A new and convenient approach for the preparation of β-cyanoethyl protected trinucleotide phosphoramidites

Article abstract of DOI:10.1039/c2ob06934b

Herein we report a convenient approach for the preparation of fully protected trinucleotide synthons to be used for the synthesis of gene libraries. The trinucleotide synthons bear β-cyanoethyl groups at the phosphate residues, and thus can be used in standard oligonucleotide synthesis without additional steps for deprotection and work-up.

Full text of DOI:10.1039/c2ob06934b

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Organic &
Biomolecular
Chemistry
Cite this: Org. Biomol. Chem., 2012, 10, 1510
www.rsc.org/obc
COMMUNICATION
A new and convenient approach for the preparation of β-cyanoethyl protected
trinucleotide phosphoramidites†
Matthäus Janczyk,a,b Bettina Appel,a Danilo Springstubbe,a Hans-Joachim Fritzc and Sabine Müller*a
Received 17th November 2011, Accepted 14th December 2011
DOI: 10.1039/c2ob06934b
of undergoing coupling reactions by the solid phase phosphora-
midite approach, and allowing for oligonucleotide deprotection
and cleavage from support in only one step.‡ In particular, our
method is characterized by the use of a pair of orthogonal pro-
tecting groups for the 5′- and 3′-hydroxyl functions, and the stan-
dard β-cyanoethyl group for protection of the phosphate moiety.
Thus, side products due to competition between the 5′- and 3′-
OH group during trinucleotide synthesis is prevented, and depro-
tection of the synthesized oligomers is achieved under mild con-
ditions, simultaneously with cleavage from the solid support,
and without application of additional protocols as required for
removal of the ortho-chlorophenyl group11–12,15,17 or alkyl
groups9,16 from the phosphate.
The synthesis strategy is outlined in Scheme 1. An N-acyl-5′-
dimethoxytrityl (DMT) protected nucleoside-3′-O-phosphorami-
dite is coupled to an N-acyl-3′-O-tert-butyldimethylsilyl
(TBDMS) protected nucleoside under standard conditions of
phosphoramidite chemistry in solution (for experimental details
see the Supplementary Information†). The resulting dinucleotide
can be extended in both the 3′- and 5′-direction. In the present
study, we decided to remove the DMT group from the 5′-
hydroxyl function followed by coupling of the 5′-deprotected
dinucleotide to an N-acyl-5′-O-DMT protected nucleoside phos-
phoramidite, to give the fully protected trinucleotide (Sup-
plementary Information†).
With 5′-DMT being the group of choice for solid phase DNA
synthesis, the search for a suitable 3′-O-protecting group
becomes a key feature of the procedure. Virnekäs et al. have
used phenoxyacetyl (PAC) for protection of the 3′-OH group.9
Removal of the PAC group requires a nucleophilic reagent,
usually hydroxyl ions. Therefore, the β-cyanoethyl group used
for protection of the phosphate moiety in standard DNA syn-
thesis was replaced with the more stable methyl group. However,
the methyl group needs to be cleaved off with thiophenol or
another thiolating agent, requiring an additional step in oligonu-
cleotide workup. Furthermore, in the mentioned study, the PAC
group was removed with methanolic ammonia, conditions that
are known to give rise to breakage of internucleotidic phospho-
triester linkages.18
Herein we report a convenient approach for the preparation
of fully protected trinucleotide synthons to be used for the
synthesis of gene libraries. The trinucleotide synthons bear
β-cyanoethyl groups at the phosphate residues, and thus can
be used in standard oligonucleotide synthesis without
additional steps for deprotection and work-up.
Over the past two decades, protein studies at the molecular and
submolecular level have greatly gained momentum, with the
focus more recently having shifted from analysis of structure–
function relationship to synthesis, i.e. the generation of proteins,
especially enzymes, with pre-deliberated, novel properties.1 The
latter development was critically facilitated by newly emerging
methods of combinatorial and evolutionary protein engineering
which combine random mutagenesis or combinatorial gene syn-
thesis with functional screening or genetic selection applied at
the phenotype level to an ensemble (“library”) of many structural
variants generated in parallel.2,3
Combinatorial gene synthesis offers the highly attractive
option of restricting structural variation to residues that are not
merely part of the structural scaffold of the protein studied, but
can, instead, be expected to be directly involved in functions
such as catalysis or others. Among a host of related methods,4–7
the use of mixtures of pre-formed trinucleotide blocks represent-
ing codons for the 20 canonical amino acids stands out as allow-
ing fully controlled partial (or total) randomization individually
at any number of arbitrarily chosen codon positions of a given
gene.8–17 Hence, the accessibility of this uniquely flexible exper-
imental regime has created substantial demand of fully protected
trinucleotide synthons of good reactivity in standard oligo-
nucleotide synthesis. Such synthons, however, are not easy to
come by. Procedures reported in the literature so far,8–17 suffer
from one or more limitations. We have developed a new method
that gives easy access to trinucleotide phosphoramidites capable
aErnst-Moritz-Arndt-Universität Greifswald, Institut für Biochemie,
Felix-Hausdorff-Str.4, 17487 Greifswald, Germany. E-mail: smueller@
uni-greifswald.de; Fax: +49 3834 864471; Tel: +49 3834 8622843
bpresent address: Laborbetriebsgesellschaft Dr Dirkes-Kersting und Dr
Kirchner mbH, Rotthauser Str. 19, 45879 Gelsenkirchen, Germany
cGeorg-August Universität Göttingen, Institut für Mikrobiologie und
Genetik, Grisebachstr.8, 37077 Göttingen, Germany
A number of reports describe the synthesis of trinucleotides
by phosphotriester chemistry with o-chlorophenyl as the phos-
phate protecting group.11–17 Here, the levulinyl group13 or the
azidomethylbenzoyl group17 has been used for 3′-OH protection,
†Electronic supplementary information (ESI) available: Experimental
procedures and analytical data. See DOI: 10.1039/c2ob06934b
1510 | Org. Biomol. Chem., 2012, 10, 1510–1513
This journal is © The Royal Society of Chemistry 2012

View Article Online
Fig. 1 MALDI-spectra of a) fully protected and b) 3′-OH-Trinucleo-
tide CAC.
lost simultaneously, such that the free 5′-and 3′-OH functions
compete in the coupling reaction. Accordingly, yields of the syn-
thesized trinucleotides were rather low, and formation of by-pro-
ducts was a strong hurdle.10 We have reinvestigated removal of
the TBDMS group from di- and trinucleotides using fluoride
ions. Traditionally, tetra-n-butylammonium fluoride (TBAF) has
been used as fluoride ion source. Under these conditions, indeed,
we observed a mixture of products, very likely resulting from
cleavage of the phosphotriester bonds (not shown). Alternatively,
we applied TEAx3HF, a reagent that is used in modern RNA
chemistry for removal of the 2′-O-TBDMS group.19 TEAx3HF
is a milder reagent compared to TBAF, allowing for specific
cleavage of the TBDMS group without attacking the phospho-
triester bond. In liquid phase, HF forms long zigzag chains with
linear F–H⋯F– units. Triethyl amine breaks off the structure and
forms the ionic compound HN(Et)3+ −F⋯(H–F)x, in which F−
based on its aggregated structure is a much softer nucleophile.20
Thus, upon treatment of the protected trinucleotide with
TEAx3HF, one main product was generated. Mass spectrometric
analysis confirmed the nature of the product being the intact tri-
nucleotide (Fig. 1, ESI Table S2†).
Next to the choice of an appropriate 3′-O-protecting group, a
second important issue is the nature of the phosphate protecting
group. It would be most desirable using the standard β-cya-
noethyl group in the trinucleotide synthons in order to avoid
additional steps during oligonucleotide work-up. As mentioned
above, most protocols for the synthesis of trinucleotides use
phosphotriester chemistry with o-chlorophenyl as the standard
phosphate protecting group. Application of phosphite triester
chemistry has been combined with methyl or ethyl groups for
phosphite protection. In all cases, special deprotection protocols
are required. We have studied the suitability of the β-cyanoethyl
group for phosphate protection in trinucleotide synthesis, in par-
ticular looking at survival of the β-cyanoethyl group under the
conditions used for the individual steps of trinucleotide prep-
aration. The main problem is the strong sensitivity of the β-cya-
noethyl group to basic conditions. Due to rather strong CH-
acidity, β-cyanoethyl is easily removed by β-elimination.
However, the acidity is only high if the neighbouring phosphor-
ous is oxidized (PV). As long as the phosphorus appears in the
PIII state, the β-cyanoethyl group is stable, even in the presence
of rather strong bases such as triethyl amine (TEA) or diisopro-
pylethyl amine (DIPEA).21 Thus, the first coupling reaction of
the N-acyl-5′-dimethoxytrityl (DMT) protected nucleoside-3′-O-
phosphoramidite to the 5′-OH group of the N-acyl-3′-O-tert-
butymdimethylsilyl (TBDMS) protected nucleoside as outlined
in Scheme 1 proceeds under standard conditions of phosphora-
midite chemistry in solution without loss of the β-cyanoethyl
Scheme 1 Synthesis of fully protected trinucleotides: i) 1 (1 mmol), 2
(0.8 mmol), MeCN, tetrazole (10 mmol), 30 min, RT, iodine solution
(0.1 M in lutidine/THF/water 1 : 2 : 1); ii) DCM with 3% trichloroacetic
acid (1.2 mmol); iii) 4 (1 mmol), 5 (1.2 mmol), MeCN, tetrazol
(10 mmol), 1 h, iodine solution (0.1 M in lutidine/THF/water 1 : 2 : 1).
or the 3′-hydroxyl function was left completely unpro-
tected.11,12,15 The latter strategy seems very convenient, and
focuses on selective reaction of an activated nucleotide with the
primary alcohol. However, not surprisingly, the authors men-
tioned the observation of by-products such as isomeric dimers
and trimers.17 This hampered all following purification steps and
if incorporated into gene libraries, frame shifts and unexpected
mutations occurred. Thus, protection of the 3′-OH group is
strongly required to avoid unwanted by-products in intermediate
and final products.
One of the first reports on the synthesis of trinucleotides
describes application of standard phosphite triester chemistry in
solution using DMT for 5′-OH protection, β-cyanoethyl for pro-
tection of the phosphite and tert-butyldimethylsilyl (TBDMS)
for 3′-OH protection.10 The TBDMS group can easily be
removed by treatment with fluoride ions. This reaction however,
was observed being accompanied by cleavage of phosphotriester
bonds.13 As an alternative, Lyttle et al.10 used 6 N hydrochloric
acid for removal of TBDMS, conditions that are rather harsh and
known to cause depurination. Furthermore, the 5′-DMT group is
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 1510–1513 | 1511

View Article Online
Table 1 Different conditions for 3′-phosphorylation of trinucleotides
2-cyanoethyl-N,N,N′,N′-tetraisopropylphosphordiamidite requires
activation by tetrazole derivatives, and the reaction is not depen-
dent on the presence of strong bases. According to the original
protocol,23 diisopropylammonium tetrazolide is used for acti-
vation of 2-cyanoethyl-N,N,N′,N′-tetraisopropylphosphordiami-
dite. Preparation of trinucleotide phosphoramidites following
this protocol, however, was again accompanied by significant
loss of the β-cyanoethyl group. Reaction of one equivalent of the
tetrazolide with one equivalent of the nucleotide produces two
equivalents diisopropylamine, which obviously is strong enough
to initiate β-elimination of the β-cyanoethyl group. In an attempt
to circumvent this problem, we replaced the tetrazolide with ben-
zylmercapto tetrazole (Table 1). This reaction produces only one
equivalent of diisopropyl amine, which is neutralized by the ben-
zylmercaptotetrazole released back after reaction. Since benzyl-
mercaptotatrazole is more strongly acidic than normal tetrazole
or ethylmercaptotetrazole,24 it acts as a perfect scavenger for di-
isopropyl amine converting it into the ammonium salt. Under
these conditions, preparation of phosphoramidites was most suc-
cessful; only traces of by-products could be detected. Following
this protocol, trinucleotide phosphoramidite synthons represent-
ing codons for all 20 amino acids were prepared. All trinucleo-
tides were successfully incorporated in short test oligo-
nucleotides (see ESI, Table S3†).
Phosphorylation
Entry reagent
Base
Activator
Yielda
n. d.
1
DIPEA
—
2
3
4
5
Pyridine
Lutidine
Collidine
—
—
—
—
n. d.
n. d.
n. d.
n. d.
The trinucleotide phosphoramidites presented here, are advan-
tageous over previously described trinucleotide synthons in
terms of quality and application. Due to the choice of a superior
3′-O-protecting group and their mild and efficient removal, no
by-products or isomeric di- and trimers were observed. Further-
more, using β-cyanoethyl as phosphate protecting group allows
for easy application of the trinucleotide phosphoramidites in
standard DNA synthesis without the need for specific protocols
during work-up and deprotection.
6
—
quantitative
a n. d.: not detectable
group. After oxidation of the phosphorous, however, it is crucial
to avoid contamination with strong bases. Cleavage of the
TBDMS group with TEAx3HF as described above is not proble-
matic, since the reagent is weakly acidic.22 According to the pro-
tocol used for 2′-O-deprotection of RNA, the reaction is
quenched by addition of water. This is not possible here, since
hydrolysis of the reagent results in a strongly acidic solution,
which would cleave off the DMT group. Therefore, we quenched
the reaction by addition of saturated bicarbonate solution with
stringent control of the pH, in order to keep neutral conditions
that leave the DMT group as well as the phosphate β-cyanoethyl
group intact. Following this procedure, 20 trinucleotides of high
purity were prepared (see ESI, Table S2†). The crucial step is the
following preparation of the trinucleotide phosphoramidite
(Table 1). We have evaluated two different ways using either 2-
cyanoethyl-N,N-diisopropyl-chlorophosphoramidite or 2-cyano-
Notes and references
‡S. Müller, M. Janczyk. Verfahren zur Herstellung von Trinukleotiden,
DE102009046981.8, PCT/EP2010/06731.3.
1 T. S. Wong, D. Roccatano and U. Schwaneberg, Environ. Microbiol.,
2007, 9, 2645–2659.
2 S. Sen, V. V. Dasu and B. Mandal, Appl. Biochem. Biotechnol., 2007,
143, 212–223.
3 N. E. Labrou, Curr. Protein Pept. Sci., 2010, 11, 91–100.
4 R. A. M. Hooft von Huijsduijnen, G. Ayala and J. F. DeLamater, Nucleic
Acids Res., 1992, 20, 919.
5 R. C. Cadwell and G. F. Joyce, PCR Methods Appl., 1992, 2, 28.
6 W. P. Stemmer, Proc. Natl. Acad. Sci. U. S. A., 1994, 91, 10747.
7 K. M. Derbyshire, J. J. Salvo and N. D. F. Grindley, Gene, 1986, 46, 145.
8 J. Sondek and D. Shortle, Proc. Natl. Acad. Sci. U. S. A., 1992, 89, 3581.
9 B. Virnekäs, L. Ge, A. Plückthun, K. C. Schneider, G. Wellnhofer and S.
E. Moroney, Nucleic Acids Res., 1994, 22, 5600.
10 M. H. Lyttle, E. W. Napolitano, B. L. Calio and L. M. Kauvar, BioTechni-
ques, 1995, 19, 274.
ethyl-N,N,N′,N′-tetraisopropylphosphordiamidite (Table
1).
11 A. Ono, A. Matsuda, J. Zhao and D. V. Santi, Nucleic Acids Res., 1995,
23, 4677.
12 A. L. Kayushin, M. D. Korosteleva, A. I. Miroshnikov, W. Kosch,
D. Zubov and N. Piel, Nucleic Acids Res., 1996, 24, 3748.
13 A. Zehl, A. Starke, D. Cech, T. Hartsch, R. Merkl and H.-J. Fritz, Chem.
Commun., 1996, 2677.
14 P. Gaytán, J. Yanez, F. Sánchez, H. Mackie and X. Soberon, Chem. Biol.,
1998, 5, 519.
15 A. Kayushin, M. Korosteleva and A. Miroshnikov, Nucleosides, Nucleo-
tides Nucleic Acids, 2000, 19, 1967.
Using 2-cyanoethyl-N,N-diisopropyl-chlorophosphoramidite
requires the presence of DIPEA functioning as scavenger for
HCl, that is generated upon reaction. As expected, preparation of
phosphoramidites following this protocol led to removal of the
β-cyanoethyl group, as analyzed by TLC. Our attempts to
replace DIPEA with less strong bases such as pyridine, lutidine
or collidine (Table 1) failed. In these cases, no formation of the
phosphoramidite could be observed. The alternative reagent
1512 | Org. Biomol. Chem., 2012, 10, 1510–1513
This journal is © The Royal Society of Chemistry 2012

View Article Online
16 J. Yanez, M. Argüello, J. Osuna, X. Soberón and P. Gaytán, Nucleic
Acids Res., 2004, 32, e158.
20 Lehrbuch der Anorganischen Chemie, eds. N. Wiberg, Walter de Gruyter,
Berlin, 1985, p. 282.
17 A. Yagodkin, A. Azhayev, J. Roivainen, M. Antopolsky, A. Kayushin,
M. Korosteleva, A. Miroshnikov, J. Randolph and H. Mackie, Nucleo-
sides, Nucleotides Nucleic Acids, 2007, 26, 473.
18 Oligonucleotide Synthesis: a Practical Approach, ed. M. J. Gait, IRL
Press, Oxford, 1984, p. 13.
21 L. J. McBride and M. H. Caruthers, Tetrahedron Lett., 1983, 24, 245.
22 D. Gasparutto, T. Livache, H. Bazin, A. M. Duplaa, A. Guy, A. Khorlin,
D. Molko, A. Roget and R. Téoule, Nucleic Acids Res., 1992, 20, 5159.
23 H. Saneyoshi, K. Tamaki, A. Ohkubo, K. Seio and M. Sekine, Tetrahe-
dron, 2008, 64, 4370.
19 S. Müller, J. Wolf and S. A. Ivanov, Curr. Org. Synth., 2004, 1, 293.
24 R. Welz and S. Müller, Tetrahedron Lett., 2002, 43, 795.
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 1510–1513 | 1513