Spontaneous aminoacylation of a RNA sequence containing a 3′-terminal 2′-thioadenosine

Article abstract of DOI:10.1002/hlca.200590233

We report the synthesis of a modified 8mer RNA sequence, (C-C-C-C-A-C-C-(2′-thio)A)-RNA 5′-(dihydrogen phosphate) (9) containing a 3′-terminal 2′-thioadenosine (Schemes 2 and 3), and its spontaneous and site-specific aminoacylation with the weakly activat

Full text of DOI:10.1002/hlca.200590233

H
ELVETICA
CHIMICA
ACTA – Vol. 88 (2005)
2897
Spontaneous Aminoacylation of a RNA Sequence Containing
a 3’-Terminal 2’-Thioadenosine
by Sébastien Porcher, Muthuppalaniappan Meyyappan, and Stefan Pitsch*
Laboratory of Nucleic Acid Chemistry, EPFL-BCH, CH-1015 Lausanne
Albert Eschenmoser zum 80. Geburtstag gewidmet
We report the synthesis of a modified 8mer RNA sequence, (C-C-C-C-A-C-C-(2’-thio)A)-RNA 5’-(dihy-
drogen phosphate) (9) containing a 3’-terminal 2’-thioadenosine (Schemes 2 and 3), and its spontaneous and
site-specific aminoacylation with the weakly activated amino acid thioester HꢀPheꢀSPh (12). This reaction,
designed in analogy to the ‘native chemical ligation’ of oligopeptides, occurs efficiently in buffered aqueous sol-
utions and under a wide range of conditions (Table). At pH values between 5.0 and 7.4, two products, the 3’-O-
monoacylated and the 3’-O,2’-S-diacylated RNA sequences 10 and 11 are formed fast and quantitatively
(Scheme 4). At pH 7.4 and 378, the 3’-O-monoacylated product 10 is formed as major product in situ by selective
hydrolysis of the O,S-diacylated precursor 11. Additionally, the preparation and isolation of the relevant 3’-O-
monoacylated product 10 was optimized at pH 5. The here presented concept could be employed for a straight-
forward aminoacylation of analogously modified tRNAs.
1. Introduction. – The ribosome-mediated, site-specific incorporation of unnatural
amino acids into proteins is a powerful tool for creating protein analogues with specific
steric, chemical, electronic, or spectroscopic properties, in vitro and in vivo. The
required aminoacylated tRNA analogues are prepared semisynthetically [1], by ribo-
zyme-catalyzed amino acid transfer from short aminoacylated RNA sequences [2],
from complementary PNA amino acid thioesters [3], or by modified aminoacyl
tRNA synthetases [4]. To facilitate the access to these important and valuable com-
pounds, we try to prepare and identify biologically active tRNA analogues that are
spontaneously aminoacylated by weakly activated amino acids, ideally under the con-
ditions of in vitro translation reactions.
Here we describe the first example of such a reaction, which consists in a spontane-
ous aminoacylation of a modified 8mer RNA sequence 9 (Scheme 3)1) that carries a 2’-
thioadenosine at the 3’-terminus, by the activated amino acid thioester S-phenyl 3-
phenyl- -thioalaninate (=(aS)-a-aminobenzenepropanethioic acid S-phenyl ester;
L
HꢀPheꢀSPh; 12). Conceptually, this experiment was designed in analogy to the so-
called ‘native chemical ligation’, which occurs between two peptides carrying an N-ter-
minal cysteine and a C-terminal thioester. Thereby, a fast intermolecular trans-thioes-
terification reaction precedes an intramolecular S ! N migration of the acyl group,
resulting in the selective formation of a new amide bond [5]. In Scheme 1, the analogy
between the two reactions is illustrated.
This sequence is identical to the 3’-terminal sequence of tRNAAla from E. coli.
1
)
© 2005 Verlag Helvetica Chimica Acta AG, Zürich

2898
HELVETICA CHIMICA ACTA – Vol. 88 (2005)
Scheme 1. Formation of an Amide Bond between a Cysteine and an Amino Acid Thioester According to the
‘Native Chemical Ligation’ [5] (top), as Compared to the Here Presented, Analogous Formation of an Ester
Bond between a 2’-Thioadenosine and an Amino Acid Thioester (bottom).
2. Syntheses. – The assembly of RNA sequences containing a 3’-terminal 2’-thio-
adenosine required the preparation of a suitably protected and immobilized building
block 6, which was prepared from the known, protected arabinonucleoside 1 [6]
(Scheme 2). First, 1 was converted with (CF3SO2)2O in pyridine/CH2Cl2 into the corre-
sponding 2’-O-(trifluoromethyl)sulfonyl derivative, which, after extraction, was treated
with 4-methoxybenzenemethanethiol/NaH in DMSO, according to a reported method
[7]. The resulting 4-methoxybenzyl thioether 2 (91% yield) was then treated with bu-
tane-1-sulfenyl chloride (prepared according to [8]) in CH2Cl2/AcOH according to [7],
resulting in the efficient formation of the butyl disulfide derivative 3 (86% yield)2).
After cleavage of the silyl protecting group with HF/pyridine, extraction, and 5’-O-
dimethoxytritylation with (MeO)2Tr-Cl in pyridine, the protected 2’-thioadenosine 4
was obtained (60% yield). The corresponding activated ester 5 was prepared with
bis(4-nitrophenyl) heptanedioate in pyridine (68% yield). Finally, 5 was immobilized
on aminoalkyl-functionalized controlled pore glass (CPG) with iPr2NEt in DMF, result-
ing in the solid support 6 with a loading of 30 mmol/g.
From 6, 2’-O-[(triisopropylsilyl)oxy]methyl(=2’-O-tom)-protected ribonucleoside
phosphoramidites [9], and a commercially available phosphate building block3), the
20
8mer RNA sequence ðꢀOÞ2Pð¼ OÞO ꢀ rðCCCCACCAd20 ðSSBuÞ Þ (8), carrying a
monophosphate group at the 5’-end and a 2’-dithioprotected 2’-deoxyadenosine at
the 3’-end, was assembled by automated synthesis under standard conditions [9], but
employing a milder oxidation agent (20 m
M
I2 in THF/pyridine/H2O 7:2 :1 instead of
100 m I2 in THF/pyridine/H2O 78 :20 :2; Scheme 3). To avoid partial decomposition
M
of the product sequence, the usually employed deprotection conditions had to be
adapted. First, the cyanoethyl protecting groups were removed by washing the crude,
i
immobilized sequence 7 with Pr2NH/MeCN 1:9 for 20 min at a flow rate of 2.5 ml/
2
)
We also prepared the corresponding phenyl and tert-butyl disulfide derivatives. However, the phenyl disul-
fide group was much too labile and was removed during the oligonucleotide assembly, and the tert-butyl
disulfide group was too stable and could never be completely cleaved under our preferred conditions.
3-[(4,4’-dimethoxytrityl)oxy]-2,2-bis(ethoxycarbonyl)propyl 2-cyanoethyl diisopropylphosphoramidite.
3
)

H
ELVETICA
CHIMICA
ACTA – Vol. 88 (2005)
2899
Scheme 2
a) 1. (CF3SO2)2O, pyridine, CH2Cl2, ꢀ158 to 108; 2. 4-methoxybenzenemethanethiol, NaH, DMSO, 108. b)
BuSCl, AcOH, CH2Cl2, 4–258. c) 1. HF·pyridine, pyridine, 258; 2. (MeO)2TrCl, pyridine, 258. d) Bis(4-nitro-
phenyl) heptanedioate, DMAP (=N,N-dimethylpyridin-4-amine), pyridine, 258. e) 1. Long-chain-alkylamino-
i
CPG, Pr2NEt, DMF, 258; 2. Ac2O, Py, 258.
min, then removal of the nucleobase protecting groups and cleavage from the solid sup-
port was achieved with 12
protecting groups were removed by 1
M
NH3 in MeOH during 6 h at 258, and finally, the 2’-O-tom
Bu4NF·3H2O/0.5 AcOH in THF during 3 h at
M
M
258. Even under these optimized conditions (required for a complete deprotection),
some by-products (ca. 30%) resulting from partial cleavage of the disulfide protecting
groups were observed. After a double chromatographic purification (1. anion-exchange
HPLC; 2. reversed-phase HPLC) and desalting, the pure product 8 was obtained in a
yield of 25% (based on solid support 6). The LC-ESI-MS of this product showed
only one signal at m/z 2611 amu (calc. for 8, 2611 amu). The HPLC and MS traces
of purified 8 are shown in Fig. 1,a and e, respectively.
3. Aminoacylation Studies. – The free thiol-substituted RNA sequence 9 was
formed in situ by reductive cleavage of the remaining disulfide protecting group of 8
(c=0.08 m
M
)4) with tris(2-carboxyethyl)phosphine (=2,2’,2’’-phosphinidynetris[acetic
acid]=TCEP; c=1–10 m
M
, depending on pH) in various aqueous buffers (pH 7.4–
3.7) at 258 (for conditions, see Table). This reaction was complete within 30 min, and
9 was obtained in a purity of >95% according to HPLC and MS analyses (Fig. 1,b
and f, resp.).
In a first set of experiments, the pH dependency of the aminoacylation reaction was
investigated (see Table and Scheme 4). Crude 9 in buffered aqueous solutions (pH 7.4–
4
)
Such a low concentration was chosen to simulate the aminoacylation of analogously modified tRNAs, which
are usually available and required only in very small quantities

2900
H
ELVETICA
CHIMICA
ACTA – Vol. 88 (2005)
Scheme 3
a) Assembly on an oligonucleotide synthesizer with 2’-O-tom-protected ribonucleosides and 3-[(4,4’-dime-
thoxytrityl)oxy]-2,2-(ethoxycarbonyl)propyl 2-cyanoethyldiisopropylphosphoramidite according to [9] but
i
employing 20 m
M
I2 in THF/pyridine/H2O 7 :2 :1 as oxidizing agent. b) 1. Pr2NH/MeCN 1:9, 20 min; 2. 12
M
NH3, MeOH, 258; 3. Bu4NF·3H2O, AcOH, THF; 4. HPLC purification. c) TCEP (=tris(2-carboxyethyl)-
phosphine) in different aq. buffers, see Table.
3.7, obtained from 8 and TCEP as described above) was treated with 12 equiv. of 125)
(added as concentrated solution in DMF) at 258, and the composition of the reaction
mixtures was analyzed by reversed-phase HPLC after 10 and 50 min (for an example,
obtained at pH 7.4, see Fig. 1,c (10 min) and d (50 min), resp.). Under all conditions, a
fast formation of two slower migrating (less polar) products was observed, which was in
agreement with the formation of the monoacylated product 10 and the diacylated prod-
uct 11, respectively. The HPLC analysis was carried out at pH 3.5, where both products
10 and 11 were hydrolytically stable6). Additional experiments at 378 and with only 3 or
6 equiv. of 12 were carried out at pH 7.4 and 5.0 (see Table). The two aminoacylated
products 10 and 11 were isolated by HPLC, and their structure was confirmed by
ESI-MS (Fig. 1,g and h)7).
5
)
Prepared from commercially available N-[(tert-butoxy)carbonyl]-
L-phenylalanine 4-nitrophenyl ester
according to [10].
All reversed-phase HPLC analyses were carried out at pH 3.5 with a 0.1
6
)
M
Et3N/H3PO4 buffer. At pH values
between 4.0 and 5.5, partial hydrolysis of the diacylated product 11, and at pH values between 5.5 and 7.5,
partial hydrolysis of both acylated products 10 and 11 was observed.
7
)
Our analytical methods (HPLC, MS, and hydrolysis studies, see below) do not show whether 10 is the 3’-O-
or the 2’-S-monoacylated product; however, since esters are thermodynamically much more stable than
analogous thioesters, and since it is well-known that the migration of acyl groups between the 2’-O and
3’-O positions of ribonucleosides is an extremely fast reaction [11], we concluded that 10 is the 3’-O-mono-
acylated derivative.

H
ELVETICA
CHIMICA
ACTA – Vol. 88 (2005)
2901
Fig. 1. a)–d) Reversed-phase HPLCtraces (detection at 260 nm) of a representative aminoacylation reaction
carried out at pH 7.4 and 258: a) disulfide-protected starting RNA sequence 8, b) 30 min after treatment with
TCEP, c) 10 min and d) 50 min after addition of 12 equiv. of HꢀPheꢀSPh (12). e)–f) Deconvoluted ESI-MS
(neg. mode) of RNA sequences 8–11 (for procedures, see Exper. Part).
At pH values between 4.7 and 5.5, and at 258, clean and almost quantitative forma-
tion of the monoacylated product 10 together with the diacylated product 11 was
observed after 50 min. At higher pH values of 6.5 and 7.4, some unidentified by-prod-
ucts were formed after 50 min (ca. 10%), but the reactions were complete already after

2902
HELVETICA CHIMICA ACTA – Vol. 88 (2005)
Table. Release and Aminoacylation of the 2’-Thioadenosine-Modified RNA Sequence 9: Exploratory Studies
under Different Conditionsa)
Conditionsb)
t 10 minc)
9 [%]
t 50 minc)
9 [%]
pH
12 [equiv.]
T [8]
10 [%]
11 [%]
10 [%]
11 [%]
7.4
6.5
5.5
5.0
4.7
3.7
7.4
7.4
5.0
5.0
5.0
5.0
5.0
12
12
12
12
12
12
6
12
6
3
25
25
25
25
25
25
25
37
25
25
37
37
37
1
3
7
10
12
43
2
18
19
19
18
18
20
24
17
29
20
27
34
29
73
71
70
70
68
36
64
72
43
23
67
53
34
5
3
37
24
10
10
10
26
78
41
33
35
32
35
41
46
64
86
87
87
55
10
42
65
47
64
60
50
<1
<1
<1
18
6
5
1
15
1
1
1
27
54
2
8
30
12
6
3
4
a) For the detailed procedure, see Exper. Part. b) Incubation of 8 (c=0.085 m
pH 7.4 and 6.5; c=2.5 m at pH 5.5, 5.0, and 4.7; c=10 m at pH 3.7) in aq. buffers (50 mM
M
; 40 ml) with TCEP (c=1.0 m
M at
M
M
each; pH 7.4,
Tris·HCl; pH 6.5, H3PO4/NaOH; pH 5.5, 5.0, and 4.7, AcOH/NaOH; pH 3.7, HCOOH/NaOH) at 258 for 30
min, followed by addition of 12 (indicated equiv. rel. to 8) in DMF (4 ml). c) t=Incubation time; quantification
by reversed-phase HPLC at 260 nm; tR (9) 14 min, tR (10) 17 min, tR (11) 18 min; % values rel. to the sum of all
detected peaks; for an example of such HPLC traces from a reaction carried out at pH 7.4 and 258, see Fig. 1,c
and d.
Scheme 4
a) TCEP in different aq. buffers, see Table. b) Addition of S-phenyl 3-phenyl-
L
-thioalaninate (HꢀPheꢀSPh;
12) in DMF to buffered aq. solutions, see Table. c) NaN3 in aq. buffer (pH 5.0), see Fig. 4.

H
ELVETICA
CHIMICA
ACTA – Vol. 88 (2005)
2903
Fig. 2. Aminoacylation reaction carried out at pH 7.4 and 378 with 9 and 12 equiv. of HꢀPheꢀSPh (12): a)
Plot of the amount of aminoacylated products (10+11) relative to the sum of all assigned RNA sequences
(9+10+11; data extracted from HPLC traces). b) Reversed-phase HPLCtrace (detection at 260 nm) of the
product mixture obtained after 2 min reaction time.
10 min at 258 (Fig. 1,c and d). With only 3 equiv. of 12, an almost quantitative amino-
acylation could be observed after 50 min at pH 5.0 and 378. Under all these conditions,
the parent RNA sequence (ꢀO)2P(=O)Oꢀr(CCCCACCA) was completely inert, indi-
cating that the thiol group is involved in, and required for the aminoacylation reaction
and that the activated amino acid is not interfering with the integrity of RNA sequen-
ces.
We then focused on two different reaction conditions: the first one typical for in
vitro translations, and the second one for the purpose of preparing and isolating amino-
acylated RNA sequences. In Fig. 2,a, the time course of a reaction carried out with 9
(c=0.08 m ) and 12 equiv. of 12 at pH 7.4 and 378 is shown. The composition of the
M
reaction mixture was analyzed by reversed-phase HPLC, and the relative amount of
products 10 and 11 was plotted against reaction time (Fig. 2,a). After 2 min already,
almost quantitative formation of the mono- and diacylated RNA sequences 10 and
11 was observed (Fig. 2,b). Only after ca. 50 min, they slowly were converted back to
the starting material 9 by hydrolytic cleavage. The plateau between 2 and 50 min indi-
cates the transient nature of the aminoacylated species 10 and 11, which are constantly
hydrolyzed and reacylated, until the activated amino acid is consumed.
The rate constants for the cleavage of the ester bond of 10 and the thioester bond of
11 at 378 and pH 7.4 (0.1 Tris·HCl) were determined by kinetic studies and are given
M
in Fig. 3,b. Under the same conditions, but at 258, the rate constant for the hydrolysis of
10 to 9 was k=0.02 minꢀ1, which is about twice the value of an adenosine nucleotide
esterified with
-Phe [12]8). To simulate the conditions of an in vitro translation reac-
L
Under these conditions, the hydrolysis rate of the activated amino acid 12 is k=0.007 minꢀ1 and 0.07 minꢀ1
at 258 and 378, respectively.
8
)

2904
HELVETICA CHIMICA ACTA – Vol. 88 (2005)
Fig. 3. a) Reversed-phase HPLCtraces (detection at 260 nm) obtained by incubating an isolated aminoacyla-
tion product mixture in 0.1 aqueous Tris·HCl (pH 7.4) at 378. b) Independantly determined kinetic parame-
M
ters of hydrolysis for the diacylated product 11 and the monocylated product 10. c) Measured (points) and
calculated (lines) composition of a hydrolysis reaction mixture obtained according to a) at pH 7.4 and 378
(for details see Exper. Part).
tion, we incubated a desalted mixture of reaction products from an acylation reaction9)
at pH 7.4 (aqueous 0.1 Tris·HCl buffer) and 378. Aliquots were removed at different
M
time intervals and analyzed by reversed-phase HPLC (for examples, see Fig. 3,a). In
Fig. 3,c, the time-dependent composition of the reaction mixture is shown together
with the curves calculated from the individually determined hydrolysis rate constants
(shown in Fig. 3,b). The hydrolysis of the diacylated 11 to the monoacylated 10 occurs
twice as fast as the hydrolysis of the latter to 9; this results in a predominant occurrence
of the relevant monoacylated RNA sequence 10 after a short incubation at 378 and pH
7.4 (Fig. 3,c).
9
)
The isolated product mixture was obtained from an acylation reaction of 9 with 12, carried out at pH 5.0 and
at 378; after 30 min, the reaction mixture was diluted with H2O and desalted on a size-exclusion cartridge by
elution with H2O. The oligonucleotide-containing eluate was stabilized with 1% AcOH and concentrated to
a smaller volume by lyophilization.

H
ELVETICA
CHIMICA
ACTA – Vol. 88 (2005)
2905
Fig. 4. Conditions for the preparation and isolation of the monoacylated RNA sequence 10 and reversed-
phase HPLCtraces (detection at 260 nm) of a) the intermediate and b) the final product mixture
An efficient acylation reaction between 9 and 12 was observed at pH 5.0 (Table),
where aminoacylated RNAs are considerably stable [12]; therefore, we decided to opti-
mize the preparation of 10 at this pH value. To minimize the formation of by-products,
we carried out the deprotection and aminoacylation reactions simultaneously. In Fig. 4,
the HPLC trace of such a reaction product is shown, which was obtained by incubation
of 8 (0.08 m ) with TCEP (30 equiv.) and 12 (5 equiv.) at pH 5.0 and 378 for 30 min.
M
This HPLC trace reveals the clean and almost quantitative formation of the aminoacy-
lated products 10 and 11. For the selective cleavage of the thioester bond (reaction 11
! 10), a variety of different conditions (pH values, nucleophiles) were investigated.
The best result was obtained by treating the reaction mixture after 30 min with
NaN3, keeping it for 30 min at 258, and isolating the products by desalting on a size-
exclusion cartridge, which removes all non-oligonucleotide components. This protocol
furnished a product mixture containing 90% of the desired monoacylated RNA
sequence 10, together with 5% of 9 and 5% of 11 (Fig. 4,b). Lyophilization of this aque-
ous eluate is possible without cleavage of the ester bond if 1% AcOH is added10).
Discussion. – In conclusion, we found that the intriguing concept of ‘native chemical
ligation’ of oligopeptides can also be applied to the straightforward acylation of 2’-thio-
nucleotides with activated amino acids. Mechanistically, it can be assumed that after
formation of the 2’-thioester derivative, a fast and almost irreversible migration of
10
)
It was possible to isolate product 10 in pure form by reversed-phase HPLC (see Exper. Part). However, such
a purification is not feasible or desirable with analogously prepared aminoacylated tRNA sequences
(76mers, available in rather small quantities), for which this method was developed.

2906
HELVETICA CHIMICA ACTA – Vol. 88 (2005)
the acyl group to the 3’-O position occurs (see Footnote 7); subsequently, the liberated
2’-SH group can then undergo another acylation reaction, resulting in formation of the
O,S-diacylated product.
At pH 7.4 and 378, the biologically relevant O-monoacylated product is formed as
major product in situ by hydrolysis of the O,S-diacylated precursor. The O-monoacy-
lated product can be prepared efficiently by acylation of the modified RNA sequence
at pH 5 and 378, followed by selective cleavage of the concomitantly formed O,S-diacy-
lated product with NaN3; its isolation is conveniently carried out by desalting on size-
exclusion cartridges, stabilization of the eluate with 1% AcOH, and lyophilization.
The here presented method could be applied to the straightforward preparation and
isolation of aminoacylated RNA sequences, or to translation systems in which a thiol-
modified tRNA acts as a catalyst, respectively. Currently, we are investigating, whether
translation systems tolerate these thiol-modified tRNAs.
This work was supported by the EPFL and by the Swiss National Science Foundation (Grant Nr. 2000-
06890).
Experimental Part
1. General. Reagents and solvents (highest purity) from various suppliers were used without further puri-
fication. Tris=2-amino-2-(hydroxymethyl)propane-1,3-diol. S-Phenyl 3-phenyl-
L
-thioalaninate (HꢀPheꢀSPh;
12) was prepared according to [10] and 2’-O-tom-protected RNA phosphoramidites according to [9]. The phos-
phoramidite building block for the introduction of the 5’-terminal monophosphate group was from Glen
Research. Size-exclusion cartridges: NAP-10 columns from Pharmacia, elution according to the manufacturer’s
instruction; Sepak cartridges from Waters, elution according to [13]. Workup implies distribution of the reaction
mixture between the indicated aq. and org. phases, drying (MgSO4) of the org. layer, and evaporation. Column
chromatography: (CC) silica gel (SiO2, Fluka); for purification of (MeO)2Tr-protected nucleosides, the SiO2 was
pre-equilibrated with 2% Et3N added to the packing solvent. TLC: precoated silica gel plates from Merck,
stained by dipping into a soln. of anisaldehyde (10 ml), conc. H2SO4 soln. (10 ml), and AcOH (2 ml) in
EtOH (180 ml) and subsequent heating. Reversed-phase HPLC: Waters Xterra RP18 (5 mm; 4.6×250 mm);
flow 1 ml/min; eluent A: 0.1
B: MeCN; elution at 258, detection at 260 nm; unless otherwise stated, a gradient A ! A/B 1:1 (30 min) was
used; tR in min. Ion-exchange HPLC: Pharmacia Source 15Q (4.6×100 mm); flow 1 ml/min; eluent A: 10 m
AcOH/NaOAc (pH 5.0) in MeOH/H2O 1 :1; eluent B: AcOH/NaOAc, 1 NaCl (pH 5.0) in MeOH/H2O 1:1,
M (Et3NH)3PO4 in H2O (pH 3.5) or 0.1M (Et3NH)OAc in H2O (pH 5.5); eluent
M
M
elution at 258, detection at 260 nm; a gradient A ! A/B 1:1 (30 min) was used. ESI-MS (pos. mode): SSQ
710 (Finnigan); measurements in MeCN/H2O/AcOH 50 :50 :1. ESI-MS (neg. mode): Q-TOF-Ultima (Micro-
mass/Waters), measurements in aq. [Me2N(Bu)H]HCO3 (12.5 mM, pH 6.3)/MeCN 1:1; the samples were intro-
duced by a syringe pump; deconvolution by MaxEnt1 software; in m/z.
2. Nucleoside Derivatives. N6-Benzoyl-2’-S-(4-methoxybenzyl)-3’,5’-O-(1,1,3,3-tetraisopropyldisiloxane-1,3-
diyl)-2’-thioadenosine (2). At ꢀ158, a soln. of 1 (260 mg, 0.44 mmol; prepared according to [6]) in CH2Cl2/pyr-
idine 1:1 (2 ml) was treated with (CF3SO2)2O (0.11 ml, 0.60 mmol). Workup after 5 h at 108 gave the crude 2’-O-
(trifluoromethyl)sulfonyl derivative of 1 (330 mg; TLC (hexane/AcOEt 1:9): Rf 0.50) as brown-orange foam,
which was dissolved in DMSO (8 ml). For the next step, a suspension of 50% NaH in mineral oil (106 mg, ca.
2.2 mmol) and 4-methoxybenzenemethanethiol (343 mg, 2.2 mmol) were added to DMSO (7 ml). After 10
min at 258, the mixture was cooled to 108, treated with the soln. of the crude 2’-O-(trifluoromethylsulfonyl deriv-
ative of 1 (330 mg; see above) in DMSO (8 ml), and stirred for 10 min at 108. After addition of AcOEt (100 ml),
extraction (1. 10% aq. citric acid soln., 2. sat. aq. NaHCO3 soln.) and CC (SiO2 (10 g), hexane/AcOEt 9 :1 !
1:1), 2 (300 mg, 91%) was obtained. Yellow foam. TLC (hexane/AcOEt 1:9): Rf 0.71. 1H-NMR (CDCl3):
0.99–1.12 (m, iPr2Si); 3.75 (s, MeO); 3.78–3.92 (m, CH2(5’)); 4.01–4.19 (m, HꢀC(4’), HꢀC(2’)); 4.93 (t,
J=6.9, HꢀC(3’)); 5.31 (s, CH2S); 6.22 (d, J=3.2, HꢀC(1’)); 6.71 (d, J=8.5, 1 arom. H); 7.12 (t, J=8.5, 2
arom. H); 7.54 (t, J=7.6, 2 arom. H); 7.62 (t, J=7.4, 1 arom. H); 8.05 (d, J=7.6, 2 arom. H); 8.14 (s, Hꢀ
C(2)); 8.80 (s, HꢀC(8)); 9.22 (br. s, NH). 13C-NMR (100 MHz, CDCl3): 13.0, 13.3, 13.5, 13.8 (4d, Me2CH);

H
ELVETICA
CHIMICA
ACTA – Vol. 88 (2005)
2907
17.3, 17.4, 17.5, 17.7, 17.8, 17.9 (6q, Me2CH); 35.5 (t, CH2S); 52.4 (d, C(2’)); 55.7 (q, MeO); 61.6 (t, C(5’)); 70.8 (d,
C(3’)); 84.6 (d, C(4’)); 90.6 (d, C(1’)); 114.2 (d, arom. C); 123.8 (s, C(5)); 128.3, 129.3 (2d, arom. C); 129.7 (s,
arom. C); 130.2, 133.2 (2d, arom. C); 134.2 (s, arom. C); 141.9 (d, C(8)); 149.9 (s, C(6)); 151.3 (s, C(4)); 153.1
(d, C(2)); 159.1 (s, MeOC), 165.1 (s, PhCO). ESI-MS: 750.31 ([M+H]+).
N6-Benzoyl-2’-(butyldithio)-2’-deoxy-3’,5’-O-(1,1,3,3-tetraisopropyldisiloxane-1,3-diyl)adenosine (3). Simul-
taneously, two solns., one containing BuSCl (250 mg, 2 mmol; prepared according to [8]) in CH2Cl2 (10 ml), and
the other 2 (300 mg, 0.4 mmol) in CH2Cl2 (10 ml), were added dropwise to a cooled (48) mixture of CH2Cl2/
AcOH 1:1 (30 ml). After addition, the ice bath was removed and the mixture stirred for 48 h at 258. Workup
and CC (SiO2 (10 g), hexane/AcOEt 4 :1 ! 1:4) gave 3 (250 mg, 86%). Light yellow foam. TLC (AcOEt/
i
CH2Cl2 1:4): Rf 0.59. 1H-NMR (400 MHz, CDCl3): 0.84 (t, J=7.0, MeCH2); 1.02–1.17 (m, Pr2Si); 1.20–1.40
(m, CH2); 1.51–1.63 (m, CH2); 2.63 (t, J=7.4, CH2S); 4.01 (dd, J=3.5, 12.5, HꢀC(5’)); 4.12 (dd, J=3.5, 12.5,
H’ꢀC(5’)); 4.16–4.23 (m, HꢀC(4’)); 4.31 (dd, J=3.2, 7.3, HꢀC(2’)); 5.23 (t, J=7.3, HꢀC(3’)); 6.43 (d,
J=3.2, HꢀC(1’)); 7.55 (t, J=7.3, 2 arom. H); 7.63 (t, J=7.3, 1 arom. H); 8.05 (br. d, J=7.4, 2 arom. H); 8.24
(s, HꢀC(8)); 8.79 (s, HꢀC(2)); 9.14 (br. s, HN-C(6)). 13C-NMR (100 MHz, CDCl3): 13.1 (q, MeCH2); 13.2,
13.5, 13.8, 13.9 (4d, Me2CH); 17.3, 17.5, 17.6, 17.8, 17.9, 18.0, 18.1, 18.2 (8q, Me2CH); 21.9 (t, MeCH2); 31.3
(t, CH2CH2); 38.7 (t, CH2S); 56.7 (d, C(2’)); 62.1 (t, C(5’)); 71.9 (d, C(3’)); 84.3 (d, C(4’)); 90.5 (d, C(1’));
128.3 (s, C(5)); 129.3 (d, arom. C); 133.2 (d, arom. C); 134.1 (s, arom. C); 142.5 (d, C(8)); 150.1 (s, C(6));
151.6 (s, C(4)); 153.1 (d, C(2)); 165.0 (s, PhCO). ESI-MS: 718.32 ([M+H]+).
N6-Benzoyl-2’-(butyldithio)-2’-deoxy-5’-O-(4,4’-dimethoxytrityl)adenosine (4). A soln. of 3 (206 mg, 0.28
mmol) in CH2Cl2 (2.2 ml) was treated with a 35% HF soln. in pyridine (2 ml; prepared from a 70% HF soln.
in pyridine (1 ml) and pyridine (1 ml)) for 1 h at 258. After workup and evaporation, the residue was dissolved
in pyridine (1.4 ml) and treated with (MeO)2TrCl (114 mg, 0.34 mmol) for 1 h at 258. Workup and CC (SiO2
(6 g), CH2Cl2 ! CH2Cl2/AcOEt 1:1) gave 4 (134 mg, 60%). Colorless foam. TLC (hexane/AcOEt 7:3): Rf
0.34. 1H-NMR (400 MHz, CDCl3): 0.80 (t, J=7.3, MeCH2); 1.17–1.30 (m, CH2); 1.38–1.40 (m, CH2); 2.37–
2.54 (m, CH2S); 3.15 (br. d, J ꢁ 2.2, OH); 3.39 (dd, J=4.0, 10.4, HꢀC(5’)); 3.49 (dd, J=4.4, 10.4, H’ꢀC(5’));
3.79 (s, 2 MeO); 4.30 (t, J=3.0, HꢀC(4’)); 4.70–4.75 (m, HꢀC(2’), HꢀC(3’)); 6.28 (d, J=8.6, HꢀC(1’)); 6.80
(d, J=8.0, 4 arom. H); 7.18–7.34 (m, 8 arom. H); 7.42 (dd, J=1.2, 8.1, 2 arom. H); 7.56 (t, J=7.4, 1 arom.
H); 7.64 (t, J=7.3, 1 arom. H); 7.98 (s, HꢀC(8)); 8.06 (d, J=7.3, 2 arom. H); 8.62 (s, HꢀC(2)); 9.06 (s, NH).
13C-NMR (100 MHz, CDCl3): 13.9 (q, MeCH2); 21.9 (t, MeCH2); 31.1 (t, CH2CH2S); 39.3 (t, CH2S); 55.7 (q,
MeO); 59.7 (d, C(2’)); 63.9 (t, C(5’)); 73.6 (d, C(3’)); 85.7 (d, C(4’)); 88.3 (d, C(1’)); 113.6 (d, arom. C); 123.8
(s, C(5)); 127.4, 128.3, 128.6, 129.3, 130.5 (5d, arom. C); 133.2, 135.9 (2s, arom. C); 142.6 (d, C(8)); 144.8 (s,
arom. C); 150.0 (s, C(6)); 152.3 (s, C(4)); 153.2 (d, C(2)); 159.0 (s, arom. C); 165.0 (s, PhCO). ESI-MS:
778.30 ([M+H]+).
N6-Benzoyl-2’-(butyldithio)-2’-deoxy-5’-O-(4,4’-dimethoxytrityl)adenosine 3’-(4-Nitrophenyl Heptane-
dioate) (5). A soln. of 4 (134 mg, 0.18 mmol) in pyridine (2 ml) was treated with DMAP (12 mg, 0.08 mmol)
and bis(4-nitrophenyl) heptanedioate (420 mg, 1.04 mmol) for 14 h at 258. Evaporation and CC (SiO2 (10 g),
hexane/AcOEt 3 :2 ! 1:9) gave 5 (122 mg, 68%). Colorless foam. TLC (hexane/AcOEt 7:3): Rf 0.53. 1H-
NMR (400 MHz, CDCl3): 0.80 (t, J=7.4, MeCH2) 1.16–1.29 (m, CH2); 1.38–1.45 (m, CH2); 1.50–1.57 (m,
CH2); 1.74–1.87 (m, 3 CH2); 2.29–2.51 (m, 2 CH2); 2.66 (t, J=7.3, CH2S); 3.45–3.54 (m, CH2(5’)); 3.79 (s, 2
MeO); 4.22–4.27 (m, HꢀC(4’)); 4.78 (dd, J=5.5, 9.1, HꢀC(2’)); 5.77 (dd, J=0.8, 5.6, HꢀC(3’)); 6.36 (d,
J=9.1, HꢀC(1’)); 6.80 (d, J=8.9, 4 arom. H); 7.04–7.10 (m, 3 arom. H); 7.10–7.16 (m, 2 arom. H); 7.21–
7.33 (m, 9 arom. H); 7.41 (dd, J=1.5, 8.2, 2 arom. H); 7.56 (t, J=7.2, 1 arom. H); 7.93 (s, HꢀC(8)); 8.07 (d,
J=7.5, 2 arom. H); 8.26 (td, J=2.0, 9.0, 2 arom. H); 8.64 (s, HꢀC(2)); 9.05 (br. s, NH). 13C-NMR (100 MHz,
CDCl3): 13.9 (q, MeCH2); 21.8 (t, MeCH2); 24.7, 24.8, 28.9, 31.1, 34.2, 34.4 (6t, CH2); 39.3 (t, CH2S); 55.6 (q,
MeO); 57.3 (d, C(2’)); 63.8 (t, C(5’)); 75.5 (d, C(3’)); 84.4 (d, C(4’)); 87.4 (d, C(1’)); 88.9 (s, arom. C); 113.7
(d, arom. C); 122.8 (s, C(5)); 125.6, 127.5, 128.3, 128.4, 128.6, 129.3, 130.5, 133.2 (9d, arom. C); 134.0, 135.8
(2s, arom. C); 142.2 (d, C(8)); 144.8, 145.7 (2s, arom. C); 150.1 (s, C(6)); 152.5 (s, C(4)); 153.3 (d, C(2));
155.8 (s, arom. C); 159.1 (s, arom. C); 164.9, 171.4, 172.6 (3s, CO). ESI-MS: 1041.29 ([M+H]+).
N6-Benzoyl-2’-(butyldithio)-2’-deoxy-5’-O-(4,4’-dimethoxytrityl)adenosine 3’-[6-(CGP-Amino)heptanoate]
i
(6). A suspension of LCAA-CPG (1.3 g; 500 Å, Millipore), 5 (122 mg, 0.12 mmol), and Pr2NEt (1.3 ml) in
DMF (5.2 ml) was shaken for 16 h at 258. After filtration, the solid was washed with DMF and CH2Cl2, sus-
pended in pyridine (1.2 ml) and Ac2O (0.8 ml), and shaken for 2 h at 258. After filtration, the solid was washed
with DMF and CH2Cl2, and dried to give 2. Loading: 30 mmol/g.
3. (C-C-C-C-A-C-C-[2’-(butyldithio)-2’-deoxy]A)-RNA 5’-(Dihydrogen Phosphate) (8). The sequence was
assembled from 60 mg of 6 by using the standard conditions for the assembly of 2’-O-tom-protected ribonucleo-
side phosphoramidites [9], but employing a modified oxidizing reagent: 20 mM I2 in THF/pyridine/H2O 7 :2 :1.

2908
HELVETICA CHIMICA ACTA – Vol. 88 (2005)
After the final detritylation, the solid support 7 was washed with iPr2NH/MeCN 1:9 for 20 min (flow rate 2.5 ml/
min). Cleavage from the solid support and deprotection was carried out with 12 NH3 in MeOH (1 ml) for 6 h at
258. The supernatant was removed by centrifugation and evaporated; the residue was treated with 1 Bu4NF·
3 H2O in THF (1 ml) and 0.5 AcOH for 3 h at 258. Then the mixture was diluted with 1 aq. Tris·HCl (1 ml; pH
M
M
M
M
7.4) and concentrated to 1 ml. After desalting on a NAP cartridge, the crude product was purified by ion-
exchange HPLC. The product-containing fractions were concentrated to ca. 40% of their initial volume and
purified by reversed-phase HPLC (pH 5.5, A/B 9 :1 ! A/B 7:3 (30 min)). The pooled, product-containing frac-
tions were evaporated to 1/3 of the volume and finally desalted on a Sepak cartridge: 30 OD of pure 8 (25% yield
based on 6). Reversed-phase HPLC (pH 3.5; Fig. 1,a): tR 19.8 (100%). ESI-MS (neg mode; Fig. 1,e): 2611 (calc.
2611).
4. Experiments Described in the Table: General Procedure for Deprotection of 8 and Aminoacylation of 9.
To 0.19 m
the indicated buffer (c=0.1
2.5, or 25 m solns., resp., see Table; prior to the addition, the pH of these TCEP stock solns. was adjusted with
aq. NaOH soln.) in the corresponding buffer (c=50 m ) and incubated at 258. After 30 min, an aliquot of 20 ml
was withdrawn and analyzed by reversed-phase HPLC (elution at pH 3.5). To the remaining reaction mixture, a
soln. of 12 in DMF (4 ml; c=10, 5, or 2.5 m , resulting in 12, 6, or 3 equiv., resp., see Table) was added and incu-
M
aq. 8 (27 ml; concentration determined spectrophotometrically at 260 nm) was added an aliquot of
M
; 27 ml) followed by an aq. TCEP soln. (6 ml; c=10, 25, or 100 m , resulting in 1.0,
M
M
M
M
bated at the indicated temp. Aliquots of 20 ml were withdrawn after 10 and 50 min and analyzed by reversed-
phase HPLC (pH 3.5). For experiments at 378, the crude 9 was pre-incubated at 378 before the addition of
12. Identical experiments with the parent RNA sequence (ꢀO)2P(=O)Oꢀr(CCCCACCA) were carried out;
according to HPLC and ESI-MS analyses, no aminoacylated or other products were formed.
Isolation and Characterization of 9. To 0.19 m
M
aq. 8 (81 ml), 0.1
M aq. AcOH/NaOH (81 ml; pH 5.0) was
added, followed by 25 m aq. TCEP (18 ml; in 0.05
M
M AcOH/NaOH, pH 5.0). After 30 min incubation at 258,
the product soln. was desalted on a NAP cartridge. Data of 9: Reversed-phase HPLC (pH 3.5): tR 14.0
(100%); Fig. 1,b. ESI-MS (neg. mode): 2523 (calc. 2523); Fig. 1,f.
5. Experiment Described in Fig. 2. To 0.19 m
27 ml), followed by 10 m aq. TCEP (6 ml; in 0.05
ml was withdrawn and analyzed by reversed-phase HPLC (elution at pH 3.5). To the remaining reaction mixture,
10 m 12 in DMF (4 ml) was added, and the soln. was incubated at 378. Aliquots were taken after 2, 10, 20, 40, 50,
M
aq. 8 (27 ml) was added aq. Tris·HCl buffer (pH 7.4, c=0.1
M;
M
M Tris·HCl buffer, pH 7.4). After 30 min at 258, an aliquot of 20
M
80, 120, and 180 min, and analyzed by reversed-phase HPLC (pH 3.5; see Fig. 2,b, for an example obtained after
a reaction time of 2 min). By integration, the ratio 9/10/11 was determined in each chromatogram and the ratio
(10+11)/(9+10+11) was plotted against the reaction time (Fig. 2,a).
6. Hydrolysis Studies Described in Fig. 3. 6.1. Hydrolysis of 10. An aq. soln. of pure 10 (c=8 m
obtained by HPLC purification and desalting as described above) was incubated for 2 min at 258 or 378, resp.
Then, 1 aq. Tris·HCl (pH 7.4; 100 ml) was added, and the incubation was continued at 258 or 378. At different
M, 900 ml;
M
time intervals, aliquots were withdrawn, treated with AcOH (! pH 3) and analyzed by reversed-phase HPLC
(pH 3.5). The integral ratios between signals of 10 and 9 were translated into individual pseudo-first-order rate
constants (Fig. 3,b): k (378)=0.075 minꢀ1 and k (258)=0.02 minꢀ1
.
6.2. Hydrolysis of 11. Desalted 11/10 3 :1 (total cꢁ8 m
M; 900 ml; obtained at pH 5.0 as described above) was
incubated for 2 min at 378. Then, 1 aq. Tris·HCl (pH 7.4; 100 ml) was added, and the incubation was continued
M
at 378. At different time intervals (2, 4, 8, 12, and 24 min), aliquots were withdrawn, treated with AcOH (! pH
3), and analyzed by reversed-phase HPLC (pH 3.5). The disappearance of the signal of 11 was translated into a
pseudo-first-order rate constant (Fig. 3,b): k (378)=0.15 minꢀ1
.
6.3. Graph Shown in Fig. 3,c. From the rates of hydrolysis k1 (11 ! 10)=0.15 minꢀ1 and k2 (10 ! 9)=0.075
minꢀ1
, the line graphs shown in Fig. 3,c were obtained according to the following equations:
[11]t =[11]0 ·exp(ꢀk1 ·t); [10]t =[([11]0 ·k1)/(k1 ꢀk2)]·[exp(ꢀk2 ·t)ꢀexp(ꢀk1 ·t)]; [9]t =[11]0 ·(1+[k1 ·exp(ꢀk2 ·t)
ꢀk2 ·exp(ꢀk1 ·t)]/[k2 ꢀk1]). The experimentally determined amounts of 9, 10, and 11, obtained at different
times (see above), are also shown in the graph (as points).
7. Hydrolysis of 12. A soln. of 12 in DMF (c=10 MM; 50 ml) was added to 1M aq. Tris·HCl (pH 7.4; 950 ml;
pre-incubated at 258 or 378). At different time intervals, aliquots were withdrawn, treated with AcOH (! pH 3),
and analyzed by reversed-phase HPLC (pH 3.5; gradient A ! B in 30 min; full loop injection). The dissapear-
ance of 12 (tR 19.3 min) was translated into individual pseudo-first-order rate constants: k (258)=0.007 minꢀ1, k
(378)=0.07 minꢀ1
.
8. Experiment Described in Fig. 4. Simultaneous Deprotection of 8 and Aminoacylation of 9, Selective Cleav-
age of 11 to 10, and Isolation and Characterization of 10. To 0.19 m aq. 8 (81 ml), 1 aq. AcOH/NaOH (81 ml;
pH 5.0) was added, and the mixture was incubated at 378 for 2 min. Then, 4 m 12 in DMF (18 ml) was added,
M
M
M

H
ELVETICA
CHIMICA
ACTA – Vol. 88 (2005)
2909
followed by 25 m
M
aq. TCEP (18 ml; in 0.05
M
AcOH/NaOH, pH 5.0). After 30 min incubation at 378, a 20-ml ali-
aq. NaN3 (480
quot was analyzed by reversed-phase HPLC (pH 3.5; see Fig. 4,a). To the remaining soln., 1.5 m
M
ml) was added and incubated for 30 min at 258. The mixture was diluted with H2O (340 ml), desalted on a NAP
cartridge, and characterized by reversed-phase HPLC (pH 3.5; Fig 4,b). Pure product 10 was obtained by
reversed-phase HPLC at pH 3.5, concentration of the product-containing fraction to ca. 60% of its initial vol-
ume, desalting on a NAP cartridge, followed by immediate addition of AcOH (1% final concentration), and lyo-
philization. Data of 10: Reversed-phase HPLC (pH 3.5): tR 16.9 (100%). ESI-MS (neg. mode): 2671 (calc. 2671);
Fig. 1,g.
Isolation and Characterization of 11. To 0.19 m
M
aq. 8 (81 ml), 0.1
M
aq. AcOH/NaOH (81 ml; pH 5.0) was
added. Then, 25 m 12 in DMF (18 ml) was added, followed by 25 m
M
M aq. TCEP (18 ml; in 0.05 M AcOH/
NaOH, pH 5.0). After 30 min incubation at 258, a 20-ml aliquot was analyzed by reversed-phase HPLC (pH
3.5), which showed 11 and 10 in a 9 :1 ratio. The remaining product soln. was desalted on a NAP cartridge
and stabilized immediately by addition of AcOH (final concentration 1%). Data of 11: Reversed-phase
HPLC (pH 3.5): tR 16.9 (10; 10%) and 18.0 (11; 90%). ESI-MS (neg. mode): 2818 (11, 90%; calc. 2819) and
2671 (10, 10%; calc. 2671); Fig. 1,h.
REFERENCES
[1] C. J. Noren, S. J. Anthony-Cahill, M. C. Griffith, P. G. Schultz, Science (Washington, D.C.) 1989, 244, 182;
J. D. Bain, E. S. Diala, C. G. Glabe, T. A. Dix, A. R. Chamberlin, J. Am. Chem. Soc. 1989, 111, 8013; M.
Lodder, S. Golovine, A. L. Laikhter, V. A. Karginov, S. M. Hecht, J. Org. Chem. 1998, 63, 794; T. Hohsaka,
Y. Ashizuka, H. Sasaki, H. Murakami, M. Sisido, J. Am. Chem. Soc. 1999, 121, 12194; T. L. Hendrickson, V.
de Crecy-Lagard, P. Schimmel, Annu. Rev. Biochem. 2004, 73, 147.
[2] N. Lee, Y. Bessho, K. Wei, J. W. Szostak, H. Suga, Nat. Struct. Biol. 2000, 7, 28; Y. Bessho, R. W. Hodgson,
H. Suga, Nat. Biotechnol. 2002, 20, 723.
[3] K. Ninomiya, T. Minohata, M. Nishimura, M. Sisido, J. Am. Chem. Soc. 2004, 126, 15984.
[4] L. Wang, A. Brock, B. Herberich, P. G. Schultz, Science (Washington, D.C.) 2001, 292, 498; T. Ashton
Cropp, P. G. Schultz, Trends Genet. 2004, 20, 625.
[5] P. E. Dawson, T. W. Muir, I. Clark-Lewis, S. B. H. Kent, Science (Washington, D.C.) 1994, 266, 776; T. W.
Muir, P. E. Dawson, S. B. H. Kent, Methods Enzymol. 1997, 289, 266.
[6] J. H. Marriott, M. Mottahedeh, C. B. Reese, Carbohydr. Res. 1991, 216, 257.
[7] A. Eleuteri, C. B. Reese, Q. Song, J. Chem. Soc., Perkins Trans. 1 1996, 2237.
[8] S. Thea, G. Cevasco Tetrahedron Lett. 1988, 29, 2865.
[9] S. Pitsch, P. A. Weiss, L. Jenny, A. Stutz, X. Wu, Helv. Chim. Acta 2001, 84, 3773.
[10] J. W. Ryan, A. Chung, ‘Antihypertensive Agents’, Eur. Patent EP 38117 A1 19811021, 1981.
[11] S. Chladek, M. Sprinzl, Angew. Chem., Int. Ed. 1985, 24, 371.
[12] A. Stutz, C. Höbartner, S. Pitsch, Helv. Chim. Acta 2000, 83, 2477.
[13] P. Wenter, S. Pitsch, Helv. Chim. Acta 2003, 86, 3955.
Received August 10, 2005