Synthesis of a 3-methyluridine phosphoramidite to investigate the role of methylation in a ribosomal RNA hairpin

Article abstract of DOI:10.1016/S0968-0896(01)00283-8

The synthesis of a 5'-O-BzH-2'-O-ACE-protected-3-methyluridine phosphoramidite is reported [BzH, benzhydryloxy-bis(trimethylsilyloxy)silyl; ACE, bis(2-acetoxyethoxy)methyl]. The phosphoramidite was employed in solid-phase RNA synthesis to generate a series of RNA hairpins containing single or multiple modifications, including the common nucleoside pseudouridine. Three 19-nucleotide hairpin RNAs that represent the 1920-loop region (G₁₉₀₆-C₁₉₂₄) of Escherichia coli 23S ribosomal RNA were generated. Modifications were present at positions 1911, 1915, and 1917. The stabilities and structures of the three RNAs were examined by using thermal melting, circular dichroism, and NMR spectroscopy. Copyright

Full text of DOI:10.1016/S0968-0896(01)00283-8

Bioorganic & Medicinal Chemistry 10 (2002) 325–332
Synthesis of a 3-Methyluridine Phosphoramidite to Investigate
the Role of Methylation in a Ribosomal RNA Hairpin
Helen M.-P. Chui,a May Meroueh,a Stephen A. Scaringeb and Christine S. Chowa,*
aDepartment of Chemistry, Wayne State University, Detroit, MI 48202, USA
bDharmacon Research Inc., Lafayette, CO80026, USA
Received 11 May 2001; accepted 1 August 2001
Abstract—The synthesis of a 50-O-BzH-20-O-ACE-protected-3-methyluridine phosphoramidite is reported [BzH, benzhydryloxy-
bis(trimethylsilyloxy)silyl; ACE, bis(2-acetoxyethoxy)methyl]. The phosphoramidite was employed in solid-phase RNA synthesis to
generate a series of RNA hairpins containing single or multiple modifications, including the common nucleoside pseudouridine.
Three 19-nucleotide hairpin RNAs that represent the 1920-loop region (G1906–C1924) of Escherichia coli 23S ribosomal RNA were
generated. Modifications were present at positions 1911, 1915, and 1917. The stabilities and structures of the three RNAs were
examined by using thermal melting, circular dichroism, and NMR spectroscopy # 2001 Elsevier Science Ltd. All rights reserved.
Introduction
new pseudouridine phosphoramidite and examined the
role of É residues at positions 1911, 1915, and 1917 of
E. coli 23S rRNA.8 The synthetic strategy involved the
use of a 50-O-silyl-20-O-orthoester protection system9,10
to generate the new phosphoramidite. In this study, a
similar protection system was employed to generate a 50-
O-BzH-20-O-ACE-3-methyluridine-30-phosphoramidite
[BzH, benzhydryloxy-bis(trimethylsilyloxy)silyl; ACE,
bis(2-acetoxyethoxy) methyl] that was compatible with
solid-phase RNA synthesis.
The natural base modification 3-methylpseudouridine
(m3É) occurs in the 1920-loop region of Escherichia coli
(E. coli) 23S ribosomal RNA (rRNA) at position 1915.1
A pseudouridine residue (É) at position 1915 appears to
be conserved among all organisms examined to date,
however, methylation of the pseudouridine at this posi-
tion may be less common.2 A number of recent cross-
linking and crystallographic studies on the ribosome
have implicated the 1920 loop of 23S rRNA (helix 69) in
playing an important role in mediating subunit associa-
tion and tRNA interactions.3À7 Many questions still
remain, however, regarding the role of pseudouridine in
this process and the influence of pseudouridine methy-
lation on the overall structure and function of 23S
rRNA.
Three hairpin RNAs were synthesized in which their
nucleotide sequences were derived from the 1920 region
of E. coli 23S rRNA. These RNAs contained a combi-
nation of uridine, 3-methyluridine, and pseudouridine
at positions 1911, 1915, and 1917. In order to under-
stand more clearly the role of methylation at position
1915, the stabilities and structures of the modified and
unmodified RNAs were examined through a variety of
biophysical methods and compared to previous studies
with RNAs containing as many as three pseudouridine
residues.8
The goal of the research described here was to probe the
specific structural and stabilizing role of the methyl
group at position 1915 in 23S rRNA by using the
nucleotide analogue 3-methyluridine (m3U). In order to
achieve this goal, the analogue was first converted into a
phosphoramidite and incorporated site-specifically into
short RNA oligonucleotides by using solid-phase
chemistry. We reported recently on the synthesis of a
Results
Synthesis of a 3-methyluridine phosphoramidite
We first prepared 30,50-O-(1,1,3,3-tetraisopropyl-1,3-
disiloxanediyl)uridine (1) (Scheme 1) according to the
*Corresponding author. Tel.: +1-313-577-2594; fax: +1-313-577-
8822; e-mail: csc@chem.wayne.edu
0968-0896/02/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(01)00283-8

326
H. M.-P. Chui et al. / Bioorg. Med. Chem. 10 (2002) 325–332
Scheme 1. (i) TIPDSCl2, pyridine, 0 ꢁC!rt; (ii) TMSCl, Et3N, CH2Cl2, 0 ꢁC!rt; (iii) N,N-dimethylformamide dimethylacetal (DMF-DMA), ben-
zene, reflux; (iv) (a) pTSA–H2O, Et3N, THF; (b) tris(2-acetoxyethoxy)orthoformate, pyridinium p-toluenesulfonate, 4-(tert-butyldimethylsilyloxy)-
3-penten-2-one, dioxane, rt for 48 h; (v) TMEDA-HF, CH3CN, 0 ꢁC ; (vi) BzH–Cl, diisopropylamine, CH2Cl2, 0 ꢁC; (vii) methyl tetraisopropyl
phosphoradiamidite, 4,5-dicyanoimidazole, CH2Cl2, rt.
literature11À14 from commercially available uridine.
Simultaneous protection of the 30- and 50-hydroxyl
groups allows for the selective addition of the 20-
hydroxyl protective group. Furthermore, the di-
siloxanediyl ester can be cleanly and efficiently removed
by exposure to TMEDA and hydrogen fluoride. We
next prepared the 20-O-trimethylsilyl derivative 2 in
quantitative yield followed by methylation of the uri-
dine moiety. Conventional methylation of base residues
typically involves the use of methyl iodide or trimethyl
phosphate, but these procedures generally require long
reaction times.15 In contrast, methylation of the base
moiety with N,N-dimethylformamide dimethylacetal16
is a convenient and facile method that takes place under
neutral conditions with shorter reaction times. This
method was used to generate 3-methyl-50-O-(1,1,3,3-tet-
raisopropyl)-1,3-disiloxanediyl)-20-O-(trimethylsilyl)ur-
idine (3) in 93% yield.
Compound 3 was treated with p-toluenesulfonic acid
and triethylamine to expose the 20-hydroxyl group
which was then reacted with tris(2-acetoxyethoxy)-
orthoformate and pyridinium-p-toluenesulfonate fol-
lowed by 4-(tert-butyldimethylsilyloxy)-3-penten-2-one
to give the 20-O-ACE protected compound 4b in 88%
yield. Subsequent removal of the 30,50-O-protective
group, silylation of the 50-OH with the BzH group, and
30-O-phosphitylation to form the phosphoramidite were
carried out as described previously.8À10 The final phos-
phoramidite product 7 was synthesized in seven steps
with an overall 27–38% yield (Scheme 1). The RNA
building block 7 containing the 50-O-BzH ether, 20-O-
ACE orthoester protective group, and 30-methyl-N,N-
diisopropylphosphoramidite was fully characterized by
1H, 13C, and 31P NMR spectroscopy, electrospray mass
spectrometry, and elemental analysis. The remaining
nucleotides (G, C, A, U, and É) were converted to
their corresponding phosphoramidites as described
previously.8À10
Synthesis and analysis of RNAs representing the
sequence from the 1920 region of E. coli 23S ribosomal
RNA
Three RNAs were synthesized in high yields (ꢀ4 mg per
1 mmol synthesis) from the building block 7 and the
remaining G, C, A, U, and É phosphoramidites. Figure
1 shows the sequences of the RNAs with the positions
of modifications (m3U or É). The natural E. coli 23S
rRNA has pseudouridine residues at positions 1911 and
Figure 1. The structures of 3-methyluridine (m3U) and pseudouridine
(É) are shown. Secondary structure representations of the synthetic
RNAs are based on the 1920-hairpin region (helix 69) of E. coli 23S
rRNA (nucleotides G1906 to C1924). The nucleotide positions have been
renumbered consecutively from G1 to C19. Positions 6 (1911), 10
(1915), and 12(1917) are modified with m 3U or É.

H. M.-P. Chui et al. / Bioorg. Med. Chem. 10 (2002) 325–332
327
19172 and a methylated derivative, m3É, at position
1915.1 In this study, uridines at positions 1911, 1915,
and 1917 were replaced with combinations of pseu-
douridine and 3-methyluridine. All RNAs will be refer-
red to by a three-nucleotide notation indicating the
composition at positions 1911, 1915, and 1917, respec-
tively (i.e., Ém3UÉ, Um3UU, and ÉUÉ). The presence
of the modified nucleosides was confirmed by P1 nucle-
ase digestion and alkaline phosphatase treatment of
each RNA, followed by HPLC analysis (data not
shown). Three other RNAs, UUU, ÉÉÉ and UÉU,
were used for comparison with the m3U-containing
RNAs. These RNAs were synthesized and studied in
previous work8 and refer to the unmodified RNA and
oligoribonucleotides containing pseudouridines at all
three positions (1911, 1915, and 1917) or one position
(1915), respectively.
and Um3UU differ by 0.1 kcal/mol). In contrast, it was
shown previously that the presence of É1915 leads to a
0.7 kcal/mol destabilization relative to U1915 when two
uridines are present at positions 1911 and 1917.8 A fur-
ther comparison of the É-containing RNAs (ÉUÉ, and
Ém3UÉ) reveals that U1915 and m3U1915 have similar
effects on RNA stability. In contrast, previous work
demonstrated that É1915 was destabilizing by 0.5 kcal/
mol if pseudouridines are present at positions 1911 and
1917.8 As summarized in Table 1, N3 methylation of
U1915 does not appear to have any effect on the stability
of the 1920-region RNA hairpin, regardless of whether
the neighboring nucleotides in the loop region (position
1917) and closing base pair (position 1911) are uridine
or pseudouridine.
The effects of N3 methylation on the structure of the
1920-region RNAs. The circular dichroism (CD) spectra
of the 1920-region hairpin RNAs were obtained in order
to compare the effects of uridine methylation on their
structures. As shown in Figure 3, the spectra of all
RNAs in this study have maxima centered around 265
nm and minima near 240 nm, which is typical for A-
form RNAs. The CD spectra for UUU and Um3UU
RNAs (Figure 3, open circles and closed circles, respec-
tively) are nearly identical, suggesting that when uri-
dines are present at residues 1911 and 1917 there is little
influence from uridine methylation at position 1915 on
the overall folded RNA structure. Similarly, compar-
ison of the CD spectra for ÉUÉ and Ém3UÉ (Figure 3,
open triangles and closed triangles, respectively) reveals
The effects of N3 methylation on the stability of the
1920-region RNAs
For the three new RNAs, Um3UU, Ém3UÉ, and ÉUÉ,
absorbance versus temperature profiles were obtained at
pH 7.0 in low salt conditions (35 mM Na+) and ana-
lyzed in terms of melting temperature (Tm), ÁHꢁ, ÁSꢁ,
and ÁGꢁ
.
Each measurement was taken in
8,17,18
37
duplicate and a range of RNA concentrations was
examined (ꢀ5–200 mM). The normalized absorbance
plots for single RNA concentrations are shown in Fig-
ure 2. The helix-to-coil transitions were independent of
RNA concentration over the range examined, thus con-
firming the existence of hairpin structures for each RNA.
The thermodynamic parameters for the three RNAs
(Um3UU, Ém3UÉ, and ÉUÉ) are given in Table 1.
no differences. In contrast, comparison of the É1911
-
and É1917-containing RNAs (Figure 3, open and closed
triangles) with U1911- and U1917-containing RNAs (Fig-
ure 3, open and closed circles) indicates differences in
the peak shapes and cross-over points (242 nm vs 250
nm for the É- and U-containing RNAs, respectively).
These results are consistent with previous studies in
which pseudouridine modification had an impact on the
folded 1920-region RNA structure.8 In both cases,
however, N3 methylation of U1915 appears to have a
negligible influence on the global RNA fold.
The thermal melting data demonstrate that the presence
of the m3U1915 methyl group has little effect on RNA
hairpin stability (see Table 1; the ÁGꢁ37 values for UUU
NMR spectroscopy was used to examine the hydrogen-
bonding interactions in the 1920-region RNAs contain-
ing methylations and pseudouridines. The 1H NMR
spectra (imino proton region) of the RNAs are shown in
Table
1. Thermodynamics
of
1920-loop
RNAs,
50-
G1906GCCGXAACYAZAACGGUC1924-30, with pseudouridine (É)
and 3-methyluridine (m3U) modifications at positions X1911, Y1915
and Z1917
,
ÁGꢁ37 (kcal/mol)a ÁHꢁ (kcal/mol) ÁSꢁ a(e.u.) Tm (ꢁC)
a
XYZ
UUUb
Um3UU
ÉUÉ
À4.9
À5.0
À5.3
À5.3
À62.2
À63.6
À61.3
À60.9
À184.8
À188.9
À180.3
À179.5
63.7
63.5
66.5
66.4
Figure 2. Representative normalized UV melting curves for the mod-
ified RNAs taken in 15 mM NaCl, 20 mM sodium cacodylate, 0.5 mM
EDTA, pH 7.0 are shown. The melting curves were obtained using an
Aviv 14DS UV–vis spectrometer. The curves for singly modified RNA
(Um3UU), pseudouridine-modified RNA (ÉUÉ), and triply modified
RNA (Ém3UÉ) are represented by closed circles, open triangles, and
closed triangles, respectively. The melting curves were normalized at
95 ꢁC.
Ém3UÉ
aConservative estimates of standard errors for ÁGꢁ37, ÁHꢁ, and ÁSꢁ
are 5%, 7%, and 8%, respectively.8,17,18 Best fits were obtained by
assuming hairpin formation (the melting profiles were concentration
independent). The buffer conditions were 15 mM NaCl, 20 mM
sodium cacodylate, and 0.5 mM Na2EDTA, pH 7.0.
bData were obtained from ref 8.

328
H. M.-P. Chui et al. / Bioorg. Med. Chem. 10 (2002) 325–332
Figures 4 and 5. The spectrum of the unmodified RNA
(UUU) (Fig. 4A) shows five imino proton resonances
that were assigned previously by using 1D NOE differ-
ence spectroscopy.8 The formation of three G-C base
pairs and one G-U mismatch are evident from the
spectrum. The loop closing U6–A14 pair (the numbering
system 1–19 has been used for NMR analysis) was not
imino resonances. One additional peak is observed at
13.3 ppm (ꢂ), which also could not be assigned based on
the 1D NOE difference spectra. Overall, these results are
consistent with the CD studies in which the Ém3UÉ
1
observed.8 The H NMR spectrum for the newly syn-
thesized Um3UU RNA (Fig. 4B) is very similar to the
UUU RNA spectrum, except for one additional weak
resonance at 10.2ppm (data not shown) that could not
be assigned based on the 1D NOE difference spectra.
1
The H NMR spectrum of UÉU was also assigned pre-
viously and is shown here for comparison (Fig. 4C).8
These data suggest that N3 methylation at position 1915
is not having much influence on the overall RNA sec-
ondary structure, in contrast to pseudouridine which
had subtle affects on the hairpin loop structure. These
results are consistent with the CD data in which
m3U1915 had little influence on the global RNA struc-
tures when uridines were present at positions 1911 and
1917, whereas the presence of É1915 led to modest
changes in the CD spectrum.8
Figure 4. The 1D imino proton (guanine N1H/uridine N3H/pseu-
douridine N1H and N3H) NMR spectra obtained at 3 ꢁC on a Varian
UNITY 500 MHz spectrometer in 30 mM NaCl, 10 mM sodium
phosphate, 0.5 mM Na2EDTA, pH 6.5 (90% H2O/10% D2O) are
shown. The RNA concentrations for UUU (A), Um3UU (B), and
UÉU (C) were 0.8 to 1 mM. Nucleotide assignments based on the 1D-
NOE difference spectroscopy (data not shown) are indicated above
each resonance with the nucleotides numbered consecutively from G1
to C19. The spectra in panels A and C were assigned previously8 and
are shown here to allow for direct comparison to Um3UU.
In contrast, the ÉUÉ RNA spectrum (Fig. 5A) exhibits
seven imino proton resonances. These resonances are
associated with five base pairs, including a É6–A14 pair.
The É6–A14 assignment is based on the additional reso-
nance at 13.1 ppm (É6N3H). The É6N1H assignment at
10.2ppm is based on a strong NOE from É6N1H to
É6H6 (data not shown). The previously assigned ÉÉÉ
RNA spectrum8 is shown in Figure 5C for comparison.
The Ém3UÉ RNA spectrum (Fig. 5B) exhibits eight
Figure 3. CD spectra of the unmodified and modified RNAs taken on
a Jasco J600 spectropolarimeter from 210 to 320 nm at ambient tem-
perature in 15 mM NaCl, 20 mM sodium cacodylate, 0.5 mM EDTA,
pH 7.0 are shown. The molar ellipticities were normalized to RNA
concentration (4.9, 7.4, 3.2, and 3.6ꢃ10À6 M in molecules of RNA
for UUU, Um3UU, ÉUÉ, and Ém3UÉ, respectively) based on an
extinction coefficient of 188,860 cmÀ1 MÀ1 for all RNAs and A260
values at 90 ꢁC. The spectra for unmodified (UUU), singly modified
(Um3UU), pseudouridine-modified (ÉUÉ), and triply modified
(Ém3UÉ) RNAs are represented by open circles, closed circles, open
triangles, and closed triangles, respectively. Each spectrum is the
average of four scans. The spectrum for UUU was obtained pre-
viously8 and is shown here to allow for direct comparison to Um3UU.
Figure 5. The 1D imino proton (guanine N1H/uridine N3H/pseu-
douridine N1H and N3H) NMR spectra for RNAs ÉUÉ (A),
Ém3UÉ (B) and ÉÉÉ (C) are shown. The experimental conditions
were the same as those described in Figure 4. The spectrum in panel C
was assigned previously8 and is shown here for comparison.

H. M.-P. Chui et al. / Bioorg. Med. Chem. 10 (2002) 325–332
329
and ÉUÉ RNAs had essentially the same global A-
form structures. In both of these RNAs, the presence of
É1911 (É6 by the NMR numbering scheme) appears to
be important for maintaining the closing base pair (É6–
A14), which may further influence the loop structure.
The presence É1917 of may also influence the loop
structure or closing base pair stability. Such effects have
been observed previously in other É-containing
RNAs.8,19,20
residue É1917 have an influence on the hairpin loop
structure as shown by differences in the shapes and
crossover points of the CD spectra (compare closed cir-
cles and closed triangles in Fig. 3) and presence of
additional resonances in the NMR spectra relative to
the Um3UU RNA (compare Figs 4B and 5B).
In summary, the 1920-region RNA with sequence
U
1911m3U1915U1917 lacks the formation of a U1911–A1919
pair as was observed previously with UÉU and UUU
RNAs.8 In contrast, RNAs with the sequences
É1911U1915É1917 and É1911m3U1915É1917 contain
a
Discussion
É1911-A1919 closing base pair that serves to stabilize the
hairpin structure. The same effect was observed with
ÉÉÉ and ÉUU RNAs.8 For both sequence contexts, U
and m3U at position 1915 are iso-energetic, whereas É
was destabilizing by ꢀ0.5 kcal/mol.8 In the present
study, the fact that methylation at the N3 position of
U1915, which is isosteric with É-N3, has no effect on the
RNA stability or secondary structure suggests that the
N3H of É1915 does not participate in hydrogen-bonding
interactions.
The properties of the natural modification m3É at posi-
tion 1915 that contribute to the E. coli 23S rRNA
structure and function are still not well understood. One
possible function of m3É1915 in the ribosome might be
to maintain a specific folded conformation of the 23S
rRNA, perhaps by stabilizing the 1920-loop structure
through its N1H or N3CH3 groups. Alternatively, this
modified nucleoside might be involved in long-range
contacts with other regions of the ribosome such as 16S
rRNA and tRNAs.7 In order to understand the con-
tribution of the N3 methyl group to the local structure
of the 1920 region, we synthesized several model RNAs
containing a 3-methyluridine derivative at position
1915. The m3U residue is isosteric with m3É on the
major groove side, such that the influence of the methyl
group alone can be assessed. The structures of the
modified RNAs were compared to an unmodified RNA
through a variety of techniques, including CD spectro-
scopy, thermal melting, and 1D imino proton NMR
spectroscopy.
Taken together, these results suggest important roles for
both É1911 and m3U1915 in modulating the RNA struc-
ture and stability. The more hydrophilic nucleotide
É1911 is stabilizing relative to U, presumably due to
additional hydrogen bonding through its N1H.8 On the
other hand, the presence of the N1H (on É) at position
1915 is destabilizing,8 whereas N3 methylation (on
m3U) has no effect on the RNA stability. These results
are in contrast to other work in which a m3U residue is
destabilizing relative to U when the modification is
located at a closing base pair and the U-N3H is
involved in hydrogen-bonding interactions.21 Results
similar to those obtained in this work were observed,
however, when a m3U residue replaced a uridine in a U-
turn motif from the anticodon loop of tRNA.20 Within
the tRNA anticodon sequence context, m3U was iso-
energetic with uridine. Ashraf and coworkers suggested
that modifications at this site in the RNA were more
likely to influence long-range interactions with other
RNAs rather than playing a role in the local secondary
folding of the anticodon loop.20 Similarly, Rife and
coworkers demonstrated that an N2-methylguanine
(m2G) modification is iso-energetic with guanine in the
context of several common secondary-structure
motifs.22 These authors suggested that the methylated
nucleoside might be introduced into natural RNAs in
order to stabilize RNA–RNA or RNA–protein interac-
tions rather than local secondary structures. The crea-
tion of a hydrophobic patch on the RNA through
methylation could play a role in specific long-range
tertiary interactions.
Modifications at positions 1911, 1915, and 1917 of the
1920 region of 23S rRNA were shown previously to
modulate the stability and local secondary structure of
the RNA.8 For example, pseudouridines either stabilize
or destabilize the structure, depending on which posi-
tions they occupy. A singly modified É1915 variant
(UÉU) was 0.7 kcal/mol less stable than the corre-
sponding unmodified RNA (UUU), and a U1911–A1919
loop-closing base pair (U6–A14 in the NMR numbering
system) was not observed in either case.8 In contrast,
only negligible effects on stability are observed in this
study for the singly modified Um3UU RNA relative to
the unmodified RNA, and the closing U1911–A1919 pair
is not observed.
This work and previous studies8 reveal that a pseu-
douridine at position 1911 is important for stabilizing
the É1911–A1919 closing base pair of the hairpin loop.
The presence of É1911 in a singly modified RNA (ÉUU)
led to a 1.0 kcal/mol stabilization and appearance of
additional resonances in the 1D imino proton spectrum
in comparison to the UUU RNA.8 Related studies have
shown that pseudouridines at loop-closing base pairs
stabilize RNA structure, possibly due to water-mediated
hydrogen-bonding interactions between the RNA-
phosphate backbone and É-N1H.19,21 The presence of
closing É-A pairs may also influence the conformation
or stability of the adjacent hairpin loops.8,19,20 In the
1920-region Ém3UÉ RNA, the É1911–A1919 pair and/or
In the 1920-region RNA, the lack of secondary struc-
ture stabilization or alteration in conformation from N3
methylation suggests that this modification is also more
likely to play a role in long-range tertiary interactions in
the ribosome. This idea is supported by cross-linking
and X-ray crystal studies on the ribosome in which the
1920 helix was shown to be involved with close contacts
with the 30S subunit and 16S rRNA or tRNA.3À5,7
A

330
H. M.-P. Chui et al. / Bioorg. Med. Chem. 10 (2002) 325–332
recent X-ray structure of the free 50S subunit revealed
that the 1920 hairpin is solvent exposed,6 and chemical
probing experiments in our laboratory have also shown
that the 1915 residue in this hairpin is solvent exposed
(data not shown). Thus, it can be concluded that the
modified base m3É1915 in E. coli 23S rRNA is more
likely to be involved in long-range interactions than in
the maintenance of local RNA conformations, although
the effects of methylation on the electronics of É could
not be assessed here. Such long-range interactions could
be important for ribosomal subunit association. The
specific role of the m3É1915 in place of the unnatural
analogue m3U1915 in the 1920-region RNA is the subject
of current investigations in our laboratories. The effi-
cient chemical synthesis of specifically N3-methylated
pseudouridine and the corresponding phosphoramidite
still remains a challenge and limitation for structure
studies on this naturally occurring modification.
a 20–60% ethyl acetate in hexane gradient to give a
1
white solid (1.96 g, 83%). H NMR (CDCl3, 400 MHz)
d (ppm): 1.00–1.10 (28H, m), 3.17 (1H, d, J=1.6 Hz),
4.00 (1H, dd, J=2.4, 13.2 Hz), 4.11 (1H, dt, J=2.4, 9.2
Hz), 4.17–4.22 (2H, m), 4.36 (1H, dd, J=4.8, 8.8 Hz),
5.69 (1H, dd, J=1.6, 8.4 Hz), 5.73 (1H, s), 7.70 (1H, d,
J=8.0 Hz), 8.87 (1H, s); 13C NMR (CDCl3, 400 MHz) d
(ppm): 12.5, 12.9 (d), 13.3, 16.8, 16.89, 16.94, 17.0,
17.21, 17.27, 17.35, 17.4, 60.3, 69.0, 75.1, 81.9, 90.9,
101.9, 140.0, 149.8, 163.1.
30, 50-O-(1,1,3,3-tetraisopropyl-1,3-disiloxanediyl)-20-O-
(trimethylsilyl)uridine (2). A solution of 1 (1.96 g, 4.03
mmol) in CH2Cl2 (45 mL) was cooled to 0 ꢁC. To this
mixture were added slowly via syringe triethylamine
(2.80 mL, 20.2 mmol) and trimethylsilyl chloride
(TMSCl) (1.54 mL, 12.1 mmol). The solution was stir-
red for 2h at 0 ꢁC then for 1.5 h at room temperature.
The resulting red solution was poured over 100 mL of
cold 1.0 M aqueous NaHCO3 and extracted twice with
CH2Cl2. The organic layers were combined, washed
with brine, dried over Na2SO4, and concentrated in
vacuo to afford a light yellow foam (2.30 g, 100%)
which was used for the next step without further pur-
ification: 1H NMR (CDCl3, 400 MHz) d (ppm): 0.19
(9H, s), 0.99–1.09 (28H, m), 3.96 (1H, dd, J=2.4, 13.6
Hz), 4.05 (1H, dd, J=4.4, 9.4 Hz), 4.12(1H, d, J=4.4
Hz), 4.16 (1H, d, J=9.6 Hz), 4.25 (1H, d, J=13.6 Hz),
5.58 (1H, s), 5.66 (1H, dd, J=1.6, 8.0 Hz), 7.97 (1H, d,
J=8.0 Hz), 9.36 (1H, s, NH); 13C NMR (CDCl3,
400 MHz) d (ppm): 0.5, 13.0, 13.1, 13.5, 13.6, 17.0, 17.1,
17.2, 17.3, 17.4, 17.6, 17.7, 17.8, 59.7, 68.3, 76.6, 81.7,
91.4, 101.4, 140.0, 150.2, 164.1. Exact mass calculated
for C24H46N2O7Si3: 558.6; found by LC-MS (ES+):
[M+Na]+ 581.2.
Experimental
General
All reactions were carried out at room temperature
under an inert atmosphere and anhydrous condition
unless other noted. Tris(acetoxyethoxy)orthoformate,
benzhydryloxy-bis(trimethylsilyloxy)silyl chloride, and
4-(tert-butyldimethylsilyloxy)-3-penten-2-one were pre-
pared as described elsewhere.9,10 All other reagents were
obtained from Aldrich (Milwaukee, WI). All solvents
for RNA purification and HPLC analysis were spectra
1
grade. H, 13C and 31P NMR spectra were obtained in
deuterated solvents on either 400 or 500 MHz spectro-
meters (Varian UNITY-400 and Unity-500, respec-
tively). The chemical shifts of the imino protons of the
1920 hairpin RNAs were reported relative to 3-(tri-
methylsilyl)propionate (Wilmad, Buena, NJ) at 0.0
ppm. Microanalysis (elemental analysis) for all com-
pounds was performed by Midwest Microlab (Indiana-
polis, IN). Nuclease digestions and HPLC analyses for
the RNAs were performed as described previously.8 The
RNA sample preparation and experimental conditions
for the thermal melting studies, circular dichroism (CD)
spectroscopy, and NMR studies were as reported
previously.8
3-Methyl-30,50 -O-(1,1,3,3-tetraisopropyl-1,3-disiloxane-
diyl)-20-O-(trimethylsilyl)uridine (3). To a solution of 2
(2.30 g, 4.12 mmol) in 25 mL benzene was added drop-
wise N,N-dimethylformamide dimethyl acetal (DMF-
DMA) (1.64 mL, 12.3 mmol). The reaction refluxed for
7 h and the solvent was evaporated. The crude product
was purified by flash chromatography on silica gel using
a 25–60% ethyl acetate in hexane gradient to afford a
1
white solid (2.20 g, 93%). H NMR (CDCl3, 400 MHz)
d (ppm): 0.2(9H, s), 0.98–1.09 (28H, m), 3.33 (3H, s),
3.96 (1H, dd, J=2.4, 14.0 Hz), 4.05–4.11 (2H, m), 4.16
(1H, dd, J=1.6, 8.8 Hz), 4.25 (1H, d, J=13.6 Hz), 5.59
(1H, s), 5.70 (1H, d, J=8.0 Hz), 7.93 (1H, d, J=8.4
Hz); 13C NMR (CDCl3, 400 MHz) d (ppm): 0.3, 12.7,
12.8, 13.2, 13.4, 16.8, 16.9, 16.96, 17.0, 17.2, 17.3, 17.4,
17.5, 27.4, 59.5, 68.1, 76.5, 81.4, 91.5, 100.6, 137.3,
150.7, 163.2. Exact mass calculated for C25H48N2O7Si3:
572.6; found by LC-MS (ES+): [M+Na]+ 595.1,
[2M+Na]+ 1167.4.
Preparation of the 50-O-BzH-20-O-ACE-3-methyl-
uridine-30-phosphoramidite
30,50-O-(1,1,3,3-tetraisopropyl-1,3-disiloxanediyl)-uridine
(1). Uridine (1.30 g, 5.35 mmol) was dried by azeotropic
removal of water with benzene in vacuo for 2h and then
dissolved in 10 mL pyridine to give a clear solution. 1,3-
Dichloro-1,1,3,3-tetraisopropyldisiloxane (TIPDSCl2)
(1.53 mL, 4.86 mmol) was added dropwise to the cooled
solution (0 ꢁC) and stirred for 6 h. The solvent was eva-
porated and the remaining residue was taken up in ethyl
acetate, washed twice with saturated NH4Cl, pH 8, and
washed once with water. The organic layers were com-
bined and washed with aqueous brine, dried over
Na2SO4, filtered, and concentrated. The crude product
was purified by flash chromatography on silica gel using
3-Methyl-30,50 -O-(1,1,3,3-tetraisopropyl-1,3-disiloxane-
diyl)uridine (4a). To a clear solution of 3 (2.12 g, 3.70
mmol) in THF (53 mL) was added p-toluenesulfonic
acid monohydrate (pTSA–H2O) (1.06 g, 5.55 mmol).
After stirring the reaction solution for 1.5 h, triethyla-
mine (1.1 mL, 8.0 mmol) was added dropwise and stir-
red for another 15 min. The reaction was quenched with

H. M.-P. Chui et al. / Bioorg. Med. Chem. 10 (2002) 325–332
331
5% aqueous NaHCO3 and extracted with ethyl acetate
(2ꢃ100 mL). The organic layers were combined, washed
with brine, dried over Na2SO4, and concentrated to give
a light yellow oil which was purified by flash chromato-
graphy on silica gel with 25–50% ethyl acetate in hexane
gradient to give a colorless thick oil. Repeated co-eva-
poration with CH2Cl2 gave a white foam (1.90 g,
100%). 1H NMR (CDCl3, 400 MHz) d (ppm): 1.00–1.09
(28 H, m), 2.84 (1H, d, J=1.2Hz), 3.33 (3H, s), 4.00
(1H, dd, J=2.4, 12.9 Hz), 4.09 (1H, ddd, J=2.4, 8.7
Hz), 4.14 (1H, d, J=5.1 Hz), 4.21 (1H, dd, J=1.8, 12.9
Hz), 4.37 (1H, dd, J=5.1, 9.1 Hz), 5.74 (1H, d, J=8.1
Hz), 5.76 (1H, s), 7.66 (1H, d, J=8.4 Hz); 13C NMR
(CDCl3, 400 MHz) d (ppm): 12.5, 12.8, 12.9, 13.3, 16.8,
16.87, 16.94, 17.0, 17.2, 17.24, 17.3, 17.4, 27.5, 60.2,
69.0, 75.3, 81.8, 91.2, 101.2, 137.5, 150.7, 162.9. Exact
mass calculated for C22H40N2O7Si2: 500; found by LC-
MS (ES+): [M+H]+ 501.
with 0.5% TMEDA in the eluents (33% ethyl acetate in
hexane, then 0–5% CH3OH in ethyl acetate) to give a
pale yellow oil (1.43 g, 93%). 1H NMR (CDCl3,
500 MHz) d (ppm): 2.07 (6H, s), 2.97 (1H, bs), 3.04 (1H,
bs), 3.31 (3H, s), 3.70–3.75 (1H, m), 3.77–3.85 (4H, m),
3.97 (1H, dd, J=2.0, 12.5 Hz), 4.12–4.14 (1H, m), 4.17–
4.26 (4H, m), 4.34 (1H, t, J=5.0 Hz), 4.66 ( 1H, t,
J=5.0 Hz), 5.45 (1H, s), 5.71 (1H, d, J=4.5 Hz), 5.78
(1H, d, J=8.5 Hz), 7.59 (1H, d, J=8.0 Hz); 13C NMR
(CDCl3, 400 MHz) d (ppm): 20.78, 20.81, 27.6, 61.9,
62.7, 62.8, 62.9, 63.2, 69.8, 75.7, 85.2, 92.0, 101.9, 112.5,
139.8, 151.1, 162.7, 170.86, 170.89. Exact mass calcu-
lated for C19H28N2O12: 476.1; found by LC-MS (ES+):
[M+Na]+ 499.1.
50 - O - [benzhydryloxy - bis(trimethylsilyloxy)silyl] - 20 - O -
[bis(2 - acetoxyethoxy)methyl-3-methyluridine (6). Solu-
tion A: to a solution of 5 (0.221 g, 0.46 mmol in 2.5 mL
CH2Cl2) was added diisopropylamine (65 mL, 0.46
mmol) and cooled to 0 ꢁC. Solution B: diisopropylamine
(156 mL, 1.1 mmol) was added dropwise to benz-
hydryloxy-bis(trimethylsiloxy)silyl chloride (BzH-Cl)
(395 mg, 0.93 mmol) in CH2Cl2 (0.8 mL) at 0 ꢁC . Ali-
quots of solution B (0.5 equiv and 0.16 equivꢃ5 aliquots)
were added to solution A at 0 ꢁC dropwise over 2min
and stirred for 10 min after each addition. The reaction
progress was monitored by TLC (hexane/ethyl acetate
1:1). Upon completion, the reaction was quenched with
5% NaHCO3 and extracted with CH2Cl2 (2100 mL).
The combined organic layers were washed with brine,
dried over Na2SO4, concentrated, and purified by flash
chromatography on silica gel with 0.5% triethylamine
(TEA) in the eluents (hexane/acetone 4:1 followed by
hexane/ethyl acetate/acetone 3:1:1) to afford a colorless
oil (343 mg, 86%). 1H NMR (CDCl3, 400 MHz) d
(ppm): 0.02 (9H, s), 0.03 (9H, s), 2.00 (3H, s), 2.02 (3H,
s), 2.98 (1H, d, J=6.0 Hz), 3.26 (3H, s, NCH3), 3.73–
3.82(5H, m), 3.93–3.98 (2H, m), 7.14–7.31 (10H, m),
7.75 (1H, d, J=8.0 Hz); 13C NMR (CDCl3, 500 MHz) d
(ppm): 1.28, 1.29, 20.57, 20.59, 27.3, 61.2, 62.6, 62.7,
62.8, 68.3, 76.7, 77.3, 83.7, 88.0, 101.3, 112.2, 126.0,
126.1, 127.1, 128.1, 137.3, 143.5, 143.6, 150.8, 162.5,
170.5, 170.6. Exact mass calculated for C38H56N2O15Si3:
864.6; found by LC-MS (ES+): [M+Na]+ 887.2.
20-O-[bis(2-acetoxyethoxy)methyl]-3-methyl-30,50-O-(1,1,3,3-
tetraisopropyl-1,3-disiloxanediyl)uridine (4b). To a 100
mL round bottom flask, compound 4a (1.85 g, 3.70
mmol in 10 mL CH2Cl2), tris(2-acetoxyethoxy)-
orthoformate (2.74 g, 8.51 mmol) and pyridinium p-
toluenesulfonate (1.11 g, 4.40 mmol) were added to give
a clear solution. The reaction was stirred for 2h, then 4-
(tert-butyldimethylsilyloxy)-3-penten-2-one (1.6 mL, 6.7
mmol) was added and stirred for 48 h. The reaction
progress was monitored by TLC with 25% acetone in
hexane containing 0.5% N,N,N0,N0-tetramethyl-
ethylenediamine (TMEDA). Upon completion, the
reaction was quenched by the addition of TMEDA (0.2
mL, 1.8 mmol) and stirred for 15 min. The resulting
brown reaction mixture was diluted with 3 mL of hex-
anes/CH2Cl2 (5:1) and purified by flash chromato-
graphy on silica gel using 0.5% TMEDA in hexane/
1
acetone (4:1) to afford a colorless oil (2.32 g, 88%). H
NMR (CDCl3, 400 MHz) d (ppm): 0.99–1.09 (28H, m),
2.05 (3H, s), 2.07 (3H, s), 3.31 (3H, s), 3.82–4.0 (7H, m),
4.16–4.30 (6H, m), 5.72(2H, d, J=8.4 Hz), 5.77 (1H, d,
J=7.2Hz), 7.84 (1H, d, J=8.4 Hz); 13C NMR (CDCl3,
400 MHz) d (ppm): 12.5, 12.8, 13.0, 13.3, 16.7, 16.9,
17.0, 17.1, 17.2, 17.3, 17.4, 20.8, 20.9, 27.4, 59.2, 60.7,
62.0, 63.1, 63.4, 67.8, 77.2, 81.7, 89.4, 101.0, 111.6,
136.8, 150.5, 162.8, 170.8, 170.9. Exact mass calculated
for C31H54N2O13Si2: 718.5; found by LC-MS (ES+):
[M+Na]+ 741.3.
50 - O - [benzhydryloxy - bis(trimethylsilyloxy)silyl] - 20 - O -
[bis(2-acetoxyethoxy)methyl]-3-methyluridine-30-(methyl-
N,N-diisopropyl)phosphoramidite (7). To a solution of 6
(1.00 g, 1.16 mmol) in CH2Cl2 (13 mL), methyl tetra-
isopropyl phosphoradiamidite (0.93 mL, 3.25 mmol),
and 4,5-dicyanoimidazole (137 mg, 1.20 mmol) were
added. The resulting cloudy solution was stirred over-
night, then quenched with 5% aqueous NaHCO3 and
extracted with CH2Cl2. The organic layers were com-
bined, dried over Na2SO4, and concentrated. The crude
product was purified two times by flash chromato-
graphy with 0.5% TEA in the eluents (20% acetone in
hexane followed by 5–10% CH2Cl2 in hexane) to afford
a colorless oil (0.60 g, 50%). Yields can be improved to
70–75% if the methyl tetraisopropyl phosphoramidite
20-O-[bis(2-acetoxyethoxy)methyl]-3-methyluridine (5).
CH3CN (6.5 mL), TMEDA (2.42 mL, 16.1 mmol) and
HF (48% aq stock solution, 0.36 mL, 11.3 mmol) were
added to a 50 mL round bottom flask at 0 ꢁC dropwise
via syringe. The HF/TMEDA mixture was stirred at
0 ꢁC for 10 min, then transferred to a light yellow solu-
tion of 4b (2.32 g, 3.23 mmol in 6.5 mL CH3CN) at 0 ꢁC
dropwise over 5 min by cannula. The resulting solution
was stirred at 0 ꢁC for 10 min then at room temperature
until no more of the starting material was apparent by
TLC analysis in 10% CH3OH in CH2Cl2. Upon com-
pletion of the reaction, the solvent was evaporated and
the residue was taken up in a mixture of hexane/CH2Cl2
5:1 (5 mL) containing 0.2% TMEDA. The crude pro-
duct was purified by flash chromatography on silica gel
3
1
reagent is freshly prepared.2
H NMR (CDCl3,
500 MHz) d (ppm) (mixture of diastereomers): 0.02–0.06
(36H, four s, Si(CH3)3), 1.15–1.20 (24H, four s,

332
H. M.-P. Chui et al. / Bioorg. Med. Chem. 10 (2002) 325–332
CH(CH3)3), 2.04 (6H, d, J=1.5 Hz, COCH3), 2.05 (6H,
s, COCH3), 3.295 and 3.298 (6H, two s, NCH3), 3.31
and 3.37 (6H, two d, J=12.5, 13.0 Hz, OCH3), 3.53–
3.63 (4H, m, CH(CH3)2), 3.72–4.36 (26H, m,
OCH2CH2, H20, 30, 40, 50, 500), 5.43 and 5.49 (2H, two d,
J=8.0 Hz, H5), 5.49 and 5.63 (2H, two s, OCH(Ph)2),
5.92and 5.93 (2H, two d, J=3.0, 4.0 Hz, H10), 7.18–
7.34 (20H, m, Ph), 7.77–7.80 (2H, two d, J=8.0 Hz,
H6); 31P NMR (CDCl3, 500 MHz) d (ppm) (mixture of
diastereomers): 151.89 and 151.92. Exact mass calcu-
lated for C43H72N3O16Si3P: 1025.4; found by LC-MS
(ES+) (CH3OH): [M+H]+ 1026.3; [M+Na]+ 1048.3.
Elemental anal. calcd for C43H72N3O16Si3P: C, 52.66;
H, 7.08; N, 4.10; found: C, 52.82; H, 6.97; N, 3.89.
Acknowledgements
This work is supported by the National Institutes of
Health (GM54632to C.S.C.) and (GM56044 S.A.S.).
We are grateful to J. SantaLucia for many helpful dis-
cussions and for use of the Aviv spectrometer. Technical
assistance was provided by M. Ksebati, L. Hryhorczuk,
D. Kitchen, and G. Haas.
References and Notes
1. Kowalak, J. A.; Bruenger, E.; Hashizume, T.; Peltier, J. M.;
Ofengand, J.; McCloskey, J. A. Nucleic Acids Res. 1996, 24,
688.
2. Ofengand, J.; Bakin, A. J. Mol. Biol. 1997, 266, 246.
3. Cate, J. H.; Yusupov, M. M.; Yusupova, G. Z.; Earnest,
T. N.; Noller, H. F. Science 1999, 285, 2095.
4. Merryman, C.; Moazed, D.; Daubresse, G.; Noller, H. F. J.
Mol. Biol. 1999, 285, 107.
5. Mitchell, P.; Osswald, M.; Brimacombe, R. Biochemistry
1992, 31, 3004.
6. Mueller, F.; Sommer, I.; Baranov, P.; Matadeen, R.;
Stoldt, M.; Wohnert, J.; Gorlach, M.; van Heel, M.; Brima-
combe, R. J. Mol. Biol. 2000, 298, 35.
7. Yusupov, M. M.; Yusupova, G. Z.; Baucom, A.; Lieber-
man, K.; Earnest, T. N.; Cate, J. H. D.; Noller, H. F. Science
2001, 292, 883.
Synthesis, deprotection, and purification of modified and
unmodified RNA oligonucleotides. Two RNAs contain-
ing a 3-methyluridine (m3U) modification at position
1915 and one pseudouridine-modified RNA were syn-
thesized chemically on a 1.0 mmol scale using poly-
styrene supports as described previously.8À10 The
sequences of the three RNAs are as follows with the
numbering based on the full length E. coli 23S
rRNA: 50-G1906GCCGU1911AACm3U1915AU1917AAC-
GGUC1924-30 (referred to as Um3UU; the names of the
RNAs correspond to the nucleotides at positions 1911,
1915, and 1917, respectively), 50- G1906GCCG-
É1911AACm3U1915AÉ1917AACGGUC1924-30 (Ém3UÉ;
where
É=pseudouridine),
and
50-G1906GCCG-
É1911AACU1915AÉ1917AACGGUC1924-30 (ÉUÉ). Each
crude RNA was divided into four microcentrifuge tubes
and deprotected by adding 400 mL of 100 mM NaOAc
buffer (pH 3.8) to each tube and heating at 60 ꢁC for 30
min. The deprotected RNA oligonucleotides were then
purified by gel electrophoresis on 20% denaturing
polyacrylamide gels followed by electroelution with an
Amicon centriluterTM and centricon 30s (Millipore Cor-
poration, Bedford, MA). The gel-purified RNA oligo-
nucleotides were desalted by dialysis against 3ꢃ4 L of
RNase-free deionized water for two days. The dialyzed
RNAs were then lyophilized to dryness and further
purified over C-18 Sep-PakTM (Waters, Milford, MA)
columns. Each Sep-PakTM column was prepared by
addition of 10 mL CH3CN followed by 10 mL RNase-
free deionized water. Approximately 10 OD units
(ꢀ300 mg) of RNA oligonucleotide was loaded onto
each column which was then washed with 5 mL
deionized water. The RNA eluted from the column in
three fractions with 60% CH3OH in water. The final
purified and desalted RNA oligonucleotides were dried
under reduced pressure in a Speed-Vac concentrator to
give a white solid. The single-stranded extinction coeffi-
8. Meroueh, M.; Grohar, P. J.; Qiu, J.; SantaLucia, J., Jr.;
Scaringe, S. A.; Chow, C. S. Nucleic Acids Res. 2000, 28, 2075.
9. Scaringe, S. A.; Wincott, F. E.; Caruthers, M. H. J. Am.
Chem. Soc. 1998, 120, 11820.
10. Scaringe, S. A. Methods Enzymol. 2000, 317, 3.
11. Markiewicz, W. T. J. Chem. Res. (M) 1979, 181.
12. Markiewicz, W. T. J. Chem. Res. (S) 1979, 2 4.
13. Markiewicz, W. T.; Biala, E.; Adamiak, R. W.; Grzesko-
wiak, K.; Kierzek, R.; Kraszewski, A.; Stawinski, J.; Wie-
´
wiorowski, M. Nucleic Acids Res. Sym. Ser. 1980, 7, 115.
14. Markiewicz, W. T.; Nowakowska, B.; Adrych, K. Tet.
Lett. 1988, 29, 1561.
15. Yamauchi, K.; Tanabe, T.; Kinoshita, M. J. Org. Chem.
1976, 41, 3691.
16. Reichman, U.; Hirota, K.; Chu, C. K.; Watanabe, K. A.;
Fox, J. J. J. Antibiot. 1977, 30, 129.
17. SantaLucia, J., Jr.; Kierzek, R.; Turner, D. H. Science
1992, 256, 217.
18. McDowell, J. A.; Turner, D. H. Biochemistry 1996, 35,
14077.
19. Durant, P. C.; Davis, D. R. J. Mol. Biol. 1999, 285, 115.
20. Ashraf, S. S.; Ansari, G.; Guenther, R.; Sochacka, E.;
Malkiewicz, A.; Agris, P. F. RNA 1999, 5, 503.
21. Yarian, C. S.; Basti, M. M.; Cain, R. J.; Ansari, G.;
Guenther, R.; Sochacka, E.; Czerwinska, G.; Malkiewicz, A.;
Agris, P. F. Nucleic Acids Res. 1999, 27, 3543.
22. Rife, J. P.; Cheng, C. S.; Moore, P. B.; Strobel, S. A.
Nucleic Acids Res. 1998, 26, 3640.
23. Barone, A. D.; Tang, J.-Y.; Caruthers, M. H. Nucleic
Acids Res. 1984, 12, 4051.
24. Richards, E. G. In Handbook of Biochemistry Molecular
Biology: Nucleic Acids; Fasman, G. D., Ed.; CRC Press:
Cleveland, OH, 1975, 596
cient (e) for each RNA (188,860 cmÀ1
M
À1) was cal-
culated as described by Richards.24 The extinction
coefficient for uridine (1.0ꢃ104 cmÀ1 MÀ1 at pH 7.0)
was used for pseudouridine and 3-methyluridine
because the nearest-neighbor extinction coefficients for
those modified bases are unknown, thus some error in
the RNA concentrations could be present.