CG base pair recognition by substituted phenylimidazole nucleosides

Article abstract of DOI:10.1039/b316077g

Four nonnatural imidazole nucleosides with different substituents were synthesized and studied for their binding to a CG Watson-Crick base pair by NMR spectroscopic techniques in an aprotic solvent. Concentration and temperature dependent measurements all

Full text of DOI:10.1039/b316077g

View Article Online / Journal Homepage / Table of Contents for this issue
CG base pair recognition by substituted phenylimidazole
nucleosides
a
b
b
Wei Wang, Maria G. M. Purwanto and Klaus Weisz*
a
Institut für Chemie, Freie Universität Berlin, Takustrasse 3, D-14195 Berlin, Germany
Institut für Chemie und Biochemie, Ernst-Moritz-Arndt-Universität Greifswald,
Soldmannstrasse 16, D-17487 Greifswald, Germany. E-mail: weisz@uni-greifswald.de
b
Received 11th December 2003, Accepted 23rd February 2004
First published as an Advance Article on the web 18th March 2004
Four nonnatural imidazole nucleosides with different substituents were synthesized and studied for their binding to a
CG Watson–Crick base pair by NMR spectroscopic techniques in an aprotic solvent. Concentration and temperature
dependent measurements allowed the determination of association constants, association enthalpies and entropies.
Strong binding was observed with analogues carrying an ureidophenyl substituent and corresponding enthalpies of
association are compatible with the anticipated formation of three hydrogen bonds to the CG base pair. In contrast,
only weak binding was observed for analogues with an aminophenyl or benzamidophenyl substituent. 2D NOE
measurements at low temperatures confirm the proposed binding mode for the high-affinity ligands but indicate
binding interactions for the weakly bound analogues different from the expected geometry.
Introduction
in Fig. 1, such analogues are able to form a specific hydrogen
bond between their endocyclic N3 nitrogen and the Watson–
Intermolecular triplex formation between single-stranded and
double-stranded nucleic acids has been shown to offer a power-
ful chemical approach for targeting specific sites within the
Crick cytosine amino group within
a complex closely
ϩ
isomorphous to the canonical TؒAT and C ؒGC base triads.
Correspondingly, the aminophenyl-substituted imidazole
nucleoside 4a should facilitate the formation of two hydrogen
1
genome. Possible applications of such a methodology include
the modulation of gene expression by blocking transcription
and thus repress protein production at the DNA level (antigene
bonds when binding a CG base pair, i.e., to H4 of cytosine and
b
the O6 carbonyl oxygen of guanosine. In addition to 4a, the
modified analogues 4b, 4c and 4d were synthesized to study the
effect of the benzoyl-, ureido- and n-butylureido-substituents
on the hydrogen bond mediated CG complex formation.
Binding of 4b should benefit from a more acidic amide proton
compared to the amino proton in 4a, thus constituting a more
effective hydrogen bond donor. Nucleosides 4c and 4d are
expected to form a third hydrogen bond to N7 of the guanine
base, thus increasing CG binding affinity.
2
strategy). In addition, by covalently attaching specific chemical
groups to the triplex-forming single strand, manipulations at
any given site of the DNA including strand cleavage seems to be
3
conceivable. Up to now, a major drawback of triplex-based
methodologies has arisen from the restricted recognition code.
Thus, only the purine bases adenine (A) and guanine (G),
positioned on the same strand of the duplex, are effectively
recognized by thymine (T) and cytosine (C) third strand bases
to form TؒAT and C ؒGC base triplets in the case of the
pyrimidine or parallel triplex motif. Pyrimidine interruptions
ϩ
4
within an all-purine tract of the double helix generally result in
triple helices of low stability due to the lack of their effective
recognition. Therefore, the design and synthesis of nonnatural
nucleoside analogues to extend the recognition to all four
possible base pairs is of major importance for any strategy that
is based on triplex formation.
Recently, much attention has been paid to the development
5
of artificial nucleobases as ligands for any given base pair.
However, due to the complex interplay of the various
interactions within a triple helical complex, predictions as for
their affinity and specificity were hardly successful. Here we
report on differently substituted phenylimidazole nucleosides
as receptors for a CG base pair. A benzamido-substituted
phenylimidazole has recently been incorporated into triplex-
Fig. 1 Proposed binding mode of substituted 4-phenylimidazole
nucleosides to a CG Watson–Crick base pair.
The synthesis of 2-unsubstituted imidazoles starting from
6
forming oligonucleotides by Dervan et al., but has been shown
an α-hydroxy or α-halogenated ketone 1 and formamide is
7
8
to bind via intercalation, contrary to what was expected. In this
based on the Bredereck imidazole synthesis recently also
6a
paper we report on a detailed thermodynamic and structural
characterization of CG binding by various novel phenyl-
imidazole nucleosides in apolar solvents in order to shed more
light on the structure–affinity relationship that is based on the
formation of specific hydrogen bonds.
employed by Griffin et al. A Vorbrüggen-type reaction of the
resulting 4-(3-nitrophenyl)-imidazole 2 with 1-O-acetyl 2,3,5-
tri-O-benzoyl-β--ribofuranose afforded the corresponding
β-nucleoside 3. The subsequent reduction of the nitro group
with Raney nickel and hydrazine as hydrogen source gave
the amino-substituted 4a (Fig. 2). Benzoylation of 4a under
standard conditions at room temperature afforded 4b.
Analogues 4c and 4d were prepared from 4a using phenyl
carbamate and n-butylisocyanate as reagents, respectively
(Fig. 3).
Results and discussion
According to molecular modeling studies, imidazole nucleo-
sides may serve as potential CG recognizing ligands. As shown
1
194
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 9 4 – 1 1 9 8
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4

View Article Online
Fig. 2 Scheme for the synthesis of 4a.
Fig. 3 Synthesis of 4b, 4c and 4d starting from 4a: (i) BzCl, pyridine,
THF, room temperature, 80%; (ii) phenyl carbamate, DMF, 60 ЊC,
2%; (iii) C H NCO, THF, 77%; RЈ = 2Ј,3Ј,5Ј-tri-O-benzoyl-β--
6
4
9
Fig. 4 NHa proton chemical shift of 4d as a function of CG
concentration at 292 K (solid squares), 268 K (open circles), 255 K
solid circles) and 245 K (open squares). Lines represent the least-
ribofuranosyl.
(
NMR titration experiments
squares fit.
The use of O-silylated 2Ј-deoxycytidine and 2Ј-deoxyguanosine
as well as O-benzoylated nucleoside analogues 4a–d not only
blocks the sugar hydroxyl groups but also greatly enhances
solubility in the methylene chloride solvent. Initial experiments
on the homoassociation of the four analogues were performed
Table 1 Summary of Kass, ∆H and ∆S for the association of nucleo-
side analogues 4a–d with a CG base pair
Ϫ1
Ϫ1
Ϫ1
Ϫ1
T /K
Kass/M
∆H/kJ mol
Ϫ8.7 ± 2
∆S/J K mol
by following the NH proton chemical shift of nucleosides 4a–d
a
4a
4b
242
15 ± 2
23 ± 3
28 ± 3
7 ± 2
12 ± 2
24 ± 2
28 ± 2
173 ± 25
289 ± 18
459 ± 30
688 ± 55
131 ± 6
316 ± 20
418 ± 40
627 ± 41
Ϫ13.5 ± 4
Ϫ26.5 ± 6
as a function of nucleoside concentration ranging from 5 to 100
mM. Homoassociation constants were determined by a non-
linear least-squares fit of the experimental data and amount
2
19
212
268
245
222
Ϫ11.5 ± 2
Ϫ24.5 ± 2
Ϫ19.4 ± 2
Ϫ1
Ϫ1
Ϫ1
to 0.8 M for 4a (290 K), 0.3 M for 4b (290 K), 18 M for 4c
(
Ϫ1
292 K) and 6 M for 4d (300 K). Disregarding some uncer-
2
10
291
79
tainties in the determination of small association constants
within the concentration range employed, homoassociation
only occurs to a small extent for all nucleosides and thus
justifies its initial neglect in the following analysis of CG com-
plexation. Also, as has been shown before, the strong
4
c
Ϫ40.8 ± 6
Ϫ25.4 ± 6
2
268
256
292
268
4
d
5
Ϫ1
binding between C and G in apolar solvents (Kass ∼ 10 M )
2
2
55
45
allows a 1 : 1 mixture of the two nucleosides to be treated as a
9
single CG species.
Binding of the nucleoside analogues to the CG base pair was
1
followed by changes in the H NMR chemical shift of the NHa
nition through hydrogen bonding at ambient temperatures.
This may be attributed to the weak amino proton donor in 4a
and steric hindrance by the bulky benzamido group in 4b (vide
infra).
proton of 4a–d. Thus, titration of the analogue (2 mM) with a
: 1 mixture of C and G (0 to 20 mM) resulted in a downfield
shift of the H signal with increasing CG concentration as
shown in Fig. 4 for the analogue 4d at four different temper-
atures. The equilibrium constant for complex formation Kass
between each of the nucleoside analogues and the CG base pair
was determined by fitting the chemical shift data as a function
of concentration to a nonlinear 1 : 1 binding isotherm. The
results of measurements at different temperatures are summar-
ized in Table 1.
1
a
Ϫ1
In contrast, association constants of 173 M (291 K) for 4c
Ϫ1
and 131 M (292 K) for 4d compare favorably with an
Ϫ1
association constant of 70 M reported for the canonical AU
base pair at ambient temperatures and indicate the formation
of three hydrogen bonds to the CG base pair. In line with
9
previous binding studies using other analogues, the ureido
group seems to be a major contributor to the strong binding of
the free nucleosides in organic solvents, however, N-alkylation
as in 4d slightly reduces the binding constant over the tem-
perature range studied.
Ϫ1
Clearly, with association constants of 15 M (242 K) and 7
(268 K) CG binding of 4a and 4b is only weak and does
not seem to allow for a specific high-affinity base pair recog-
Ϫ1
M
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 9 4 – 1 1 9 8
1195

View Article Online
We have also performed reverse titrations by adding the
analogue (0–20 mM) to a 2 mM solution of the CG base pair at
2
86 K. Based on the proposed binding mode, the non-Watson–
Crick bound cytosine amino proton H is in all cases expected
b
to undergo a downfield shift upon titration due to the form-
ation of a hydrogen bond to the imidazole N-3. However, the
observation of cytosine amino signals is severely hampered by
the overlap of resonances, especially with the nucleoside
analogue in excess. We have therefore employed specifically
15
4
- N labeled cytidine, enabling the use of a one-dimensional
15
1
H– N HMQC experiment as a filter that effectively eliminates
all other proton resonances. The downfield shifts observed in
such experiments for the C H amino resonance when adding
b
4
a–d to the CG mixture were again fitted to a 1 : 1 binding
isotherm (Fig. 5). Association constants Kass for the complex-
ation of nucleoside analogues with the CG base pair were found
to agree within the estimated margin of error with the associ-
ation constants obtained from the former titrations (data not
shown). Only for 4c, having the largest self-association constant
of all analogues, a more noticeable deviation of about 40% in
the heteroassociation constant at 286 K is observed for the
two independent titrations and may be attributed to a non-
negligible degree of self-association.
Fig. 6 Van’t Hoff plot of the association constants Kass for 4aؒCG
(
(
solid circles), 4bؒCG (open circles), 4cؒCG (solid squares) and 4dؒCG
open squares) base triple formation in CD Cl . Lines represent the
2
2
least-squares fit.
the enthalpy values not only reflect the formation of hydrogen
bonds but also include the enthalpy of solvation, i.e., the entire
enthalpy difference between solvated complex and solvated
monomers. However, assuming only small differences between
the enthalpy of solvation for the dimer and monomers, Ϫ∆H
should reasonably well reflect the hydrogen bond energy.
Ϫ1
Consequently, values found between Ϫ8 and Ϫ12 kJ mol for
4
aؒCG and 4bؒCG base triple formation point to two very weak
hydrogen bond contacts or, alternatively, to only one hydrogen
bond interaction between the analogue and the base pair. Much
larger enthalpic contributions of Ϫ24 kJ mol and Ϫ19 kJ
Ϫ1
Ϫ1
mol are determined for 4cؒCG and 4dؒCG, respectively.
Assuming the formation of three hydrogen bonds in these
Ϫ1
cases, the heat of formation averages to about 7 kJ mol per
hydrogen bond. Although smaller than expected for a single
hydrogen bond of medium strength, these values seem
reasonable given the non-cooperative nature of the contacts
and a weakening of individual interactions upon introducing
additional interaction sites without perfect complementarity.
As expected, entropies decrease on complex formation,
Fig. 5 Representative plot of the cytosine H amino proton chemical
b
Ϫ1
Ϫ1
Ϫ1
Ϫ1
shift in a CG base pair as a function of the 4d nucleoside concentration
at 286 K. The solid line represents the least-squares fit with limiting
chemical shifts indicated by the dotted lines.
ranging from Ϫ13 J K mol for 4aؒCG to Ϫ41 J K mol
for 4cؒCG. Because a loss in entropy can be partly associated
with a restricted conformational flexibility upon CG com-
plexation, the rigid analogue 4a with a non-acylated amino
substituent is expected to lose less degrees of freedom on
binding compared to the other analogues. Also, 4d with its
n-butyl substituent might strongly favor a Z-conformation thus
restricting amide C–N bond rotations even in the uncomplexed
monomer and reducing the entropic penalty for the association
when compared to 4c. It must be cautioned, however, that the
entropies determined from the temperature dependence of
association constants are subject to a relatively large degree of
uncertainty. Also, the interpretation of experimental entropies
of association at a molecular level is not straightforward
and should necessarily also include effects from solvation and
desolvation which may be significant.
The limiting chemical shift at c4a–d
∞ for the cytosine
amino proton H as also determined from the curve fitting in
b
these reverse titrations (see Fig. 5) corresponds to its shift in a
trimolecular complex. Because the chemical environment of
this proton is expected to be similar in all complexes, it can be
used as a direct measure of the local strength and extent of
hydrogen bonding in the various associates. Upon titrating 4d
and 4c to the CG base pair, limiting chemical shifts of 6.8 ppm
and 6.5 ppm for C H in 4dؒCG and 4cؒCG were determined.
b
In contrast, smaller downfield shifts upon complexation with
limiting chemical shifts of 6.1 ppm and 5.8 ppm for the amino
proton signal in 4bؒCG and 4aؒCG may point to the coexistence
of additional binding modes with altered interaction sites for
the latter two complexes. Unlike the non-Watson–Crick bound
C H amino proton, the Watson–Crick bound H proton signal
2D NOE measurements
b
a
experienced only a minor shift upon adding 4a–d, indicating
that the CG base pair is not disrupted upon binding of the
analogue.
For a more detailed structural characterization of the com-
plexes, 2D NOE experiments on 1 : 1 mixtures of the non-
natural nucleosides and the CG base pair were performed at
2
4
13 K in methylene chloride. A portion of the spectrum for the
dؒCG base triple is shown in Fig. 7. Due to the low temper-
Enthalpy and entropy
In order to analyze the complexation data in terms of enthalpic
and entropic contributions, the temperature dependent associ-
ation constants were subjected to a van’t Hoff analysis. A plot
atures employed for the measurements, equilibria are largely
shifted toward the intact base triad. Again, typical contacts
observed between guanine and cytosine imino and amino
protons for Watson–Crick base pairing indicate that the base
pair is not disrupted in the presence of the analogue. Most
notably, however, the spectrum displays several intermolecular
NOE connectivities between base protons of 4d and the CG
Ϫ1
of lnKass as a function of T is shown in Fig. 6 for the various
complexes and corresponding enthalpies and entropies of
association that have been determined by fitting the experi-
mental data are summarized in Table 1. It has to be noted that
1
196
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 9 4 – 1 1 9 8

View Article Online
information on the thermodynamics of hydrogen bond
mediated association, but also on the binding mode, pre-
requisite for establishing useful structure–affinity relationships.
Our data indicate that, in contrast to the high-affinity binding
of analogues carrying an ureidophenyl group, analogues with
an unsubstituted or benzoylated aminophenyl substituent only
exhibit poor affinity and specificity toward a CG base pair. This
can be attributed to the amino group being a weak hydrogen
bond donor and the benzamido group giving rise to steric
clashes with the guanine purine ring. Within a triple helical
complex the benzamido-modified analogue was previously
shown to bind through preferential intercalation rather than
through the formation of specific hydrogen bonds. Clearly,
more studies are needed to gain a better understanding of the
contribution of additional non-hydrogen bond interactions in
the base pair recognition of nucleic acids. However, although
eliminating additional interactions in macromolecular systems,
the association of free nucleosides may provide useful inform-
ation on the binding of ligands within a biological context.
Experimental
General methods
NMR experiments were performed on a Bruker AMX500
1
spectrometer. H chemical shifts in methylene chloride were
referenced relative to CHDCl (δH = 5.32 ppm). Temperatures
2
were calibrated with a standard solution of 4% CH OH in
3
CD OD and adjusted by a Eurotherm Variable Temperature
3
Unit to an accuracy of ±1.0 ЊC. Concentration dependent
chemical shifts were fitted with an appropriate equation by
employing the Marquardt–Levenberg algorithm.
Fig. 7 Portion of a 2D NOE spectrum of a 1 : 1 mixture of 4d
and a CG base pair in CD Cl at 213 K (mixing time = 200 ms).
Intermolecular crosspeaks between 4d and the base pair are marked by
arrows and indicated in the structure at the top; R = 3Ј,5Ј-di-O-
Materials
2
2
Reagents of the highest quality available were purchased
from Sigma-Aldrich, Deisenhofen, Germany. [4- N]-2Ј-deoxy-
cytidine was prepared from unlabeled 2Ј-deoxyuridine as
described previously. Free cytidine and guanosine were
O-silylated using triisopropylsilyl chloride and subsequently
purified by HPLC. All reactions were controlled by TLC on
silica gel plates (Merck silica gel 60 F254). If necessary, solvents
were dried by standard procedures prior to use.
15
(
triisopropylsilyl)-β--2Ј-deoxyribofuranosyl, RЈ
=
2Ј,3Ј,5Ј-tri-O-
benzoyl-β--ribofuranosyl.
10
base pair. Thus, as seen in Fig. 7, intermolecular NOE cross-
peaks between cytosine H5 and the 4d AH2 proton as well as
between the Watson–Crick bound cytosine amino proton H4a
and 4d AH2 and BH2 protons are compatible with the
proposed binding mode of the nucleoside analogue to the CG
base pair. In addition, a crosspeak between H8 of guanine
and the α-methylene protons of the n-butyl residue of 4d
corroborates the formation of a third hydrogen bond to N7
of guanosine. Likewise, corresponding NOE contacts C NHa–
1-(2Ј,3Ј,5Ј-Tri-O-benzoyl-ꢀ-D-ribofuranosyl)-4-(3-nitro-
phenyl)imidazole 3. A mixture of 4-(3-nitrophenyl)imidazole
6a
2
(4.7 g, 24.84 mmol), 1-O-acetyl-2,3,5-tri-O-benzoyl-β--
ribofuranose (12.75 g, 24.78 mmol) and potassium nonaflate
(21.13 g, 62.38 mmol) were co-distilled to dryness three times
each with 30 mL absolute acetonitrile (freshly distilled) and
finally dissolved in 250 mL absolute acetonitrile. After addition
of hexamethyldisilazane (HMDS, 1.71 mL, 8.10 mmol) and
trimethylchlorosilane (TCS, 6.4 mL, 49.56 mmol) under an
inert atmosphere the mixture was stirred for 24 hours under
reflux. Within the first 4 hours TCS (4 mL total) and HMDS
(1 mL total) were added in regular time intervals. After addition
of 150 mL acetonitrile and 200 mL methylene chloride the
4
c AH2, C H5–4c AH2 and C NH –4c BH2 are also in line with
a
the corresponding binding mode of 4c. Expected inter-
molecular crosspeaks for the non-Watson–Crick bound C H4b
amino proton were not observed possibly due to significant
broadening of the signal at this temperature.
In contrast to 4c and 4d, we only found a single weak inter-
molecular NOE contact to the CG base pair for both 4a and 4b,
namely, a contact between BH2 of 4a or 4b and H5 of cytosine
(
not shown). Without any additional intermolecular crosspeak,
the NOE data do not allow an unambiguous characterization
of the 4aؒCG and 4bؒCG geometries again reflecting the weak
binding of the two analogues to the base pair. However, the
presence of a C H5-4a/b BH2 contact suggests the formation of
complexes with binding modes different from what is expected.
In fact, molecular models of a tentative 4bؒCG base triplet
indicate that it does not allow coplanarity of the bases due
to a steric clash between the benzamido group of 4b and the
guanine purine ring.
mixture was extracted with a saturated NaHCO solution. The
3
organic phase was washed with a saturated NaCl solution,
dried over Na SO and the solvent was removed under vacuum.
2
4
Extraction of the solid in hexane/AcOEt 1 : 1 and removal of
the solvent afforded a crude product, which was purified by
column chromatography (SiO , CH Cl /MeOH 100 : 0 to
2
2
2
1
95 : 5); yield: 1.6 g (10%); H NMR (250 MHz, 293 K, DMSO):
δ (ppm) = 4.71–4.90 (3H, m, H4Ј, H5Ј, H5Љ), 5.94 (2H, m, H2Ј,
H3Ј), 6.45 (1H, d, H1Ј), 7.41–8.17 (19H, m, ArH), 8.26 (1H, m,
H4 Phenyl), 8.52 (1H, s, H2 Phenyl).
Conclusion
1
-(2Ј,3Ј,5Ј-Tri-O-benzoyl-ꢀ-D-ribofuranosyl)-4-(3-amino-
phenyl)imidazole 4a. To a solution of 1-(2Ј,3Ј,5Ј-tri-O-benzoyl-
β--ribofuranosyl)-4-(3-nitrophenyl)imidazole (1.60 g,
The detailed NMR spectroscopic analysis of CG binding with
differently substituted imidazole nucleosides not only gives
3
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 9 4 – 1 1 9 8
1197

View Article Online
2
.5 mmol) in anhydrous methylene chloride, Raney nickel
removed under vacuum and the crude product was purified
(
3.6 g) and hydrazine (11 mL, 0.35 mmol) were added. After the
using column chromatography and HPLC (SiO ; CH Cl /
MeOH 97 : 3); yield: 270 mg, (77%); H NMR (250 MHz, 293
2
2
2
1
suspension was stirred for 24 hours, the catalyst was filtered off
and the solvent was removed. The product was purified by
column chromatography (SiO , CH Cl /MeOH 97.5 : 2.5);
yield: 0.56 g (37%); H NMR (250 MHz, 293 K, CDCl ):
δ (ppm) = 3.61 (2H, s, NH ), 4.55–4.96 (3H, m, H4Ј, H5Ј, H5Љ),
K, CDCl ): δ (ppm) = 0.87 (3 H, t, CH ), 1.22–1.50 (4 H, m,
3
3
CH CH ), 3.22 (2 H, q, N–CH ), 4.54–4.79 (3 H, m, H4Ј,
2
2
2
2
2
2
1
H5Ј, H5Љ), 5.58 (1 H, t, NH), 5.84 (2 H, m, H2Ј, H3Ј), 6.1
(1 H, d, H1Ј), 7.08–8.10 (22 H, m, ArH and NH); HRMS
(EI, 80 eV 150 ЊC): m/z = 702.26745; calculated: 702.268965
3
2
5
.92 (2H, m, H2Ј, H3Ј), 6.14 (1H, d, H1Ј), 6.51–8.21 (21H, m,
ϩ
ArH). (Found: C, 69.4; H, 5.1; N, 6.6. Calc. for C H N O : C,
[C H N O ] .
35
29
3
7
40 38
4
8
6
9.65; H, 4.8; N, 7.0%).
Acknowledgements
Synthesis of 1-(2Ј,3Ј,5Ј-tri-O-benzoyl-ꢀ-D-ribofuranosyl)-4-(3-
We thank the Fonds der Chemischen Industrie for financial
benzamidophenyl)imidazole 4b. 320 mg 4a was co-distilled to
dryness three times each with 15 mL absolute THF (freshly
distilled) and then dissolved in 60 mL absolute THF. 0.24 mL
pyridine and 0.2 mL (0.172 mmol) freshly distilled benzoyl
chloride were added under argon and the reaction mixture was
stirred at room temperature for 3 days. After completion of
the reaction, the solvent was removed under vacuum and the
product was purified by column chromatography (SiO2;
CH Cl /MeOH 98 : 2) and HPLC (SiO ; CH Cl /MeOH 98.5 :
support.
References
1
D. Praseuth, A. L. Guieysse and C. Hélène, Biochim. Biophys. Acta,
999, 1489, 181–206.
(a) L. J. Maher III, B. Wold and P. B. Dervan, Science, 1989, 245,
725–730; (b) J. C. Hanvey, M. Shimizu and R. D. Wells, Nucleic
Acids Res., 1990, 18, 157–161.
1
2
2
2
2
2
2
1
3
4
5
S. A. Strobel, L. A. Doucette-Stamm, L. Riba, D. E. Housman and
P. B. Dervan, Science, 1991, 254, 1639–1642.
1
.5); yield: 300 mg (80%); H NMR (250 MHz, 293 K, CDCl ):
3
δ (ppm) = 4.56–4.81 (3 H, m, H4Ј, H5Ј, H5Љ), 5.89 (2 H, m, H2Ј,
H3Ј), 6.1 (1 H, d, H1Ј), 7.15–8.35 (26 H, m, ArH); HRMS (EI,
N. T. Thuong and C. Hélène, Angew. Chem., Int. Ed. Engl., 1993, 32,
6
66–690.
8
0 eV 150 ЊC): m/z = 707.22583; calculated: 707.226765
ϩ
42 33 3 8
(a) C.-Y. Huang and P. S. Miller, J. Am. Chem. Soc., 1993, 115,
10456–10457; (b) S. Sasaki, S. Nakashima, F. Nagatsugi, Y. Tanaka,
M. Hisatome and M. Maeda, Tetrahedron Lett., 1995, 36, 9521–
[C H N O ] .
9
524; (c) T. E. Lehmann, W. A. Greenberg, D. A. Liberles,
C. K. Wada and P. B. Dervan, Helv. Chim. Acta, 1997, 80, 2002–
022; (d ) F. Lecubin, R. Benhida, J.-L. Fourrey and J.-S. Sun,
Synthesis of 1-(2Ј,3Ј,5Ј-tri-O-benzoyl-ꢀ-D-ribofuranosyl)-4-(3-
ureidophenyl)imidazole 4c. 0.1 g 4a was dissolved in anhydrous
DMF (0.5 mL) at 60 ЊC. To this solution was added 0.24 g (1.76
mmol) phenyl carbamate and the reaction mixture was stirred
at 60 ЊC until TLC indicated that the educt had been consumed
2
Tetrahedron Lett., 1999, 40, 8085–8088; (e) D. Guianvarc’h,
R. Benhida, J.-L. Fourrey, R. Maurisse and J.-S. Sun, Chem.
Commun., 2001, 1814–1815; ( f ) for reviews see S. O. Doronina and
J.-P. Behr, Chem. Soc. Rev., 1997, 63–71; (g) D. M. Gowers and
K. R. Fox, Nucleic Acids Res., 1999, 27, 1569–1577; (h) M. G. M.
Purwanto and K. Weisz, Curr. Org. Chem., 2003, 7, 427–446.
6 (a) L. C. Griffin, L. L. Kiessling, P. A. Beal, P. Gillespie and
P. B. Dervan, J. Am. Chem. Soc., 1992, 114, 7976–7982; (b)
L. L. Kiessling, L. C. Griffin and P. B. Dervan, Biochemistry, 1992,
31, 2829–2834.
(
72 h). The mixture was evaporated to dryness and then purified
by column chromatography (SiO ; CH Cl /MeOH 95 : 5); yield:
2
2
2
1
1
4
50 mg (62%); H NMR (250 MHz, 293 K, CDCl ): δ (ppm) =
3
.56–4.82 (3 H, m, H4Ј, H5Ј, H5Љ), 5.65 (2 H, s, NH ), 5.90 (2 H,
2
m, H2Ј, H3Ј), 6.05 (1 H, d, H1Ј), 7.02–8.08 (22 H, m, ArH and
NH); HRMS (EI, 80 eV 150 ЊC): m/z = 629.17657; calculated:
ϩ
7 K. M. Koshlap, P. Gillespie, P. B. Dervan and J. Feigon, J. Am.
Chem. Soc., 1993, 115, 7908–7909.
6
29.179815 [C H N O –NH ] .
36
30
4
8
3
8
9
(a) H. Bredereck and G. Theilig, Chem. Ber., 1953, 86, 88–96; (b)
G. Theilig, Chem. Ber., 1953, 86, 96–109; (c) M. R. Grimmet, Adv.
Heterocycl. Chem., 1970, 12, 103–183.
Synthesis of 1-(2Ј,3Ј,5Ј-tri-O-benzoyl-ꢀ-D-ribofuranosyl)-4-(3-
n-butylureidophenyl)imidazole 4d. 300 mg 4a was co-distilled to
dryness three times each with 15 mL absolute THF and then
dissolved in 40 mL absolute THF. After addition of 0.15 mL
1.38 mmol) n-butylisocyanate under argon, the mixture was
refluxed for 1.5 h and stirred for an additional 24 h at room
(a) S. C. Zimmerman and P. Schmitt, J. Am. Chem. Soc., 1995, 117,
1
0769–10770; (b) E. Mertz, S. Mattei and S. C. Zimmerman, Org.
Lett., 2000, 2, 2931–2934; (c) D. Lengeler and K. Weisz, Tetrahedron
Lett., 2001, 42, 1479–1481.
(
1
0 G. Kupferschmitt, J. Schmidt, T. Schmidt, B. Fera, F. Buck and
temperature. After completion of the reaction, the solvent was
H. Rüterjans, Nucleic Acids Res., 1987, 15, 6225–6241.
1
198
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 1 9 4 – 1 1 9 8