Synthesis and antisense properties of 2′-O-(2S-methoxypropyl)-RNA- modified gapmer antisense oligonucleotides

Article abstract of DOI:10.1002/cmdc.201402099

To ascertain whether increasing hydrophobicity can enhance the activity of second-generation antisense oligonucleotides (ASOs) in muscle, we investigated the antisense properties of 2′-O-(2S-methoxypropyl)-RNA (2S-MOP)-modified ASOs. Synthesis of the 2S-MOP 5-methyl uridine phosphoramidite was accomplished on a multi-gram scale by Lewis-acid-catalyzed ring opening of 5′-O-tert-butyldiphenylsilyl ether-protected 2,2′-anhydro-5-methyl uridine with 2S-methoxy-1-propanol. Synthesis of the 2S-MOP 5-methyl cytidine nucleoside from the corresponding 5-methyl uridine nucleoside was accomplished by formation and displacement of a 4-triazolide intermediate with aqueous ammonia. 2S-MOP-modified oligonucleotides were prepared on an automated DNA synthesizer and showed similar enhancements in duplex thermal stability as 2′-O-methoxyethyl RNA (MOE)-modified oligonucleotides. 2S-MOP-containing antisense oligonucleotides were evaluated in Balb-c mice and showed good activity for decreasing the expression levels of scavenger receptor B1 (Srb1) and phosphatase and tensin homologue (PTEN) mRNA in liver and muscle tissue.

Full text of DOI:10.1002/cmdc.201402099

CHEMMEDCHEM
COMMUNICATIONS
DOI: 10.1002/cmdc.201402099
Synthesis and Antisense Properties of 2’-O-(2S-Methoxy-
propyl)-RNA-Modified Gapmer Antisense Oligonucleotides
Jinghua Yu, Sanjay K. Pandey, Hetal Khatri, Thazha P. Prakash, Eric E. Swayze, and
Punit P. Seth*[a]
To ascertain whether increasing hydrophobicity can enhance
the activity of second-generation antisense oligonucleotides
(ASOs) in muscle, we investigated the antisense properties of
2’-O-(2S-methoxypropyl)-RNA (2S-MOP)-modified ASOs. Synthe-
sis of the 2S-MOP 5-methyl uridine phosphoramidite was ac-
complished on a multi-gram scale by Lewis-acid-catalyzed ring
opening of 5’-O-tert-butyldiphenylsilyl ether-protected 2,2’-an-
hydro-5-methyl uridine with 2S-methoxy-1-propanol. Synthesis
of the 2S-MOP 5-methyl cytidine nucleoside from the corre-
sponding 5-methyl uridine nucleoside was accomplished by
formation and displacement of a 4-triazolide intermediate with
aqueous ammonia. 2S-MOP-modified oligonucleotides were
prepared on an automated DNA synthesizer and showed simi-
lar enhancements in duplex thermal stability as 2’-O-methoxy-
ethyl RNA (MOE)-modified oligonucleotides. 2S-MOP-contain-
ing antisense oligonucleotides were evaluated in Balb-c mice
and showed good activity for decreasing the expression levels
of scavenger receptor B1 (Srb1) and phosphatase and tensin
homologue (PTEN) mRNA in liver and muscle tissue.
the treatment of metabolic, cardiovascular, cancer, and orphan
diseases.
PS-modified gapmer ASOs generally show best activity in
liver, which is the primary site of accumulation of this class of
molecules.[5] However, chemical modifications that improve
ASO activity in tissues, such as skeletal muscle, could be ex-
tremely useful for developing oligonucleotide-based therapeu-
tics for the treatment of muscular dystrophies.[8] Towards this
end, we recently showed that gapmer ASOs modified with hy-
drophobic modifications such as S-constrained ethyl bridged
nucleic acid (BNA) (S-cEt, 3)[9] or tricyclo-DNA[10] show improved
antisense pharmacology in tissues other than liver.[11] To ascer-
tain whether the antisense properties of MOE oligonucleotides
in extrahepatic tissues could be similarly enhanced by increas-
ing the hydrophobicity of the 2’-O-methoxyethyl substituent,
we synthesized 2’-O-(2S-methoxy)-propyl RNA (2S-MOP, 4) and
examined the antisense properties of this and other modified
gapmer ASOs (Figure 1B).
Crystal structures of a MOE-modified RNA duplex showed
that the MOE side chain is extensively hydrated and projects
into the minor groove of the modified duplex.[12] As a result,
Second-generation antisense oligonucleotides (ASOs) are chi-
meric oligonucleotides that promote degradation of the tar-
geted RNA through the RNase H mechanism.[1,2] These ASOs
are typically fully modified with the phosphorothioate (PS, 1)[3]
linkage and have a central window of 8–16 DNA nucleotides
flanked on either end with 2’-O-methoxyethyl RNA (MOE, 2)
nucleotides.[4] The PS linkage improves ASO metabolic stability
and also enhances ASO binding to plasma proteins. This pre-
vents the ASO from being excreted in the urine and facilitates
ASO distribution into peripheral tissues.[5] The central DNA gap
region serves as the substrate for RNase H, while the flanking
MOE nucleotides enhance ASO affinity for complementary RNA
and further stabilize the ASO from nuclease-mediated metabo-
lism (Figure 1A).[6] Second-generation ASOs represent the most
advanced oligonucleotide therapeutic platform in the clinic.
One second-generation ASO, Kynamro, was recently approved
by the US Food and Drug Administration (FDA) for the treat-
ment of familial hypercholesterolemia.[7] In addition, over 30
ASOs are at various stages of development in clinical trials for
[a] J. Yu, Dr. S. K. Pandey, H. Khatri, Dr. T. P. Prakash, Dr. E. E. Swayze,
Dr. P. P. Seth
Isis Pharmaceuticals, Inc.
2855 Gazelle Court, Carlsbad, CA 92010 (USA)
E-mail: pseth@isisph.com
Figure 1. A) Pictorial representation of the RNase H antisense mechanism of
gene silencing using 2’-modified gapmer antisense oligonucleotides (ASOs);
B) Structures of PS DNA (1), MOE (2), S-cEt (3), 2S-MOP (4) nucleotides.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201402099
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 2040 – 2044 2040

CHEMMEDCHEM
COMMUNICATIONS
www.chemmedchem.org
the MOE side chain creates a sphere of hydration around the
backbone PS linkage. This decreases nonspecific binding of the
ASO to extra- and intracellular proteins and greatly mitigates
the nonspecific toxicities of PS DNA ASOs resulting from pro-
miscuous protein binding.[13] The crystal structure also showed
that the oxygen atom of the MOE side chain exists in a gauche
orientation relative to the 2’-oxygen atom on the RNA furanose
ring. This interplay between the MOE side chain and the
2’-oxygen atom further rigidifies the conformation of the fura-
nose in the C3-endo sugar pucker and improves affinity for
complementary RNA.
Synthesis of the 5-Me-cytidine phosphoramidite 11 was ac-
complished from nucleoside 8 as shown in Scheme 2 using
a slight modification of a previously described procedure.[16]
We hypothesized that introducing
a methyl group in
a stereo-defined manner at the 2-position of the MOE side
chain could enhance the gauche effect leading to further en-
hancement of RNA affinity and metabolic stability towards nu-
clease-mediated degradation.[14] For our initial evaluation, we
decided to focus on the 2S-MOP isomer 4, where the 2-methyl
group was not expected to interfere with the backbone hydra-
tion and was expected to project away from the nucleobase
(Figure 2).
Figure 2. Structures of MOE (2) and 2S-MOP (4) nucleotides in the RNA-like
C3’-endo conformation. A) In MOE, the gauche oxygen atoms trap water
around PS backbone; B) In 2S-MOP, the methyl group reinforces gauche
effect and increases hydrophobicity in minor groove.
Scheme 1. Gram-scale synthesis of 2S-MOP thymidine phosphoramidite 9.
Reagents and conditions: a) Me3Al in heptane (1.5 equiv), 1308C, 66 h, pres-
sure vessel, (42%, 50 g scale); b) Et3N·3HF, Et3N, THF; c) DMTrCl, pyr, RT, 16 h
(65%, two steps); d) N-methyl-imidazole (0.25 equiv), tetrazole (0.8 equiv),
DMF, RT, 8 h, (85%).
The industrial-scale synthesis of MOE pyrimdine nucleosides
involves a Lewis-acid-mediated ring opening of 2,2’-anhydro
nucleosides, such as 5, using methoxyethanol as the solvent.[15]
This process can be conveniently carried out on a multi-kilo-
gram scale with minimal protecting group manipulations. We
envisaged using a similar strategy for the synthesis of the 2S-
MOP thymidine nucleoside 7 (Scheme 1). We anticipated using
optically pure alcohol 6 to effect ring opening of the anhydro
nucleoside (5) in the presence of an appropriate Lewis acid. Al-
cohol 6 was first prepared from the corresponding commer-
cially available ester by reduction with lithium aluminum hy-
dride (Supporting Information).
Reaction of anhydro nucleoside 5 with alcohol 6 in the pres-
ence of trimethylaluminum as the Lewis acid provided nucleo-
side 7 in good yield on a 50 g scale. The tert-butyldiphenylsilyl
ether (TBDPS) group in 7 was removed using triethylamine tri-
hydrofluoride in THF[16] followed by re-protection of the 5’-hy-
droxy group as the 4,4’-dimethoxytrityl (DMTr) ether to provide
nucleoside 8 in good overall yield (65% over two steps). A
phosphitylation reaction provided phosphoramidite 9, which
was now ready for automated oligonucleotide synthesis.
Scheme 2. Gram-scale synthesis of 2S-MOP 5-Me-cytidine phosphoramidite
11. Reagents and conditions: a) Et3N, TMSCl, MeCN, RT, 30 min; b) 1,2,4-tria-
zole, POCl3, 0oC to RT, 1 h; c) Aq NH3, RT, 12 h; d) Bz2O, DMF, RT, 12 h (86%,
four steps); e) N-methyl-imidazole (0.25 equiv), tetrazole (0.8 equiv), DMF, RT,
8 h, (85%).
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 2040 – 2044 2041

CHEMMEDCHEM
COMMUNICATIONS
www.chemmedchem.org
The 3’-hydroxy group in 8 was first transiently protected as the
trimethylsilyl (TMS) ether followed by reaction of the thymine
nucleobase with phosphorus oxychloride and 1,2,4-triazole to
provide the corresponding 4-triazolide. Displacement of the
4-triazolide with aqueous ammonia followed by protection of
the exocyclic amino group as the benzoate ester provided the
5’-O-DMTr-protected 5-Me-cytosine nucleoside (10) in good
yield. A phosphitylation reaction then provided the desired
phosphoramidite (11).
Table 2. Sequence of antisense oligonucleotides (ASOs) targeting mouse
scavenger receptor B1 (Srb1) and phosphatase and tensin homologue
(PTEN) mRNA used for mouse experiment.
ASO Modification Target
Sequence (5’–3’)[a]
Tm [8C][b]
C1
C2
C3
MOE
2S-MOP
S-cEt
Srb1
Srb1
Srb1
GmCTTmCAGTmCATGAmCTTmCmCTT
GmCTTmCAGTmCATGAmCTTmCmCTT
TmCAGTmCATGAmCTTmC
69.7
69.0
59.0
D1
D2
D3
MOE
2S-MOP
S-cEt
PTEN mCTGmCTAGmCmCTmCTGGATTTGA
PTEN mCTGmCTAGmCmCTmCTGGATTTGA
PTEN mCTTAGCACTGGCmCT
67.9
67.1
62.2
Next, we evaluated the ability of 2S-MOP to stabilize oligo-
nucleotide duplexes with complementary RNA in thermal de-
naturation experiments, and the results are given in Table 1.
[a] Base code—A=adenine, G=guanine, T=thymine, mC=5-methyl cy-
tosine; bold and/or underlined letters indicate modified nucleotides;
[b] Tm values were measured in 10 mm sodium phosphate buffer (pH 7.2)
containing 100 mm NaCl and 0.1 mm EDTA using complementary RNA.
Table 1. Duplex-stabilizing properties of 2’-O-methoxyethyl RNA (MOE)-
and 2’-O-(2S-methoxypropyl)-RNA (2S-MOP)-modified oligonucleotides
versus complementary RNA.
more potent than longer ASOs.[9] The longer ASOs C1 and C2
showed higher duplex thermal stability as compared with the
shorter ASO C3 when paired with complementary RNA strands
(Table 2).
Oligonucleotide
Modification
Sequence (5’–3’)[a]
DTm [8C][b]
A1
A2
A3
DNA
MOE
2S-MOP
GCGTTTTTTGCG
GCGTTTTTTGCG
GCGTTTTTTGCG
control
+0.9
+1.1
ASOs C1, C2 and C3 were evaluated in an animal experi-
ment for their ability to inhibit Srb1 mRNA expression in liver,
diaphragm, quadricep, and gastrocnemius muscle. Balb-c mice
were injected subcutaneously twice weekly for three weeks
with ASOs C1, C2 and C3 (25 mgkgÀ1). The animals were then
sacrificed 72 hours after the last dose, the liver, quadricep, gas-
trocnemius, and diaphragm tissues were collected, homogen-
ized and analyzed for changes in Srb1 mRNA levels as com-
pared with saline-treated controls. Plasma transaminase levels
and organ weights were also recorded after sacrifice to assess
tolerability. ASOs C1 and C2 showed excellent activity for de-
creasing Srb1 mRNA in liver and diaphragm, but reduced activ-
ity in quadriceps, heart, and gastrocnemius muscle. In contrast,
the shorter ASO (C3) showed excellent activity in all tissues ex-
amined (Figure 3a). All ASOs were well tolerated with no ad-
verse effects on plasma alanine transaminase (ALT) and blood
urea nitrogen (BUN) levels, or organ and body weights (Fig-
ure 3b–d).
B1
B2
B3
DNA
MOE
2S-MOP
CCAGTGATATGC
CCAGTGATATGC
CCAGTGATATGC
control
+1.7
+1.9
[a] Base code—A=adenine, G=guanine, T=thymine, mC=5-methyl cy-
tosine; bold and/or underlined letters indicate modified nucleotides;
[b] Change in Tm value relative to the appropriate control; values were
measured in 10 mm sodium phosphate buffer (pH 7.2), containing
100 mm NaCl and 0.1 mm EDTA. Sequence of RNA complement 5’-r(AG-
CAAAAAACGC)-3’ for A1–A3 and 5’-r(GCAUAUCACUGG)-3’ for B1–B3.
MOE (2) and 2S-MOP (4) nucleosides were incorporated into
two oligonucleotide sequences (A and B) on an automated
DNA synthesizer using standard phosphoramidite chemistry
(Supporting Information). Evaluation of 2S-MOP-modified oli-
gonucleotides A3 (DTm +1.18C/mod.) and B3 (DTm +1.98C/
mod.) in Tm experiments revealed that the additional methyl
group in the S configuration did not interfere with the ability
of the modification to stabilize oligonucleotide duplexes in
comparison with the corresponding MOE-modified oligonucle-
otides A2 (DTm +0.98C/mod.) and B2 (DTm +1.78C/mod.).
We prepared MOE, 2S-MOP and S-cEt gapmer ASOs C1, C2
and C3, respectively (Table 2), targeting mouse scavenger re-
ceptor B1 (Srb1) mRNA for evaluation in animals. Srb1 is ubiqui-
tously expressed in all tissues, and ASOs targeting this mRNA
have been used previously by us to profile oligonucleotide
chemical modifications in animal experiments.[11] ASOs C1 and
C2 were fully PS-modified 20-mers with a 10-base DNA gap
region flanked on either end with five MOE or 2S-MOP nucleo-
tides. For ASO C2, we used one incorporation of MOE-guanine
at the 5’-end of the ASO because of unavailability of 2S-MOP-
guanine phosphoramidite. ASO C3 also had the same 10-base
DNA gap as C1 and C2, which is flanked on each side with
two S-cEt nucleotides. The use of higher affinity S-cEt nucleo-
tides permits the use of shorter ASOs, which at times can be
To examine whether 2S-MOP could enhance activity in
muscle for a different sequence or target, we investigated
ASOs D1, D2 and D3 targeting another ubiquitously expressed
gene, phosphatase and tensin homologue (PTEN).[17] Balb-C
mice were injected subcutaneously with ASOs D1 and D2
(50 mgkgÀ1) and D3 (25 mgkgÀ1) twice weekly for three
weeks. Animals were sacrificed 72 hours after the last dose,
and liver, quadricep, gastrocnemius and diaphragm tissues
were collected, homogenized and analyzed for changes in
phosphatase and tensin homologue (PTEN) mRNA levels as
compared with saline-treated controls. Plasma transaminase
levels and organ weights were recorded after sacrifice to
assess tolerability. We observed a greater than 90% decrease
in PTEN mRNA in liver for all ASOs. However, 2S-MOP-modified
ASO D2 showed better activity in diaphragm (*p <0.0286) and
quadricep (ns=not significant) as compared with the corre-
sponding MOE-modified ASO, D1. In contrast, D1 showed
better downregulation of PTEN in gastrocnemius (*p <0.0286)
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 2040 – 2044 2042

CHEMMEDCHEM
COMMUNICATIONS
www.chemmedchem.org
Figure 3. Summary of activity and tolerability data for MOE, 2S-MOP and S-cEt antisense oligonucleotides (ASOs) targeting mouse Srb1. Mice (Balb-C, n=4/
group) were injected subcutaneously with ASOs C1, C2 and C3 (25 mgkgÀ1 twice a week for three weeks) targeting mouse scavenger receptor B1 (Srb1). The
mice were sacrificed 72 h after the last ASO dose, tissues were collected and weighed, and plasma chemistries were recorded. A) Selected tissues were homo-
genized and analyzed for the change in Srb1 mRNA content relative to saline-treated controls. B) Alanine transaminase (ALT) and blood urea nitrogen (BUN)
levels and C) organ weights were recorded post-sacrifice. D) Body weights during the three-week study showing minimal differences from saline-treated
groups. All data are expressed as mean ÆSEM.
and heart (*p <0.0286). In general, D1 and D2 showed activity
comparable with D3 in all tissues tested. However, unlike the
Srb1 study where ASOs C1, C2 and C3 were dosed at
25 mgkgÀ1/twice weekly for three weeks, D1 and D2 were
dosed at 50 mgkgÀ1 while D3 was dosed at 25 mgkgÀ1 twice
weekly for three weeks. All ASOs were well tolerated with no
changes in organ weights (Figure 4b) or other parameters of
tolerability.
in multiple extrahepatic tissues. Presumably, the lower molecu-
lar weight and decreased charge facilitates more efficient re-
lease of shorter ASOs from endosomal and lysosomal compart-
ments. Our data further highlight the complex interplay be-
tween ASO length, modification pattern, PS content, and over-
all charge for activity in animal experiments.
In summary, we report the synthesis and antisense activity
of a hydrophobic MOE RNA analogue for antisense applica-
tions. In general, the increased hydrophobicity imparted by
the 2S-MOP modification did not translate to improved activity
in all extrahepatic tissues as compared with the parent MOE
ASOs. However, we did observe improved activity in dia-
phragm and quadricep muscle for the 2S-MOP ASO D2 target-
ing mouse PTEN, suggesting that sequence and target may
play some role in determining ASO activity in muscle. Interest-
ingly, the shorter cEt ASOs generally showed the best activity
Experimental Section
Experimental procedures for preparing phosphoramidites 9
and 11 from compound 5, oligonucleotide synthesis, Tm meas-
urements, and methods for the animal experiments are provid-
ed in the Supporting Information, available via http://
dx.doi.org/10.1002/cmdc.201402099.
For in vivo experiments, the Institutional Animal Care and Use
Committee (IACUC) approved all procedures.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 2040 – 2044 2043

CHEMMEDCHEM
COMMUNICATIONS
www.chemmedchem.org
Keywords: antisense
oligonucleotides
·
gapmers
·
methoxyethyl RNA derivatives · methylation
[1] H. Wu, W. F. Lima, S. T. Crooke, J. Biol. Chem. 1999, 274, 28270–28278.
[2] C. F. Bennett, E. E. Swayze, Annu. Rev. Pharmacol. Toxicol. 2010, 50, 259–
293.
[3] F. Eckstein, Antisense Nucleic Acid Drug Dev. 2000, 10, 117–121.
[4] P. Martin, Helv. Chim. Acta 1995, 78, 486–504.
[5] R. S. Geary, Expert Opin. Drug Metab. Toxicol. 2009, 5, 381–391.
[6] C. F. Bennett in Antisense Drug Technology: Principles, Strategies and Ap-
plications, 2nd ed. (Ed.: S. T. Crooke), CRC Press, Boca Raton, 2007,
pp. 273–304.
[7] R. G. Lee, J. Crosby, B. F. Baker, M. J. Graham, R. M. Crooke, J. Cardiovasc.
Transl. Res. 2013, 6, 969–980.
[8] T. M. Wheeler, A. J. Leger, S. K. Pandey, A. R. MacLeod, M. Nakamori, S. H.
Cheng, B. M. Wentworth, C. F. Bennett, C. A. Thornton, Nature 2012,
488, 111–115.
[9] P. P. Seth, A. Siwkowski, C. R. Allerson, G. Vasquez, S. Lee, T. P. Prakash,
E. V. Wancewicz, D. Witchell, E. E. Swayze, J. Med. Chem. 2009, 52, 10–
13.
[10] D. Renneberg, C. J. Leumann, J. Am. Chem. Soc. 2002, 124, 5993–6002.
[11] S. Murray, D. Ittig, E. Koller, A. Berdeja, A. Chappell, T. P. Prakash, M. Norr-
bom, E. E. Swayze, C. J. Leumann, P. P. Seth, Nucleic Acids Res. 2012, 40,
6135–6143.
[12] M. Teplova, G. Minasov, V. Tereshko, G. B. Inamati, P. D. Cook, M. Mano-
haran, M. Egli, Nat. Struct. Biol. 1999, 6, 535–539.
[13] S. P. Henry, T.-W. Kim, K. Kramer-Strickland, T. A. Zanardi, R. A. Fey, A. A.
Levin in Antisense Drug Technology: Principles, Strategies and Applica-
tions, 2nd ed. (Ed.: S. T. Crooke), CRC Press, Boca Raton, 2007, pp. 327–
364.
[14] V. Kotikam, V. A. Kumar, Tetrahedron 2013, 69, 6404–6408.
[15] B. S. Ross, Q. Song, M. Han, Nucleosides Nucleotides Nucleic Acids 2005,
24, 815–818.
[16] P. P. Seth, G. Vasquez, C. A. Allerson, A. Berdeja, H. Gaus, G. A. Kinberger,
T. P. Prakash, M. T. Migawa, B. Bhat, E. E. Swayze, J. Org. Chem. 2010, 75,
1569–1581.
[17] M. Butler, R. A. McKay, I. J. Popoff, W. A. Gaarde, D. Witchell, S. F. Murray,
N. M. Dean, S. Bhanot, B. P. Monia, Diabetes 2002, 51, 1028–1034.
Figure 4. Summary of activity and tolerability data for MOE, 2S-MOP and S-
cEt antisense oligonucleotides (ASOs) targeting mouse PTEN. Mice (Balb-C,
n=4/group) were injected subcutaneously with ASOs D1, D2 (50 mgkgÀ1
)
and D3 (25 mgkgÀ1) targeting mouse PTEN, twice a week for three weeks.
The mice were sacrificed 72 h after the last ASO dose, and tissues were col-
lected and weighed. A) Selected tissues were homogenized and analyzed for
the change in PTEN mRNA content relative to saline-treated controls.
B) Organ weights were recorded post-sacrifice. All data are expressed as
mean ÆSEM.
Received: April 1, 2014
Published online on May 28, 2014
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 2040 – 2044 2044