TABLE 2. HPLC/MS Analysis of Deprotected Oligonucleotidesa
oxidation extraction largely removes this material from the IL
mediated synthesis.
synthetic
support
retention
time (min)
% total
area
low resolution
MS (-ve mode)
sequence
ApT
In this study, a new strategy to synthesize oligonucleotides
using an ionic liquid as a soluble support has been demonstrated.
Oligonucleotides up to tetrameric species have been synthesized
and have been shown to be comparable to products generated
using standard automated DNA synthesis techniques. Overall,
the solution based ILSS method provides a novel route to
IL
CPG
IL
CPG
IL
CPG
IL
CPG
IL
CPG
IL
26.3
26.3
16.7
16.8
19.0
18.9
25.3
25.3
29.9
29.5
32.4
31.9
99.3
96.1
93.3
89.5
97.1
95.1
91.7
92.5
92.1
94.0
87.3
90.5
554.2
554.2
530.2
530.2
570.2
570.2
545.2
545.2
849.1
849.2
1153.2
1153.2
CpT
GpT
1
31
TpT
construct and characterize (by TLC, H NMR, P NMR, and
MS techniques) oligonucleotides during chain assembly as well
as facile, precipitation based purifications of the intermediates
during chain elongation. At this point, the method has been
demonstrated at the 100-250 µmol scale and should be scalable.
Compared to the solid phase method, it offers the potential
advantage of reduced cost of the relatively expensive solid
supports (CPG, OligoPrep, NittoPhase, etc.). Compared to solu-
tion-phase synthesis using soluble polymer supports, such as
poly(ethylene glycol) (PEG, MW ) 5000-20 000), the ILSS
method with compound 1 (MW ) 214) as the support has a
much better loading capacity. Of course, while we have
demonstrated that the ILSS approach is compatible with standard
oligonucleotide synthesis chemistries, it remains to be deter-
mined if the approach is applicable to longer oligomers.
TpTpT
TpTpTpT
CPG
a
IL ) prepared with ionic liquid support; CPG ) prepared with
controlled pore glass with gene machine; % total area for IL based
oligonucleotides excludes peak area due to IL peaks.
methodology employed, there are small losses associated with
the manipulations involved in the purifications of the com-
pounds. These losses are likely fixed, and larger-scale synthesis
should show higher proportional recoveries.
The oligonucleotides synthesized above in solution have been
compared to the same sequences synthesized on controlled pore
glass (CPG; 1 µmol scale), a commonly used solid support, and
the two systems have been deprotected in parallel. Complete
deprotection of the desired oligonucleotides was achieved by
treating them with concentrated ammonium hydroxide/ethanol
for 48 h at room temperature or 16 h at 60 °C. These conditions
ensure complete cleavage of the cyanoethyl protecting group,
the ionic liquid moiety, any protection of the exocyclic amines
of the bases (Ade, Cyt, and Gua), and the monosuccinate linker.
The oligonucleotides were isolated by removal of the ethanol
and ammonium hydroxide solution under vacuum, redissolution
in water (the solid support is simply settled by centrifugation
for the CPG supported oligomers), and then chromatographic
purification by ion-pairing reverse phase HPLC, anion-exchange
HPLC, or polyacrylamide gel electrophoresis.
Experimental Section
DMTTSucc-IL (3). Compound 2 (1 g, 1.63 mmol), substituted
imidazolium tetrafluoroborate 1 (0.38 g, 1.76 mmol), and DMAP
(
0.052 g, 0.41 mmol) were placed in a dry, nitrogen-purged 100
mL round-bottom flask. To this mixture was added dicyclohexyl-
carbodiimide (DCC) (0.68 g, 3.3 mmol), followed by dry CH CN
20 mL). The reaction mixture was stirred for 3 days at room
3
(
temperature. TLC analysis in 9:1 chloroform:methanol showed the
formation of a more polar product. When the reaction was stopped,
the insoluble DCU byproduct was allowed to settle, and the reaction
mixture was filtered and washed with CH CN several times. The
3
solvent was evaporated, and the residue was again washed with
ether to remove any unreacted DCC and finally collected and dried
under vacuum. The product 3 was obtained as a light brown foam
The products of the ILSS procedure have been compared via
LCMS to those obtained through automated solid-phase syn-
1
(
9
1.1 g, 83% yield). H NMR (acetone-d
6
): δ 10.03 (1H, s, NH),
.10 (1H, s, CH), 7.83 (1H, s, CH), 7.71 (1H, s, CH), 7.60 (1H, s,
CH), 7.50-6.92 (14H, m, Ar-H), 6.33 (1H, t, J ) 8 Hz, CH),
.50 (1H, d, J ) 6 Hz, CH), 4.55-4.51 (2H, m, CH ), 4.70-4.66
1
thesis techniques (Table 2). In all cases, the retention times of
the IL generated material correlate well with the CPG generated
oligomer. The purities of the oligomers prepared by the two
methods are comparable. For the dimers, the IL method gave
products of better purities, whereas for the trimers and tetramers,
the CPG method gave slightly purer products. The origin of
impurities appears to be different in the two methods. In the
ILSS generated trimer and tetramer, the impurities are due to
the presence of small amounts of “n - 1” peaks, which are
likely due to the incomplete detritylation at the stage prior to
coupling. This tritylated material would not be removed during
the extraction and would be deblocked in the detritylation after
coupling giving the n - l oligomer. Thus it is extremely
important to ensure complete detritylation at each step. On the
other hand, thymidine nucleoside is visible in several of the
HPLC traces of oligomers synthesized on CPG. Its presence
5
2
(2H, m, CH ), 4.15 (1H, br s, CH), 4.05 (3, s, CH ), 3.80 (3H, s,
2
3
CH
3
), 3.50-3.40 (2H, m, CH
2
), 2.70 (4H, br s, 2 CH ), 2.61-2.40
2
+
2 3 43 4 9
(2H, m, CH ), 1.42 (3H, s, CH ). ESI-MS for C40H N O : calcd
7
23.3, found 723.4.
HO
T
Succ-IL (4). To a solution of 3 (2.96 g, 3.65 mmol) in
dichloromethane (200 mL) was added 3% v/v TFA in dichlo-
romethane or CH CN (100 mL). During the addition of TFA, the
solution became reddish-orange, and stirring was maintained for
0 min. The product was precipitated from 1:9 ethyl acetate:ethyl
3
2
ether, filtered, redissolved in a minimum amount of the acid
solution, precipitated again, and filtered. The precipitate was rinsed
with 1:9 ethyl acetate:ethyl ether, recovered from the filter by
3
dissolving in CH CN, and evaporated under reduced pressure,
yielding compound 4 as a light brown foam (1.88 g, 96% yield).
1
H NMR (DMSO-d
.10 (1H, s, CH), 7.76 (1H, s, CH), 7.72 and 7.45 (1H, s, CH),
7.70 (1H, s, CH), 6.15 (1H, m, CH), 5.27 and 5.19 (1H, m, CH),
6
): δ 11.40 and 11.39 (total of 1H, 2 s, NH),
9
arises from incomplete coupling or partial detritylation of the
solid support prior to the initial capping step.24 Though this
4
.60 and 3.60 (2H, m, CH
CH
unresolved m, 2 CH
2
), 4.45 (2H, br s, CH
2
), 4.39 (2H, br s,
), 2.61-2.60 (4H,
), 1.77 (3H, s, CH ).
is normally present in materials synthesized on CPG, the post-
2
), 4.24-3.93 (1H, m, CH), 3.86 (3, s, CH
3
2
), 2.40-2.16 (2H, m, CH
2
3
(
23) Hadden, C. E.; Martin, G. E.; Krishnamurthy, V. V. Magn. Reson.
Chem. 2000, 38, 143-147.
24) Pon, R. T.; Usman, N.; Ogilvie, K. K. Biotechniques 1988, 6, 768-
75.
+
ESI-HRMS for C20
H
27
N
4
O
8
: required 451.18289, found 451.18234.
DMT
(
TpTSucc-IL (6a). Compound 4 (0.24 g, 0.45 mmol), thymidine
7
phosphoramidite 5a (0.57 g, 0.77 mmol), and dicyanoimidazole
J. Org. Chem, Vol. 71, No. 20, 2006 7909