Efficient synthesis of antisense phosphorothioate oligonucleotides using a universal solid support

Article abstract of DOI:10.1016/j.tet.2006.02.040

It is demonstrated that solid support containing a novel universal linker could be efficiently used to synthesize both phosphorothioate oligodeoxyribonucleotides and second-generation 2′-O-methoxyethyloligoribonucleotides with high yield and quality as judged by ion-pair-liquid chromatography-electrospray mass spectroscopy, ³¹P NMR and reversed phase HPLC. Analysis of oligonucleotides shows quality being superior to that produced with standard succinyl-linker solid supports, without contamination of materials resulting from linker or support backbone decomposition.

Full text of DOI:10.1016/j.tet.2006.02.040

Tetrahedron 62 (2006) 4528–4534
Efficient synthesis of antisense phosphorothioate oligonucleotides
using a universal solid support
*
R. Krishna Kumar, Andrei P. Guzaev, Claus Rentel and Vasulinga T. Ravikumar
Isis Pharmaceuticals, 2282 Faraday Avenue, Carlsbad, CA 92008, USA
Received 28 December 2005; revised 10 February 2006; accepted 15 February 2006
Available online 7 March 2006
Abstract—It is demonstrated that solid support containing a novel universal linker could be efficiently used to synthesize both
phosphorothioate oligodeoxyribonucleotides and second-generation 20-O-methoxyethyloligoribonucleotides with high yield and quality as
judged by ion-pair-liquid chromatography–electrospray mass spectroscopy, 31P NMR and reversed phase HPLC. Analysis of
oligonucleotides shows quality being superior to that produced with standard succinyl-linker solid supports, without contamination of
materials resulting from linker or support backbone decomposition.
q 2006 Published by Elsevier Ltd.
1. Introduction
benzoyluridine-50-O-succinyl, attached to CPG.10 Since then
several groups have reported various analogs of universal
linkers.11–26 These supports employ nucleosidic material,
which does not get incorporated into oligonucleotide chains
and hence goes to waste. Alternatively, some non-nucleoside-
based universal supports have also been proposed, but
cleavage of oligomers is either tedious or involves use of
salts or other conditions. In addition, in our hands, evaluation
of at least three of reported universal linker procedures when
repeated and analyzed carefully by ion-pair-liquid chromato-
graphy–electrospray mass spectroscopy (a technique known
to reveal small amounts (!1%) of adducts or linker
molecules), showed the quality of oligonucleotides is not
the same as compared to standard method and was
contaminated with the pendent linker molecule. Also, there
was considerable (ca. 10%) amount of oligonucleotide still
attached to solid support making it an inefficient approach.
These limitations have so far, outweighed the added
convenience of one support material for every sequence,
and pre-derivatized supports are still predominantly
preferred.
Synthesis of oligonucleotides and their analogs have under-
gone revolutionary changes in the last several years.1–3
Currently, for small- as well as large-scale, the most widely
used approach both by academic institutions and by
pharmaceutical companies is the solid-supported synthesis
utilizing b-cyanoethyl protected phosphoramidites of various
nucleosides. The synthesis is performed in automated
DNA/RNA synthesizer machines using controlled pore
glass (CPG) or Amersham Biosciences’ HL30 or PS200
polymeric solid support. The 30-nucleoside is linked to solid
support generally through a succinyl-linker to give support-
bound nucleoside. The synthesis consists of stepwise coupling
of individual monomeric units followed by oxidation of
resulting phosphite triester with iodine to give the phosphate
triester. In case of phosphorothioate analogs sulfurization is
performed and the reagent of choice is phenylacetyl disulfide
(PADS)4,5 or 3H-1,2-benzodithiole-3-one-1,2-dioxide
(Beaucage reagent).6,7 Thus, four solid supports are required
for synthesis of oligodeoxyribonucleotides. The advancement
of antisense therapeutics has resulted in use of RNA–D-
NA–RNA chimeric molecules wherein the wings consists
of 20-O-alkyl RNA nucleotides (20-O-methoxyethyl or 2-O-
methyl).8,9 This has necessitated the need to have a large
number of pre-derivatized polymer supports. To obviate this
need, several research groups have proposed the concept of a
universal solid support. Gough et al. were the first to propose a
universal support containing a linker, 3-anisoyl-20-(30-O-
Herein, we have redesigned the vicinal diol-based linker
system so that the linker remains firmly bound to support
during the deprotection and cleavage steps of DNA
synthesis. Thus, any DNA molecule that is released from
support is not contaminated with linker molecule.
2. Results and discussion
Keywords: Oligonucleotide; Succinate; Detritylation; Universal solid
support.
Early on we realized that a major drawback in design of
various universal linker molecules is the presence of a
hydroxyl group that was derivatized with a cleavable group
*
Corresponding author. Tel.: C1 760 603 2412; fax: C1 760 603 4655;
e-mail: vravikumar@isisph.com
0040–4020/$ - see front matter q 2006 Published by Elsevier Ltd.
doi:10.1016/j.tet.2006.02.040

R. K. Kumar et al. / Tetrahedron 62 (2006) 4528–4534
4529
such as succinyl or oxalyl. Instead, if the linker molecule
possesses a non-hydroxyl functionality (e.g., carboxyl) that
could be directly attached/coupled to amino-derivatized
solid support to form an amide bond, then after oligomer-
ization, the first step during ammonia deprotection is the
unmasking of the protecting group on neighbouring
hydroxyl group followed by intra-molecular attack on
phosphate/phosphorothioate group similar to base-catalyzed
hydrolysis of RNA leading to release of oligonucleotide.
Thus, any oligonucleotide released from solid support has to
be of good quality since the linker molecule will be still
attached to solid support. In addition, if synthesis were to be
efficient we have to demonstrate that there is no measurable
amount of oligonucleotide still attached to solid support.
Thus both quality and quantity (yield) issues could be
addressed successfully if the molecule is designed
appropriately.
during ring opening during loading is capped by forming an
amide bond with n-propyl amine using HATU and HOBt
coupling condition. A loading of 90 mmol/g was obtained
while using HL30 support (recommended loading level by
manufacturer).
2.2. Synthesis of phosphorothioate oligodeoxyribo-
nucleotide
To demonstrate this methodology, a 20-mer phosphoro-
thioate oligodeoxyribonucleotide [PS-d(GTTCTCGCTGG
TGAGTTTCA), ISIS 3521] was chosen as an example.
Three syntheses were carried out (standard nucleoside
containing succinate linked support as control and universal
linker attached support in duplicate). The results are
summarized in Table 1.
2.3. Investigation on yield of oligonucleotide
2.1. Synthesis of universal linker and loading to support
To determine if any oligonucleotide was still attached to
support leading to loss of yield, the support after cleaving
and washing thoroughly to remove the synthesized
oligonucleotide, was incubated with aqueous methylamine
at 55 8C for 14 h. The solution after filtration of solid
support was analyzed by HPLC and also by UV
measurement. No detectable level of oligonucleotide was
observed. Subsequently, the support was dried thoroughly
under high vacuum and then tested for presence of DMT
group by the usual acid treatment (p-toluene sulphonic acid
in CH3CN). No orange color was observed. This clearly
indicates that release of oligonucleotide is quantitative
(Scheme 2).
In our earlier initial communication, we proposed a novel
compound 4 as an efficient and high quality yielding
universal linker molecule.27 Synthesis and loading of this
molecule 4 is shown in Scheme 1. The synthesis starts with
bis-hydroxylation of olefin 1 according to literature
procedure. No significant changes were made to this
procedure after several attempts were made to improve
yield and product obtained was used further without any
additional purification. Chemoselective protection of diol 2
with 4,40-dimethoxytrityl chloride happened to be tricky.
The molecule was highly soluble in water in spite of having
a large lyphophilic DMT group (possibly due to presence of
two carboxyl and one hydroxyl groups) and care should be
taken to obtain reasonable yield (22%). The overall yield of
two steps starting from olefin 1 was low (!10%) but due to
inexpensive nature of starting materials, it did not
discourage us from scaling up this molecule. We were
able to scale up approximately 0.5 kg in a single batch.
Treatment of the DMT compound 3 with acetic anhydride in
pyridine gave the acetoxy protected cyclic anhydride 4 as
colorless foam in almost quantitative yield. Loading of the
universal linker molecule 4 to solid support was carried out
in pyridine with amino-derivatized support to give the
appropriate loading. The free carboxyl group generated
2.4. Investigation on increased level of (nL1)-mer
The quality of an oligonucleotide could be measured and
quantitated by various analytical methods. Depending upon
the technique used, full or partial information could be
obtained. Strong anion exchange chromatography, mass
spectrometry, and 31P NMR spectroscopy may be used for
quantitative assessment of PO-content. Capillary gel
electrophoresis (CGE) could be used for quantitation of
deletion sequences [(nK1)-mers]. For an analytical method
to be good, it should have been demonstrated for its
necessary accuracy, precision, linearity, range, selectivity
and ruggedness for use in routine testing. In our laboratories,
we use state-of-the-art ion-pair high performance liquid
chromatography–mass spectroscopy (IP-LC–MS) technique
as a specific, accurate, and sensitive means of quantitating
oligonucleotides containing deletion sequences, depuri-
nated sequences and 30-terminal phosphorothioate mono-
ester within a matrix of PS-oligonucleotides. Removal of
deletionmers [(nK1)-mers; internal and terminal] on
preparative scale using chromatographic separation tech-
nology is difficult to achieve without significant yield loss.
Incomplete detritylation, potential re-tritylation (due to
reversible reaction), incomplete coupling followed by
incomplete capping and some other unknown mechanisms
are a potential source of formation of (nK1)-mers. The
extent of (nK1)-mer formation is frequently used as a
quality measure of the performance of detritylation
conditions.
Scheme 1. Synthesis of universal linker loaded support: (a) (i) OsO4/H2O2/
H2O/acetone/tBuOH; 30 8C, 24 h; (ii) DMT chloride, pyridine; rt, 20 h;
(b) (i) Ac2O/NMI/Py; (c) (i) Py, rt; (ii) HATU/HOBT/MeCN/Py; (iii)
nPrNH2/MeCN; (iv) Ac2O/NMI/Py.

4530
R. K. Kumar et al. / Tetrahedron 62 (2006) 4528–4534
Table 1. Comparison of oligonucleotides synthesized using standard succinate and universal linker supports
Support used for synthesis
Crude yield
(mg/mmol)
Crude full
length (%)
(RP HPLC)
Purified full
length (%)
(IP-LC–MS)
(Depurinated
species) (%)
(IP-LC–MS)
P]S:P]O
31P NMR)
P]S:P]O
(IP-LC–MS)
(
dA succinate (0849-149)
Universal linker (0849-150)
Universal linker (0849-151)
6.2
5.9
6.4
70
77
77
82.4
86.6
87.2
5.1
2.2
2.3
99.66:0.34
99.62:0.38
99.60:0.40
99.59:0.41
99.64:0.36
99.61:0.39
Depurinated species include n-G, n-A/n-GCH2O, n-ACH2O, 30-TPT.
phosphoramidite synthon (15 equiv), extended contact time
(20 min), use of recycling during coupling and various other
efforts did not lead to any measurable improvement.
2.5. Optimization on first detritylation step
There are at least two ways of removing the DMT group
from the universal linker molecule efficiently viz., use of
excess deblock solution and extending the contact time.
After several experiments aimed at optimized detritylation
condition, we found that efficient removal of DMT group
could be achieved by using the same volume of acid
solution but slowing down the delivery pump to double the
contact time. Alternatively, the detritylation cycle could be
repeated one more time (twice the deblock volume and
time). With this optimized detritylation condition, we
resynthesized the phosphorothioate oligonucleotide (ISIS
3521), purified and analyzed by 31P NMR, RP-HPLC and
IP-LC–MS. Comparable results were achieved between the
two oligonucleotides synthesized (Table 3).
Scheme 2. Investigation of any oligonucleotide attached to support leading
to yield loss.
Analyses of the synthesized phosphorothioate oligonucleo-
tides by IP-LC–MS clearly indicate that removal of DMT
group from the secondary hydroxyl group of the universal
linker molecule is slow as shown by increased levels of dA
(as compared to oligonucleotide synthesized using succinate
loaded support) (Table 2). We reasoned that all other cycles
and conditions being equal, the increased level of dA could
come from inefficient detritylation during the first cycle.
This kind of precise information would not have been
possible with other kinds of techniques. In addition, we have
shown earlier that increased depurination occurs when base-
protected deoxyadenosine has an electron withdrawing
group attached at 30-position (succinate group) as compare0d
to having a phosphate/thioate group. A majority of this 3 -
terminal depurinated species undergoes elimination
followed by fragmentation during ammonia incubation
step leading to formation of 30-terminal phosphorothioate
monoester (30-TPT).28 This possibility is eliminated or
substantially reduced while using universal linker molecule
(Table 2). We also confirmed that increased levels of dA is
not due to inefficient coupling as use of large excess of
Table 3. IP-LC–MS analysis of oligonucleotide synthesized using
optimized detritylation cycle for universal linker support (0830-10Z
using nucleoside-loaded support; 0830-11Zusing universal linker loaded
support)
Species
0830-10 (%)
0830-11 (%)
n [ISIS 3521]
P]O
n-dG
85.8
6.8
0.4
0.5
90.0
6.0
0.3
0.6
n-dA
T
n-dC
n-G
n-A/n-GCH2O
n-ACH2O
30-TPT
Sum
0.9
0.5
0.4
1.0
2.4
1.3
100.0
0.5
0.4
0.2
0.7
0.7
0.6
100.0
2.6. Synthesis of 20-O-methoxyethyl modified RNA
chimera phosphorothioate oligonucleotides
Table 2. IP-LC–MS analysis of oligonucleotide synthesized using standard
succinate and universal linker supports
Due to instability of wild-type DNA, phosphorothioate
oligonucleotides, where one of the non-bridging oxygens of
the internucleotide phosphate is formally replaced by a
sulfur atom, are currently the modification of choice for
design and development of therapeutic drugs. To further
increase the therapeutic value of these phosphorothioate
drugs, several nucleoside modifications have been investi-
gated. Of these, 20-O-methoxyethyl (MOE) modified
oligoribonucleotide chimera has been selected in our
laboratories (at Isis Pharmaceuticals) and multiple drugs
are in various stages of human clinical trials against a
variety of diseases.
Species
0829-149 (%)
0829-150 (%)
0829-151 (%)
n [ISIS 3521]
P]O
n-dG
84.6
7.0
1.2
0.5
2.0
0.7
0.2
0.7
84.8
5.5
1.0
4.9
1.4
0.8
0.1
0.7
85.1
4.9
1.3
4.9
1.6
0.8
0.1
0.5
n-dA
T
n-dC
n-G
n-A/n-GCH2O
n-ACH2O
30-TPT
Sum
2.2
0.9
100.0
0.5
0.3
100.0
0.5
0.3
100.0

R. K. Kumar et al. / Tetrahedron 62 (2006) 4528–4534
4531
Table 4. Comparison of oligonucleotides synthesized using succinate and universal linker supports
Support used for synthesis
Crude yield
(mg/mmol)
Crude full length (%)
(RP HPLC)
Purified full length (%)
(IP-LC–MS)
(nK1) (%)
(IP-LC–MS)
P]S:P]O
(IP-LC–MS)
MOE meC succninate (0830-45)
Universal linker (0830-40)
Universal linker (0830-43)
6.76
7.60
7.13
77
74
70
91.1
90.0
91.2
2.9
3.0
3.2
3.8
4.0
2.9
To demonstrate the applicability of universal linker
molecule for synthesis of MOE oligonucleotides, a 20-mer
the phosphate/phosphorothioate diester charged backbone.
Subsequent treatment with concentrated aqueous
ammonium hydroxide, removes the acetyl protection from
the vicinal hydroxyl group. Intra-molecular attack on
adjacent phosphate/phosphorothioate center followed by
cyclization releases the 30-hydroxyl oligonucleotide
(Scheme 3).
phosphorothioate,
[PS-MOE-(GCTCC)-d(TTCCAC-
TGAT)-MOE-(CCTGC)-30] (ISIS 113715) where deoxycy-
tidine and MOE cytidine have 5-methyl substitution was
chosen as an example. For control experiment, MOE meC
succinate loaded PS200 Primer support (200 mmol/g) was
used. Similar cycle conditions were employed for oligonu-
cleotide synthesis like phosphorothioate oligodeoxyribonu-
cleotide including optimized extended condition for the first
cycle involving removal of DMT group from the universal
linker molecule. The results are summarized in Table 4. In
addition, a portion of crude material obtained from each
synthesis was purified by C18 reversed phase HPLC, the
final DMT removed and then analyzed by ion-pair liquid
chromatography–electrospray mass spectrometry (IP-
LC–MS) (Table 5). The level of n-MOE meC (first base
attached to support) and n-MOE meU (since both
deletionmers have same mass) compare well between
three experiments indicating that overall quality of
oligonucleotide obtained using universal linker attached
support is good with no detectable levels of any 30-
modifications. The sample was analyzed by 31P NMR,
analytical RP-HPLC, and IP-LC–MS.
2.8. Economic and quality impact of using universal
linker solid support
There are both direct and indirect cost savings while using
universal linker loaded solid support. While it is difficult to
precisely quantify the savings, overall there is definitely
substantial economic benefit by switching over to one
inventory of this solid support instead of at least eight
different (four deoxy and four MOE) supports. In addition,
while using succinate loaded supports, in particular
containing 5-methyl MOE cytidine nucleoside, we have
observed instability of the benzoyl group as evidenced by
formation of longer formation, possibly arising out of
branching from the exocyclic amine group.29 Currently, we
don’t know if this happens during loading protocol or upon
storage of the support at room temperature. We do not
observe this longer formation while using universal linker
loaded support. Besides these benefits, one major advantage
for therapeutic applications is that this universal linker
loaded solid supports are no longer considered as starting
materials but rather as raw materials. Eliminating a starting
material in drug development is a considerable advantage
towards cost reduction (both direct and indirect costs) and
more than compensates the additional cost involved in
performing one more cycle containing an amidite as
compared to nucleoside-loaded supports.
Table 5. IP-LC–MS analysis of oligonucleotides synthesized using MOE
meC succinate and universal linker supports
Species
0830-45 (%)
0830-40 (%)
0830-43 (%)
n [ISIS 113715]
P]O
n-dG
93.2
3.8
0.0
0.1
0.3
0.5
0.8
92.9
4.0
0.1
0.1
0.4
0.2
1.0
94.0
2.9
0.1
0.1
0.4
0.3
1.3
n-dA
n-T/n-5medC
n-MOE G
n-MOE meU/n-MOE
meC
n-G
0.0
0.8
0.3
0.2
0.1
0.7
0.4
0.1
0.1
0.6
0.2
0.0
3. Summary and conclusions
n-A/n-GCH2O
n-ACH2O
30-TPT
A novel, conformationally pre-organized non-nucleosidic
universal solid support for oligonucleotide synthesis has
been developed. The solid support featured two chemically
equivalent hydroxy groups locked in syn-periplanar orien-
tation and orthogonally protected with 4,40-dimethoxytrityl
and acetyl groups. The solid support was extensively tested
in preparation 0 of phosphorothioate analogs containing
200-deoxy and 2 -O-methoxyethylnucleoside residues at the
3 -terminus. Upon completion of oligonucleotide chain
assembly, the support-bound oligonucleotide material was
treated with concentrated ammonium hydroxide, which
removed the O-acetyl protection. The deprotected hydroxyl
group then affected the transesterification of a phos-
phorothioate linkage between the solid support and the 30-
terminal nucleoside residue to result in a facile release of the
oligonucleotide to solution. In addition, the solid support
containing this universal linker molecule is stable at room
Sum
100.0
100.0
100.0
Subsequently, multiple oligonucleotides (both deoxy and
MOE) were synthesized at different scales, various supports
and different synthesizers (Amersham Biosciences Akta 10
and Akta 100 DNA/RNA synthesizer); yield and quality
were found to be equivalent or slightly better compared to
succinate loaded supports.
2.7. Mechanism of release of oligonucleotide from
support
A reasonable mechanism for release of oligonucleotide is
depicted in Scheme 3. At the end of oligonucleotide
synthesis, treatment of the support with triethylamine:aceto-
nitrile removes the cyanoethyl group and generates

4532
R. K. Kumar et al. / Tetrahedron 62 (2006) 4528–4534
Scheme 3. Mechanism for release of oligonucleotide from support.
temperature for extended period of time (at least 1 year) as
shown by consistent good quality of oligonucleotides
produced with it. Extended detritylation condition to remove
the DMT group further indicates that this group is very stable.
(OsO4) as described in literature and used without any
additional purification.30
4.1.2. (1a,2a,3a,4a,5a,6a)-5-Hydroxy-6-(4,40-dimethoxy-
trityloxy)-7-oxabicyclo[2.2.1]heptane-2,3-dicarboxylic
acid (3). 4,40-Dimethoxytrityl chloride (38.75 g,
114.5 mmol) was added in aliquots to a solution of compound
2 (16.95 g, 77.5 mmol) in anhydrous pyridine (200 mL) over
a period of 6 h. The reaction mixture was stirred overnight at
room temperature. The solvent was evaporated, and the
residue taken up in ethyl acetate (1 L) and 1 M aqueous
triethylammonium acetate (100 mL). The organic solution
was washed with 1 M aqueous triethylammonium acetate
(120 mL), diluted with methanol (100 mL), dried over
sodium sulfate, and evaporated. The residue was dissolved
in ethyl acetate (250 mL) and treated with ether (750 mL). A
colorless amorphous solid, which precipitates was collected,
washed with ether, and dried to give pure product as a free
4. Experimental
4.1. Materials and methods
Anhydrous acetonitrile (water content!0.001%) was
purchased from Burdick and Jackson (Muskegon, MI). 50-
O-Dimethoxytrityl-30-N,N-diisopropylaminoe-30-O-(2-
cyanoethyl) phosphoramidites (T, dAbz, dCbz, dGibu) were
purchased from Amersham Pharmacia Biotech, Milwaukee,
WI. Toluene was purchased from Gallade, Escondido, CA.
Dichloroacetic acid was purchased from Clariant Life
Sciences. All other reagents and dry solvents were
purchased from Aldrich and used without further purifi-
cation. Primer support HL30 and PS200 was obtained from
Amersham Biosciences, Uppsala, Sweden. 1H-Tetrazole
was purchased from American International Chemical,
Natick, MA. Phenylacetyl disulfide (PADS) was purchased
from Acharya Chemicals, Dombivli, India. 31P NMR
spectra were recorded on a Unity-200 spectrometer (Varian,
Palo Alto, CA) operating at 80.950 MHz. Capillary gel
electrophoresis was performed on a eCAP ssDNA 100 Gel
Capillary (47 cm) on a P/ACE System 5000 using Tris/
borate/7 M urea buffer (all Beckman), running voltage
14.1 kV, temperature 40 8C. Thin-layer chromatography
was performed on silica gel 60F-254 (Merck) plates and
compounds were detected under shorter-wavelength
UV light.
1
acid (27.5 g, 56%). H NMR (pyridine-d5): d 7.87 (2H, m);
7.72 (2H, m); 7.66 (2H, m); 7.34 (2H, m); 7.25 (1H, m); 6.96
(4H, m); 5.35 (1H, d, JZ1.2 Hz); 4.39 (1H, d, JZ6.4 Hz);
4.23 (1H, d, JZ6.4 Hz); 3.85 (1H, d, JZ1.2 Hz); 3.68 (3H,
s); 3.65 (3H, s); 3.26 (1H, d, JZ9.2 Hz); 3.05 (1H, d, JZ
9.2 Hz). 13C NMR (100.573 MHz, DMSO-d6): d 172.1,
171.9, 158.5, 145.7, 136.7, 136.2, 130.0, 129.9, 129.1, 128.1,
127.9, 127.8, 126.9, 87.4, 83.6, 82.1, 76.3, 74.5, 55.3, 46.8,
46.5. HRESMS: calcd for C29H27O9 (MK), 519.1655; found,
519.1663.
4.1.3. (1a,2a,3a,4a,5a,6a)-5-Acetoxy-6-(4,40-dimethoxy-
trityloxy)-7-oxabicyclo[2.2.1]heptane 2,3-dicarban-
hydride (4). Compound 3 (1.57 g, 3.0 mmol) was treated
with acetic anhydride (3.0 g) and pyridine (15 mL) for 3 h
at room temperature. The mixture was concentrated and
co-evaporated with pyridine (5!15 mL) to give the title
compound as colorless foam. Due to instability nature of
molecule (being an anhydride), further purification was not
4.1.1. Preparation of universal linker 1a,2a,3a,
4a,5a,6a)-5,6-dihydroxy-7-7-oxabicyclo[2.2.1]hetane-
2,3-dicarboxylic acid (2). The title compound was
synthesized by bis-hydroxylation of (3aR,4S,7R,7aS)-rel-
3a,4,7,7a-tetrahydro-4,7-epoxyisobenzofuran-1,3-dione, 1
with hydrogen peroxide in presence of osmium tetroxide
1
attempted and was used as such in the next step. H NMR
(pyridine-d5): d 7.73 (2H, m); 7.60–7.55 (overlaps with a
solvent peak, m); 7.40 (2H, m); 7.30 (1H, m); 7.00 (4H, m);

R. K. Kumar et al. / Tetrahedron 62 (2006) 4528–4534
4533
5.71 (1H, d, JZ6.4 Hz); 5.21 (1H, s); 4.52 (1H, d, JZ
6.4 Hz); 3.91 (1H, d, JZ7.2 Hz); 3.74 (3H, s); 3.73 (3H, s);
3.67 (1H, s); 3.63 (1H, d, JZ7.2 Hz); 2.26 (3H, s). 13C
NMR (100.573 MHz, DMSO-d6): d 176.0, 173.3, 158.5,
145.4, 136.6, 136.2, 130.1, 129.9, 129.1, 128.3, 127.8,
127.8, 126.9, 87.4, 83.6, 82.1, 76.3, 74.5, 55.3, 46.8, 46.5,
20.9. IP-LC–MS: calcd for C30H27O9 (MK), 531.161;
found, 531.204.
solutions were prepared using anhydrous CH3CN (ca.
˚
10 ppm) and were dried further by addition of activated 4 A
molecular sieves (w50 g/L). Details of synthesis cycle are
given in Table 6. At the end of each synthesis, the support was
thoroughly dried to determine the crude weight yield, treated
with a solution of triethylamine–CH3CN (1/1, v/v) at room
temperature for 2 h to remove the b-cyanoethyl protecting
groups,34 then treated with 30% aqueous ammonium
hydroxide solution for 12 h at 55 8C to effect release from
support and base deprotection. Yield (expressed in mg of
oligonucleotide/mmol of support),35 31P NMR and analytical
RP-HPLC (full length determination) data were collected for
each synthesis. The crude material obtained from each
synthesis was purified by C18 reversed phase HPLC, the
final DMT removed and then analyzed by ion-pair liquid
chromatography–electrospray mass spectrometry (IP-
LC–MS). The final product after lyophilization was obtained
as a colorless hygroscopic solid (yield: 0.56–0.62 g).
4.2. Loading of universal linker molecule (4) to HL30
solid support
Amino-derivatized HL30 primer support (4.0 g, 0.51 mmol)
was gently shaken with compound 4 (1.39 g, 2.55 mmol) in
pyridine (17 mL) for 4 h. The suspension was filtered, and
solid support was washed with pyridine (3!20 mL). The
solid support was additionally washed with ethyl acetate,
dried, and capped by treating with a mixture of Ac2O–
pyridine–N-methylimidazole–THF (10/10/10/70, v/v/v/v)
for 3 h at room temperature. Finally, the support was
washed with acetonitrile (CH3CN), ethyl acetate and dried.
Loading of 5 (94G0.4 mmol/g) was determined by the
standard DMT assay.
Table 6. Synthesis parameters of cycle used on pharmacia OligoPilot II
synthesizer
Step
Reagent
Volume Time
(mL)
(min)
Detritylation
Coupling
3% Dichloroacetic acid/toluene 72
Phosphoramidite (0.2 M), 1H-tetra- 10, 15
1.5
5
4.3. Capping of solid support
zole (0.45 m) in acetonitrile
PADS (0.2 M) in 3-picoline–
CH3CN (1/1, v/v)
Sulfurization
Capping
36
3
2
The solid support from previous step (1.0 g) was treated
with 0.2 M HATU and 0.15 M HOBT in CH3CN–pyridine
(4/1, 6 mL) for 5 min. The liquid phase was removed, and
solid support was treated with 0.5 M n-propylamine in
CH3CN (5 mL) for 15 min. The solid support was then
washed with CH3CN (5!10 mL) and capped with a
mixture of Ac2O–pyridine–N-methylimidazole–THF
(10/10/10/70, v/v/v/v) for 8 h at room temperature. Finally,
the solid support 5 was washed with CH3CN (50 mL), ethyl
acetate (50 mL) and dried. Loading of 5 (90G0.4 mmol/g)
was determined by standard DMT assay.
Ac2O/pyridine/CH3CN,
NMI/CH3CN
24, 24
4.6. HPLC analysis and purification of oligonucleotides
Analysis and purification of oligonucleotides by reversed
phase high performance liquid chromatography (RP-HPLC)
was performed on a Waters Novapak C18 column (3.9!
300 mm) using a Waters HPLC system (600E System
Controller, 996 Photodiode Array Detector, 717 Auto-
sampler). For analysis an acetonitrile (A)/0.1 M triethylam-
monium acetate gradient was used: 5–35% A from 0 to
10 min, then 35–40% A from 10 to 20 min, then 40–95% A
from 20 to 25 min, flow rateZ1.0 mL/min/50% A from 8
to 9 min, 9 to 26 min at 50% flow rateZ1.0 mL/min,
tR(DMT-off) 10–11 min, tR(DMT-on) 14–16 min. The
DMT-on fraction was collected and was evaporated in
vacuum, redissolved in water and the DMT group was
removed as described below.
4.4. Loading of universal linker molecule (4) to PS200
and controlled pore glass solid supports
Loading of universal linker molecule (4) to amino-
derivatized supports like PS200 and controlled pore glass
(CPG) were similar to the above described protocol for
HL30 except that reagents were adjusted accordingly to
obtain loadings of 200G5 and 45G5 mmol/g, respectively.
4.5. Oligonucleotide synthesis
4.7. Dedimethoxytritylation
All syntheses were performed on a Amersham Biosciences
OligoPilot II DNA/RNA synthesizer at approximately
160 mmol scale in a 6.33 mL fixed column using
b-cyanoethyl phosphoramidite synthons (1.75 equiv,
0.2 M in CH3CN). 1H-Tetrazole (0.45 M in CH3CN) was
used as activator and phenylacetyl disulfide (PADS) (0.2 M
in 3-picoline/CH3CN 1:1, v/v) as sulfur transfer reagent.31
3% Dichloroacetic acid in toluene was used for removal of
acid-labile dimethoxytrityl group32,33 Capping reagents
were made to the recommended Amersham Biosciences
receipe: Cap A: N-methylimidazole–CH3CN (1/4 v/v), Cap
B: acetic anhydride–pyridine–CH3CN (2/3/5, v/v/v). Solid
supports loaded with universal linker molecule at recom-
mended loading levels were used. Amidite and tetrazole
An aliquot (30 mL) was transferred into an Eppendorff tube
(1.5 mL), and acetic acid (50%, 30 mL) was added. After
30 min at room temperature sodium acetate (2.5 M, 20 mL)
was added, followed by cold ethanol (1.2 mL). The mixture
was vortexed and cooled in dry ice for 20 min. The
precipitate was spun down with a centrifuge, the supernatant
was discarded and the precipitate was rinsed with ethanol
and dried under vacuum.
4.8. MS sample preparation
HPLC-purified and dedimethoxytritylated oligonucleotide
was dissolved in 50 mL water, ammonium acetate (10 M,

4534
R. K. Kumar et al. / Tetrahedron 62 (2006) 4528–4534
trials against APOB-100, PTP-1B, VLA4, TRPM2, survivin,
STAT-3. eIF-4E, etc. for the treatment of a variety of diseases
such as cancer, psoriasis, diabetes, asthma, arthritis, multiple
sclerosis, etc.
5 mL) and ethanol were added and vortexed. The mixture
was cooled in dry ice for 20 min and after centrifugation the
precipitate was isolated. This procedure was repeated two
more times to convert the oligonucleotide to the ammonium
form.
10. Gough, G. R.; Brunden, M. J.; Gilham, P. T. Tetrahedron Lett.
1983, 24, 5321–5324.
4.9. IP-HPLC–MS analysis
11. deBear, J. S.; Hayes, J. A.; Koleck, M. P.; Gough, G. R.
Nucleosides Nucleotides 1987, 6, 821–830.
HPLC with UV and MS detection was performed using an
HPLC system consisting of a binary pump, a degasser, a
column oven and a variable wavelength UV detector. After
passing through the UV detector, the column eluate was
introduced directly into a single quadrupole, electrospray
mass spectrometer. Samples were separated using a C18
column (3 mm, 2!150 mm) and eluted at 0.2 mL/min with
a gradient of CH3CN in 5 mM tributylammonium acetate
(TBAA) as the ion-pairing agent.
12. Schwartz, M. E.; Breaker, R. R.; Asteriadis, G. T.; Gough, G. R.
Tetrahedron Lett. 1995, 36, 27–30.
13. Lyttle, M. H.; Hudson, D.; Cook, R. M. Nucleic Acids Res.
1996, 24, 2793–2798.
14. Lyttle, M. H.; Dick, D. J.; Hudson, D.; Cook, R. M.
Nucleosides Nucleotides 1999, 18, 1809–1824.
15. Scheuer-Larsen, C.; Rosenbohm, C.; Joergensen, T. J. D.;
Wengel, J. Nucleosides Nucleotides 1997, 16, 67–80.
16. Kumar, P.; Gupta, K. C. Nucleic Acids Res. 1999, 27, e2.
17. Nelson, P. S.; Muthini, S.; Vierra, M.; Acosta, L.; Smith, T. H.
Biotechniques 1997, 22, 752–756.
18. Kumar, P.; Mahajan, S.; Gupta, K. C. J. Org. Chem. 2004, 69,
6482–6485.
Supplementary data
19. Kumar, P.; Gupta, K. C. React. Funct. Polym. 1999, 41,
197–204.
Supplementary data associated with this article can be found,
in the online version, at doi:10.1016/j.tet.2006.02.040.
Supplementary material contains copies of 31P (D2O) NMR,
reversed-phase HPLC and IP-LC–MS analysis of phos-
phorothioate oligonucleotides (ISIS 3521 and ISIS 113715)
made using standard nucleoside succinate loaded support and
universal linker loaded support.
20. Azhayev, A. V. Tetrahedron 1999, 55, 787–800.
21. Kumar, P.; Gupta, K. C. Helv. Chim. Acta. 2003, 86, 59–64.
22. Kumar, P.; Dhawan, R.; Chandra, R.; Gupta, K. C. Nucleic
Acids Res. 2002, 30, e130.
23. Scheuer-Larson, C.; Rosenbohm, C.; Jorgensen, T. J. D.;
Wengel, J. Nucleosides Nucleotides 1997, 16, 67–80.
24. Nelson, P. S.; Muthini, S.; Kent, M. A.; Smith, T. H.
Nucleosides Nucleotides 1997, 16, 1951–1959.
25. Azhayev, A. V.; Antopolsky, M. L. Tetrahedron 2001, 57,
4977–4986.
References and notes
26. Zheng, X.; Gaffney, B. L.; Jones, R. A. Nucleic Acids Res.
1997, 25, 3980–3983.
1. Beaucage, S. L.; Iyer, R. P. Tetrahedron 1992, 48, 2223–2311.
2. Beaucage, S. L.; Iyer, R. P. Tetrahedron 1993, 49, 1925–1963.
3. Beaucage, S. L. In Oligonucleotide Synthesis In Comprehensive
Natural Product Chemistry; Barton, D. H. R., Nakanishi, K.,
Kool, E. T., Eds.; DNA and Aspects of Molecular Biology;
Pergamon, 1999; p 105.
27. Guzaev, A. P.; Manoharan, M. J. Am. Chem. Soc. 2003, 125,
2380–2381.
28. Ravikumar, V. T.; Kumar, R. K.; Capaldi, D. C.; Turney, B.;
Rentel, C.; Cole, D. L. Org. Proc. Res. Dev. 2003, 7, 259–266.
29. Kurata, C.; Bradley, K.; Gaus, H.; Luu, N.; Cedillo, I.;
Ravikumar, V. T.; Van Sooy, K.; McArdle, J. V.; Capaldi,
D. C. Bioorg. Med. Chem. Lett. 2006, 16, 607–614.
30. Daniels, R.; Fischer, J. L. J. Org. Chem. 1963, 28, 320–322.
31. Krotz, A. H.; Gorman, D.; Mataruse, P.; Foster, C.; Godbout,
J. D.; Coffin, C. C.; Scozzari, A. N. Org. Proc. Res. Dev. 2004,
8, 852–858.
4. Cheruvallath, Z. S.; Wheeler, P. D.; Cole, D. L.; Ravikumar,
V. T. Nucleosides Nucleotides 1999, 18, 485–492.
5. Cheruvallath, Z. S.; Carty, R. L.; Moore, M. N.; Capaldi, D. C.;
Krotz, A. H.; Wheeler, P. D.; Turney, B. J.; Craig, S. R.;
Gaus, H. J.; Scozzari, A. N.; Cole, D. L.; Ravikumar, V. T.
Org. Proc. Res. Dev. 2000, 4, 199–204.
32. Krotz, A. H.; Cole, D. L.; Ravikumar, V. T. Bioorg. Med.
Chem. 1999, 7, 435–439.
6. Iyer, R. P.; Egan, W.; Regan, J. B.; Beaucage, S. L. J. Am.
Chem. Soc. 1990, 112, 1253–1254.
33. Krotz, A. H.; Carty, R. L.; Moore, M. N.; Scozzari, A. N.;
Cole, D. L.; Ravikumar, V. T. Green Chem. 1999, 277–281.
34. Capaldi, D. C.; Gaus, H.; Krotz, A. H.; Arnold, J.; Carty, R. L.;
Moore, M. N.; Scozzari, A. N.; Lowery, K.; Cole, D. L.;
Ravikumar, V. T. Org. Proc. Res. Dev. 2003, 7, 832–838.
35. Experience has taught us that yields expressed in terms of
weight/mmol of support are more reliable than those expressed
in terms of optical density/mmol.
7. Iyer, R. P.;Phillips,L. R.;Egan, W.;Regan,J. B.;Beaucage, S. L.
J. Org. Chem. 1990, 55, 4693–4699.
8. Altmann, K.-H.; Dean, N. M.; Fabbro, D.; Freier, S. M.;
Geiger, T.; Haner, R.; Husken, D.; Martin, P.; Monia, B. P.;
Muller, M.; Natt, F.; Nicklin, P.; Phillips, J.; Pieles, U.;
Sasmor, H.; Moser, H. E. Chimia 1996, 50, 168–176.
9. A number of second-generation phosphorothioate oligo-
nucleotides are in various stages of pre-clinical and clinical