High-yield solution-phase synthesis of di- and trinucleotide blocks assisted by polymer-supported reagents

Article abstract of DOI:10.1021/ol0511777

(Chemical Equation Presented) A new solution-phase phosphoramidite approach is reported for oligonucleotide synthesis employing recyclable solid-supported reagents. It uses polyvinyl pyridinium tosylate as the activator of a nucleoside-3′-O-phosphoramidit

Full text of DOI:10.1021/ol0511777

ORGANIC
LETTERS
2005
Vol. 7, No. 16
3485-3488
High-Yield Solution-Phase Synthesis of
Di- and Trinucleotide Blocks Assisted
by Polymer-Supported Reagents
Ce´cile Dueymes,† Andreas Scho
Aude-Emmanuelle Navarro,† Albert Meyer,† Meinolf Lange,‡ Jean-Louis Imbach,†
Fritz Link,‡ Francois Morvan,† and Jean-Jacques Vasseur*,†
1
nberger,†,‡ Ilaria Adamo,†
¸
ERT “Oligonucle´otides: Me´thodologie, Valorisation”, UMR 5625 CNRS-UM2
UniVersity Montpellier II, 34095 Montpellier, France, and Girindus AG,
33790 Halle-Kuensebeck/Westf, Germany
Vasseur@uniV-montp2.fr
Received May 19, 2005
ABSTRACT
A new solution-phase phosphoramidite approach is reported for oligonucleotide synthesis employing recyclable solid-supported reagents. It
uses polyvinyl pyridinium tosylate as the activator of a nucleoside-3 -O-phosphoramidite in the coupling step with a 5 -OH nucleoside or
′
′
dinucleotide. The resulting phosphite triester was either sulfurized or oxidized using polystyrene-bound trimethylammonium tetrathionate or
periodiate. This method avoids complicated purification steps, as excess reagents are easily removed by filtration.
The synthesis of large amounts of oligonucleotides and their
phosphorothiate analogues is crucial for their development
and their manufacture as therapeutics.1 With the help of large-
scale automatic DNA synthesizers, synthesis on solid support
based on the phosphoramidite approach2 is the standard way
to produce oligonucleotides. Though very efficient, this
process is not able to afford very large (multikilogram to
ton) quantities in a few single batches.3 Costly raw materials,
solid supports, and technical materials partly explain this
shortcoming.4 Thus, the investigation of alternative synthetic
methodologies is a matter of particular importance. A few
approaches combining solid phase and solution phase have
been tentatively explored.5-7
Scale-up of a solution-phase synthesis is more predictable
and less expensive than a solid-phase synthesis but requires
purification and isolation processes after each reaction step.
To overcome these limitations, we have designed a new
approach where polymer-bound reagents promote chemical
transformations of the oligonucleotide which is in solution.8
Indeed, solid-supported reagents reduce or even eliminate
purification steps since they are simply removed from the
(5) Mihaichuk, J. C.; Hurley, T. B.; Vagle, K. E.; Smith, R. S.; Yegge,
J. A.; Pratt, G. M.; Tompkins, C. J.; Sebesta, D. P.; Pieken, W. A. Org.
Proc. Res. DeV. 2000, 4, 214.
† University of Montpellier II.
‡ Girindus AG.
(1) Reese, C. B. Tetrahedron 2002, 58, 8893.
(2) Beaucage, S. L.; Iyer, R. P. Tetrahedron 1992, 48, 2223.
(3) Bonora, G. M.; Zaramella, S.; Ravikumar, V. Croat. Chim. Acta 2001,
74, 779.
(4) Sanghvi, Y. S; Guo, Z.; Pfundheller, H. M.; Converso, A. Org. Proc.
Res. DeV. 2000, 4, 175.
(6) Bonora, G. M.; Biancotto, G.; Maffini, M.; Scremin, C. L. Nucleic
Acids Res. 1993, 21, 1213.
(7) Wo¨rl, R.; Ko¨ster, H. Tetrahedron 1999, 55, 2941.
(8) Adamo, I.; Dueymes, C.; Scho¨nberger, A.; Imbach, J.-L.; Meyer, A.;
Morvan, F.; Debart, F.; Vasseur, J.-J.; Lange, M.; Link, F. WO Patent
013154, 2004.
10.1021/ol0511777 CCC: $30.25
© 2005 American Chemical Society
Published on Web 07/12/2005

reaction mixtures by filtration.9 This approach was applied
to the synthesis of several di- and trinucleotides with oxo-
and thiono-phosphotriester internucleosidic linkages.
Solid-phase synthesis using solid-phase reagents and
catalysts, also called inverse solid-phase synthesis10 or hybrid
solid/solution-phase synthesis,11 has found numerous ap-
plications in combinatorial chemistry for high-throughput
screening12-14 and in peptide chemistry.15 In contrast, only
a few examples have been reported in the oligonucleotide
field. For instance, poly(3,5-diethylstyrene)-sulfonyl chloride
was used as the coupling agent to create dinucleoside
phosphodiester bonds according to the out-of-date phos-
phodiester method.16,17 Sulfonic acid resins were proposed
as alternatives of 80% acetic acid to promote detritylation
of 5′-O-DMTr phosphodiester base-deprotected oligonucle-
otides.18
nucleosides 4 assisted by 1 or 2 (10 equiv) led to desired
phosphite triesters 5 (δ 140 ppm) with concomitant disap-
pearance of 4, provided that the shaking of the reaction
mixture was efficient (Scheme 1). However, as the reactions
Scheme 1. Phosphoramidite Coupling Assisted by Supported
Pyridinium Polymersa
Phosphoramidite Coupling. Our examination focused on
the phosphoramidite method. Tetrazole is the most exten-
sively employed activator of the reaction between the 5′-
hydroxyl function of nucleosides and nucleoside 3′-O-
phosphoramidites leading to a phosphite triester internucleo-
sidic linkage. Our concern is that it needs to be of high purity,
involving dangerous sublimation. Because of its potential
to explode,19 tetrazole is not currently sold as a solid powder.
Our inspiration came from previous works on pyridinium
acidic salts that were efficiently used as safer and cheaper
activators.19,20 Polystyrene sulfonic acid resins (DOWEX
50W X8 or Amberlist 15) employed as pyridinium salt 1
and polyvinylpyridinium tosylate 2 were tested as phos-
phoramidite activators. One advantage of polymers 1 and 2
is that they are or could be easily prepared from com-
mercially available resins and regenerated at the end of the
reaction by simple washings. Their loading was determined
by elemental analysis. The coupling reactions were followed
by 31P NMR and by HPLC. It can be noted that polyvinyl
pyridinium hydrochloride was used for the preparation of
nucleoside 3′-O-phosphoramidites starting from 3′-OH nu-
cleosides and 2-cyanoethyl-N,N,N′,N′-tetraisopropyl phos-
phoramidite.4 The low yields obtained in these reactions were
explained by the low availability of the pyridinium functions
on the polymer.
a PV ) polyvinyl, PS ) polystyrene. B1, B2 ) T, 2-N-isobutyryl
G, 4-N-benzoyl C, 6-N-benzoyl A. DMTr ) 4,4′-dimethoxytrityl,
Lev ) Levulinyl,. CNE ) 2-cyanoethyl.
performed with resins 1 gave no reproducible results
concerning the reaction times from one batch to another, our
attention was focused on more reliable polymer 2 (1-2 h of
coupling). The phosphite triesters 5 were contaminated with
the excess of starting phosphoramidites 3 and with cyano-
ethyl H-phosphonate diesters 6 (δ 8 ppm) resulting from
partial hydrolysis of 3 due to the water content of the resin.
All our efforts to get rid of this residual water by extensive
drying of the polymers (washing, coevaporation, high
vacuum) were inefficient. To facilitate the purification (see
below), the unreacted 3 was completely converted to 6 by
addition of water without damaging 5 before the polymers
were removed by filtration. The crude mixtures were directly
used for the next step.
Oxidation and Sulfurization. Dinucleotide phosphotri-
esters and their thiono-phosphotriester analogues are obtained
from phosphite triesters by oxidation and sulfurization,
respectively. In solid-phase DNA synthesis, the oxidation
could be performed in aqueous medium with conventional
I2/H2O or in anhydrous medium with tert-butylhydroperoxide
for example.21 The periodate anion immobilized on an ion-
exchange resin is able to oxidize triphenylphosphane in
triphenylphosphane oxide.22 In contrast to periodate salts,
the efficiency of which is limited by their insolubility in
nonpolar solvents, the polymer-supported periodate can be
used in a diversity of solvents.23
In our case, the coupling between phosphoramidites 3 (δ
148-150 ppm) in excess (1.5 equiv) and 5′-hydroxyl
(9) Ley, S. V.; Baxendale, I. R.; Bream, R. N.; Jackson, P. S.; Leach,
A. G.; Longbottom, D. A.; Nesi, M.; Scott, J. S.; Storer, R. I.; Taylor, S.
J. J. Chem. Soc., Perkin Trans. 1 2000, 3815.
(10) Sherrington, D. C. J. Polym. Sci., Part A: Polym. Chem. 2001, 39,
2364.
(11) Jaunzem, J.; Kashin, D.; Scho¨nberger, A.; Kirschning, A. Eur. J.
Org. Chem. 2004, 3435.
(12) Thompson, L. A. Curr. Opin. Chem. Biol. 2000, 4, 324.
(13) Kirschning, A.; Monenschein, H.; Wittenberg, R. Angew. Chem.,
Int. Ed. 2001, 40, 650.
(14) Bhalay, G.; Dunstan, A.; Glen, A. Synlett 2000, 12, 1846.
(15) Filip, S. V.; Lejeune, V.; Vors, J.-P.; Martinez, J.; Cavelier, F. Eur.
J. Org. Chem, 2004, 1936.
(16) Rubinstein, M.; Patchornik, A. Tetrahedron Lett. 1972, 13, 2881.
(17) Rubinstein, M.; Patchornik, A. Tetrahedron 1975, 31, 1517.
(18) Iyer, R. P.; Jiang, Z.; Yu, D.; Tan, W.; Agrawal, S. Synth. Comm.
1995, 25, 3611.
(21) Hayakawa, Y.; Uchiyama, M.; Noyori, R. Tetrahedron Lett. 1986,
27, 4191.
(19) Eleuteri, A.; Capaldi, D. C.; Krotz, A. H.; Cole, D. L.; Ravikumar,
V. T. Org. Proc. Res. DeV. 2000, 4, 182.
(20) Beier, M.; Pfleiderer, W. HelV. Chim. Acta 1999, 82, 879.
(22) Harrison, C. R.; Hodge, P. J. Chem. Soc., Perkin Trans. 1 1982,
509.
(23) Drewry, D. H.; Coe, D. M.; Poon, S. Med. Res. ReV. 1999, 19, 97.
3486
Org. Lett., Vol. 7, No. 16, 2005

containing organic solutions. We took advantage of this ionic
character to support the tetrathionate ion on anion exchange
polymer resins. The sulfurizing reagents were easily prepared
from cross-linked poly 4-vinyl pyridine or attached to the
quaternary ammonium cations of the polymeric backbone
of Amberlyst A-26 by simple stirring of an aqueous solution
of potassium tetrathionate with the resins. The pyridinium
tetrathionate was not considered further because the ionic
bond between tetrathionate and pyridinum ions was not
sufficiently stable. The elemental analysis of polymer-
supported reagent 10 indicated a loading of 1.81 mmol of
tetrathionate per gram of resin without potassium content,
signifying that the dianion was bound to two quaternary
ammonium cations on the resin. Generally, 5 molar equiv
of 10 was able to sulfurize dinucleotide phosphite triesters
5 (δ 140 ppm) to give rise to thiono-phosphotriesters 11 (δ
68 ppm) in 2-3 h.
Scheme 2. Oxidation of Phosphite Triester Linkages Using a
Polymer-Supported Periodate
At the end of the sulfurization, the resin 12 was filtered
off. Elemental analysis of 12 was in agreement with the
conversion of 1 equiv of tetrathionate into trithionate still
bound to the resin. This is in accord with the mechanism
hypothesized by Efimov28 and proposed in Scheme 3,
The periodate anion was immobilized to the polymeric
backbone of the anion-exchange resin Amberlyst A-26 by
simple interaction with quaternary ammonium cations.24 This
reagent 7 (2 equiv) oxidizes crude mixtures obtained during
the coupling reaction to provide the desired phosphotriester
dinucleosides 8 (δ -2 ppm) contaminated with the 3′-
phosphodiester nucleosides 9 resulting from the oxidation
of the H-phosphonate diesters 6 (δ 0 ppm). Completion of
the reaction was observed within 2 h. The resin was filtered
off and the solvent evaporated.
Scheme 3. Sulfurization of Phosphite Triester Linkages Using
a Polymer-Supported Tetrathionate
There are many sulfur-transfer reagents that have been
developed in oligonucleotide chemistry; the most extensively
employed is the expensive 3-H-1,2-benzodithiol-3-one 1,1-
dioxide25 known as Beaucage’s reagent. However, during
sulfurization, it produces a sulfoxide byproduct that acts as
an oxidizing reagent, giving rise to the contamination of the
desired thiono-phosphotriesters with unwanted phosphotri-
esters.
Recently, the sulfurizing agent xanthane hydride26 was
bound to a polymer, and the resulting reagent was able to
sulfurize nucleoside phosphite triesters27 but was not applied
to oligonucleotides. Although efficient, one drawback of this
supported reagent is the time and the cost of its elaboration.
Reagents ionically attached to organic supports are cheaper
than covalently linked species. Moreover, they are regener-
able.
Several years ago, it was shown that inorganic tetrathionate
is an effective, while sluggish, sulfurizing agent of phosphite
triesters in oligonucleotide chemistry.28 Because of its ionic
nature and consequently its insolubility in nonpolar solvents,
its use as a pyridinium salt was recommended in water-
involving nucleophilic attack of the phosphorus atom of the
phosphite triester 5 on the disulfide bond of the tetrathionate
dianion, leading to a phosphonium intermediate, and followed
by nucleophilic attack by the resulting thiosulfate anion at
the sulfonyl sulfur atom of the phosphonium, giving rise to
the thiono-phosphotriester 11 and the supported trithionate
resin 12. It could be pointed out that the H-phosphonate
diester side product 6 was not sulfurized. When desired (see
delevunylation), the sulfurization was followed, after filtra-
tion of 12, with the oxidation of 6 into the polar phosphodi-
ester 9 promoted by the periodate resin 7. The thiono-
(24) Warning: The preparation is graded as explosive by both the Method
A14 of ChemG [Chemicals Act] and the German Law of Explosive
Materials [2.SprengV].
(25) Iyer, R. P.; Phillips, L. R.; Egan, W.; Regan, J. B.; Beaucage, S. L.
J. Org. Chem. 1990, 55, 4693.
(26) Tang, J. Y.; Han, Y.; Tang, X. J.; Zhang, Z. Org. Proc. Res. DeV.
2000, 4, 194.
(27) Zhang, Z.; Han, Y.; Tang, J. X.; Tang, J. Y. Tetrahedron Lett. 2002,
43, 4347.
(28) Efimov, V. A.; Kalinkina, A. L.; Chakhmakhcheva, O. G.; Schmaltz,
T. S.; Jayaraman, K. Nucleic Acids Res. 1995, 23, 4029.
Org. Lett., Vol. 7, No. 16, 2005
3487

conventional aqueous NaHCO3 workup allowed the removal
of side-products 5′-OH-3′-cyanoethyl phosphodiesters 13
from desired 5′-OH phosphotriesters 14 and 5′-OH H-
phosphonate diester 15 from 5′-OH thiono-phosphotriesters
16. Finally, the trityl residues were withdrawn by precipita-
tion of the crude dinucleotides in diethyl ether, and com-
pounds 14 and 16 were obtained without chromatography.
Scheme 4. Detritylation and Delevunylation of Dinucleotide
Oxo and Thiono-phosphotriesters
Typical processes were run on a 1-10 mmol scale. The
dinucleoside 5′-OH-3′-O-Lev oxo- and thiono-phosphotri-
ester yields for the overall procedure were 81-96%, and
their purity, determined by reverse-phase HPLC, was 90-
96%. The identity of the compounds was confirmed by 31P
NMR and by MALDI-TOF MS. Several trinucleotides were
obtained from 5′-OH dinucleotides 14 and 16 using exacly
the same protocole: coupling, hydrolysis, filtration, oxidation
or sulfurization, filtration, detritylation, aqueous workup, and
precipitation.
To allow elongation from the 3′-side of building blocks,
the delevulinylation reaction was performed from fully
protected crudes containing oxo- or thiono-phosphotriesters
8 or 11, respectively. Before doing it, an additional oxidation
of the side products 6 into phosphodiester 9 was performed
after the sulfurization step (see above). The 3′-O-levulinyl
protecting group is usually selectively removed with a
solution of hydrazine in acetic acid/pyridine.30 We tested here
the ability of a polymer-supported hydrazine 17 to remove
it. This resin was prepared from the strong acid cation
exchanger Amberlyst 15 by washing it with an aqueous
solution of hydrazine. The deprotection of the 3′-OH function
of the 5′-O-DMTr building blocks to afford compounds 18
and 19 was complete within 3 h.
phosphotriester 11 was not affected by this additional
oxidative treatment, and 6 was removed from the reation
mixture by adsorption during filtration on a small pad of
silica gel.
We have developed a chemical approach to construct short
oligonucleotide building blocks using polymer-supported
reagents that allow us to prepare them without chromatog-
raphy purification. These blocks could be either converted
into phosphoramidites or H-phosphonates for further elonga-
tion in solution or on solid support. All the polymer-
supported reagents employed in this study result from ionic
interactions with the organic support. Consequently, they
could be regenerated and reused, reducing their cost and their
impact on the environment.
Detritylation and Delevulinylation. As stated above,
strong acid cation exchangers were proposed as acidic
substitutes for the detritylation of 5′-O-DMTr-deprotected
oligonucleotides in water.18 One main advantage is that the
trityl cation formed during the acidic treatment was trapped
by the resin so that the oligonucleotide was obtained pure
after filtration. The use of “in house” sulfonic acid resins in
organic solvents (dichloromethane-methanol 97.5:2.5, v/v)
was also described for nucleoside chemistry purposes.29 The
reactivity was significantly dependent on the nature of the
polymers. It was reported that during the course of the
detritylation, depurination was not observed.
All our attempts to replace usual acidic treatments for
detritylation with a huge assortment of sulfonic acid resins
failed. The DMTr cation was not fully trapped, but the main
inconvenience was the degradation of the compounds. In our
approach, the crude mixtures obtained from the oxidation
or sulfurization were treated with 2% toluenesulfonic acid
in CH2Cl2-MeOH, 7:3 v/v, at 0 °C for 0.5-1 h. Then,
Acknowledgment. This work was supported by Ministe`re
de l’Education Nationale, de la Recherche et de la Tech-
nologie, France, and by Girindus AG, Germany.
Supporting Information Available: Selected experi-
mental procedures. This material is available free of charge
via the Internet at http://pubs.acs.org.
OL0511777
(29) Patil, S. V.; Mane, R. B.; Salunkhe, M. M. Synth. Comm. 1994, 24,
2423.
(30) van Boom, J.; Burgers, P. M. J. Tetrahedron Lett. 1976, 52, 4875.
3488
Org. Lett., Vol. 7, No. 16, 2005