A conformationally preorganized universal solid support for efficient oligonucleotide synthesis

Article abstract of DOI:10.1021/ja0284613

A novel, conformationally preorganized nonnucleosidic universal solid support for oligonucleotide synthesis was developed. The solid support featured two chemically equivalent hydroxy groups locked in syn-periplanar orientation and orthogonally protected with 4,4′-dimethoxytrityl and acetyl groups. The solid support was extensively tested in the preparation of oligonucleotides and their phosphorothioate analogues containing 2′-deoxy, 2′-O-methyl, 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 hydroxy group then effected the transesterification of a phosphate linkage between the solid support and the 3′-terminal nucleoside residue to result in a facile release of the oligonucleotide to solution. The kinetics of the release process was studied in a continuous flow of concentrated aqueous ammonium hydroxide at a temperature of 300.15 K. Optimal conditions for the release of oligonucleotides depending on the chemistry of the backbone and 3′-terminal nucleoside residue were formulated. Copyright

Full text of DOI:10.1021/ja0284613

Published on Web 02/08/2003
A Conformationally Preorganized Universal Solid Support for Efficient
Oligonucleotide Synthesis
Andrei P. Guzaev* and Muthiah Manoharan
Department of Medicinal Chemistry, Isis Pharmaceuticals, 2292 Faraday AVenue, Carlsbad, California 92008
Received September 9, 2002 ; E-mail: aguzaev@isisph.com
Scheme 1. Preparation of the Solid Support 5a
The automated synthesis of oligonucleotides is often carried out
on solid supports containing 3′-terminal nucleosides attached via a
readily cleavable ester linkage. One limitation of this approach is
that a minimum of four solid supports is required for the preparation
of unmodified oligodeoxynucleotides. For the synthesis of novel
oligonucleotides, a continuously growing number of supports
carrying other modified nucleosides and 3′-terminal modifiers are
required. In a more straightforward approach, a 3′-terminal nucleo-
side 3′-O-phosphoramidite is coupled to the hydroxy group of a
universal linker, a support-bound mono-acylated 1,2-diol, via a 3′-
O-phosphodiester or -triester moiety.1 Exposure of an assembled
oligonucleotide to aqueous ammonium hydroxide or other bases
removes the acyl protection. The released hydroxy group transes-
terifies the phosphate moiety so that the oligonucleotide is dephos-
phorylated at the 3′-terminus and a derivative of ethylene phosphate
is formed as a side product. This approach eliminates the need for
preparation of sequence-specific solid supports.
-3
a
(a) OsO4/H2O2/H2O/acetone/ether/tBuOH; (b) Ac2O/pyridine; (c) (1)
aminoalkyl CPG/Py, (2) Ac2O/NMI/Py, (3) HATU/HOBT/MeCN/Py then
nPrNH2/MeCN, (4) Ac2O/NMI/Py.
Scheme 2. Synthesis of Oligonucleotides 11-25 on Solid Support
5
and the Structure of Solid Support-bound Oligonucleotide 27
The common limitation of the reported universal linkers is that
the release of oligonucleotides requires prolonged heating with
1,2
concentrated ammonium hydroxide or it occurs under orthogonal
conditions.3 Moreover, because their structure-activity relationship
has not been extensively studied, further improvement of these
reagents has been retarded.
a
In the hydrolysis of RNA, one of the key factors governing the
reaction rate is the distance between the 2′-O and P(V) atoms.4
We hypothesized that locking the two vicinal C-O bonds of an
alkyl 2-hydroxyethyl phosphate in a syn-periplanar conformation
might reduce the distance between the 2-O and P(V) atoms and
hence increase the rate of phosphodiester hydrolysis.
To test the hypothesis, a solid support 5 that featured two
protected hydroxy groups in the vicinal positions with the required
syn-periplanar orientation was synthesized as depicted in Scheme
1
and tested in the oligonucleotide synthesis (Scheme 2). The
protected 20-mer phosphodiester (PO) oligonucleotides 11-14, their
phosphorothioate (PS) analogues 15-18, and chimeric oligonucle-
otides 19-25 (Scheme 2 and Table 1) were assembled on the solid
support 5 using the phosphoramidite building blocks 28-30 (Chart
1).
The support-bound oligonucleotides 6 (Scheme 2) were treated
with concentrated aqueous ammonium hydroxide for 6 h at room
temperature, and the liquid phase was then heated for 8 h at 55 °C.
The product distribution in the deprotection mixtures was analyzed
by reverse-phase HPLC and ES MS to prove that the intended
compounds 11-25 (Table 1) were indeed present in the mixtures
as the predominant products. It is important to note that neither
the corresponding oligonucleotide-3′-phosphates nor any other 3′-
modified products could be detected in the reaction mixtures.
Comparison of the crude yields of 11-18 and 21-24 with those
obtained by synthesis on conventional 3′-O-succinyl nucleoside
solid supports did not reveal any significant differences.
The kinetics of the release of the oligonucleotides 11-25 from
5
the solid support 6 was next studied by the continuous flow method
6
with appropriate modifications. Essentially, concentrated aqueous
ammonium hydroxide (14.3 M, 27.1%) was passed through the solid
support 6 at a constant flow rate and at a temperature of 300.15 K,
and the UV absorbance of the eluate containing the released 11-
25 was recorded as a function of time. The data obtained for the
release of 14 are shown in Figure 1.
2380
9
J. AM. CHEM. SOC. 2003, 125, 2380-2381
10.1021/ja0284613 CCC: $25.00 © 2003 American Chemical Society

C O MMU N I C A T I O N S
constant k ) 1.3 × 10 s-1 (12.9 times more slowly), which
strongly supported the positive effect of conformational preorga-
nization in 9.6
-5
Table 1. Time Required for 95% Release of Oligonucleotides
1
1-25 from the Universal Solid Support 6 with 14.3 M Aqueous
Ammonium Hydroxide at 300.15 K
oligonucleotide
Because of the high complexity of the kinetic scheme, it was
not possible to extract all of the rate constants involved. To provide
guidelines for the practical applications of 5 in oligonucleotide
synthesis, the time required for 95% release of the oligonucleotides
1-25 from 6 was measured (Table 1).
Comparison of these release times demonstrated that compounds
compound
B
R3
X
95% release, min
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
A
G
C
T
A
G
C
T
H
H
H
H
H
H
H
H
MOE
MOE
MOE
MOE
MOE
MOE
OMe
O
O
O
O
S
S
S
S
O
O
S
S
S
238
272
251
256
343
456
365
382
180
129
161
212
159
182
123
1
1
9-25 bearing 2′-O-alkylribonucleoside residues at the 3′-terminus
were released 1.5-2.3 times faster than the corresponding 2′-deoxy
counterparts 12 and 14-18. The release of 20 was comparable to
that of 25, which suggested that the rate acceleration could be
attributed to the presence of the 2′-oxygen rather than to the
structure of the 2′-O-substituent. One might argue that the introduc-
tion of the 2′-alkoxy group to the 3′-terminal nucleoside residue
G
5-Me-U
A
G
5-Me-C
5-Me-U
U
S
O
a
decreased the pK value of the 3′-hydroxy group, which facilitated
the departure of the 3′-terminal nucleoside.
Throughout the series, a positive “thio-effect” was observed; that
is, PO oligonucleotides were released 1.2-1.7 times faster than
the corresponding PS analogues. No substantial dependence of the
release time on the structure of the 3′-terminal base moiety was
observed. However, compounds 12, 16, 19, and 22 where B ) G
were released 6-39% more slowly than the corresponding com-
pounds with other bases, and, thus, for each type of backbone and
sugar modification studied, these oligonucleotides represented the
worst case scenario.
Chart 1. Phosphoramidite Building Blocks 28-30
In conclusion, the universal support 5 is fully compatible with
the conditions of oligonucleotide synthesis. On treatment with
aqueous ammonium hydroxide, the oligonucleotides are quantita-
tively released with favorable kinetics.
Supporting Information Available: Experimental procedures for
2-5 and 27, spectral data for 2-4, HPLC profiles and ES mass spectra
for 11-25 (PDF). This material is available free of charge via the
Internet at http://pubs.acs.org.
References
(
1) Universal solid supports derived from uridine: (a) Crea, R.; Horn, T.
Nucleic Acids Res. 1980, 8, 2331-2348. (b) Schwartz, M. E.; Breaker,
R. R.; Asteriadis, G. T.; Gough, G. R. Tetrahedron Lett. 1995, 36, 27-
Figure 1. The rate of the release of 14 from 6 (O) as a function of the
extent of the release of 14 (14.3 M aqueous ammonium hydroxide, 300.15
K). The linear fit is shown as a red line.
3
0. (c) Lyttle, M. H.; Dick, D. J.; Hudson, D.; Cook, R. M. Nucleosides
Nucleotides 1999, 18, 1809-1824.
(
2) Universal supports derived from cis-3,4-dihydroxytetrahydrofuran: (a)
Scott, S.; Hardy, P.; Sheppard, R. C.; McLean, M. J. In InnoVation
Perspect. Solid-Phase Synthesis; Epton, R., Ed.; Mayflower Worldwide:
Birmingham, U.K., 1994; pp 115-124. (b) Scheuer-Larson, C.; Rosen-
bohm, C.; Jorgensen, T. J. D.; Wengel, J. Nucleosides Nucleotides 1997,
As suggested in Scheme 2, the solid support 6 may rapidly
undergo two concurrent reactions to give the intermediates 7 and
1
6, 67-80. (c) Nelson, P. S.; Muthini, S.; Kent, M. A.; Smith, T. H.
8
1
. The phosphotriester 7 may rapidly release the oligonucleotides
1-25 or it may form 9 concurrently. Compound 8 can only be
Nucleosides Nucleotides 1997, 16, 1951-1959. (d) Nelson, P. S.; Muthini,
S.; Vierra, M.; Acosta, L.; Smith, T. H. BioTechniques 1997, 22, 752-
756. (e) Azhayev, A. V. Tetrahedron 1999, 55, 787-800.
converted to 9, which then releases 11-25 to the solution at a
slower rate. Indeed, the plot in Figure 1 demonstrates that 14 was
released via two concurrent processes. A faster step, which may
be attributed to the release of 14 (ca. 15%) via hydrolysis of 7,
was virtually complete in less than 20 min, while the majority of
oligonucleotide material on the solid phase was believed to be
accumulated in 9. The release of 14 via hydrolysis of 9 obeyed the
kinetic law for a pseudo-first-order reaction in solution with the
(
3) Universal supports derived from 3-amino-1,2-propanediol: (a) Lyttle, M.
H.; Hudson, D.; Cook, R. M. Nucleic Acids Res. 1996, 24, 2793-2798.
(b) Azhayev, A. V.; Antopolsky, M. L. Tetrahedron 2001, 57, 4977-
4986.
(
4) (a) Oivanen, M.; Kuusela, S.; L o¨ nnberg, H. Chem. ReV. 1998, 98, 961-
990. (b) Silnikov, V. N.; Vlassov, V. V. Russ. Chem. ReV. 2001, 70, 491-
508.
(
5) Connors, K. A. Chemical Kinetics: The Study of Reaction Rates in
Solution; Wiley-VCH: Weinheim, Fed. Rep. Germany, 1990; pp 177-
180.
-
5
-1
(6) See the Supporting Information for a detailed discussion.
apparent rate constant k of 16.8 × 10
s . By comparison, 14
was released from the reported 271 (Scheme 2) with the rate
c
JA0284613
J. AM. CHEM. SOC.
9
VOL. 125, NO. 9, 2003 2381