecule
ligands, or mono-disperse peptides and polymers with a defined
chemical structure allows the generation of homogeneous ON
conjugates for structure-activity relationship investigations and
further preclinical development [11]. Arginine or lysine rich cell-
penetrating peptides, glycoclusters, lipids, (linear or branched)
Abbreviations: ON, oligonucleotide; PEG, polyethylene glycol; CPG, controlled
pore glass; LCAA, long chain amino alkyl; TCEP, tris(2-carboxyethyl)phosphin; TBP,
tri-n-butylphosphine; HOBt, 1-hydroxybenzotriazole; DIPEA, N,N-diisopropyle-
thylamine;
hexafluorophosphate.
* Corresponding author.
BOP,
benzotriazol-1-yloxy)
tris(dimethylamino)phosphonium
E-mail address: johannes.winkler@univie.ac.at (J. Winkler).
http://dx.doi.org/10.1016/j.ejmech.2016.05.001
0223-5234/© 2016 Elsevier Masson SAS. All rights reserved.

M. Dirin et al. / European Journal of Medicinal Chemistry 121 (2016) 132e142
133
polyamines, (linear or branched) poly- or oligo-ethylene-glycol
chains and numerous small molecules have all received interest
as possible oligonucleotide ligands [12].
In order to circumvent the problems associated with solution-
phase approaches, i.e. the cumbersome purification and its
limited applicability to lipophilic ligands, we evolved the method to
allow on-bead coupling. Besides enabling facile purification, this
method is also applicable to a wider range of ligands including
those not easily soluble in aqueous buffers [10,30].
Although conventional postsynthetic coupling reactions on the
30- and 50- positions of oligonucleotides in solution seem to be
straightforward, there are some serious limitations in many cases,
particularly for ligands with poor solubility in aqueous solutions.
Often, multistep deprotection and activation procedures with
subsequent purification steps are necessary, and electrostatic in-
teractions of some ligands (particularly basic peptides) with oli-
gonucleotides often hamper the efficiency of the coupling reactions
[13]. These standard methods, typically making use of amino- or,
less frequently, thiol-linkers introduced during oligonucleotide
synthesis, are limited to attaching a single ligand at a terminal
nucleotide position. Small molecules such as cholesterol or carbo-
hydrates can also be tethered to the 50-position of oligonucleotides
by using respective phosphoroamidite building blocks [14,15].
Attachment on the 20-position via linkers attached to the hydroxyl
group of modified ribose nucleotides offer more versatile conju-
gation at any nucleotide within the sequence as well as multiple
attachments within a single sequence [16]. Ligands are usually
attached to the cleaved and purified oligonucleotide by solution-
phase fragment coupling. In addition, stepwise solid-phase syn-
thesis of peptide-oligonucleotide conjugates is possible, starting
either with assembly of the ON followed by on-line ligand synthesis
or vice versa [10,17]. The use of appropriately derivatized phos-
phoramidite building blocks carrying the desired ligand is another
possibility to generate conjugates with small molecule ligands. In
this strategy, ligands are tethered to the oligonucleotides via
phosphate linkages, and it has been recently widely employed
particularly for preparing single- and double stranded GalNAc
conjugates for hepatocyte targeting [8,18e21].
There are a variety of chemical linkages used to join ligands with
oligonucleotides, including amides, hydrazines, oximes, and many
more [10,11,22]. Thiols afford an orthogonal reaction site to form
ON-ligand conjugates through thioether, thioester or disulfide
linkages [17,23]. Due to its ease of formation and their favorable
stability profile for therapeutic applications, native ligation via a
disulfide bond is perhaps the most extensively explored linkage in
oligonucleotide peptide bioconjugations [22,24,25]. Disulfide link-
ages between ligands and oligonucleotides have been predomi-
nantly formed by employing post-synthetic approaches [25].
Although the exact extent and the kinetics of cleavage within
endosomes are not fully elucidated, disulfide linkages are designed
to at least partially release the delivered oligonucleotide from the
ligand within endosomal compartments after successful cellular
internalization [26]. Basic molecules are believed to contribute to
subsequent endosomal escape through the proton sponge effect
[27]. Linker cleavability is crucial for at least conjugates with large
ligands such as PEGylated oligonucleotides and peptide- or protein-
oliognucleotide conjugates, because bulky substituents impede
interaction with the respective effector nucleases [28].
2. Results and discussion
To enable efficient formation of disulfide linked oligonucleotide
conjugates and facilitate removal of ligand educts and reaction side
product, we developed a feasible method for fragment coupling on
the controlled pore glass (CPG) solid support. A resin-bound, still
protected oligonucleotide with a free thiol group facilitates the
coupling of lipophillic ligands in organic solvents. This is of
particular importance considering the high hydrophilicity and
anionic character of oligonucleotides which renders them unfa-
vorable for conjugation reactions with other charged ligands due to
aggregation [31]. In contrast, the resin-bound ON is practically
neutrally charged since the internucleosidic oxygen atom is still
protected as cyanoethyl ether. For the sake of achieving high
coupling efficiencies, the ligand is usually used in considerable
excess, which complicates the purification when the conjugation
reaction is carried out in solution. In contrast, on a solid support,
the unreacted conjugate group and the side products are easily
removed by simple filtration, which markedly facilitates purifica-
tion procedures.
A crucial point is the necessity of orthogonal cleavage of the
thiol protection group and the linkage compatibility with the final
global deprotection and cleavage step which is usually achieved in
concentrated aqueous ammonia. The stability of peptide amide
bonds during standard deprotection/cleavage is well documented
[32]. However, when preparing disulfide linked conjugates, the
oligonucleotide deprotection/cleavage step has to be adapted due
to cleavage of the disulfide bond under those basic conditions.
In order to enable on-resin coupling, an adequate orthogonal
oligonucleotide protection/deprotection strategy had to be devel-
oped. Among the conventional sulfur protecting groups adapted to
ON synthesis and post-synthetic in-solution coupling, the S-trityl
group has found widespread use due to its stability towards
phosphoramidite ON synthesis strategy and subsequent depro-
tection/cleavage and conjugation procedures, particularly with
respect to RNA ligation reactions [33]. However, with respect to
solid-phase conjugation approach, this strategy bears important
shortcomings; the standard argentometric deprotection of the tri-
tyl group followed by regeneration produces insoluble side prod-
ucts which cannot be thoroughly washed off the resin. As a result,
poor reagent diffusion results in drastic reduction of yields during
the subsequent coupling steps [33,34]. We chose the tert-butyl
sulfanyl group because it possesses favorable characteristics for
post-synthetic on-bead deprotection as well as adequate stability
for withstanding the conditions of phosphoramidite based oligo-
nucleotide synthesis. In contrast to solid-phase coupling, the risk of
the generation of homodimeric oligonucleotides is minimized.
By using S-sulfonate derivatized ligands [29], the required
deprotection and activation steps of other disulfide-building pro-
cedures are avoided, as is the risk of homodimerization. We
selected a highly cationic peptide and an amphiphilic ethylene
glycol ligand for coupling to prove the use of the method for
compounds with different solubility profile. In particular, basic
peptides are difficult to attach to oligonucleotide because of their
high charge density, which makes them prone to aggregation and
hampers the use of aqueous solutions. The S-sulfonate group
withstands the deprotection conditions of the Fmoc peptide syn-
thesis strategy thus enabling peptide synthesis products to directly
Solid-phase fragment coupling offers an alternative method for
the stepwise synthesis on a single support. In this approach, the
entire conjugate is assembled on a single support, either by
conjugation of a prefabricated conjugate group or by assembling
the conjugate moiety by a stepwise process prior to or after the ON
synthesis [9]. The advantage of this method compared to solution
fragment coupling is the application to a broad spectrum of ligands
including lipophilic molecules and the ease of separating excess
reagents and byproducts. We have recently communicated a facile
post-synthetic disulfide formation method for solution based
fragment coupling [29]. An S-sulfonate protected cysteine is uti-
lized as the complementary ligand part to attach a peptide ligand
onto the 20-thioethyl arm of the ON in aqueous buffered solutions.

134
M. Dirin et al. / European Journal of Medicinal Chemistry 121 (2016) 132e142
undergo thiolysis for bioconjugations. The free deprotected thiol
group of the resin-bound ON can thus directly react with the S-
sulfonate protected thiol function of the ligand. For oligolysine
peptides, we attached a cysteine at the N-terminus as a tether. For
facilitating coupling, we utilized the more lipophilic S-methane
sulfonate group for a short ethylene glycol ligand.
For synthesis of the nucleotide building block, the hydroxyl
group of the 20-side chain (4) [29] was transiently converted to a
bromide to produce a key compound (5) capable of evolving into
various building blocks. We performed conversion of the bromide 5
to the desired activated disulfide terminal methylsulfonyl-
protected thiol (6), followed by replacement of the methyl-
sulfonyl group with a more stable protecting group for oligonu-
cleotide synthesis, the tert-butyl sulfanyl group (7).
stability and a certain degree of enhanced intracellular uptake. Such
conjugates can easily be used in combination with other chemical
modifications or delivery systems that improve circulation times
and biodistribution.
To evaluate on-bead coupling for amphilic and hydrophobic,
positively charged ligands, we coupled a short polyethylene glycol
chain and several oligolysine peptides [29] to the oligonucleotide.
Therefore, a short ethylene glycol fragment consisting of five
ethylene glycol units was derivatized with a terminal S-meth-
anesulfonyl protected thiol (Fig. 2).
2.2. Deprotection of the thioalkyl tether and coupling of the ligands
The removal of the 30- and 50-silyl groups was quantitatively
carried out with triethylamine trihydrofluoride (8), thereby mini-
mizing the risk of partial exocyclic amine and S-tert-butyl-sulfanyl
side chain cleavage. Subsequent 5'-dimethoxytritylation yielded
the nucleoside building block 9, which was purified by recrystal-
lization. This key intermediate can be easily phosphitylated and
thus allows incorporation of 20-attachment sites at any position
within an oligonucleotide chain.
We attached the nucleoside to a CPG solid support via a succinyl
spacer (10) for use at the 30-terminal site. The first adenosine of a
model sequence directed against bcl-2 was exchanged for a
modified 20-O-thioethyl grafted adenosine. A long chain alkyl
amine controlled pore glass (LCAA-CPG) solid phase loaded with
the thioethylene modified adenosine 10 was successfully applied in
automated phosphodiester oligonucleotide synthesis using stan-
dard phosphoramidite coupling protocols.
Deprotection of the tert-butyl sulfanyl can be carried out either
with
a
phosphorous-containing reagent, preferably tris(2-
carboxyethyl)phosphin (TCEP) or tri-n-butylphosphine (TBP), or a
sulfur-containing reducing agent such as dithiothreitol or ethane-
thiol [44,45]. Due to the high functionalization of the ONs and in
order to decrease the risk of side reactions or cleavage from the
resin, we opted for the phosphorous-based reduction. Hence, the
20-thiol was deprotected after successful ON assembly using trib-
utylphosphine (Fig. 3). After thorough washing to remove the
reducing agent, the conjugation of the ethylene glycol chain was
achieved by addition of a 4-fold excess of the corresponding ligand
in anhydrous acetonitrile. The same reaction conditions produced
only moderate yields for oligolysines, which were thus applied in a
mixture of acetonitrile and aqueous Tris buffer. Triple repetitions of
the coupling step with 10 equiv. oligolysines were necessary for
complete conversion. The deprotection and disulfide formation
steps were monitored using the colorimetric Ellman's test [46].
Although a rather large excess of the peptides needed to be
employed, the unreacted surplus can easily be recovered from the
reaction solution.
2.1. Preparation of peptide and ethylene glycol ligands for coupling
The ligands were chosen based on two main strategies applied
for improving the pharmacokinetic properties of the ONs: charge-
neutralizers with polycationic molecules such as basic peptides
[12], and ethylene glycol chains as used in the Stealth™ technology
[35]. It has been demonstrated that use of basic peptides not only
partially neutralizes the multiple negative charges of the oligonu-
cleotide backbone, but also protects the oligonucleotide in acidic
environments of endosomal compartments and facilitates endo-
somal escape [36,37].
2.3. Cleavage and full deprotection of the conjugate
Due to the relatively high susceptibility of disulfide bond
cleavage in aqueous alkaline conditions which carry the risk of a b-
elimination reaction, the standard oligonucleotide deprotection
and cleavage step was replaced with treatment with methanolic
ammonia (Fig. 3). After cleavage, deprotection, and desalting, all
products were analyzed by HPLC and gel electrophoresis
(supporting information) for purity and their structures and
integrity were verified by high-resolution mass spectrometry
(Table 1). All conjugates were easily soluble in aqueous solution at
We prepared oligolysines P1-4 (Fig. 1) by conventional Fmoc/tBu
strategy on a Rink amide resin [29]. To provide a thiol-terminated
linker, sulfo-cysteine was sequentially incorporated at the N-ter-
minus of the oligolysine amide chains (3e6 L-lysine residues, Fig.1).
Polyethylene glycol (PEG) conjugation has been shown to pro-
long the half-life of the ON conjugates in plasma [28]. Long and
short polyethylene glycol chains have been demonstrated to
minimize the enzymatic degradability of the ONs and limit their
unspecific interactions with serum proteins and non-targeted cells
[38]. For antisense and siRNA applications, it has been shown that
an increasing length of ethylene glycol chains reduces the gene
silencing effect, likely by interfering with binding to the effector
nucleases [39]. Shorter strands, on the other hand are well tolerated
by the gene silencing machinery and do not reduce silencing effi-
ciency [40,41]. Furthermore it has been shown that independently
of the molecular weight, branched PEG was less efficient than linear
PEG. Several studies have demonstrated that long chains are not
necessarily a prerequisite for repellency towards proteins [42,43] In
order to avoid clearance by the mononuclear phagocytic system,
short chains are sufficient. Although only longer PEG chains can
efficiently rescue oligonucleotide from rapid renal elimination in
the absence of plasma protein binding, conjugates with shorter
ethylene glycol chains have utility in inferring an increasing
stock concentrations of 100 mM. The conjugates produced using this
method showed at least 90% purity without any additional purifi-
cation after resin cleavage (supporting information). The oligoly-
sine conjugates 10e13 are identical with the compounds generated
using solution-phase fragment coupling and their analytical prop-
erties reported earlier [29].
Ethylene glycol conjugate 14 was characterized by reversed-
phase HPLC which showed a single product with a longer reten-
tion time than the oligonucleotide educt (supporting information).
Polyacrylamide gel electrophoretic analysis showed a slightly lower
migration caused by the increase of molecular weight. High-
resolution mass spectrometry analyses proved the identity of all
bioconjugates (Table 1). Circular dichroism spectroscopy of the
conjugates indicated the retention of B-type DNA structure,
showing that the attachment of the basic peptides caused no pro-
found alteration of the spatial structure of the nucleic acids with
only slight differences in the far UV region (supporting
information).

M. Dirin et al. / European Journal of Medicinal Chemistry 121 (2016) 132e142
135
Fig. 1. Preparation of the nucleoside building block 10 for attachment to CPG and oligonucleotide synthesis. Optimized procedures resulted in a total yield of 32%.
Fig. 2. Synthesis of the pentaethylene glycol ligand 17 for on-bead coupling to 20-thiol-modified oligonucleotides.
2.4. Evaluation of gene-silencing properties
Successful gene silencing was measured by western blotting and
quantification of Bcl-2 protein bands (Fig. 4). All experiments were
conducted with naked ONs with phosphodiester backbones and 1:1
stoichiometric mixtures of the corresponding isosequential ON and
the ligand as controls to verify the conjugation-dependent
The gene-silencing activity of the conjugates was examined in
the absence of any transfection enhancing agent in a melanoma
tumor model with the antiapoptotic Bcl-2 protein as the target.

136
M. Dirin et al. / European Journal of Medicinal Chemistry 121 (2016) 132e142
Fig. 3. Oligonucleotide synthesis and solid-phase coupling to PEG5 and oligolysine peptides.
Table 1
Sequences and analytical data of the conjugates 10e14. Oligonucleotide sequence is TTCTCCCAGCGTGCGCCA (no modifications apart from the 30-nucleotide), and the indicated
ligands were attached to the 20-linker at the 30-terminal nucleotide. High-resolution mass spectra were acquired in negative-ion mode after desalting in 1 M ammonium
acetate solution. (A: methanol; B: 50 mM triethylammonium acetate buffer, pH 7).1: 28e33% solvent A in 30 min; flow rate ¼ 1 mL minꢀ1 at 25 ꢁC; column: Thermo Scientific
BDS Hypersil C18 (150 ꢂ 4 mm).2: 15e70% solvent A in 35 min; flow rate ¼ 1 mL minꢀ1 at 25 ꢁC; column: Phenomenex Synergi 4u Fusion-RP 80A (150 ꢂ 4.60 mm).
Compound Sequence
HPLC retention time Mass found/calculated
(min)
10
11
12
13
14
Oligonucleotide-
K3
Oligonucleotide-
K4
Oligonucleotide-
K5
6.111
6.261
6.481
6.951
[M-9H]9ꢀ: 664.5683 (664.5871)
[M-9H] 9ꢀ: 678.8310 (678.8285)
[M-9H] 9ꢀ: 693.0599 (693.0699)
[M-10H]10ꢀ: 636.6681 (636.4793)
Oligonucleotide-
K6
Oligonucleotide- 24.422
PEG5
[M-7H]7ꢀ: 820.8690 (820.7092); [M-8H]8ꢀ: 718.1360 (717.9946); [M-9H]9ꢀ: 638.2386 (638.1054); [M-10H]10ꢀ
574.3033 (574.1941)
:
silencing activity. We observed a concentration-dependent effect,
which for the oligolysine conjugates was in direct relation with
increasing peptide chain length as reported earlier [29]. Among the
conjugates 10e14, the hexalysine conjugate 13 showed the stron-
gest effect (9-fold reduction) even at a 100 nM concentration
(Fig. 4A). Neither the wild-type ON nor a complex of the corre-
sponding oligolysine and ON had any significant effect on bcl-2
levels, underlining the importance of covalent attachment of the
basic peptide to the phosphodiester ON for these conjugates.
Treatment of cells with conjugate 14 (PEG5 ligand) resulted in
statistically significant down regulation of Bcl-2 expression
(Fig. 4B) at both 100 and 1000 nM concentrations. Co-
administration of uncoupled PEG5 ligand and the ON at resulted
in less efficient, but also significant down-regulation. This effect
might be explained by a certain destabilizing impact of the
amphiphilic ethylene-glycol chain on the cell membrane, which is
also present when co-administrating the two components.
For toxicity determination, 607B cells treated with the conju-
gates were assayed for their impact on cell proliferation and in a
caspase cleavage assay (Fig. 5). No reduction in cellular viability or
apoptosis induction was detected for any of the compounds 10e14
in concentrations as high as 1
mM. In contrast, an isosequential

M. Dirin et al. / European Journal of Medicinal Chemistry 121 (2016) 132e142
137
Fig. 5. Cell viability (A) and caspase activity (B) of oligolysine-oligonucleotide disulfide
conjugates 10e14 (A) and oligoethylene-oligonucleotide conjugate 14 (B), after
transfection-free application on 607B cell line after 24 h incubation. Results are re-
ported as mean values with n ¼ 3 s.d.
Fig. 4. Bcl-2 expression level of oligolysine-oligonucleotide disulfide conjugates 10e13
(A) [29] and PEG5-oligonucleotide conjugate 14 (B), after naked application on 607B
cell line after 72 h incubation followed by protein extraction and western blotting.
Results are reported as mean values with n ¼ 3 s.d.
bioconjugates enables the coupling of a large range of ligands to
oligonucleotides with quick and facile purification. The represen-
tative conjugates that were synthesized demonstrate the feasibility
of conjugation of both highly charged and amphiphillic ligands and
prove the sufficient stability during the modified cleavage and
deprotection procedure. The prepared phosphodiester oligonucle-
otide conjugates show promising in vitro target down regulation in
the absence of any transfection enhancing agent while being non-
toxic.
phosphorothioate oligonucleotide resulted in concentration-
dependent inhibition of viability and an increase in caspase acti-
vation at both 100 nM and 1 mM concentrations. The documented
intrinsic toxicity of phosphorothioate oligonucleotides results in
inhibition of cellular proliferation which is independent on down
regulation of the anti-apoptotic bcl-2 protein [47,48]. Although bcl-
2 plays a pivotal role in apoptosis regulation, its depletion by gene
silencing is not sufficient for increased apoptosis rates without an
additional trigger through chemo- or radiotherapy [49,50].
3. Conclusion
4. Materials and methods
The reported on-bead method for assembly of disulfide linked
Synthesis reagents and solvents were obtained from

138
M. Dirin et al. / European Journal of Medicinal Chemistry 121 (2016) 132e142
SigmaeAldrich (St. Louis, MO, USA); Bachem (Bubendorf,
Switzerland), or Fisher Scientific (Waltham, MA, USA) and were
used without further purification. Dimethylformamide (Fisher) was
dried by refluxing over calcium hydride followed by vacuum
distillation. Peptide synthesis reagents and amino acids were pur-
chased from Bachem and SigmaeAldrich. Oligonucleotide synthe-
sis reagents, solvents and nucleoside building blocks were supplied
by Proligo (SAFC, SigmaeAldrich). Thin-layer chromatography was
carried out on Merck 60 F254 aluminum-coated silica gel plates.
Product spots were visualized by UV shadowing (254 nm).
For column chromatography, Acros 0.035e0.070 mm, 60 Å silica
gel was used. NMR spectra were recorded on a Varian 500 MHz or
on a Bruker 200 MHz spectrometer. High-resolution mass spectra
were collected on a Bruker Daltonics microTOF-Q II mass spec-
trometer with an ESI source with an accuracy of 5 ppm. HPLC was
performed on a Shimadzu LC-20 AD system equipped with a UV/VIS
detector (260 nm) or a Merck Hitachi D-7000 with the UV/VIS
detector set at 220 nm. CD spectra were recorded on a Jasco J-810
spectropolarimeter. Gel electrophoresis and blotting reagents and
buffers were supplied by Serva (Heidelberg, Germany), and blots
were digitalized using a Molecular Imager ChemiDoc™ XRS System
and processed with Quantity One Software (both Bio-Rad Labora-
tories, Carlsbad, CA, USA). Absorbance and fluorescence measure-
ments of apoptosis and viability assays were measured on a TECAN
82.4 (C-20), 70.7 (20-OCH2), 69.9 (C-30), 61.6 (C-50), 36.2 (i-prCH),
29.8 (BrCH2), 26.0 (tert-BuCH3), 25.7 (tert-BuCH3), 19.2 (i-prCH3),
18.4 (tert-BuC), 18.1 (tert-BuC), - 4.9 (SiCH3), - 5.4 (SiCH3).
ESI-HRMS m/z [MþH]þ: calculated for C28H51BrN5O5Si2
672.2606, found 672.2632.
4.1.3. 30-,50-Bis(tert-butyldimethylsilyl)-N6-isobutyryl-adenosine-
20-O-(ethyl)-methanesulfonothioate (6)
To a solution of 5 (600 mg, 0.89 mmol) in 15 mL anhydrous
dimethylformamide sodium methanethiosulfonate (0.131 g,
0.98 mmol) was added in an argon atmosphere. The reaction
mixture was heated at 70 ꢁC until thin-layer chromatography
revealed completion of the reaction (16 h).
The solvent was then removed under vacuum and the residue
was taken up in ethyl acetate, washed with water and brine and
dried over sodium sulphate. The organic phase was evaporated to
obtain 0.60 g (96%) of pale yellow crystals as the desired product
which was used without further purification.
4.1.4. Rf 0.29 (tert-butyl methyl ether)
1H NMR (500 MHz, CDCl3):
d
¼ 8.70 (s, 1H, H-2), 8.53 (br s, 1H,
NH), 8.37 (s, 1H, H-8), 6.18 (d, J ¼ 3.2 Hz, 1H, H-10), 4.52 (t, J ¼ 5.1 Hz,
1H, H-30), 4.32 (dd, J ¼ 3.2, 2.9 Hz, 1H, H-20), 4.14 (dt, J ¼ 2.9, 2.5 Hz,
1H, H-40), 4.01 (dd, J ¼ 11.7, 2.8 Hz, 1H, H-50/1), 3.95 (t, J ¼ 5.7 Hz,
2H, OCH2), 3.78 (dd, J ¼ 11.7, 2.8 Hz, 1H, H-50/2), 3.37 (dd, J ¼ 9.5,
5.4 Hz, 2H, SCH2), 3.34 (s, 3H, -SO2CH3), 3.21 (septet, J ¼ 6.5 Hz, 1H,
i-prCH), 1.31 (d, J ¼ 7.0 Hz, 6H, i-prCH3), 0.93 (s, 9H, tert-BuCH3),
0.92 (s, 9H, tert-BuCH3), 0.11 (s, 12H, SiCH3).
€
Infinite M200 Pro (Tecan, Grodig, Austria).
4.1. Synthesis of the modified nucleoside
4.1.1. 30-,50-Bis(tert-butyldimethylsilyl)-20-O-(2-bromo-ethyl)-N6-
isobutyryl-adenosine (5)
13C NMR (125 MHz, CDCl3):
d
¼ 175.9 (CON), 152.6 (C-2), 150.7
(C-4), 149.3 (C-6), 141.3 (C-8), 122.4 (C-5), 87.0 (C-10), 84.8 (C-40),
82.8 (C-20), 69.6 (C-30), 69.5 (20-OCH2), 61.5 (C-50), 50.6 (-SO2CH3),
36.2 (SCH2), 36.2 (i-prCH), 26.0 (tert-BuCH3), 25.7 (tert-BuCH3),
19.2 (i-prCH3), 18.5 (tert-BuC), 18.1 (tert-BuC), - 4.5 (SiCH3), - 5.4
(SiCH3).
A stirred solution of 4 (2.10 g, 3.44 mmol) [29] in 11 mL anhy-
drous dichloromethane and 4 mL triethylamine was cooled to 0 ꢁC
in an argon atmosphere. Methanesulfonyl chloride (0.43 g,
3.80 mmol) was diluted in 4 mL dichloromethane and added
dropwise via a syringe. The reaction mixture was stirred at room
temperature for 2 h and subsequently evaporated to dryness. The
orange oily residue was taken up in ethyl acetate and washed with
1 M sodium bicarbonate solution, water and brine and dried over
sodium sulphate followed by evaporation of the solvent under
vacuum.
The residue was then dissolved in 15 mL anhydrous tetrahy-
drofuran and mixed with a solution of ultra-dry lithium bromide
(0.90 g, 10.32 mmol) in anhydrous THF in an argon atmosphere.
After heating the reaction mixture at 60 ꢁC for 6 h, thin-layer
chromatography showed complete conversion of the starting ma-
terial. Afterwards, the solvent was removed and the pale yellow
residue was redissolved in ethyl acetate.
ESI-HRMS m/z [MþH]þ: calculated for C29H54N5O7S2Si2
704.2998, found 704.2935.
4.1.5. 30-,50-Bis(tert-butyldimethylsilyl)-20-O-(2-tert-
butyldisulfanyl-ethyl)-N6-isobutyryl-adenosine (7)
2-Methyl-2-propanethiol (0.092 g, 1.02 mmol) was added
dropwise to a cooled (ice bath) solution of 6 (0.60 g, 0.85 mmol) in
5 mL dichloromethane containing 2 equiv of (0.24 mL) triethyl-
amine. After stirring for 6 h (TLC) at room temperature, the solvents
were removed under vacuum. The residue was taken up in ethyl
acetate and washed with water and brine followed by drying over
sodium sulphate. The solution was thereafter evaporated and the
residue was chromatographed on a silica gel column. The title
compound was eluted with a 0e50% gradient of tert-butyl methyl
ether in petroleum ether to obtain 0.57 g (94%) of a white powder.
The organic phase was washed with water and brine, dried over
sodium sulphate and chromatographed on a silica gel column
eluted with a gradient of 0e70% tert-butyl methyl ether in petro-
leum ether. The corresponding fractions were pooled and evapo-
rated under vacuum to produce 2.11 g (91%) of a pale yellow
powder as the product.
4.1.6. Rf 0.35 (1:1 tert-butyl methyl ether/petroleum ether)
1H NMR (500 MHz, CDCl3):
d
¼ 8.72 (s, 1H, H-2), 8.51 (br s, 1H,
NH), 8.34 (s, 1H, H-8), 6.17 (d, J ¼ 4.1 Hz,1H, H-10), 4.50 (dd, J ¼ 5.1,
4.7 Hz, 1H, H-30), 4.39 (t, J ¼ 4.4 Hz, 1H, H-20), 4.13 (ddd, J ¼ 7.9, 6.9,
3.2 Hz, 1H, H-40), 4.00 (dd, J ¼ 11.4, 3.8 Hz, 1H, H-50/1), 3.84e3.76
(m, 3H, OCH2 and H-50/2), 3.21 (septet, J ¼ 6.5 Hz, 1H, i-prCH), 3.37
(dd, J ¼ 7.0, 6.6 Hz, 2H, SCH2),1.31 (d, J ¼ 6.6 Hz, 6H, i-prCH3),1.28 (s,
9H, S-tertBuCH3), 0.93 (s, 9H, Si-tert-BuCH3), 0.92 (s, 9H, Si-tert-
BuCH3), 0.11 (s, 12H, SiCH3), 0.12 (s, 3H, SiCH3), 0.10 (s, 9H, SiCH3).
4.1.2. Rf 0.35 (1:20 MeOH/CH2Cl2)
1H NMR (500 MHz, CDCl3):
d
¼ 8.72 (s, 1H, H-2), 8.60 (br s, 1H,
NH), 8.36 (s, 1H, H-8), 6.19 (d, J ¼ 3.8 Hz,1H, H-10), 4.51 (dd, J ¼ 5.35
and 3.8, 1H, H-30), 4.41 (dd, J ¼ 4.9, 3.8 Hz, 1H, H-20), 4.14 (dt, J ¼ 5.7,
2.8 Hz,1H, H-40), 4.00 (dd, J ¼ 11.7, 3.5 Hz,1H, H-50/1), 3.97e3.87 (m,
2H, OCH2), 3.78 (dd, J ¼ 11.7, 2.9 Hz, 1H, H-50/2), 3.44 (t, J ¼ 6.3 Hz,
1H, BrCH2), 3.19 (septet, J ¼ 6.0 Hz, 1H, i-prCH), 1.30 (d, J ¼ 6.9 Hz,
6H, i-prCH3), 0.92 (s, 18H, tert-BuCH3), 0.11 (s, 3H, SiCH3), 0.10 (s,
9H, SiCH3).
13C NMR (125 MHz, CDCl3):
d
¼ 175.9 (CON), 152.6 (C-2), 150.9
(C-4), 149.2 (C-6), 141.6 (C-8), 122.5 (C-5), 87.1 (C-10), 84.2 (C-40),
82.2 (C-20), 69.9 (C-30), 69.5 (20-OCH2), 61.8 (C-50), 47.9 (S-tertBuC),
39.7 (SCH2), 36.2 (i-prCH), 29.8 (S-tertBuCH3), 26.0 (Si-tert-BuCH3),
25.7 (Si-tert-BuCH3), 19.2 (i-prCH3), 18.4 (Si-tert-BuC), 18.1 (Si-tert-
BuC), -4.5 (SiCH3), -4.8 (SiCH3), -5.40 (SiCH3), -5.42 (SiCH3).
13C NMR (125 MHz, CDCl3):
(C-4), 149.3 (C-6), 141.5 (C-8), 122.5 (C-5), 87.2 (C-10), 85.0 (C-40),
d
¼ 176.0 (CON), 152.6 (C-2), 150.8

M. Dirin et al. / European Journal of Medicinal Chemistry 121 (2016) 132e142
139
ESI-HRMS m/z [MþH]þ: calculated for C32H60N5O5S2Si2
81.9 (C-20), 69.8 (C-30), 69.3 (20-OCH2), 62.8 (C-50), 55.2 (OMe), 48.2
(S-tertBuC), 39.9 (SCH2), 36.2 (i-prCH), 29.8 (S-tertBuCH3), 19.2 (i-
prCH3).
714.3569, found 714.3601.
4.1.7. 20-O-(2-Tert-butyldisulfanyl-ethyl)-N6-isobutyryl-adenosine
(8)
ESI-HRMS m/z [MþNa]þ: calculated for C41H49N5NaO7S2
810.2966, found 810.2970.
To a solution of 7 (0.25 g, 0.35 mmol) in 2 mL tetrahydrofuran
triethylamine trihydrofluoride 97% (0.23 g, 1.40 mmol) was added
dropwise and the mixture was stirred at room temperature for 4 h
until thin-layer chromatography showed complete deprotection of
the starting material.
The solvent was then evaporated and the residue was taken up
in ethyl acetate, washed with water and brine and dried over so-
dium sulphate. The organic phase was removed to produce 0.170 g
(98%) of the desired product as a pale yellow foam which was used
without further purification.
4.1.11. 20-O-(2-Tert-butyldisulfanyl-ethyl)-3’-succinyl-50-O-
dimethoxytrityl-N6-isobutyryl-adenosine (10)
Compound 9 (0.060 g, 0.076 mmol) was dissolved in 2 mL
anhydrous dichloromethane containing 3 equiv (0.032 mL) trie-
thylamine in an argon atmosphere. The reaction mixture was
cooled in an ice bath before succinic anhydride (0.015 g, 0.15 mmol)
was added.
The reaction mixture was stirred for 1 h at 0 ꢁC and allowed to
come slowly to room temperature. After 2 h, thin-layer chroma-
tography revealed complete conversion of the starting material.
Solvents were removed under vacuum, the residue redissolved in
ethyl acetate and washed with saturated bicarbonate solution,
water, brine and dried over sodium sulphate. The solution was then
evaporated to dryness under vacuum and the residue was recrys-
tallized from tert-butyl methyl ether/n-hexane to produce 0.066 g
(98%) of the title compound as a pale yellow foam.
4.1.8. Rf 0.31 (1:15 MeOH/CH2Cl2)
1H NMR (500 MHz, CDCl3):
d
¼ 8.70 (s, 1H, H-2), 8.61 (br s, 1H,
NH), 8.05 (s, 1H, H-8), 6.13 (d, J ¼ 11.7 Hz, 1H, OH-50), 5.94 (d,
J ¼ 7.6 Hz, 1H, H-10), 4.83 (dd, J ¼ 7.6, 4.4 Hz, 1H, H-20), 4.58 (d,
J ¼ 4.4, 1H, H-30), 4.38 (s, 1H, H-40), 3.97 (d, J ¼ 13.00 Hz, 1H, H-50/1),
3.79 (d, J ¼ 12.00 Hz, 1H, H-50/2), 3.77e3.67 (m, 2H, OCH2), 3.24
(septet, J ¼ 6.00 Hz, 1H, i-prCH), 3.14 (s, 1H, OH-50), 2.76e2.68 (m,
2H, SCH2), 1.31 (d, J ¼ 7.00 Hz, 6H, i-prCH3), 1.28 (s, 9H, S-
tertBuCH3).
4.1.12. Rf 0.17(1:20 MeOH/CH2Cl2)
1H NMR (500 MHz, CDCl3):
d
¼ 8.90 (br s, 1H, NH), 8.63 (s, 1H, H-
13C NMR (125 MHz, CDCl3):
d
¼ 176.0 (CON), 152.0 (C-2), 150.0
2), 8.16 (s, 1H, H-8), 7.40 (d, J ¼ 7.60 Hz, 2H, Ph-2,6), 7.30 (d,
J ¼ 8.8 Hz, 4H, Ar-2,6), 7.27e7.16 (m, 3H, Ph-3,4,5), 6.81 (d,
J ¼ 7.60 Hz, 4H, Ar-3,5), 6.12 (d, J ¼ 6 Hz,1H, H-10), 5.47 (dd, J ¼ 4.8,
3.5, 1H, H-30), 4.93 (t, J ¼ 5.7 Hz, 1H, H-20), 4.36 (dt, J ¼ 6.9, 3.5 Hz,
1H, H-40), 3.78 (s, 3H, OMe), 3.84e3.67 (m, 2H, OCH2), 3.53 (dd,
J ¼ 10.7, 3.5 Hz, 1H, H-50/1), 3.41 (dd, J ¼ 10.8, 3.8 Hz, 1H, H-50/2),
3.14 (septet, J ¼ 6.7 Hz, 1H, i-prCH), 2.81e2.66 (m, 6H, succin-CH2
and SCH2), 1.299 (d, J ¼ 7.0 Hz, 3H, i-prCH3), 1.295 (d, J ¼ 6.9 Hz, 3H,
i-prCH3), 1.23 (s, 9H, S-tertBuCH3).
(C-4), 150.1 (C-6), 143.1 (C-8), 123.7 (C-5), 89.4 (C-10), 88.0 (C-40),
81.3 (C-20), 70.9 (C-30), 68.7 (20-OCH2), 63.3 (C-50), 48.3 (S-tertBuC),
39.8 (SCH2), 36.2 (i-prCH), 29.8 (S-tertBuCH3), 19.2 (i-prCH3), 19.1
(i-prCH3).
ESI-HRMS m/z [MþH]þ: calculated for C20H32N5O5S2 486.1839,
found 486.1843.
4.1.9. 20-O-(2-Tert-butyldisulfanyl-ethyl)-50-O-dimethoxytrityl-N6-
isobutyryl-adenosine (9)
13C NMR (125 MHz, CDCl3):
d
¼ 176.0 (CON), 171.3 (COO), 158.6
4,40-Dimethoxytrityl chloride (0.058 g, 0.17 mmol) was added to
a cooled solution of 8 (0.075 g, 0.15 mmol) in 1 mL of anhydrous
dichloromethane containing 3 equiv triethylamine (0.065 mL). The
reaction mixture was allowed to warm up to room temperature and
stirred for 16 h until thin-layer chromatography showed comple-
tion of the reaction.
(Ar-4), 152.7 (C-2), 151.3 (C-4), 149.5 (C-6), 144.2 (Ph-1), 141.2 (C-8),
135.4 (Ar-1), 135.4 (Ar-1), 130.1 (Ar-2,6), 128.1 (Ph-2,6), 128.0 (Ph-
3,5), 127.1 (Ph-4), 122.4 (C-5), 113.3 (Ar-3,5), 86.9 (trityl-C), 86.7 (C-
10), 82.2 (C-40), 80.1 (C-20), 71.5 (C-30), 69.7 (20-OCH2), 62.9 (C-50),
55.2 (OMe), 48.9 (S-tertBuC), 39.8 (SCH2), 36.1 (i-prCH), 29.8 (S-
tertBuCH3), 29.1 (succin-CH2), 19.2 (i-prCH3).
After quenching the reaction for 0.5 h in the presence of 0.5 mL
of methanol, the solvents were removed under vacuum and the
residue was redissolved in ethyl acetate, washed with saturated
sodium bicarbonate solution, water, brine and dried over sodium
sulphate. The organic phase was then evaporated and the resulting
yellow foam was recrystallized from tert-butyl methyl ether/n-
hexane to give 0.098 g (83%) of the title compound as a pale yellow
foam.
ESI-HRMS m/z [MþNa]þ: calculated for C45H53N5NaO10S2
910.3126, found 910.3126.
4.2. Loading of the 20-modified nucleoside 10 onto the LCAA-CPG
To a solution of 10 (20.90 mg, 26.76
thylformamide (0.5 mL) 1-hydroxybenzotriazole (HOBt, 4.05 mg,
29.96 mol) and N,N-diisopropylethylamine (DIPEA, 11.1 mg,
85.6 mol) were added. The mixture was then transferred to a re-
mmol) in anhydrous dime-
m
m
4.1.10. Rf 0.27(1:40 MeOH/CH2Cl2)
action vessel containing LCAA-CPG (200 mg) followed by addition
of benzotriazol-1-yloxy)tris(dimethylamino)phosphonium hexa-
fluorophosphate (BOP, 13.25 mg, 29.96 mmol). The slurry was
1H NMR (500 MHz, CDCl3):
d
¼ 8.65 (s, 1H, H-2), 8.46 (br s, 1H,
NH), 8.20 (s, 1H, H-8), 7.43 (d, J ¼ 7.60 Hz, 2H, Ph-2,6), 7.32 (dd,
J ¼ 8.9, 1.6 Hz, 4H, Ar-2,6), 7.29e7.22 (m, 3H, Ph-3,4,5), 6.81 (d,
J ¼ 8.60 Hz, 4H, Ar-3,5), 6.18 (d, J ¼ 3.8 Hz,1H, H-10), 4.63 (dd, J ¼ 4.7,
4.1 Hz, 1H, H-20), 4.52 (dd, J ¼ 11.1, 5.4, 1H, H-30), 4.24 (dt, J ¼ 8.5,
4.1 Hz, 1H, H-40), 4.05 (ddd, J ¼ 10.4, 7.3, 5.4 Hz, 1H, OCH2/1),
3.89e3.87 (m, 1H, OCH2/1), 3.79 (s, 3H, OMe), 3.53 (dd, J ¼ 10.70,
3.2 Hz, 1H, H-50/1), 3.43 (dd, J ¼ 10.8, 3.20 Hz, 1H, H-50/2), 3.17
(septet, J ¼ 6.81 Hz, 1H, i-prCH), 2.90e2.86 (m, 2H, SCH2), 1.32 (d,
J ¼ 7.00 Hz, 6H, i-prCH3), 1.28 (s, 9H, S-tertBuCH3).
shaken gently at room temperature in an argon atmosphere. After
16 h, the resin was filtered off and washed three times with
dimethylformamide, methanol and dichloromethane, successively,
and was air-dried.
The free amino groups were subsequently capped by addition of
a mixture of acetic anhydride/pyridine/1-methylimidazole/anhy-
drous THF (1/1/1.5/15) the resin under an argon atmosphere. After
2 h, the resin was washed three times with acetonitrile, methanol
and dichloromethane and air-dried for a short time before being
completely dried in an exsiccator overnight. Loading was deter-
mined by adding 3% trichloroacetic acid in dichloromethane and
determination of the absorption at 498 nm of the supernatant and
13C NMR (125 MHz, CDCl3):
(C-2), 150.9 (C-4), 149.3 (C-6), 144.4 (Ph-1), 141.4 (C-8), 135.6 (Ar-1),
135.6 (Ar-1), 130.1 (Ar-2,6), 128.2 (Ph-2,6), 127.9 (Ph-3,5), 127.0 (Ph-
4), 122.6 (C-5), 113.2 (Ar-3,5), 87.1 (C-10), 86.6 (trityl-C), 84.1 (C-40),
d
¼ 175.8 (CON), 158.6 (Ar-4), 152.6

140
M. Dirin et al. / European Journal of Medicinal Chemistry 121 (2016) 132e142
was estimated to be 51.6
m
mol gꢀ1
.
loading buffer (95% formamide). As tracer, a mixture of bromo-
phenol blue and xylene cyanol was used, and the gel was run in 1x
TBE buffer until the dyes reached about one third and two thirds,
respectively, of the migration distance of the gel. The gel was
stained in methylene blue solution (0.002% in TBE), destained in
water, and digitalized in a GS-710 densitometric scanner (Bio-Rad).
4.3. Oligonucleotide synthesis, bioconjugation, and analysis
Oligodeoxynucleotides were synthesized with phosphodiester
bonds on a Polygen 10 column DNA synthesizer using standard
phosphoramidite chemistry protocols using the DMT-off mode.
After completed synthesis, the desired amount of the resin-
oligonucleotide was transferred to a micro test tube for the
consecutive deprotection and coupling steps.
4.4. In vitro gene silencing activity
Gene silencing assays were performed as described previously
[29]. In short, 607B cells were seeded in 24-well tissue culture
plates (Greiner, bio-one, Kremsmünster, Austria). After 24 h, the
cells were incubated with the indicated concentrations of oligo-
nucleotides or conjugates, and after 72 h incubation, the culture
medium was then removed and the cells were lysed at 4 ꢁC. Ali-
The loaded oligonucleotide-LCCA-CPG (10 mg) was slurried in
tetrahydrofuran (200
mL) followed by addition of tri-N-butylphos-
phine (4 equiv). After 5 min, one
mL doubly-distilled water was
added and the reaction mixture was agitated gently for 2 h at 25 ꢁC.
Subsequently, the resin was filtered off and washed three times
with tetrahydrofuran, methanol and dichloromethane and dried in
an evacuated exsiccator overnight. The assessment of deprotection
and the optimization of the reaction were carried out employing
the Ellman test.
quots of cell lysates containing 20 mg of total protein were separated
on a 15% SDS-polyacrylamide gel and blotted onto a PVDF mem-
brane. The blots were developed with the primary antibodies
directed either against actin (SigmaeAldrich, 1:5000) and Bcl-2
(Zymed, 1:500) and with HRP-coupled secondary antibodies in
TBST (anti-rabbit for actin, anti-mouse for Bcl-2, 1:100,000 dilu-
tion). Bands were detected by chemiluminescence using the
ImmunStar Western C Kit (Bio-Rad) and a Chemi-Doc gel imager
(Bio-Rad). The protein bands were then analyzed utilizing the
Quantity One™ (version 4.6.3) software package (Bio-Rad).
The LCAA-CPG from the previous step (deprotection) was sus-
pended in a mixture of PEG5 (3 equiv) in 200
mL anhydrous
acetonitrile followed by the addition of DIPEA (6 equiv). The reac-
tion mixture was agitated gently at room temperature for 4 h and
the resin was filtered off, washed three times with acetonitrile and
dichloromethane and air-dried. Optimization of reaction conditions
and reaction time was carried out by performing the Ellman test.
The LCAA-CPG-oligonucleotide-SH was suspended in a mixture
of 2:1 ACN/TriseHCl (pH 7.5) buffer followed by addition of the
peptides [29]. The reaction mixture was agitated gently at room
temperature overnight and the resin was filtered off, washed three
times with 2:1 ACN/H2O, ACN and dichloromethane and air-dried.
Optimization of reaction conditions and reaction time was carried
out by the Ellman test.
The resin was suspended in a screw-capped vial in concentrated
methanolic ammonia solution (7 N) and heated for 16 h at 55 ꢁC to
afford cleavage from the solid support and removal of the pro-
tecting groups. After cooling down to room temperature and
removing the excess of ammonia, the solvent was evaporated on a
SpeedVac. The residue was then redissolved in a small volume
4.5. Cell viability assay
Cell viability and caspase activites were tested as described [29].
In short, 607B cells (2 ꢂ 104 cells per well) were seeded in trans-
parent (viability) or black (caspase activity) 96-well microplates
(Greiner bio-one). After 24 h, oligonucleotides and conjugates were
added at the indicated concentrations and the plate was incubated
for another 24 h. Subsequently, the EZ4U proliferation assay kit
(Biomedica) was used according to the manufacturer's instructions
to quantify cell viability/proliferation. The absorbance at 465 nm
was recorded with 620 nm as reference wavelength. For the cas-
pase activity readout, lysis buffer was added and the plates were
incubated 20 min on ice. Assay buffer containing the peptide sub-
strate Ac-DEVD-AMC was added and fluorescence was recorded at
room temperature (excitation: 360 nm; emission: 460 nm; slit
width: 5 nm) until optimal signal to noise ratios were reached. Cell
viability and apoptosis rates are presented in relation to untreated
cells.
(50e100 mL) of doubly distilled water and precipitated from iso-
propanol (1 mL). After overnight incubation at ꢀ20 ꢁC, the sus-
pension was centrifuged at 16,000 g for 15 min at 4 ꢁC. The
supernatant was removed and the pellet was air-dried. After
redissolving in water and dilution to 100 mM stock solutions, the
purity of the oligonucleotides was checked by HPLC. If required, the
raw mixture was passed through a column of Sephadex™ G-25 for
desalting. Oligonucleotide conjugates with a purity of at least 90%
(as integrated from HPLC traces) were used without further
purifications.
The overall efficiency of the process was assessed by HPL
chromatography. Analytical HPLC of the oligolysine conjugates was
carried out according to method A using a Thermo Scientific BDS
Hypersil C18 column (150 ꢂ 4 mm) and 260 nm UV detection on a
Shimadzu (LC-20 AD) HPLC instrument. A gradient of 28e33%
solvent A (A: methanol; B: 50 mM triethylammonium acetate
buffer, pH 7) was run in 30 min at a flow rate of 1 mL minꢀ1 at 25 ꢁC.
For the OEG5-conjugate method B was applied on a Shimadzu
(LC-20 AD) HPLC instrument equipped with a UV/VIS detector
4.6. Statistical analysis
Statistical significance of differences was determined by using
the student's t-test (SigmaPlot 12.5). P values < 0.05 were consid-
ered to be significant.
Disclosure
A patent application comprising some of the described meth-
odology and structures has been filed.
Acknowledgment
€
(260 nm) using
a
Phenomenex Synergi 4u Fusion-RP 80A
We thank Martina Koberl and Sandra Haselmaier-Auernigg for
(150 ꢂ 4.60 mm; 4
m
) column running a gradient of 15e70% solvent
support with mass spectrometric analyses.
A (A: methanol; B: 50 mM triethylammonium acetate, pH 7) in
35 min at a flow rate of 1 mL minꢀ1 at 25 ꢁC.
Appendix A. Supplementary data
For gel electrophoresis, a urea polyacrylamide gel (20% acryl-
amide,1x Tris-borate buffer TBE) was prepared and run in a Bio-Rad
Mini Protean II apparatus. Samples were mixed 1:10 in formamide
Supplementary data associated with this article can be found in
theonlineversion, athttp://dx.doi.org/10.1016/j.ejmech.2016.05.001.

M. Dirin et al. / European Journal of Medicinal Chemistry 121 (2016) 132e142
141
M.E. Østergaard, M.T. Migawa, E.E. Swayze, P.P. Seth, Solid-phase synthesis of
50-triantennary N-acetylgalactosamine conjugated antisense oligonucleotides
using phosphoramidite chemistry, Bioorg. Med. Chem. Lett. 25 (2015)
4127e4130, http://dx.doi.org/10.1016/j.bmcl.2015.08.019.
These data include MOL files and InChiKeys of the most important
compounds described in this article.
[21] T.P. Prakash, J. Yu, M.T. Migawa, G.A. Kinberger, W.B. Wan, M.E. Østergaard,
R.L. Carty, G. Vasquez, A. Low, A. Chappell, K. Schmidt, M. Aghajan, J. Crosby,
H.M. Murray, S.L. Booten, J. Hsiao, A. Soriano, T. Machemer, P. Cauntay,
S.A. Burel, S.F. Murray, H.J. Gaus, M.J. Graham, E.E. Swayze, P.P. Seth,
Comprehensive structure activity relationship of triantennary N-Acetylga-
lactosamine conjugated antisense oligonucleotides for targeted delivery to
hepatocytes, J. Med. Chem. 59 (2016) 2718e2733, http://dx.doi.org/10.1021/
acs.jmedchem.5b01948.
[22] R.L. Juliano, Peptide-oligonucleotide conjugates for the delivery of antisense
and siRNA, Curr. Opin. Mol. Ther. 7 (2005) 132e136.
[23] A. Astriab-Fisher, D.S. Sergueev, M. Fisher, B. Ramsay Shaw, R.L. Juliano,
Antisense inhibition of P-glycoprotein expression using peptide-
eoligonucleotide conjugates, Biochem. Pharmacol. 60 (2000) 83e90, http://
dx.doi.org/10.1016/S0006-2952(00)00310-5.
[24] K.T. Hastings, P. Cresswell, Disulfide reduction in the endocytic pathway:
immunological functions of gamma-interferon-inducible lysosomal thiol
reductase, Antioxid. Redox Signal. 15 (2011) 657e668, http://dx.doi.org/
10.1089/ars.2010.3684.
[25] N. Venkatesan, B.H. Kim, Peptide conjugates of oligonucleotides: synthesis
and applications, Chem. Rev. 106 (2006) 3712e3761, http://dx.doi.org/
10.1021/cr0502448.
[26] C.D. Austin, X. Wen, L. Gazzard, C. Nelson, R.H. Scheller, S.J. Scales, Oxidizing
potential of endosomes and lysosomes limits intracellular cleavage of
disulfide-based antibodyedrug conjugates, Proc. Natl. Acad. Sci. U. S. A. 102
(2005) 17987e17992, http://dx.doi.org/10.1073/pnas.0509035102.
[27] Z.u. Rehman, D. Hoekstra, I.S. Zuhorn, Mechanism of polyplex- and Lipoplex-
mediated delivery of nucleic acids: real-time visualization of transient
References
[1] T. Yamamoto, M. Nakatani, K. Narukawa, S. Obika, Antisense drug discovery
and development, Future Med. Chem. 3 (2011) 339e365, http://dx.doi.org/
10.4155/fmc.11.2.
[2] D. He, E. Wagner, Defined polymeric materials for gene delivery, Macromol.
Biosci. 15 (2015) 600e612, http://dx.doi.org/10.1002/mabi.201400524.
[3] K. Jain, P. Kesharwani, U. Gupta, N.K. Jain, Dendrimer toxicity: let's meet the
challenge, Int. J. Pharm. 394 (2010) 122e142, http://dx.doi.org/10.1016/
j.ijpharm.2010.04.027.
[4] C. Lorenzer, M. Dirin, A.-M. Winkler, V. Baumann, J. Winkler, Going beyond the
liver: progress and challenges of targeted delivery of siRNA therapeutics,
J. Control. Release 203 (2015) 1e15, http://dx.doi.org/10.1016/
j.jconrel.2015.02.003.
[5] R.L. Juliano, K. Carver, C. Cao, X. Ming, Receptors, endocytosis, and trafficking:
the biological basis of targeted delivery of antisense and siRNA oligonucleo-
tides, J. Drug Target. 21 (2013) 27e43, http://dx.doi.org/10.3109/
1061186X.2012.740674.
[6] T. Coelho, D. Adams, A. Silva, P. Lozeron, P.N. Hawkins, T. Mant, J. Perez,
J. Chiesa, S. Warrington, E. Tranter, M. Munisamy, R. Falzone, J. Harrop,
J. Cehelsky, B.R. Bettencourt, M. Geissler, J.S. Butler, A. Sehgal, R.E. Meyers,
Q. Chen, T. Borland, R.M. Hutabarat, V.A. Clausen, R. Alvarez, K. Fitzgerald,
C. Gamba-Vitalo, S.V. Nochur, A.K. Vaishnaw, D.W.Y. Sah, J.A. Gollob, O.B. Suhr,
Safety and efficacy of RNAi therapy for transthyretin amyloidosis, N. Engl. J.
Med. 369 (2013) 819e829, http://dx.doi.org/10.1056/NEJMoa1208760.
[7] J.K. Nair, J.L.S. Willoughby, A. Chan, K. Charisse, M.R. Alam, Q. Wang,
M. Hoekstra, P. Kandasamy, A.V. Kel'in, S. Milstein, N. Taneja, J. O'Shea,
S. Shaikh, L. Zhang, R.J. van der Sluis, M.E. Jung, A. Akinc, R. Hutabarat,
S. Kuchimanchi, K. Fitzgerald, T. Zimmermann, T.J.C. van Berkel, M.A. Maier,
K.G. Rajeev, M. Manoharan, Multivalent N-acetylgalactosamine-conjugated
siRNA localizes in hepatocytes and elicits robust RNAi-mediated gene
silencing, J. Am. Chem. Soc. 136 (2014) 16958e16961, http://dx.doi.org/
10.1021/ja505986a.
[8] K.G. Rajeev, J.K. Nair, M. Jayaraman, K. Charisse, N. Taneja, J. O'Shea,
J.L.S. Willoughby, K. Yucius, T. Nguyen, S. Shulga-Morskaya, S. Milstein,
A. Liebow, W. Querbes, A. Borodovsky, K. Fitzgerald, M.A. Maier,
M. Manoharan, Hepatocyte-specific delivery of siRNAs conjugated to novel
non-nucleosidic trivalent N-acetylgalactosamine elicits robust gene silencing
in vivo, ChemBioChem 16 (2015) 903e908, http://dx.doi.org/10.1002/
cbic.201500023.
membrane destabilization without endosomal lysis, ACS Nano
7 (2013)
3767e3777, http://dx.doi.org/10.1021/nn3049494.
[28] J. Winkler, Therapeutic oligonucleotides with polyethylene glycol modifica-
tions, Future Med. Chem. 7 (2015) 1721e1731, http://dx.doi.org/10.4155/
fmc.15.94.
[29] M. Dirin, E. Urban, B. Lachmann, C.R. Noe, J. Winkler, Concise postsynthetic
preparation of oligonucleotide-oligopeptide conjugates through facile disul-
fide bond formation, Future Med. Chem.
7
(2015) 1657e1673, http://
dx.doi.org/10.4155/fmc.15.109.
ꢀ
[30] F. Maurel, F. Debart, F. Cavelier, A.R. Thierry, B. Lebleu, J.-J. Vasseur, E. Vives,
Toward high yield synthesis of peptideeoligonucleotide chimera through a
disulfide bridge: a simplified method for oligonucleotide activation, Bioorg.
Med. Chem. Lett. 15 (2005) 5084e5087, http://dx.doi.org/10.1016/
j.bmcl.2005.07.086.
[31] Y. Singh, P. Murat, E. Defrancq, Recent developments in oligonucleotide
conjugation, Chem. Soc. Rev. 39 (2010) 2054e2070, http://dx.doi.org/10.1039/
B911431A.
[9] K. Lu, Q.-P. Duan, L. Ma, D.-X. Zhao, Chemical strategies for the synthesis of
peptideꢀoligonucleotide conjugates, Bioconjugate Chem. 21 (2010) 187e202,
http://dx.doi.org/10.1021/bc900158s.
[32] J. Winkler, E. Urban, C. Noe, Oligonucleotides conjugated to short lysine
chains, Bioconjugate Chem. 16 (2005) 1038e1044, http://dx.doi.org/10.1021/
bc049729d.
[33] S. Jin, C.V. Miduturu, D.C. McKinney, S.K. Silverman, Synthesis of amine- and
thiol-modified nucleoside phosphoramidites for site-specific introduction of
biophysical probes into RNA, J. Org. Chem. 70 (2005) 4284e4299, http://
dx.doi.org/10.1021/jo050061l.
[34] P.G.M. Wuts, T.W. Greene, Greene's Protective Groups in Organic Synthesis,
fourth ed., John Wiley & Sons, Inc., Hoboken, New Jersey, 2007.
[35] M.L. Immordino, F. Dosio, L. Cattel, Stealth liposomes: review of the basic
science, rationale, and clinical applications, existing and potential, Int. J.
Nanomed. 1 (2006) 297e315.
[36] P. Jarver, I. Mager, U. Langel, In vivo biodistribution and efficacy of peptide
mediated delivery, Trends Pharmacol. Sci. 31 (2010) 528e535, http://
dx.doi.org/10.1016/j.tips.2010.07.006.
[37] K. Ezzat, S.E. Andaloussi, E.M. Zaghloul, T. Lehto, S. Lindberg, P.M. Moreno,
J.R. Viola, T. Magdy, R. Abdo, P. Guterstam, R. Sillard, S.M. Hammond,
M.J. Wood, A.A. Arzumanov, M.J. Gait, C.I. Smith, M. Hallbrink, U. Langel,
PepFect 14, a novel cell-penetrating peptide for oligonucleotide delivery in
solution and as solid formulation, Nucleic Acids Res. 39 (2011) 5284e5298,
http://dx.doi.org/10.1093/nar/gkr072.
[38] Y. Ikeda, Y. Nagasaki, Impacts of PEGylation on the gene and oligonucleotide
delivery system, J. Appl. Polym. Sci. 131 (2014) 40293, http://dx.doi.org/
10.1002/app.40293.
[39] F. Iversen, C. Yang, F. Dagnæs-Hansen, D.H. Schaffert, J. Kjems, S. Gao, Opti-
mized siRNA-PEG conjugates for extended blood circulation and reduced
urine excretion in mice, Theranostics 3 (2013) 201e209, http://dx.doi.org/
10.7150/thno.5743.
[40] Z. Gaziova, V. Baumann, A.-M. Winkler, J. Winkler, Chemically defined poly-
ethylene glycol siRNA conjugates with enhanced gene silencing effect, Bioorg.
[10] H. Lonnberg, Solid-phase synthesis of oligonucleotide conjugates useful for
delivery and targeting of potential nucleic acid therapeutics, Bioconjugate
Chem. 20 (2009) 1065e1094, http://dx.doi.org/10.1021/bc800406a.
[11] J. Winkler, Oligonucleotide conjugates for therapeutic applications, Ther.
Deliv. 4 (2013) 791e809, http://dx.doi.org/10.4155/tde.13.47.
[12] H. Margus, K. Padari, M. Pooga, Cell-penetrating peptides as versatile vehicles
for oligonucleotide delivery, Mol. Ther. 20 (2012) 525e533, http://dx.doi.org/
10.1038/mt.2011.284.
[13] C.-P. Chen, L.-R. Zhang, X.-B. Peng, S.-Q. Wang, L.-H. Zhang, A concise method
for the preparation of peptide and arginine-rich peptide-conjugated antisense
oligonucleotide, Bioconjugate Chem. 14 (2003) 532e538.
[14] S.M. Gryaznov, D.H. Lloyd, Modulation of oligonucleotide duplex and triplex
stability via hydrophobic interactions, Nucleic Acids Res. 21 (1993)
5909e5915, http://dx.doi.org/10.1093/nar/21.25.5909.
€
€
€
€
€
€
[15] J. Makila, S. Jadhav, A. Kiviniemi, M. Kakela, H. Liljenback, P. Poijarvi-Virta,
€
T. Laitala-Leinonen, H. Lonnberg, A. Roivainen, P. Virta, Synthesis of multi-
galactose-conjugated 20-O-methyl oligoribonucleotides and their in vivo im-
aging with positron emission tomography, Bioorg. Med. Chem. 22 (2014)
6806e6813, http://dx.doi.org/10.1016/j.bmc.2014.10.034.
[16] J. Winkler, K. Saadat, M. Diaz-Gavilan, E. Urban, C. Noe, Oligonucleotide-
polyamine conjugates: influence of length and position of 2 0-attached poly-
amines on duplex stability and antisense effect, Eur. J. Med. Chem. 44 (2009)
670e677, http://dx.doi.org/10.1016/j.ejmech.2008.05.012.
[17] D.A. Stetsenko, M.J. Gait, Efficient conjugation of peptides to oligonucleotides
by “native ligation”, J. Org. Chem. 65 (2000) 4900e4908, http://dx.doi.org/
10.1021/jo000214z.
[18] T. Yamamoto, M. Sawamura, F. Wada, M. Harada-Shiba, S. Obika, Serial
incorporation of a monovalent GalNAc phosphoramidite unit into hepatocyte-
targeting antisense oligonucleotides, Bioorg. Med. Chem. 24 (2016) 26e32,
http://dx.doi.org/10.1016/j.bmc.2015.11.036.
[19] T.P. Prakash, M.J. Graham, J. Yu, R. Carty, A. Low, A. Chappell, K. Schmidt,
C. Zhao, M. Aghajan, H.F. Murray, S. Riney, S.L. Booten, S.F. Murray, H. Gaus,
J. Crosby, W.F. Lima, S. Guo, B.P. Monia, E.E. Swayze, P.P. Seth, Targeted de-
livery of antisense oligonucleotides to hepatocytes using triantennary N-
acetyl galactosamine improves potency 10-fold in mice, Nucleic Acids Res. 42
(2014) 8796e8807, http://dx.doi.org/10.1093/nar/gku531.
Med.
Chem.
22
(2014)
2320e2326,
http://dx.doi.org/10.1016/
j.bmc.2014.02.004.
[41] N. Shokrzadeh, A.-M. Winkler, M. Dirin, J. Winkler, Oligonucleotides conju-
gated with short chemically defined polyethylene glycol chains are efficient
antisense agents, Bioorg. Med. Chem. Lett. 24 (2014) 5758e5761, http://
dx.doi.org/10.1016/j.bmcl.2014.10.045.
[42] P. Vermette, L. Meagher, Interactions of phospholipid- and poly(ethylene
[20] T.P. Prakash, W. Brad Wan, A. Low, J. Yu, A.E. Chappell, H. Gaus, G.A. Kinberger,

142
M. Dirin et al. / European Journal of Medicinal Chemistry 121 (2016) 132e142
glycol)-modified surfaces with biological systems: relation to physico-
chemical properties and mechanisms, Colloids Surf. Biointerfaces 28
(2003) 153e198, http://dx.doi.org/10.1016/s0927-7765(02)00160-1.
[47] J. Winkler, M. Stessl, J. Amartey, C. Noe, Off-target effects related to the
B
phosphorothioate modification of nucleic acids, ChemMedChem 5 (2010)
1344e1352, http://dx.doi.org/10.1002/cmdc.201000156.
[43] M.P. Monopoli, D. Walczyk, A. Campbell, G. Elia, I. Lynch, F. Baldelli Bombelli,
K.A. Dawson, Physicalꢀchemical aspects of protein corona: relevance to
in vitro and in vivo biological impacts of nanoparticles, J. Am. Chem. Soc. 133
(2011) 2525e2534, http://dx.doi.org/10.1021/ja107583h.
[44] D. Orain, J. Ellard, M. Bradley, Protecting groups in solid-phase organic syn-
thesis, J. Comb. Chem. 4 (2002) 1e16, http://dx.doi.org/10.1021/cc0001093.
[45] I. Coin, M. Beyermann, M. Bienert, Solid-phase peptide synthesis: from stan-
dard procedures to the synthesis of difficult sequences, Nat. Prot. 2 (2007)
3247e3256, http://dx.doi.org/10.1038/nprot.2007.454.
[48] M. Dirin, J. Winkler, Influence of diverse chemical modifications on the ADME
characteristics and toxicology of antisense oligonucleotides, Expert Opin. Biol.
Ther. 13 (2013) 875e888, http://dx.doi.org/10.1517/14712598.2013.774366.
[49] E.M. Anderson, P. Miller, D. Ilsley, W. Marshall, A. Khvorova, C.A. Stein,
L. Benimetskaya, Gene profiling study of G3139- and Bcl-2-targeting siRNAs
identifies a unique G3139 molecular signature, Cancer Gene Ther. 13 (2006)
406e414, http://dx.doi.org/10.1038/sj.cgt.7700901.
[50] M. Stessl, C.R. Noe, J. Winkler, Off-target effects and safety aspects of phos-
phorothioate oligonucleotides, in: V.A. Erdmann, J. Barciszewski (Eds.), From
Nucleic Acids Sequences to Molecular Medicine, RNA Technologies, Springer
Verlag, Berlin, 2012, pp. 67e83.
[46] A. Aitken, M. Learmonth, Estimation of disulfide bonds using Ellman's reagent,
in: J.M. Walker (Ed.), The Protein Protocols Handbook, Humana Press, 2002,
pp. 595e596.