5-(Benzylmercapto)-1H-tetrazole as activator for 2′-O-TBDMS phosphoramidite building blocks in RNA synthesis

Article abstract of DOI:10.1016/S0040-4039(01)02274-2

An improved method for the preparation of 5-(benzylmercapto)-1H-tetrazol as activator in RNA synthesis is described. Reaction of benzylthiocyanate with azide delivers the corresponding tetrazole with 72% yield under optimised conditions. We demonstrate that 5-(benzylmercapto)-1H-tetrazol is a superior activator of 2′-O-TBDMS phosphoramidite building blocks. Compared to routinely used 1H-tetrazol, application of a 0.25 M 5-(benzylmercapto)-1H-tetrazol solution in acetonitrile allows for higher coupling yields (>99%), lower coupling times (3 min) and reduced excess of phosphoramidites in solution over the solid-phase nucleotides (8-fold).

Full text of DOI:10.1016/S0040-4039(01)02274-2

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 795–797
5-(Benzylmercapto)-1H-tetrazole as activator for 2%-O-TBDMS
phosphoramidite building blocks in RNA synthesis
Ru¨diger Welz and Sabine Mu¨ller*
Humboldt-Universita¨t zu Berlin, Institut fu¨r Chemie, Fachinstitut fu¨r Organische und Bioorganische Chemie, Brook-Taylor-Str. 2,
12489 Berlin, Germany
Received 29 October 2001; accepted 29 November 2001
Abstract—An improved method for the preparation of 5-(benzylmercapto)-1H-tetrazol as activator in RNA synthesis is described.
Reaction of benzylthiocyanate with azide delivers the corresponding tetrazole with 72% yield under optimised conditions. We
demonstrate that 5-(benzylmercapto)-1H-tetrazol is a superior activator of 2%-O-TBDMS phosphoramidite building blocks.
Compared to routinely used 1H-tetrazol, application of a 0.25 M 5-(benzylmercapto)-1H-tetrazol solution in acetonitrile allows
for higher coupling yields (>99%), lower coupling times (3 min) and reduced excess of phosphoramidites in solution over the
solid-phase nucleotides (8-fold). © 2002 Elsevier Science Ltd. All rights reserved.
The chemical synthesis of RNA oligomers is less facile
than DNA synthesis. In contrast to DNA monomer
building blocks, ribose phosphoramidites have a 2%-
hydroxyl group, which requires additional protection
and which may sterically interfere with the coupling
reaction. Among the 2%-O-protecting groups which have
activation reagents, which are more powerful than the
routinely used 1H-tetrazole.3,4
We have explored the potential of 5-(benzylmercapto)-
1H-tetrazole as activating reagent in RNA synthesis
(Scheme 1). Even though, this tetrazole derivative was
mentioned before in connection with 2%-O-[(triisopropyl-
silyl)oxy]methyl (TOM) monomers,2 to our knowledge
it has never been evaluated with the commercially
widely available and more routinely used TBDMS
building blocks. Using 400 ml of a 0.25 M solution of
this tetrazole derivative in acetonitrile (corresponding
to a 100-fold excess) and a 8-fold excess of dissolved
been
suggested,1
the
(tert-butyl)-dimethylsilyl
(TBDMS) group has found widest application. The
coupling of 2%-O-TBDMS protected monomers under
standard tetrazole activation takes 12 to 15 min with
yields in the range of 97%. Strategies for improvement
of RNA coupling yields involve either new 2%-O-pro-
tecting groups which are sterically less demanding,2 or
Scheme 1. RNA phosphoramidite coupling reaction with 5-(benzylmercapto)-1H-tetrazole activation. Pac: phenoxyacetyl, Ac:
acetyl, i-Pr-Pac: 3-isopropyl-phenoxyacetyl
Keywords: RNA synthesis; activator; 5-(benzylmercapto)-1H-tetrazole.
* Corresponding author. Tel.: ++30 2093 8393 (office), 7511 (lab); fax: ++30 2093 7266; e-mail: sabine.mueller.2@rz.hu-berlin.de
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)02274-2

796
R. Welz, S. Mu¨ller / Tetrahedron Letters 43 (2002) 795–797
pling yields generally is dependent on 5-(benzylmer-
capto)-1H-tetrazole rather than on the nature of the
2%-O-protecting group. Thus, in our opinion both pro-
tecting groups can be used with a small advantage on
the TBDMS side, 2%-O-TBDMS protected phosphor-
amidite building blocks are somewhat cheaper and
more widely available.
phosphoramidites over the solid-phase nucleotide, we
obtained average coupling yields >99%. HPLC traces of
the crude synthesis products clearly show a main peak
for each oligoribonucleotide trityl-off (Fig. 1, RNA 1, 2
and 3 represent a 25, 29 and 42 mer, respectively).
These results demonstrate that the postulated problem
in TBDMS chemistry due to sterical interfering of the
TBDMS group with the coupling reaction obviously
can be overcome by the more potent activator. It is not
clear from the literature,2 what coupling yields have
been obtained applying TOM chemistry with usual
tetrazole activation. Possibly the enhancement of cou-
Comparison of 5-(benzylmercapto)-1H-tetrazole to
other known activators revealed the following order in
coupling yields of 2%-O-TBDMS phosphoramidites: 5-
(benzylmercapto)-1H-tetrazole>5-(ethylmercapto)-1H-
tetrazole>4,5-dicyanoimidazole>1H-tetrazole. The high
activation potential of 5-(benzylmercapto)-1H-tetrazole
is very likely based on its higher acidity (1H-tetrazole:
pKa=4.89; 5-(ethylmercapto)-1H-tetrazole: pKa=4.28;
5-(benzylmercapto)-1H-tetrazole: pKa=4.08).5 Thus, in
the first step of activation the trivalent phosphorus is
more readily protonated and subsequently the coupling
reaction is accelerated. Due to the aromatic substituent,
hydrophilicity of 5-(benzylmercapto)-1H-tetrazole is
considerably decreased, which allows for preparation of
an absolutely water-free activator solution.
5-(Benzylmercapto)-1H-tetrazole can easily be synthe-
sised in the laboratory, which is another advantage in
terms of saving costs for the otherwise rather expensive
activator reagents. Typically, it is prepared from reac-
tion of benzylthiocyanate with sodium azide and
ammonium chloride as acidic catalyst. Protocols pro-
vided in the literature, however, yield only 27%5,7 or
48%8 of the product with the latter being raised to 63%
by switching from 25 to 105 mmol scale.8 In order to
find a strategy which allows for fast, inexpensive and
efficient preparation of 5-(benzylmercapto)-1H-tetra-
zole we screened a variety of reaction conditions (Table
1).
The major problem is to dissolve both, the organic
benzylthiocyanate and the inorganic salts (sodium azide
and ammonium chloride). Thus, either DMF or an
organic solvent/water mixture has to be used. DMF is
difficult to remove after reaction and usually is present
in trace amounts in the final product. Solvent/water
mixtures also are disadvantageous, since benzylthio-
cyanate undergoes classical nucleophilic substitution
with NaN3 as a side reaction, that is even more facili-
tated with the azide anion in aqueous media.7 To
circumvent these problems, initially we carried out the
reaction with lithium azide, which is better soluble in
organic solvents than its sodium analogue. For the
same purpose ammonium acetate was used instead of
ammonium chloride as a source of proton. The reaction
was carried out over 6 h in absolute dioxan at elevated
temperatures and delivered 5-(benzylmercapto)-1H-tet-
razole in 65% yield (Table 1, line 1). Since the prepara-
tion of lithium azide is rather time consuming and
expensive,10 we reinvestigated the synthesis of 5-(ben-
zylmercapto)-1H-tetrazole with sodium azide and
ammonium chloride as hydrazonic acid source. Under
phase-transfer conditions as reported by LeBlanc and
Figure 1. HPLC traces of crude RNA 1
CCG AAA GGU GAA CCA G), 2 (AUC UCU CUU UCU
CCC UGA CAG UCC AGA AA) and 3 (ACC AGA GAA
ACA CAG CUA AGG AUG AAA GUC UAU GCU GGU
AUA UUA CCU GGU). Synthesis of oligomers was carried
out on 1-mmol scale with a Pharmacia GeneAssembler as
reported previously6 with the following changes: activation
reagent: 0.25 M solution of (5-benzylmercapto)-1H-tetrazole
in acetonitrile, coupling time: 3 min, 8-fold excess of phos-
6 (CAG AAA UCA
6
6
phoramidites. Average coupling yields were 99.5% for 1
6 ,
99.2% for 2, 99.1% for 3. HPLC-conditions: Ion exchange
6
6
HPLC on PL-SAX 1000 (Polymer Laboratories) with buffer
A: 20 mM sodium acetate (pH 6.5)/20% CH3CN and buffer
B: 20 mM sodium acetate (pH 6.5)/20% CH3CN, 1 M NaCl,
flow rate: 1 ml/min at 0%B for 2 min, followed by a linear
gradient of 0–100% B over 40 min (16 and 26 ), or on MonoQ
(Pharmacia) with buffer A: 10 mM Na3PO4 (pH 11.5) and
buffer B: 10 mM Na3PO4 (pH 11.5), 1M NaCl, flow rate: 1
ml/min at 0%B over 5 min and a linear gradient of 0–100%B
over 50 min (36 ).

R. Welz, S. Mu¨ller / Tetrahedron Letters 43 (2002) 795–797
797
Table 1. Screening of reaction conditions for the synthesis of 5-(benzylmercapto)-1H-tetrazole. All reactions were carried out
at 0.1 mol scale
c
Solvent
Azide
Catalyst
T/°C
Time (h)
Yield (%)
1
2
3
4
5
6
Dioxan
LiN3
NH4OAc
90
70
90
90
70
70
6
96
6
6
24
96
65
43
30
62
63
72b
Toluene/H2O (70:30)
Dioxan/H2O (80:20)
Dioxan/H2O (80:20)
Dioxan/H2O (80:20)
Dioxan/H2O (80:20)
NaN3
NaN3
NaN3
NaN3
NaN3
NH4Cl, HDTMABa
NH4Cl,
NH4Cl, raschig rings
NH4Cl, raschig rings
NH4Cl, raschig rings
a Hexadecyltrimethylammonium bromide (phase-transfer catalyst).
b Repeated preparation of 5-(benzylmercapto)-1H-tetrazole at 0.1 to 0.5 mol delivered the product independent of the reaction scale with 67 to
72% yield.
Jursic,8 at 0.1 mol scale we were able to produce the
tetrazole derivative in 43% yield (Table 1, line 2). Using
80% aqueous dioxan as solvent we found, that sodium
azide is still only poorly soluble, which might account
for the observed low yield (Table 1, line 3). We carried
out the reaction in the presence of Raschig rings in
order to increase the reactivity of the azide by mechan-
ical activation. As shown in the Table 1, yields are
considerably improved (lines 4–6). Reaction over 6 h at
90°C delivered the product with 62% yield, that could
be improved to 72% by applying longer reaction times
and lower temperatures. The working up procedure
involves repeated extraction of the aqueous reaction
mixture with organic solvent.9 When the reaction was
carried out under phase-transfer conditions,8
HDTMAB was found to disturb phase separation mak-
ing the extraction less efficient. Thus, the protocol
provided in this article is advantageous in terms of
higher yields as well as easier handling.
D.; Workman, C.; Sweedler, D.; Gonzalez, C.; Scaringe,
S.; Usman, N. Nucleic Acids Res. 1995, 23, 2677–2684.
4. Vargeese, C.; Carter, J.; Yegge, J.; Krivjansky, S.; Settle,
A.; Kropp, E.; Peterson, K.; Pieken, W. Nucleic Acids
Res. 1998, 26, 1046–1050.
5. Lieber, E.; Enkoji, T. J. Org. Chem. 1961, 26, 4472–4479.
6. Schmidt, C.; Welz, R.; Mu¨ller, S. Nucleic Acids Res.
2000, 28, 886–894.
7. Finnegan, W. G.; Henry, R. A.; Lofquist, R. J. Am.
Chem. Soc. 1958, 80, 3908–3911.
8. LeBlanc, B. W.; Jursic, B. S. Synth. Commun. 1998, 28,
3591–3599.
9. 74.5 g Benzylthiocyanate (0.5 mol), 40 g sodium azide
(1.3 equiv.) and 69.5 g ammonium chloride (1.3 equiv.)
were mixed with 300 ml of 80% aqueous dioxane.
Raschig rings (glas, 4 mm, ø 4 mm) were added and the
mixture was vigorously stirred for 96 h at 70°C with a
strong magnetic stirrer. After cooling-down the solution
was acidified to pH 2 with conc. HCl followed by addi-
tion of 100 ml water. The mixture was extracted three
times with 200 ml CHCl3. The combined organic layers
were re-extracted with 3×100 ml 5% NaHCO3, and the
aqueous extracts again acidified to pH 2 with conc. HCl
affording 65 g of 5-(benzylmercapto)-1H-tetrazole as a
white powder (72% yield). The analytical data allowed for
application of the product in oligonucleotide synthesis
without further purification. The protocol for preparation
of 5-(benzylmercapto)-1H-tetrazole in the presence of
LiN3 and anhydrous dioxan is analogous, but proceeds
without Raschig rings. Analytical data: MS (EI): m/z
192.1 (M=192.24 g/mol). EA: theoretical (%) C, 49.98;
H, 4.19; N, 29.14; S, 16.68. Found ( 0.3%) C, 50.10; H,
In summary our results show, that 5-(benzylmercapto)-
1H-tetrazole can be prepared from cheap reagents in a
rather short time. Using 5-(benzylmercapto)-1H-tetra-
zole in RNA synthesis, coupling times are reduced and
also the required amount of monomer building blocks
as well as the activator is decreased compared to tradi-
tional methods. This is an important factor concerning
costs and time for chemical preparation of oligoribonuc-
leotides and should be of great interest to any chemist
involved in RNA synthesis.
References
1
4.06; N, 29.01; S, 16.60. H NMR (CD3CN): l 4.26 (2H,
s, CH2), 7.05 (5H, m, aromatic); 13C NMR (CD3CN): l
38.2 (CH2), 127.7/128.1/128.8 (CH), 137.0 (phenyl C),
160.6 (tetrazole).
1. Ogilvie, K. K.; Sadana, K. L.; Thompson, E. A.; Quil-
liam, M. A.; Westmore, J. B. Tetrahedron Lett. 1974, 15,
2861–2863.
10. Brauer, G. Handbuch der Pra¨parativen Anorganischen
Chemie in drei Ba¨nden, Bd.1; Ferdinand Enke Verlag:
Stuttgart, 1975; p. 457.
2. Pitsch, S.; Weiss, P. A.; Wu, X.; Ackermann, D.; Honeg-
ger, T. Helv. Chim. Acta 1999, 82, 1753–1761.
3. Wincott, F.; DiRenzo, A.; Shaffer, C.; Grimm, S.; Tracz,