Gene silencing activity of siRNA molecules containing phosphorodithioate substitutions

Article abstract of DOI:10.1021/cb300078e

Chemically synthesized small interfering RNAs (siRNAs) have been widely used to identify gene function and hold great potential in providing a new class of therapeutics. Chemical modifications are desired for therapeutic applications to improve siRNA efficacy. Appropriately protected ribonucleoside-3′-yl S-[β-(benzoylmercapto)ethyl]pyrrolidino- thiophosphoramidite monomers were prepared for the synthesis of siRNA containing phosphorodithioate (PS2) substitutions in which the two non-bridging oxygen atoms are replaced by sulfur atoms. A series of siRNAs containing PS2 substitutions have been strategically designed, synthesized, and evaluated for their gene silencing activities. These PS2-siRNA duplexes exhibit an A-form helical structure similar to unmodified siRNA. The effect of PS2 substitutions on gene silencing activity is position-dependent, with certain PS2-siRNAs showing activity significantly higher than that of unmodified siRNA. The relative gene silencing activities of siRNAs containing either PS2 or phosphoromonothioate (PS) linkages at identical positions are variable and depend on the sites of modification. 5′-Phosphorylation of PS2-siRNAs has little or no effect on gene silencing activity. Incorporation of PS2 substitutions into siRNA duplexes increases their serum stability. These results offer preliminary evidence of the potential value of PS2-modified siRNAs.

Full text of DOI:10.1021/cb300078e

Articles
pubs.acs.org/acschemicalbiology
Gene Silencing Activity of siRNA Molecules Containing
Phosphorodithioate Substitutions
Xianbin Yang,*,† Malgorzata Sierant,‡ Magdalena Janicka,‡ Lukasz Peczek,‡ Carlos Martinez,§
Tom Hassell,§ Na Li,† Xin Li,† Tianzhi Wang,∥ and Barbara Nawrot‡
†AM Biotechnologies LLC, 12521 Gulf Freeway, Houston, Texas 77034, United States
‡Department of Bioorganic Chemistry, Centre of Molecular and Macromolecular Studies, Polish Academy of Sciences, Lodz, Poland
§Sigma Life Science, 9186 Six Pines, The Woodlands, Texas 77380, United States
∥The Sealy Center for Structural Biology & Molecular Biophysics, University of Texas Medical Branch, Galveston, Texas 77555,
United States
S
* Supporting Information
ABSTRACT: Chemically synthesized small interfering RNAs (siRNAs)
have been widely used to identify gene function and hold great potential
in providing a new class of therapeutics. Chemical modifications are
desired for therapeutic applications to improve siRNA efficacy.
Appropriately protected ribonucleoside-3′-yl S-[β-(benzoylmercapto)-
ethyl]pyrrolidino-thiophosphoramidite monomers were prepared for the
synthesis of siRNA containing phosphorodithioate (PS2) substitutions
in which the two non-bridging oxygen atoms are replaced by sulfur
atoms. A series of siRNAs containing PS2 substitutions have been
strategically designed, synthesized, and evaluated for their gene silencing
activities. These PS2-siRNA duplexes exhibit an A-form helical structure similar to unmodified siRNA. The effect of PS2
substitutions on gene silencing activity is position-dependent, with certain PS2-siRNAs showing activity significantly higher than
that of unmodified siRNA. The relative gene silencing activities of siRNAs containing either PS2 or phosphoromonothioate (PS)
linkages at identical positions are variable and depend on the sites of modification. 5′-Phosphorylation of PS2-siRNAs has little or
no effect on gene silencing activity. Incorporation of PS2 substitutions into siRNA duplexes increases their serum stability. These
results offer preliminary evidence of the potential value of PS2-modified siRNAs.
The substitution of both non-bridging phosphate oxygen
atoms with sulfur atoms gives rise to a phosphorodithioate
(PS2) internucleotide linkage, which like natural phosphoro-
diester linkages of RNA is achiral at phosphorus. The PS2
substitution is a very attractive RNA analogue because it is a
closely related mimic of natural RNA and should have other
biochemical and biophysical properties similar to phosphoro-
diester linkages of RNA. In addition, it has been shown that
dinucleoside phosphorodithioates (PS2-dimers) have high
nuclease resistance.12 Moreover, a hammerhead ribozyme
with one PS2 substitution at the cleavage site was evaluated
for its cleaving rate to the expected product.13 Further, PS2-
dimers have shown a great resistance to alkaline degradation
when compared to natural RNA derivatives.12 Over the past 20
years, the only reported PS2-RNA chemistries are the
thiophosphoramidite solution-phase synthesis of PS2-dimers,12
the H-phosphonothioate solid-phase syntheses of a 12-mer
oligoribonucleotide containing a single PS2 substitution,13 and
NA interference (RNAi) has emerged as a novel
Rmechanism that is activated in mammalian cells by small
interfering RNAs (siRNAs), which hold great potential in
providing a new class of therapeutics.1 Although unmodified
siRNAs have been used with success for gene silencing,
chemical modifications of one or both strands are desired for
research and pharmaceutical applications in order to enhance
potency and to improve pharmacokinetic properties.2−4
A
variety of chemical modifications have been evaluated for their
effects on RNAi activity including selected phosphate backbone
modifications such as phosphoromonothioates (PS)5,6 and
boranophosphates.7 These modifications involve substitution of
a single non-bridging phosphate oxygen atom with either a
sulfur atom or a borane group. This renders the internucleotide
linkage nuclease resistant, but unlike natural RNA, both the PS-
and borane-modified phosphorus centers become chiral.
Synthesis of the PS-RNA by standard phosphoramidite
methodology generates a mixture of unresolvable diastereo-
meric oligomers possibly having variable biochemical, bio-
physical, and biological properties.8 While stereocontrolled
synthesis of P-chiral PS9,10 or boranophosphates11 represents
one possible solution to this problem, another lies in the
synthesis of modifications that are achiral at phosphorus.
Received: February 21, 2012
Accepted: April 18, 2012
Published: April 18, 2012
© 2012 American Chemical Society
1214
dx.doi.org/10.1021/cb300078e | ACS Chem. Biol. 2012, 7, 1214−1220

ACS Chemical Biology
Articles
a 15-mer oligoribouridylate phosphorodithioate bearing ex-
clusively PS2 substitutions.14
In this paper, we report the synthesis of protected
ribonucleoside-3′-yl S-[β-(benzoylmercapto)ethyl]pyrrolidino-
thiophosphoramidite monomers 1a−d (Figure 1) and their
Synthesis of RNAs Containing PS2 Substitutions. In
our previous studies, we identified and verified an active
unmodified siRNA sequence that effectively silences the
BACE1 (β-site APP-cleaving enzyme 1) in cellular models
(siRNA-1 in Supplementary Table 1S).18 We chose this
sequence for initial tests of siRNA molecules containing the
PS2 substitutions. Thus, 21 siRNAs (PS2-3−24, Supplementary
Table 1S) were designed to contain up to four PS2
substitutions strategically placed at positions considered
important for silencing activity. In addition, PS2-siRNAs
containing 5′-phosphate were chemically synthesized to
evaluate the gene silencing effect of the 5′- hydroxyl and 5′-
phosphate of PS2-siRNAs since the 5′-phosphate plays a critical
role for RNAi activity.19−21 The PS2-RNAs (Supplementary
Table 1S) were synthesized on an Expedite 8909 DNA/RNA
Synthesizer using commercial 5′-DMT-2′-O-TBDMS nucleo-
side (ABz, CAc, GAc, and U) phosphoramidite monomers as well
as thiophosphoramidites 1a−d as described in Methods. The
average stepwise coupling efficiency of all phosphoramidites
including thiophosphoramidites was about 97% as estimated by
the DMT-cation assay. After deprotection, all of the modified
RNAs were isolated by FPLC according to a previously
described protocol for purifying PS2-DNAs.22 The PS2-RNAs
were desalted using reverse-phase HPLC to yield the PS2-RNA
final products. The structures of the PS2-RNAs were confirmed
by ESI-MS. The purity of all RNAs and correctness of double
strand structures of PS2-siRNAs were assessed by denaturing
20% polyacrylamide (PAGE) and agarose gel electrophoresis
analyses, respectively.
Figure 1. Synthesis of thiophosphoramidites 1a−d: (i) tris-
(pyrrolidino)phosphine, 1H-tetrazole; (ii) 1-(trimethylsilyl)imidazole;
(iii) ethanedithiol monobenzoate, 1H-tetrazole. Abbreviations: BZ
=
AdeBz (a), CytAc (b), GuaAc (c), Ura (d); DMT = dimethoxytrityl; Ph
= phenyl; TBDMS = tert-butyldimethylsilyl.
use for the solid-phase preparation of RNA molecules
containing between one and four PS2 substitutions. These
modified RNAs are used as components of a novel type of
siRNA molecule. We envisaged that the addition of PS2
modifications to certain positions of the siRNA duplex would
increase nuclease resistance and enhance RNAi activity.
Moreover, we characterize the gene silencing activity of these
PS2-siRNA duplexes. Finally, we analyze their conformational
structures and evaluate the serum stability of PS2-siRNAs. To
the best of our knowledge these are the first studies evaluating
biological properties of siRNA duplexes containing the PS2
substitutions.
RNAi Activity of PS2-siRNAs. The impact of introducing
the PS2 substitution(s) in the siRNA duplex was investigated in
a GFP/RFP dual fluorescence assay (DFA).23 The gene
silencing results are summarized in Supplementary Table 1S
and Figure 2 (the level of a target gene expression is presented
as the % of negative control). Since the PS2-modified duplexes
exhibit slightly decreased thermal stability (Supplementary
Table 1S), the introduction of the PS2 modification(s) at the
3′-end of the sense strand was hypothesized to favor duplex
unwinding and therefore enhance gene silencing activity. Thus,
we designed PS2-3 having one PS2 substitution at position 20
and PS2-4 having two PS2 substitutions at positions 19 and 20
of the sense strand (counting from the 5′-end). Interestingly,
PS2-3 showed slightly lower RNAi activity compared to that of
wild-type siRNA-1 (39.4% versus 35.0%), while PS2-4 showed
significantly enhanced gene silencing activity (17.2% versus
35.0%; p < 0.001). PS2-6 having four PS2 substitutions at
positions 5, 10, 15, and 20 showed a significant decrease in
RNAi activity when compared to the wild-type siRNA-1 (76.7%
versus 35.0%, p < 0.001). PS2-7 having three PS2 substitutions
at positions 5, 12, and 20 (two identical positions with PS2-6)
showed gene silencing activity comparable to that of the wild-
type siRNA-1, while PS2-8 having two PS2 substitutions at
position 10 (the sense strand cleavage site) and position 20
showed moderately reduced RNAi activity when compared to
siRNA-1. In contrast, PS2-9 having two PS2 substitutions at
positions 3 and 12 showed significantly better gene silencing
when compared to siRNA-1 (18.7% versus 35.0%, p < 0.001).
Therefore, we conclude that the PS2 substitution effects in the
sense strand are position-dependent. This has been shown in
previous studies where various chemical modifications were
added to siRNAs to determine the biochemical properties
required for RNAi.2,4,23−27
RESULTS AND DISCUSSION
■
Preparation of Protected Ribonucleoside Thiophos-
phoramidites. The PS2-RNAs were synthesized using a solid-
phase thiophosphoramidite method because of the speed and
simplicity of this approach. Currently, the only practical
method for solid-phase synthesis of PS2-DNA uses commer-
cially available 2′-deoxyribonucleoside thiophosphoramidites
(www.glenresearch.com).15,16 The appropriately protected
ribonucleoside-3′-yl S-[β-(benzoylmercapto)ethyl]pyrrolidino-
thiophosphoramidite monomers 1a−d (Figure 1) were
prepared in ca. 75−85% yield.
We chose to use the β-(benzoylmercapto)ethyl sulfur
protecting group because this is in the context of PS2-
containing DNAs.17 In addition, it was reported to help reduce
the PS contamination in the final products when compared to
the use of β-cyanoethyl or 2,4-dichlorobenzyl.17 Moreover, the
β-(benzoylmercapto)ethyl sulfur protecting group was reported
to have better stability in solution than the 2,4-dichlorobenzyl
group. Like the β-cyanoethyl group, the removal of the β-
(benzoylmercapto)ethyl group can also be conveniently carried
out during a simple base-deprotection treatment to give the
final RNA products. The 31P NMR spectra of these
thiophosphoramidites 1a−d showed the expected two singlets
between δ 160 and δ 180 ppm for two diastereoisomers
(Supplementary Figure 1S). Since the 2′-deoxyribonucleoside
thiophosphoramidites were reported to be degraded and
rearranged to the Arbuzov-type product16,17 on silica gel
during column chromatography, the ribonucleoside thiophos-
phoramidites were used in the PS2-RNA synthesis without
further purification. This approach was successful, indicating
that further purification of the thiophosphoramidites is
unnecessary.
1215
dx.doi.org/10.1021/cb300078e | ACS Chem. Biol. 2012, 7, 1214−1220

ACS Chemical Biology
Articles
positions, is close to the mean value of the BACE1-GFP
expression levels for PS2-14 (28.0% reading) and PS2-15
(35.4% reading) having a single PS2 substitution at the 10th
and 11th position, respectively. It appears that the PS2
substitutions for these phosphates have a potential to improve
the gene silencing activity. Thus, on the basis of the above
results, one could expect some positive effects on RNAi activity
for siRNAs having PS2 substitutions on the antisense strand.
SiRNA duplexes PS2-5, 10, 17, and 18 containing one to four
PS2 substitutions in the antisense strand showed no gene
silencing activity toward BACE1 in the assay depicted in Figure
2. These results may be due to a couple of plausible factors.
The recent Archaeoglobus fulgidus (Af Piwi)/siRNA duplex
crystal structure33 indicates that the phosphate groups from
the first phosphate (between the first and second bases from
the 5′ end) to the fourth phosphate in the antisense strand
closely interact with the residues conserved in the Piwi and Ago
subfamilies. Since PS2 has a stronger affinity for protein
residues than phosphate in general,34,35 PS2 substitutions at
these four phosphates might alter the interaction between the
four phosphates and residues of the Piwi protein, which could
possibly inhibit processes in which this part of the siRNA is
engaged.33,36 In addition, the coordination of one of the oxygen
atoms of the second phosphate (between the second and third
bases) to a magnesium ion33 is vital for RNAi activity. This
interaction is probably disrupted when the sulfur atom is
substituted for the oxygen atom because the soft sulfur atom of
the PS2 is known to coordinate poorly with hard magnesium
cations.19,30,31,33,37 Since PS2-5, 10, 17, and 18 all contain at
least one PS2 substitution from the first to the fourth
phosphate of the antisense strand, their gene silencing activity
could be affected by the altered affinity between the siRNA
duplex and the Piwi protein. As well, the gene silencing activity
of PS2-5 and 10 could further be affected by the disruption of a
magnesium cation−phosphate interaction. In the following
paragraph, we observed that PS-5 and PS-10 show partial
recovery of RNAi activity when directly compared with their
counterparts PS2-5 and PS2-10, respectively. These results
confirm the importance of the second and third phosphates of
the antisense strand in the RNAi process. More systematic
studies would be required to fully determine the influence of
each phosphate of the antisense strand on the silencing
properties of the siRNA duplex.
Rana and co-workers reported that RNAi activity absolutely
requires A-form helix formation between target mRNA and its
guiding antisense strand.4,21 To determine the overall structures
of PS2-siRNAs, CD spectra were collected for all of the PS2-
siRNA duplexes 3−23 and compared with the CD spectrum of
the wild-type siRNA-1. All of the CD spectra of the PS2-
siRNAs are similar to the spectra of the unmodified duplex and
are consistent with the typical A-type structure of double-
stranded RNA (a maximum of the positive Cotton effect at 268
nm and crossover point at 240 nm). This result is in contrast
with earlier studies that demonstrated significant structural
perturbations from PS2 substitutions in DNA duplexes.34 Since
the CD spectra indicate that PS2-siRNA is globally similar in
structure to the wild-type siRNA (Figure 3), the observed
different RNAi activity caused by the PS2 substitution is likely
to depend on the interaction of the phosphate functions with
metal ions and amino acid residues of the RISC proteins as
well. Our results suggest that PS2 substitutions have crucial
interactions with the active RISC complex.
Figure 2. Gene silencing activity of PS2-siRNAs. Right panel: schemes
of siRNA duplexes, where the colored orange balls represent PS2
substitutions. Left panel: PS2-siRNA activities were tested in a DFA as
described in Methods. BACE-GFP/RFP levels were normalized to
cells transfected with negative control siRNA in each experiment
(Control). The orange bars represent PS2 modifications in the sense
strand, and the blue bars represent PS2 modifications in the antisense
strand. Error bars represent the standard deviation. The results are also
tabulated in Supplementary Table 1S.
The antisense strand of siRNA has been reported to be more
sensitive to chemical modifications than the sense
strand.2,5,26−28 This is most likely due to the importance of
this strand for incorporation into the RISC, nucleation and
binding with target mRNA, as well as mRNA cleavage.29−31 To
determine the impacts of PS2 modification(s) in the antisense
strand, PS2 modifications at several positions were also
examined. Modified siRNAs PS2-11, 12, 13, 19, 20, 21, 22,
and 23 exhibited equal or moderately reduced gene silencing
activity when compared to the wild-type siRNA-1. These PS2-
siRNA duplexes containing one to three PS2 substitutions
decreased the BACE1 expression to the range of 37.2−60.8%
versus 35.0% for the wild-type siRNA-1. Next, we evaluated the
gene silencing activity of PS2-14, 15, and 16 siRNAs where PS2
substitutions are opposite to the cleavage site of the target
mRNA, which is between the 10th and 11th residues from the
5′-end of the antisense strand.32 Based on a model derived from
the crystal structure of a bound antisense strand,20,33 these
phosphates do not directly interact with the Ago2 protein. Our
experiments showed that the PS2 substitutions for these
phosphates have similar or slightly enhanced gene silencing
activity (Figure 2, Supplementary Table 1S) when compared
with the wild-type siRNA-1. Importantly, PS2-14 having only
one PS2 substitution between nucleosides 10th and 11th (just
opposite to the cleavage site) slightly improved the gene
silencing activity when compared to the wild-type siRNA-1
(28.0% for PS2-14 versus 35.0%; p = 0.05). Interestingly, the
BACE1-GFP expression level for PS2-16 (32.7% reading),
having two PS2 substitutions at both the 10th and 11th
1216
dx.doi.org/10.1021/cb300078e | ACS Chem. Biol. 2012, 7, 1214−1220

ACS Chemical Biology
Articles
their counterparts PS2-5 and PS2-10, respectively. These results
indicate the importance of the interactions between the
phosphates of the antisense strand and the Argonaute protein
in the RNAi process.
5′-Terminal Phosphate of the PS2-siRNA. During the
processing of long dsRNA by Dicer, endogenous siRNAs are
generated with phosphate groups at their 5′- ends. Synthetic
siRNAs usually display the 5′-OH group, and to become
incorporated into the RISC and mediate target mRNA cleavage,
the guide strand of the duplex has to be phosphorylated. The
importance of the 5′-terminal phosphate of the guide strand has
been demonstrated by recent crystallographic structural studies
of the Ago2 guide strand binding pocket.20,30 The 5′-end
phosphate is critical for the binding of the guide strand with
Ago2 and its incorporation into the RISC. Additionally, it has
been shown that chemical or enzymatic pre-phosphorylation of
the guide strand is generally unnecessary because the synthetic
siRNAs are phosphorylated by endogenous kinases directly
after transfection into the cells.41,42 However, some chemical
modifications introduced at the 5′-ends of the RNA duplex can
interfere with the endogenous kinase activity and disturb the
phosphorylation process. In these cases the chemical pre-
phosphorylation of the guide strand at the 5′-end can rescue the
siRNA activity.42 Because our PS2-5 duplex has PS2
modifications located near the 5′-end of the antisense strand
(at the second and third positions), we wondered whether the
loss of activity is caused by inhibition of kinase activity by the
presence of the PS2 substitution. To test our suspicions, we
chemically synthesized variants of the PS2-5 siRNA having a 5′-
terminal phosphate on just the antisense strand (PS2-5-P) or
on both the sense and antisense strands (PS2-5-PP;
Supplementary Table 2S). We transfected these phosphory-
lated PS2-5 siRNAs into HeLa cells and assessed the RNAi
activity in comparison to nonphosphorylated PS2-5 siRNA. We
did not observe any improvement of the PS2-5 activity (Figure
5). Similar studies were performed using phosphorylated
versions of the PS2-4, 9, and 16, as well as siRNA-1. The 5′-
phosphorylation had little or no effect on the activity of any of
these siRNAs. These results suggest that the free 5′-hydroxyl of
the wild-type siRNA or PS2-siRNA duplex is first phosphory-
Figure 3. Circular dichroism (CD) spectra of wild-type siRNA
(siRNA-1) and PS2-siRNAs: PS2-14, 4, 11, and 10 with 1, 2, 3, and 4
PS2 substitutions, respectively.
Phosphorodithioate (PS2) versus Phosphoromono-
thioate (PS) Substitutions. PS modifications have been
used extensively to increase the stability of antisense
oligonucleotides38,39 and were also examined for their RNAi
activity.4−6,40 However, since the phosphorus center with PS
substitution is chiral, the PS-siRNA synthesized by the standard
phosphoramidite chemistry results in a mixture of unresolvable
diastereomeric oligomers. These may potentially have variable
biochemical, biophysical, and biological properties. Since the
PS2-siRNAs are achiral at phosphorus, we hypothesized that
the introduction of PS2 linkages into siRNA might be able to
improve the RNAi activity when compared to PS-siRNA. We
have therefore synthesized several PS-siRNAs (PS-4, 5, 9, 10,
and 16) in which the PS2 linkages of PS2-4, 5, 9, 10, and 16
were replaced by PS linkages. We then compared gene silencing
activity of the PS2- and PS-model siRNAs in HeLa cells using a
DFA assay (shown in Figure 4). We found that both PS2-4 and
Figure 4. Silencing activity of PS2-siRNAs and PS-siRNAs targeting
the BACE1 gene. Percent of BACE-GFP fusion gene expression in
cells treated with unmodified or modified siRNA at 5 nM. The results
of each individual experiment were normalized to the negative control
siRNA. Error bars represent the standard deviation. PS2-siRNA is
represented by red bars, and PS-siRNA is represented by green bars.
Unmodified and control siRNA are represented by light black bars.
Figure 5. Effect of 5′-end phosphorylation of siRNAs on BACE1 gene
silencing. Percent of BACE-GFP gene expression was determined in
cells treated with unmodified or modified siRNA at 5 nM
concentration. The results of each individual experiment were
normalized to negative control siRNA without 5′-end phosphate.
Error bars represent the standard deviation. Native siRNAs without 5′-
end phosphate are represented by light black bars. PS2-siRNAs are
represented by red bars. SiRNA-1 and PS2-siRNAs with 5′-end
phosphate (P) at the antisense strand are represented by green bars.
SiRNA-1 and PS2-siRNAs with 5′- end phosphates (PP) at both the
sense and antisense strands are represented by purple bars.
PS2-9 showed statistically significant increase in gene silencing
activity (p < 0.001) when compared with that of PS-4 and PS-9.
The decreased gene silencing activities for the PS-siRNAs may
be caused by the mixture of isomers that may have variable
biological effects. Both siRNA PS2-16 and PS-16 have RNAi
activity very similar to that of the wild-type siRNA-1. This
observation may indicate that the interactions between these
phosphates and the protein are not critical. The most
interesting results were observed from PS-5 and PS-10, which
show partial recovery of RNAi activity when compared with
1217
dx.doi.org/10.1021/cb300078e | ACS Chem. Biol. 2012, 7, 1214−1220

ACS Chemical Biology
Articles
lated by endogenous kinases and then enters the RNAi
pathway, which is consistent with the published results.12 The
results also demonstrated that the free 5′-hydroxyl of the PS2-
siRNA duplex is sufficient for RNAi activity in mammalian cells.
Stability of PS2-siRNAs in Serum. The identification of
highly nuclease-resistant siRNAs is a key concern for in vivo
applications, and great efforts have been applied to enhance
stability by chemical modification.4,18,27,43,44 Since the above
experiments showed that PS2-siRNAs such as PS2-4 and PS2-9
could effectively enhance RNAi activity compared to that of the
wild-type siRNA-1, it was of interest to study the serum stability
of PS2-siRNAs as well as to compare it with serum stability of
the PS-siRNAs. To carry out such studies, unmodified or
modified dsRNAs (Supplementary Table 3S) were incubated in
10% fetal bovine serum (FBS) at 37 °C for up to 24 h. At
various time points, siRNAs were extracted, separated on 4%
agarose gels under nondenaturing conditions, stained with
ethidium bromide, and quantified. The observed stability of the
PS2-siRNA duplexes was consistently higher than that of their
counterpart PS-siRNAs, while nuclease stability of both PS2-
siRNA and PS-siRNA was higher than that of the unmodified
siRNA (Figure 6). The increased nuclease stability of the PS2-
structure comparable to that of the unmodified siRNA
counterpart. The siRNAs containing PS2 substitutions in either
the sense strand or the antisense strand were evaluated for their
gene silencing activity in the DFA system. The PS2
modifications were poorly tolerated in the seed region of the
antisense strand but were well tolerated in the central part of
the antisense strand. In contrast, the sense strand tolerated
most of the PS2 modifications located in all tested positions,
with significantly improved gene silencing activity in several
positions, e.g., PS2-4 (PS2 in positions 19 and 20) and PS2-9
(PS2 in positions 3 and 12). 5′-Phosphorylation of the PS2-
siRNAs did not substantially improve gene silencing activity.
This means that PS2 modification does not disturb the
endogenous phosphorylation process. The biophysical proper-
ties of the PS2-siRNA modification, reflected in its enhanced
serum stability, will hopefully translate to improved in vivo
potency.
METHODS
■
Preparation of the Thiophosphoramidites 1a−d. 5′-O-(4,4′-
Dimethoxytrityl)-2′-O-(tert-butyldimethylsilyl) ribonucleosides (2a−d,
1.0 mmol, Figure 1) were phosphitylated by means of tris-
(pyrrolidino)phosphine (130 μL, 1.0 mmol)17,35 in the presence of
1H-tetrazole (0.35 mmol) in anhydrous dichloromethane (15 mL)
containing a spatula of 3 Å molecular sieves for 15 min at RT.
Trimethylsilylimidazole (15 μL, 0.1 mmol) was then added to the
reaction mixture, and the reaction was stirred for 5 min. The resulting
bis(pyrrolidino)phosphite intermediates were converted into the
desired thiophosphoramidites (1a−d) by treatment in situ with
monobenzoylethanedithiol (220 μL, 1.3 mmol)17,35 and additional
1H-tetrazole (2.7 mmol). After an aqueous workup to remove 1H-
tetrazole and tetrazolide salts, the fully protected RNA thiophosphor-
amidites (1a−d, Figure 1) were isolated by precipitation from hexane
in 75−85% yield at more than 90% purity as assessed by 31P NMR
(Figure 1S). 1a (BZ = AdeBz): yield, 85%, 31P NMR, δ = 166.7, 174.2
ppm. 1b (BZ = CytAc): yield, 75%, 31P NMR, δ = 168.4, 175.2. ppm. 1c
(BZ = GuaAc): yield, 87%, 31P NMR, δ = 163.8, 175.3 ppm. 1d (BZ =
Ura): 77%, 31P NMR, δ = 170.01, 177.1 ppm.
Synthesis of RNA Containing PS2 Substitutions. All PS2-
modified RNAs were synthesized on 1000 Å controlled pore glass T
support (resulting in a thymidine deoxyribonucleoside at the 3′ end).
The standard 2′-O-TBDMS RNA phosphoramidites were used for
incorporation of A, C, G, and U residues. Each standard nucleotide
was coupled using approximately 120 μL of a 0.1 M solution of the
appropriate phosphoramidite in anhydrous acetonitrile. Each PS2
substitution was coupled using approximately 250 μL of a 0.15 M
solution of the appropriate thiophosphoramidite in anhydrous
acetonitrile, except that G thiophosphoramidite (1c) was prepared
in anhydrous acetonitrile containing 10% anhydrous dichloromethane.
All coupling times were 10 min. A 0.05 M solution of EDITH reagent
in anhydrous acetonitrile was used as a sulfurizing agent to oxidize the
Figure 6. Serum stability of PS2-siRNA and PS-siRNA in comparison
to unmodified siRNA. The sequences of PS2-siRNA and siRNAs,
listed in the Supplementary Table 3S, were incubated with 10% fetal
bovine serum (FBS) in PBS buffer at 37 °C. Aliquots were collected
after 0.5, 1, 2, 3, 4, 6, 8, and 24 h, separated in 4% agarose gel, and
analyzed using G-box gel visualization system (SynGene). Graph
represents results from two independent experiments. Horizontal axis
represents time (h). Vertical axis represents remaining siRNA (%).
Bars represent the standard deviation.
siRNAs should have implications for in vivo application, e.g., in
mouse models or in therapeutic applications of siRNA.
However, like other studies for most chemistries, we find the
PS2-siRNA serum stability to be positively correlated with the
level of the RNA modifications, but highly modified siRNAs
generally displayed poor silencing. The number of PS2
modification added to siRNA to increase serum stability must
be chosen prudently in order to maintain or possibly improve
high RNAi silencing activity.
Summary. In this report we present a method of using
thiophosphoramidites with a β-(benzoylmercapto)ethyl pro-
tecting group16,17 on sulfur for the first solid-phase synthesis of
the PS2-RNAs. By automating this process, 21 PS2-siRNAs
containing from one to four PS2 linkages were prepared, and
their identity and purity were confirmed by gel electrophoresis
and mass spectrometry. CD spectra indicated an A-form helical
internucleosidic thiophosphite to the phosporodithiotriester.45
A
solution of 0.02 M I2 in THF/pyridine/water was used to oxidize
internucleosidic phosphites to phosphorotriesters. The average
stepwise coupling efficiencies of all thiophosphoramidites were about
97% as estimated by the dimethoxytrityl cation assay. After completion
of the synthesis, the solid support was suspended in ammonium
hydroxide/methylamine (AMA) solution (prepared by mixing 1
volume of ammonium hydroxide (28%) with 1 volume of 40%
aqueous methylamine) and heated at 65 °C for 15 min to release the
product from the support and to complete the removal of all
protecting groups except the TBDMS group at the 2′-position. The
solid support was filtered, and the filtrate was concentrated to dryness.
The obtained residue was resuspended in 115 μL of anhydrous DMF
and then heated for 5 min at 65 °C to dissolve the crude product. TEA
(60 μL) was added to each solution, and the solutions were mixed
gently. TEA·3HF (75 μL) was added to each solution, and the tubes
were then sealed tightly and incubated at 65 °C for 2.5 h. The reaction
1218
dx.doi.org/10.1021/cb300078e | ACS Chem. Biol. 2012, 7, 1214−1220

ACS Chemical Biology
Articles
was quenched with 1.75 mL of DEPC-treated water. The samples were
analyzed by UV−Vis spectroscopy, and the OD readings were
recorded.
Aliquots of 10 μL were collected after 0.5, 1, 2, 3, 4, 6, 8, and 24 h,
diluted in 1x loading buffer (Fermentas), frozen in liquid nitrogen, and
kept at −70 °C until analysis. A similar control reaction without FBS
was performed for each siRNA as well. Samples were separated in 4%
low melting agarose (NuSieve GTG Agarose, Cambrex) under
nondenaturing conditions. Gels were analyzed using a G-box gel
visualization system and quantified by GeneTools 4.00 software
(SynGene).
Purification and Characterization of RNAs and PS2-RNAs.
Purification was performed on an Amersham Biosciences P920 FPLC
instrument fitted with a Mono Q 10/100 GL column, based on a
previously described procedure.22 The buffers were prepared with
DEPC-treated water, and their compositions were as follows: Buffer A:
25 mM Tris-HCl, 1 mM EDTA, pH 8.0; Buffer B: 25 mM Tris-HCl, 1
mM EDTA, 1 M NaCl, pH 8.0. The FPLC gradient profile was as
follows: 0−100% B over 50 min, 100% B for 5 min, 100−0% B over 5
min. Purified fractions were pooled and desalted via RP-HPLC on an
Amersham Biosciences P900 fitted with a PRP-1 Column (Hamilton
Co., Reno, NV). The buffers used were as follows: Buffer A: 0.1 M
NH4OAc in DEPC-treated water, pH 7.0; Buffer B: HPLC grade
acetonitrile. The gradient run was as follows: 0−35% B over 25 min,
35−0% B over 1 min. After desalting, the samples were frozen at −80
°C for 1 h, lyophilized, and dissolved in DEPC-treated water. All the
RNA, PS-RNA and PS2-RNA oligonucleotides were characterized by
denaturing polyacrylamide gel electrophoresis. Representative PS2-
RNAs were further characterized by ESI-MS.
ASSOCIATED CONTENT
* Supporting Information
This material is available free of charge via the Internet at
■
S
http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: xianbin.yang@thioaptamer.com.
■
Notes
The authors declare no competing financial interest.
Preparation of SiRNA Duplexes. Assembly of siRNA duplexes
was performed in phosphate buffered saline (PBS) by heating the
equivalent mixture of RNA oligonucleotides coding the sense and
antisense strands of siRNA at 95 °C for 2 min followed by slow
cooling to room temperature (over 2 h). The assembly of the resulting
duplexes was confirmed by a 4% agarose gel electrophoresis.
Circular Dichroism (CD) Measurements. CD spectra were
recorded on a CD6 dichrograph (Jabin-Yvon, Longjumeau, France)
using cells with 0.5 cm path length, 2 nm bandwidth, and 1−2 s
integration time. Each spectrum was smoothed with a 25-point
algorithm (included in the manufacturer′s software, version 2.2.1) after
averaging of at least three scans. The spectra from 200 to 340 nm were
recorded at 25 °C in the same buffer (100 mM NaCl, 0.1 mM EDTA,
10 mM TRIS, pH 7.4) as in the melting experiments. The
concentration of the two complementary RNA oligonucleotides was
ca. 2 μM.
Cell Line Culture and Transfection Conditions. HeLa cells
were cultured in RPMI 1640 medium (Gibco) supplemented with
10% heat-inactivated FBS (Gibco), 100 U/mL penicillin, and 100 μg/
mL streptomycin (Polfa) at 37 °C and 5% CO2 in 96-well black well
plates with transparent bottom (Perkin-Elmer). Before transfection,
the culture medium was replaced with fresh antibiotic-free medium.
HeLa cells were cotransfected with 15 ng/well pDsRed2-N1 plasmid
DNA (BD Biosciences), 70 ng/well pBACE-GFP-plasmid DNA,23,44
and the appropriate siRNA at 5.0 nM final concentration in OPTI-
MEM1 medium (Gibco). Transfection was performed using Lipofect-
amine 2000 (Invitrogen) at a 2:1 ratio (2 μL of Lipofectamine 2000
per 1 μg of nucleic acids). The cells were incubated for 6 h, and then
the medium containing the transfection mixture was replaced with the
fresh culturing medium with antibiotics. After 48 h incubation the cells
were washed three times with PBS and lysed overnight with a 1:3
mixture of NP-40 buffer (150 mM NaCl, 1% Nonidet P-40, 50 mM
Tris-HCl, pH 7, protease inhibitor cocktail, Roche) and PBS at 37 °C.
Cell lysates were used for fluorescence determination. Mock
transfection control was included for each experiment.
ACKNOWLEDGMENTS
■
We thank D. Gorenstein, R. Durland, C. Lam, and W. He for
their helpful discussions. Support from the U.S. National
Institutes of Health (R43GM084552, R44GM084552, and
R43CA141842 to X.Y.) and from the Polish Ministry of Science
and Higher Education (PBZ-MNiSW-07/I/2007 to B.N. and N
N204 540039/2010-2013 to M.S.) is gratefully acknowledged.
REFERENCES
■
(1) Elbashir, S. M., Harborth, J., Lendeckel, W., Yalcin, A., Weber, K.,
and Tuschl, T. (2001) Duplexes of 21-nucleotide RNAs mediate RNA
interference in cultured mammalian cells. Nature 411, 494−498.
(2) Shukla, S., Sumaria, C. S., and Pradeepkumar, P. I. (2009)
Exploring chemical modifications for siRNA therapeutics: a structural
and functional outlook. ChemMedChem 5, 328−349.
(3) Behlke, M. A. (2006) Progress towards in vivo use of siRNAs.
Mol. Ther. 13, 644−670.
(4) Chiu, Y. L., and Rana, T. M. (2003) siRNA function in RNAi: a
chemical modification analysis. RNA 9, 1034−1048.
(5) Braasch, D. A., Jensen, S., Liu, Y., Kaur, K., Arar, K., White, M. A.,
and Corey, D. R. (2003) RNA interference in mammalian cells by
chemically-modified RNA. Biochemistry 42, 7967−7975.
(6) Li, Z. Y., Mao, H., Kallick, D. A., and Gorenstein, D. G. (2005)
The effects of thiophosphate substitutions on native siRNA gene
silencing. Biochem. Biophys. Res. Commun. 329, 1026−1030.
(7) Li, P., Sergueeva, Z. A., Dobrikov, M., and Shaw, B. R. (2007)
Nucleoside and oligonucleoside boranophosphates: chemistry and
properties. Chem. Rev. 107, 4746−4796.
(8) Lebedev, A. V., and Wickstrom, E. (1996) The chirality problem
in P-substituted oligonucleotides. Perspect. Drug Discovery Des. 4, 17−
40.
(9) Wada, T., Fujiwara, S., Sato, T., Oka, N., and Saigo, K. (2004)
Stereocontrolled synthesis of phosphorothioate RNA by the
oxazaphospholidine approach. Nucleic Acids Symp. Ser. (Oxford), 57−
58.
(10) Sierzchala, A., Okruszek, A., and Stec, W. J. (1996)
Oxathiaphospholane method of stereocontrolled synthesis of
diribonucleoside 3′,5′-phosphorothioates. J. Org. Chem. 61, 6713−
6716.
(11) Wada, T., Maizuru, Y., Shimizu, M., Oka, N., and Saigo, K.
(2006) Stereoselective synthesis of dinucleoside boranophosphates by
an oxazaphospholidine method. Bioorg. Med. Chem. Lett. 16, 3111−
3114.
Dual Fluorescence Assay (DFA). Fluorescence values of BACE-
GFP fusion protein and red fluorescent protein (RFP) were measured
using a Synergy HT reader (BIO-TEK); data quantification was
performed using KC4 software. Excitation and emission wavelengths
were as follows: EGFP λEx = 485/20 nm, λEm = 528/20 nm; RFP λEx
=
530/25 nm, λEm = 590/30 nm. The siRNA activity was calculated as
the ratio of EGFP to RFP fluorescence values, averaged over four
independent repetitions. The relative level of fluorescence (EGFP/
RFP) in control cells transfected with pBACE-GFP, pDsRed2-N1 (BD
Biosciences), and control nonsilencing siRNA23 was taken as the
reference (100%).
Serum Stability Assay. Unmodified (PO) and modified (PS2 or
PS) siRNAs (100 pmols/reaction) were diluted in the PBS buffer
(without Ca2+/Mg2+) with 10% FBS (Gibco) and incubated at 37 °C.
(12) Petersen, K. H., and Nielsen, J. (1990) Chemical synthesis of
dimer ribonucleotides containing internucleotidic phosphorodithioate
linkages. Tetrahedron Lett. 31, 911−914.
1219
dx.doi.org/10.1021/cb300078e | ACS Chem. Biol. 2012, 7, 1214−1220

ACS Chemical Biology
Articles
(13) Derrick, W. B., Greef, C. H., Caruthers, M. H., and Uhlenbeck,
O. C. (2000) Hammerhead cleavage of the phosphorodithioate
linkage. Biochemistry 39, 4947−4954.
(14) Greef, C. H., Seeberger, P. H., and Caruthers, M. H. (1996)
Synthesis of phosphorodithioate RNA by the H-phosphonothioate
method. Tetrahedron Lett. 37, 4451−4454.
(15) Wiesler, W. T., Marshall, W. S., and Caruthers, M. H. (1993)
Synthesis and purification of phosphorodithioate DNA. Methods Mol.
Biol. 20, 191−206.
(16) Yang, X., and Mierzejewski, E. (2010) Synthesis of nucleoside
and oligonucleoside dithiophosphates. New J. Chem. 34, 805−819.
(17) Wiesler, W. T., and Caruthers, M. H. (1996) Synthesis of
phosphorodithioate DNA via sulfur-linked, base-labile protecting
groups. J. Org. Chem. 61, 4272−4281.
(18) Sierant, M., Kubiak, K., Kazmierczak-Baranska, J., Warashina,
M., Kuwabara, T., and Nawrot, B. (2009) Evaluation of BACE1
silencing in cellular models. Int. J. Alzheimer's Dis. 2009, 1−10.
(19) Wang, Y., Sheng, G., Juranek, S., Tuschl, T., and Patel, D. J.
(2008) Structure of the guide-strand-containing argonaute silencing
complex. Nature 456, 209−213.
(20) Parker, J. S., Roe, S. M., and Barford, D. (2005) Structural
insights into mRNA recognition from a PIWI domain-siRNA guide
complex. Nature 434, 663−666.
(21) Rana, T. M. (2007) Illuminating the silence: understanding the
structure and function of small RNAs. Nat. Rev. Mol. Cell. Biol. 8, 23−
36.
(22) Yang, X., Hodge, R. P., Luxon, B. A., Shope, R., and Gorenstein,
D. G. (2002) Separation of synthetic oligonucleotide dithioates from
monothiophosphate impurities by anion-exchange chromatography on
a mono-q column. Anal. Biochem. 306, 92−99.
(23) Sipa, K., Sochacka, E., Kazmierczak-Baranska, J., Maszewska, M.,
Janicka, M., Nowak, G., and Nawrot, B. (2007) Effect of base
modifications on structure, thermodynamic stability, and gene
silencing activity of short interfering RNA. RNA 13, 1301−1316.
(24) Allerson, C. R., Sioufi, N., Jarres, R., Prakash, T. P., Naik, N.,
Berdeja, A., Wanders, L., Griffey, R. H., Swayze, E. E., and Bhat, B.
(2005) Fully 2′-modified oligonucleotide duplexes with improved in
vitro potency and stability compared to unmodified small interfering
RNA. J. Med. Chem. 48, 901−904.
(32) Elbashir, S. M., Martinez, J., Patkaniowska, A., Lendeckel, W.,
and Tuschl, T. (2001) Functional anatomy of siRNAs for mediating
efficient RNAi in Drosophila melanogaster embryo lysate. EMBO J. 20,
6877−6888.
(33) Parker, J. S., Roe, S. M., and Barford, D. (2004) Crystal structure
of a PIWI protein suggests mechanisms for siRNA recognition and
slicer activity. EMBO J. 23, 4727−4737.
(34) Volk, D. E., Yang, X., Fennewald, S. M., King, D. J., Bassett, S.
E., Venkitachalam, S., Herzog, N., Luxon, B. A., and Gorenstein, D. G.
(2002) Solution structure and design of dithiophosphate backbone
aptamers targeting transcription factor NF-kappaB. Bioorg. Chem. 30,
396−419.
(35) Yang, X., Fennewald, S., Luxon, B. A., Aronson, J., Herzog, N.
K., and Gorenstein, D. G. (1999) Aptamers containing thymidine 3′-
O-phosphorodithioates: synthesis and binding to nuclear factor-
kappaB. Bioorg. Med. Chem. Lett. 9, 3357−3362.
(36) Parker, J. S., Parizotto, E. A., Wang, M., Roe, S. M., and Barford,
D. (2009) Enhancement of the seed-target recognition step in RNA
silencing by a PIWI/MID domain protein. Mol. Cell 33, 204−214.
(37) Ma, J. B., Yuan, Y. R., Meister, G., Pei, Y., Tuschl, T., and Patel,
D. J. (2005) Structural basis for 5′-end-specific recognition of guide
RNA by the A. fulgidus Piwi protein. Nature 434, 666−670.
(38) Stein, C. A. (1996) Phosphorothioate antisense oligodeox-
ynucleotides: questions of specificity. Trends Biotechnol. 14, 147−149.
(39) Stein, C. A., and Cheng, Y. C. (1993) Antisense oligonucleo-
tides as therapeutic agents--is the bullet really magical? Science 261,
1004−1012.
(40) Grunweller, A., Wyszko, E., Bieber, B., Jahnel, R., Erdmann, V.
A., and Kurreck, J. (2003) Comparison of different antisense strategies
in mammalian cells using locked nucleic acids, 2′-O-methyl RNA,
phosphorothioates and small interfering RNA. Nucleic Acids Res. 31,
3185−3193.
(41) Weitzer, S., and Martinez, J. (2007) hClp1: a novel kinase
revitalizes RNA metabolism. Cell Cycle 6, 2133−2137.
(42) Weitzer, S., and Martinez, J. (2007) The human RNA kinase
hClp1 is active on 3′ transfer RNA exons and short interfering RNAs.
Nature 447, 222−226.
(43) Jackson, A. L., Burchard, J., Leake, D., Reynolds, A., Schelter, J.,
Guo, J., Johnson, J. M., Lim, L., Karpilow, J., Nichols, K., Marshall, W.,
Khvorova, A., and Linsley, P. S. (2006) Position-specific chemical
modification of siRNAs reduces “off-target” transcript silencing. RNA
12, 1197−1205.
(25) Prakash, T. P., Allerson, C. R., Dande, P., Vickers, T. A., Sioufi,
N., Jarres, R., Baker, B. F., Swayze, E. E., Griffey, R. H., and Bhat, B.
(2005) Positional effect of chemical modifications on short
interference RNA activity in mammalian cells. J. Med. Chem. 48,
4247−4253.
(44) Sierant, M., Sobczak, M., Janicka, M., Paduszynska, A., and
Piotrzkowska, D. (2010) Biological and physicochemical character-
ization of siRNAs modified with 2′,2′-difluoro-2′-deoxycytidine
(gemcitabine). New J. Chem. 34, 918−924.
(45) Ma, M. Y., Dignam, J. C., Fong, G. W., Li, L., Gray, S. H., Jacob-
Samuel, B., and George, S. T. (1997) Evaluation of 3-ethoxy-1,2,4-
dithiazoline-5-one (EDITH) as a new sulfurizing reagent in
combination with labile exocyclic amino protecting groups for solid-
phase oligonucleotide synthesis. Nucleic Acids Res, 25, 3590−3593.
(26) Nawrot, B., and Sipa, K. (2006) Chemical and structural
diversity of siRNA molecules. Curr. Top. Med. Chem. 6, 913−925.
(27) Bramsen, J. B., Laursen, M. B., Nielsen, A. F., Hansen, T. B.,
Bus, C., Langkjaer, N., Babu, B. R., Hojland, T., Abramov, M., Van
Aerschot, A., Odadzic, D., Smicius, R., Haas, J., Andree, C., Barman, J.,
Wenska, M., Srivastava, P., Zhou, C., Honcharenko, D., Hess, S.,
Muller, E., Bobkov, G. V., Mikhailov, S. N., Fava, E., Meyer, T. F.,
Chattopadhyaya, J., Zerial, M., Engels, J. W., Herdewijn, P., Wengel, J.,
and Kjems, J. (2009) A large-scale chemical modification screen
identifies design rules to generate siRNAs with high activity, high
stability and low toxicity. Nucleic Acids Res. 37, 2867−2881.
(28) Dowler, T., Bergeron, D., Tedeschi, A. L., Paquet, L., Ferrari, N.,
and Damha, M. J. (2006) Improvements in siRNA properties
mediated by 2′-deoxy-2′-fluoro-beta-D-arabinonucleic acid (FANA).
Nucleic Acids Res. 34, 1669−1675.
(29) Hall, A. H., Wan, J., Shaughnessy, E. E., Ramsay Shaw, B., and
Alexander, K. A. (2004) RNA interference using boranophosphate
siRNAs: structure-activity relationships. Nucleic Acids Res. 32, 5991−
6000.
(30) Wang, Y., Juranek, S., Li, H., Sheng, G., Wardle, G. S., Tuschl,
T., and Patel, D. J. (2009) Nucleation, propagation and cleavage of
target RNAs in Ago silencing complexes. Nature 461, 754−761.
(31) Wang, Y., Juranek, S., Li, H., Sheng, G., Tuschl, T., and Patel, D.
J. (2008) Structure of an argonaute silencing complex with a seed-
containing guide DNA and target RNA duplex. Nature 456, 921−926.
1220
dx.doi.org/10.1021/cb300078e | ACS Chem. Biol. 2012, 7, 1214−1220