Phosphorylating reagent-free synthesis of 5′-phosphate oligonucleotides by controlled oxidative degradation of their 5′-end

Article abstract of DOI:10.1021/ol300542s

The 5′-phosphorylated oligonucleotides (5′-pONs) are currently synthesized using expensive and sensitive modified phosphoramidite reagents. In this work, a simple, cost-effective, efficient, and automatable method is presented, based on the controlled oxidation of the 5′-terminal alcohol followed by a β-elimination reaction. The latter reaction leads to the removal of the terminal 5′-nucleoside and subsequent formation of the 5′-phosphate moiety. Thus, chemical phosphorylation of oligonucleotides (DNA or RNA) is achieved without using modified phosphoramidites.

Full text of DOI:10.1021/ol300542s

ORGANIC
LETTERS
2012
Vol. 14, No. 8
2030–2033
Phosphorylating Reagent-Free Synthesis of
50-Phosphate Oligonucleotides by Controlled
Oxidative Degradation of Their 50-End
ꢀ
Corinne Sallamand, Audrey Miscioscia, Remy Lartia,* and Eric Defrancq
ꢀ
Departement de Chimie Moleculaire, UMR CNRS 5250, Universite Joseph Fourier,
BP 53, 38041 Grenoble Cedex 9, France
ꢀ
ꢀ
remy.lartia@ujf-grenoble.fr
Received March 2, 2012
ABSTRACT
The 50-phosphorylated oligonucleotides (50-pONs) are currently synthesized using expensive and sensitive modified phosphoramidite reagents.
In this work, a simple, cost-effective, efficient, and automatable method is presented, based on the controlled oxidation of the 50-terminal alcohol
followed by a β-elimination reaction. The latter reaction leads to the removal of the terminal 50-nucleoside and subsequent formation of the
50-phosphate moiety. Thus, chemical phosphorylation of oligonucleotides (DNA or RNA) is achieved without using modified phosphoramidites.
Oligonucleotides containing a 50-terminal phosphate
group have found applications in several domains, ranging
from PCR processes,1 gene construction,2 cloning,3
mutagenesis,4 to conjugation reactions.5 The small inter-
fering RNAs (siRNAs), which have demonstrated great
promise in therapeutics for selective inhibition of gene
expression are short 50-pONs.6 The 50-pONs are thus an
important part of the biologist0s toolbox. Such oligonucleo-
tides are routinely prepared by automated synthesis by using
on-support reactions with a modified phosphoramidite
reagent. The importance of this modification is highlighted
by the impressive number of reagents devoted to chemical
phosphorylation that have been developed,7ꢀ9 and some of
them are also commercially available.8ꢀ10
However, despite this wide choice, these reagents are
closely related since they all belong to the phosphoramidite
family. Hence, they share common disadvantages: poor
(7) (a) Celebuski, J. E.; Chan, C.; Jones, R. A. J. Org. Chem. 1992, 57,
5535. (b) van Heden Noort, G. J.; Verhagen, C. P.; van der Horst, M. G.;
Overkleeft, H. S.; van der Marel, G. A.; Filippov, D. V. Org. Lett. 2008,
10, 4461. (c) Meyer, A.; Bouillon, C.; Vidal, S.; Vasseur, J.-J.; Morvan,
F. Tetrahedron Lett. 2006, 47, 8867. (d) Ausın, C.; Grajkowski, A.;
(1) Barany, F. Proc. Natl. Acad. Sci. U. S. A. 1991, 88, 189.
(2) Wosnick, M. A.; Barnett, R. W.; Vicentini, A. M.; Erfle, H.;
Elliott, R.; Sumner-Smith, M.; Mantei, N.; Davies, R. W. Gene 1987, 60,
115.
(3) Sambrook, J.; Fritsch, F.; Maniatis, T. Molecular Cloning: A
Laboratory Manual; Cold Spring Harbor Press: Cold Spring Harbor, NY,
1989.
ꢀ
Cieslak, J.; Beaucage, S. L. Org. Lett. 2005, 7, 4201. (e) Leuck, M.; Vagle,
K. E.; Shawn Roach, J.; Wolter, A. Tetrahedron Lett. 2004, 45, 321.
(f) Lartia, R.; Asseline, U. Tetrahedron Lett. 2004, 45, 5949. (g) Thuong,
N. T.; Chassignol, M. Tetrahedron Lett. 1987, 28, 4157. (h) Uhlmann, E.;
Engels, J. Tetrahedron Lett. 1986, 27, 1023. (i) Cooke, L. A.; Frauendorf,
C.; Gıˆlea, M. A.; Holmes, S. C.; Vyle, J. S. Tetrahedron Lett. 2006, 47,
719. (j) Connolly, B. A. Tetrahedron Lett. 1986, 27, 463.
€
(8) Guzaev, A.; Salo, H.; Azhayev, A.; Lonnberg, H. Tetrahedron
1995, 51, 9375.
(4) Fritz, H.-J. DNA Cloning: A Practical Approach; IRL Press:
Oxford, 1985.
(5) Grimm, G. N.; Boutorine, A. S.; Helene, C. Nucleosides, Nucleo-
tides Nucleic Acids 2000, 19, 1943.
(9) Horn, T.; Urdea, M. S. Tetrahedron Lett. 1986, 27, 4705.
(10) (a) Groody, E. US Patent 5,071,074, 1991. (b) Guzaev, A.;
Azhayev, A.; Lonnberg, H. US Patent 5,959,090, 1999. (c) Urdea, M. S. H. T.
US Patent 5, 332,845, 1994. (d) Vagle, K.; Leuck, M.; Wolter, A. U.S. Pat. Appl.
0230047, November 18, 2004.
(6) (a) Hamilton, A. J.; Baulcombe, D. C. Science 1999, 286, 950.
(b) Bernstein, E.; Caudy, A. A.; Hammond, S. M.; Hannon, G. J. Nature
2001, 409, 363. (c) Gaynor, G. W.; Campbell, B. J.; Cosstick, R. Chem.
Soc. Rev. 2010, 39, 4169.
€
r
10.1021/ol300542s
Published on Web 04/05/2012
2012 American Chemical Society

Scheme 1. Reaction Pathwaya,b
a CNE = cyanoethyl; gray balls refer to controlled pore glass.
b The supported methylthiomethylated ONs linked to the CPG support are not shown for clarity.
stability upon dilution and production by multistep synthe-
sis. Commercialy available reagents are expensive, and
the incorporation of a phosphate group at the 50 end of a
20-mer DNA is as expensive as the synthesis of the 20-mer
moiety itself (see the Supporting Information). Consider-
ing the biological importance of 50-pONs, there is an
urgent need to develop cost-effective phosphorylation
methods in order to expand their use in laboratories.
It has been shown that drugs, such as neocarzinostatin11
and esperamicin,12 or artificial nucleases13 are able to
cleave DNA, affording a single strand DNA terminated
by a 50-phosphate group. Further insights into the mole-
cular mechanism reveal that the release of the 50-pON is
achieved through the oxidation of the 50-carbon to alde-
hyde followed by a β-elimination reaction.14
Consequently, it could be envisioned that the oxidation
of the 50 primary alcohol into aldehyde in the supported
oligonucleotidic chain could be an efficient and cheap
method to generate 50-pON without the use of modified
phosphoramidite. Other groups have previously reported
the formation of 50-phosphate group using β-elimination
reaction from an aldehydic ON, but these methods still
required a modified amidite reagent and an extra step
before the final basic cleavage.15
In this work, we report a facile method to oxidize the
50-hydroxyl group on support-bound ONs affording the
corresponding 50-aldehyde. Subsequent treatment in basic
conditions leads to the removal of the last nucleoside
through a β-elimination mechanism and formation of the
Figure 1. RP-HPLC chromatogrammes of crude reaction mix-
ture of 5 pT3 obtained using (a) commercialy phosphorylation
0
reagent, (b) the present method, and (c) coinjection of both
mixtures. Methylthiomethylated adduct is indicated by an
asterisk (vide infra).
(11) (a) Kappen, L. S.; Goldberg, I. H. Biochemistry 1983, 22, 4872.
(b) Kappen, L. S.; Goldberg, I. H.; Liesch, J. M. Proc. Natl. Acad. Sci.
U. S. A. 1982, 79, 744.
(12) Christner, D. F.; Frank, B. L.; Kozarich, J. W.; Stubbe, J.;
Golik, J.; Doyle, T. W.; Rosenberg, I. E.; Krishnan, B. J. Am. Chem.
Soc. 1992, 114, 8763.
50-pON. Most interestingly, the method is fully compatible
with automated DNA and RNA synthesis.
ꢀ
(13) (a) Pratviel, G.; Pitie, M.; Bernadou, J.; Meunier, B. Angew.
Chem., Int. Ed. Engl. 1991, 30, 702. (b) Angeloff, A.; Dubey, I.; Pratviel,
G.; Bernadou, J.; Meunier, B. Chem. Res. Toxicol. 2001, 14, 1413.
(14) Navacchia, M. L.; Montevecchi, P. C. Org. Biomol. Chem 2006,
4, 3754.
(15) (a) Nadeau, J. G.; Singleton, C. K.; Kelly, G. B.; Weith, H. L.;
Gough, G. R. Biochemistry 1984, 23, 6153. (b) Banmvarth, W.; Wippier,
J. Helv. Chim. Acta 1990, 73, 1139.
A large number of reagents have been reported for
oxidizing alcohol to aldehyde. Preliminary assays using
hypervalent iodine reagent such as IBX or DessꢀMartin
periodinane gave unacceptable amounts of byproducts.
We then turned toward Moffat’s reagent, which consists of
Org. Lett., Vol. 14, No. 8, 2012
2031

Table 1. Phosphorylation Yields for DNA Oligonucleotides (Entries 1ꢀ15) and RNA Oligonucleotides (Entries 16ꢀ21)
ESI-MS data (MW)b
entry
synthesized ONa
expected
observed
yield (%)c
1
(T) pT GTT
1259.2
3362.5
3362.5
3362.5
3362.5
3272.5
3272.5
3392.6
3434.6
5370.8
6346.1
4623.8
5826.0
4912.8
7117.2
3384.3
3384.3
5824.6
2621.3
3802.5
4582.5
1258.7
3362.8
3363.0
3363.0
3362.9
3272.9
3272.9
3392.9
3435.2
5372.2
6347.3
4625.0
5827.0
4914.0
7119.4
3384.5
3384.5
5848.5e
2623.1
3804.9
4584.0
82
80
84
80
81
77
82
83
72
85
87
72
80
76
69
81
79
82
72
89
60
2
(A) pTT TTT TTT TTT
3
(T) pTT TTT TTT TTT
4
(G) pTT TTT TTT TTT
5
(C) pTT TTT TTT TTT
6
(T) pTCT CCT TCC CT
7
(C) pTCT CCT TCC CT
8
(T) pATT TAC TAA AT
9
(T) pGAC GAT CGT TA
10
11
12
13
14
15
16d
17d
18d
19d
20d
21d
(T) pCCT CTC TTT CTC TCT TTC
(T) pCCA TAT CCA ATT CAC ATA CTC
(T) pCAG CTA GAC CAT GCA
(T) pCAT ACA TGA ACA TAC ACT A
(T) pCGA CAT CGA CAT CGC A
(T) pTCA GAT ACT TAG CAT GGA CA A CA
(U) pUU UUU UUU UUU
(C) pUU UUU UUU UUU
(U) pUCU CCU UUC CUC UCU UUC U
(U) pGUG UGU GC
(C) pAUA CAU CCA AUU
(U) pGGU UAG UUG UGG UA
a Base in parentheses refers to the oxidized 50-nucleobase of the starting ODN released during reaction. b Obtained after deconvolution. c Determined
from RP-HPLC profiles of the crude reaction mixture. d Oligoribonucleotides. e Only Naþ adducts were observed.
a mixture of a carbodiimide derivatives and an acid in
DMSO. It has been successfully used for the synthesis of
50-aldehydic nucleosides.16 Preliminary experiments re-
vealed that for handling purposes, diisopropylcarbodii-
mide (DIC) and dichloroacetic acid (DCA) were the best
candidates.
complete conversion of the hydrated aldehyde to 50-pON
(Scheme 1 and the Supporting Information).
The structure of the resulting product was confirmed
by satisfactory 31P NMR and mass data (see the Supporting
Information). In particular, the 31P NMR spectrum showed
the appearance of a new peak at 3.83 ppm attributed to the
phosphomonoester group in addition of the peak corre-
The tetrathymidylate bound to CPG-solid support0was
chosen for the model reaction because the expected 5 pT3
productiseasilyresolvedby RP-HPLCfromtheparentT4.
The reaction mixture was neither stirred nor heated to
ensure a future technological transfer toward automated
DNA synthesizer. Different reaction conditions (i.e., differ-
ent ratios and concentrations of DIC and DCA reagents
and reaction times) were studied. It was found that the
best DIC/DCA ratio is 6:1 (concentrations in reactant of
0.6 and 0.1 M for DIC and DCA, respectively). Use of higher
concentrations did not enhance the yields (see the Support-
ing Info0rmation). We observed a ca. 90% yield for conver-
sion to 5 pT3 in less than 30 min at rt. Moreover, the putative
aldehyde intermediate (Scheme 1) could be trapped by
reacting it with an oxyamine derivative (see the Supporting
Information). It was also observed that heating of the
reaction mixture to 45 °C leads to a decrease in reaction yield.
The β-elimination reaction was concomitant to the final
deprotection step of the support-bound ONs and gave
identical results irrespective of conditions used (i.e., 30%
NH4OH, MA (40% aq MeNH2), 50mMK2CO3 in MeOH).
Evaporation to dryness was found necessary to achieve
sponding to the phosphodiester bonds at ꢀ1.0 ppm.17 The
0
HPLC profile of a coinjection of 5 pT3 synthesized using
commercialy available phosphoramidite9,10c and the present
method showed a single peak (see Figure 1), thus suggesting
that the two products are identical.
Whatever the sequence and the conditions used, a
small amount (2ꢀ5%) of 50-methylthiomethylated pro-
duct (Scheme 1) was also isolated18 (see the Supporting
Information). This byproduct was obtained by reaction
between the 50-alcohol of the CPG-bound ON and the
[CH3;SdCH2]þ cation formed upon elimination from
the DCI adduct of DMSO.16a It should be noted that
this unavoidable impurity could interfere with some
applications.
In order to assess the chemospecificity of this reaction,
the possibility of nucleobase degradation was evaluated.
Purines,19 and particulary contiguous guanines,20 are
(17) Cui, Z.; Zhang, B. Helv. Chim. Acta 2007, 90, 297.
(18) It should be noted that this byproduct is specific to the present
phosphorylation method, compared to those described in refs 7ꢀ10, and
could be used as a “marker”.
ꢀ
(19) Crespo-Hernandez, C. E.; Close, D. M.; Gorb, L.; Leszczynski, J.
J. Phys. Chem. B 2007, 111, 5386.
(20) Saito, I.; Nakamura, T.; Nakatani, K.; Yoshioka, Y.; Yamaguchi,
K.; Sugiyama, H. J. Am. Chem. Soc. 1998, 120, 12686.
(16) (a) Pfitzner, K. E.; Moffatt, J. G. J. Am. Chem. Soc. 1965, 87,
5661. (b) Pfitzner, K. E.; Moffatt, J. G. J. Am. Chem. Soc. 1965, 87, 5670.
2032
Org. Lett., Vol. 14, No. 8, 2012

especially prone to oxidation. Various ONs containing the
four different nucleobases as well as guanine tracks were
prepared, subjected to the aforementionned oxidation
conditions, deprotected, purified, and subjected to enzy-
matic digestion by alkaline phosphatase and nuclease P1
enzymes. RP-HPLC profiles of the resulting mixture were
identical to those obtained from nonoxidized ONs (see the
Supporting Information), thus suggesting the absence of
nucleobase degradation.
The method reported therein was then adapted for
automated ONs synthesizer. The results were initially
poorly reproducible because of the high viscosity of
DMSO solutions, which prevented regular flow through
the fine tubing of the synthesizer. This was resolved by
adding 20% acetonitrile to DMSO. DIC and DCA solu-
tions were found stable for weeks upon dilution in DMSO.
Bottles were clipped on unused amidite positions, and
reagents were mixed in the main valve block by simulta-
neous argon positive pressure (see the Supporting Infor-
mation for modified script). Several 50-phosphorylated
DNA and RNA sequences were then synthesized to vali-
date the method developed in this work (see Table 1). The
nature of the released 50-base was found to have no
influence on reaction yields (Table 1, consider entries
2ꢀ5, entries 6ꢀ7, and entries 16ꢀ17), and for practical
reasons, T (or U) was used as final nucleobase. Yields were
determined from crude reaction mixture by UV monitor-
ing of RP-HPLC analysis. They are consistently good
for short sequences (ca. <15 bases). For some longer
sequences (see entries 14, 15, and 21), slighty lower yields
were obtained.
phosphoramidites. It is fully automatable and does not
need an extra step before nor after final basic deprotection.
Therefore, we believe this method could be valuable in
particular for high throughput synthesis of 50-pONs or for
their synthesis in bulk quantities.21
With the exception of Kool’s work on direct ON
iodination,22 little work has been done on direct functio-
nalization of ONs. Modification of ONs at their 50-end is
almost always achieved by use of the modified phosphor-
amidite. We hope our work will constitute a new approach
for the production of modified ODN. Notably, we are
currently working on the use of the aldehydic intermediate
as a versatile key compound toward other modified ONs.
Acknowledgment. The Nanobio program is acknowl-
edged for providing synthesis and purification facilities.
The authors thank Dr. J.-F. Constant and Dr. P. Murat
(both at University of Grenoble) for fruitful scientific
ꢀ
discussions and R. Gueret (University of Grenoble) for
running ESI-MS analysis. GRAVIT (Grenoble Alpes
Valorisation de l0Innovation et de la Technologie) is aknow-
ledged for financial funding. Dr. Y. Singh (University of
New Jersey) is acknowledged for careful proof-reading.
Supporting Information Available. Enlarged version of
Figure 1. Reaction yield optimization experiments
(reagent concentrations and reaction time). Evidence
for aldehydic intermediate and methylthiomethyl bypro-
duct formation. Enzymatic degradation experiments.
Estimation of costs for the two methods. Modified script
for ABI3400 synthesizer. ESI-MS and RP-HPLC analy-
sis for phosphorylated ONs. This material is available
free of charge via the Internet at http://pubs.acs.org.
The present method uses particulary cheap and
stable reagents contrary to the commercialy available
(21) Defrancq, E.; Lartia, R. U.S. Pat. Appl. 0112285, May 12, 2011.
(22) Miller, G. P.; Kool, E. T. Org. Lett. 2002, 4, 3599.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 8, 2012
2033