The supramolecular helical architecture of 8-oxoinosine and 8-oxoguanosine derivatives

Article abstract of DOI:10.1002/chem.200601126

The 8-oxoguanosine derivative 1 and the 8-oxoinosine derivative 2b, with appropriate substituents on their ribose moieties, form hexagonal lyotropic mesophases in hydrocarbon solvents. Small-angle X-ray scattering analysis of a film of 1 and of the mesophase of 2b, and NMR and CD spectra of isotropic solutions of 2b, indicate that in both cases the supramolecular structures adopted are continuous helices formed by a hydrogen-bond network between the heterocyclic bases. Notably, while derivative 2b, which bears large substituents on its ribose moiety, undergoes self-assembly and mesophase formation, oxoinosine 2a, with only decanoyl groups on its ribose moiety, does not. This may be ascribed to the reduced amphiphilic properties of the latter and the absence of aromatic groups.

Full text of DOI:10.1002/chem.200601126

FULL PAPER
DOI: 10.1002/chem.200601126
The Supramolecular Helical Architecture of 8-Oxoinosine and
8-Oxoguanosine Derivatives
Stefano Lena,[a] Mauro A. Cremonini,[b] Francesco Federiconi,[c] Giovanni Gottarelli,*[a]
Carla Graziano,[a] Luca Laghi,[b] Paolo Mariani,[c] Stefano Masiero,[a]
Silvia Pieraccini,[a] and Gian Piero Spada*[a]
Abstract: The 8-oxoguanosine deriva-
tive 1 and the 8-oxoinosine derivative
2b, with appropriate substituents on
their ribose moieties, form hexagonal
lyotropic mesophases in hydrocarbon
solvents.Small-angle X-ray scattering
analysis of a film of 1 and of the meso-
phase of 2b, and NMR and CD spectra
of isotropic solutions of 2b, indicate
that in both cases the supramolecular
structures adopted are continuous heli-
ces formed by a hydrogen-bond net-
work between the heterocyclic bases.
Notably, while derivative 2b, which
bears large substituents on its ribose
moiety, undergoes self-assembly and
mesophase formation, oxoinosine 2a,
with only decanoyl groups on its ribose
moiety, does not.This may be ascribed
to the reduced amphiphilic properties
of the latter and the absence of aro-
matic groups.
Keywords: liquid crystals · oxogua-
nosine · oxoinosine · self-assembly ·
supramolecular chemistry
Introduction
bases, guanine (G) appears to be the most versatile as, be-
sides forming the standard GC pair, it is able to self-assem-
ble to give different supramolecular architectures.[4] In par-
ticular, depending on the substituents present and the exper-
imental conditions, lipophilic G-derivatives can generate
quartets,[5] ribbonlike structures,[6] and, in the presence of
metal templates, G-quartet-based aggregates (Figure 1).
Both columnar structures and ribbons can form liquid-
crystalline gels with different topologies.Furthermore, col-
umnar structures have recently been proposed as possible
ion channels,[7] whereas ribbons on surfaces show interesting
electrical properties.[8]
The construction of complex structures based on weak inter-
actions is one of the major goals of supramolecular chemis-
try;[1] on the other hand, helical structures occur widely in
Nature and are responsible for some fundamental biological
functions.It is therefore natural that several research groups
have focused their studies on the building of supramolecular
helices.[2]
The use of nucleobases to build up complex supramolec-
ular structures is being increasingly investigated.[3] Of these
In a recent paper, we reported that 8-oxoguanosine (8-
oxoG; 1), bearing lipophilic substituents on the sugar
moiety, forms liquid-crystalline gels in hydrocarbon sol-
vents.[9] 8-Oxoguanosine possesses a rich array of hydrogen-
bond donor and acceptor sites, and identifying their possible
interactions in the self-assembly process constitutes a diffi-
[a] S.Lena, Prof.G.Gottarelli, C.Graziano, Dr.S.Masiero,
Dr.S.Pieraccini, Prof.G.P.Spada
Alma Mater Studiorum–Università di Bologna
Dipartimento di Chimica Organica “A.Mangini”
Via San Giacomo 11, 40126 Bologna (Italy)
Fax : (+39)051-209-5688
E-mail: gottarel@alma.unibo.it
GianPiero.Spada@unibo.it
[b] Dr.M.A.Cremonini, Dr.L.Laghi
Alma Mater Studiorum - Università di Bologna
Dipartimento di Scienze degli Alimenti
Piazza Goidanich 60, 47023 Cesena (Italy)
[c] Dr.F.Federiconi, Prof.P.Mariani
Università Politecnica delle Marche
Dipartimento di Scienze Applicate ai Sistemi Complessi
Via Ranieri 65, 60131 Ancona (Italy)
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
Chem. Eur. J. 2007, 13, 3441 – 3449
ꢀ 2007 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
3441

Figure 2.Donor/acceptor groups of the 8-oxoguanine moiety (a) and the
helical supramolecular architecture (b).
ple, 8-oxoI can self-assemble to give a ribbon, a quartet, or
a continuous helix (Figure 3).
Herein, we present supporting evidence for the helical
structure assignment by characterizing the solid-state film
obtained from derivative 1 by means of small-angle X-ray
scattering (SAXS), and we describe the self-assembly of
lipophilic 8-oxoI derivatives 2a and 2b, and a liquid-crystal-
G
line gel formed by 2b in hydrocarbon solvents.Notably,
while derivative 2b, bearing large substituents on its ribose
moiety, undergoes self-assembly and mesophase formation,
Figure 1.Donor/acceptor sites of the guanine moiety (a), the G-quartet
arrangement (b), the columnar G-quartet ion-directed architecture (c),
and two different ribbon-like assemblies (d, e) in the absence of cations.
cult task.However, we have obtained strong evidence that
the exocyclic amino group is not directly involved in the
self-assembly process.[9] Even so, several self-assembled
structures can be hypothesized: a ribbon, a cyclic quartet,
and a continuous helix.Of the aforementioned structures,
we identified the helical architecture as the most populated
one for 8-oxoG 1, both in solution and in the liquid-crystal-
line phases (Figure 2).
8-Oxoinosine derivatives (8-oxoI; 2) do not possess the
exocyclic amino group and should thus be suitable for
checking whether the supramolecular structure assigned to
8-oxoG is correct, even though such subtle modifications
may clearly influence the self-assembly processes.In princi-
Figure 3.Donor/acceptor groups of the 8-oxoI moiety (a) and three possi-
ble supramolecular architectures: a ribbon (b), a cyclic quartet (c), and a
continuous helix (d).
3442
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2007, 13, 3441 – 3449

Helical Self-Assembly of an 8-Oxoinosine
FULL PAPER
8-oxoinosine 2a, with only decanoyl groups on its ribose
moiety, does not.This can be attributed to the reduced am-
phiphilic properties of the latter.
distance of about 4.5 Š), which indicates the disordered con-
formation of the lipophilic substituents.
From the unit cell, the cross-sectional radius of the helix
can be determined by assuming that the columns have a cir-
cular section with radius R and are infinite in length.The re-
lationship between the cross-sectional area of the cylinder
and the two-dimensional hexagonal unit cell surface is given
by Equation (2):[9–11]
Results and Discussion
Small-angle X-ray scattering of 8-oxoG 1: As it proved im-
possible to obtain X-ray quality crystals of the series of 8-
oxoguanosine derivatives, we studied the SAXS of the film
obtained by spin-drying of a solution of derivative 1 in
chloroform.
p
pR2 ¼ ð 3=2Þa2cv,G
ð2Þ
The X-ray diffraction pattern is shown in Figure 4.Ac-
cording to our previous work,[9] the data are consistent with
where cv,G is the volume concentration of the 8-oxoguanine
base inside the unit cell volume.In this special case, in
which the solvent is absent, cv,G corresponds to the volume
fraction of the 8-oxoguanine residue (VG = 372 Š3) with re-
spect to the molecular volume (Vmol = 1182 Š3) calculated
from standard atomic dimensions, that is cv = 0.327, giving
a radius of 9.9 Š.
The structural analysis was further developed by consider-
ing the intensities of the Bragg peaks observed in the low-
angle diffraction region.The two-dimensional electron-den-
sity profile can thus be obtained by calculating the two-di-
mensional Fourier distribution [Eq.(3)]: [11,12]
X
1ðx,yÞ ¼
h,kðꢁFh,kcosðQðxÞh,kxÞcosðQðxÞh,kyÞÞ
ð3Þ
Figure 4.X-ray diffraction profile obtained from a dry film of 1.
where Fh,k is the Fourier coefficient of the peak at position
Qh,k, and is related to the peak intensity by Fh,k (Ih,k/mh,k)1/2,
where mh,k is the multiplicity of the h,k reflection.The phase
information for each diffraction order is either positive or
negative for a centrosymmetric scattering length density
profile, such as that pertaining to the hexagonal phase.The
correct sign of the Fourier coefficients Fh,k has been ob-
tained by considering a geometrical model of the helix.[12]
For the sake of simplicity, this model was assumed to be cy-
lindrically symmetric.The unit cell was split into four re-
gions: the central cavity, the 8-oxoguanine region, the sugar
residue region, and the hydrocarbon region in which the lip-
ophilic substituents (and possibly also the hydrocarbon sol-
vent) are located, as shown in Figure 5.The model was ob-
tained by adding uniform disks of appropriate density.The
Fourier transform of this 2D model is given by the continu-
ous function F(Q) [Eq.(4)]: [12]
the presence of a hexagonal columnar phase.In particular,
the low-angle diffraction region is characterized by a series
of at least seven narrow peaks with spacing ratios in the
p
p
p p
p
p
order 1: 3: 4: 7: 9: 12: 13, which index according to a
two-dimensional hexagonal lattice of p6m symmetry.[10]
From the Bragg spacings Qh,k, the unit cell dimension a
(that is, the inter-axial distance between the aggregates) has
been obtained from Equation (1):[10]
p
p
a ¼ ð4p= 3Þ ðh2 þ k2ꢀhkÞ=Qh,k
ð1Þ
where h and k are the Miller indices of the observed Bragg
reflections.In full agreement with the concentration de-
pendence observed in hexane,[9] and as the film is dry, a
rather small unit cell dimension of 32.9 Š was derived.
Experimental evidence for the columnar nature of the
2
FðQÞ ¼ 2ðR1 ð11ꢀ12ÞJ1ðQR1Þ=QR1
phase has been obtained by analysis of the high-angle dif-
[9,10]
2
fraction region.In accordance with previous research,
narrow band is observed at about Q
a
þR2 ð12ꢀ13ÞJ1ðQR2Þ=QR2
þR3 ð13ꢀ10ÞJ1ðQR3Þ=QR3Þ=ðR1 ð1ꢀ12Þ
þR2 ð12ꢀ13Þ þ R3 ð13ꢀ10ÞÞ
ꢀ1
ð4Þ
= 1.81 Š (see
2
2
Figure 4).This band is related to the nature of the order
inside the structure elements, which are helical columns
composed of 8-oxoG residues stacked almost perpendicular-
ly to the column axis.From the position of this peak, a
repeat distance between neighboring 8-oxoguanosines of
3.45 Š is obtained.
2
2
where J1 is the first-order Bessel function of the first kind;
11, 12, and 13 are the electron densities of the central cavity,
the 8-oxoguanine, and the sugar residue, respectively, and 10
is the average electron density of the hydrocarbon region.
Moreover, R1 is the radius of the central hole, R2 is the
The high-angle region is also characterized by a large
ꢀ1
band, centered at Q = 1.4 Š (corresponding to a repeat
Chem. Eur. J. 2007, 13, 3441 – 3449
ꢀ 2007 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
3443

G.Gottarelli, G.P.Spada et al.
ence of the hole in the central region of the helix, the ap-
proximately constant density corresponding to the tetramer-
ic contour region, and the flat electron density in regions far
away from the cylinder surfaces.The outer radius of the 8-
oxoguanine shell can be directly measured from the map as
10.5 Š. The very good agreement between the helix radius
calculated from the unit cell dimension and molecular vol-
umes (9.9 Š; see above) and that measured from the map is
noteworthy.
Synthesis and self-assembly of 8-oxoI 2: In the helix pro-
posed for 8-oxoG, the 2-amino group does not participate
directly in the hydrogen-bond framework between the bases
that generate the helical structure.[9] We have therefore syn-
thesized two lipophilic 8-oxoinosines to see if we could still
obtain assembled structures and liquid-crystalline gels.De-
rivatives 2a and 2b were prepared according to Scheme 1.
Commercial adenosine (3) was brominated at the 8-posi-
tion with bromine in sodium acetate buffer.The resulting 8-
bromoadenosine 4 was diazotized with NaNO2 in acetic acid
at room temperature and the resulting diazonium salt was
hydrolyzed in situ to afford 8-bromoinosine 5.Derivative 5
was then converted to triacetyl 8-oxoinosine 6 by substitu-
tion of bromine with acetate in acetic acid/acetic anhydride
at 1608C and subsequent hydrolysis of the resulting enol
ester.Subsequent transesterification with sodium methoxide
in methanol at 508C gave 8-oxoinosine 7.This compound
was esterified with the appropriate anhydride or aroyl chlo-
ride in MeCN at room temperature to afford 2a and 2b, re-
spectively.
In contrast to derivative 1, which exhibits two liquid-crys-
talline phases in hydrocarbon solvents,[9] solutions of deriva-
tive 2a do not show any birefringence up to about 80% w/w
in either hexane or chloroform.Furthermore, in the same
solvents, the circular dichroism spectra of 2a show only
weak signals corresponding to the 8-oxoinosine chromo-
phore (see Supporting Information).The shape and intensity
of the CD spectrum suggest the absence of a helicoidal
order in the supramolecular structure.Finally, the 1H NMR
spectrum in CDCl3 is quite similar to that in DMSO (in
which H-bonded architectures are not expected to exist);
that is to say, there is no notable shift of the resonances or
variation of band widths with the change of solvent.
Figure 5.Section of the cylindrically symmetric geometrical model used
to represent the aggregate.
outer radius of the 8-oxoguanine shell, and R3 is the outer
radius of the shell within which the sugar residues are locat-
ed.Dimensions and scattering length densities used in the
model, as derived from molecular models and chemical
compositions, were as follows: R1 = 1.5 Š, R2 = 7 Š, R3 =
12 Š, 11 = 0.33 eŠꢀ3, 12 = 0.46 eŠꢀ3, 13 = 0.47 eŠꢀ3, and
10 = 0.27 eŠꢀ3.The corresponding F(Q) continuous func-
tion is shown in the Supporting Information.Even though
the soft and disordered nature of these systems may serious-
ly affect the model structure,[11] the derived continuous func-
tion can be used to set the signs of the experimental Fourier
coefficients in order to directly calculate the electron density
map by applying Equation (3).
The electron density distribution is shown in Figure 6.The
circular 8-oxoG contour can be readily appreciated, but
other important features are also evident, such as the pres-
The different behavior of 2a and 1 may be attributed to
solvophobic effects.The lyomesomorphism requires that the
molecules are amphiphilic: the driving force of the self-as-
sembly derives from the segregation of the different parts of
the molecules, such that in the self-assembled species, the
lyophilic part is exposed to the solvent, while the lyophobic
part is not in contact with the solvent.The absence of the
amino group in 2a decreases the lyophobic character of the
aromatic base and this reduces the amphiphilic properties of
2a and hence its propensity to give large supramolecular
structures and consequently lyotropic mesophases.
Figure 6.Calculated electron density distribution for a dry film of 1.The
normalized Fourier coefficients are as follows: F1,0
=
0.313; F2,1
=
ꢀ0.100; F2,0 = ꢀ0.163; F3,1 = ꢀ0.090; F3,0 = ꢀ0.080; F4,2 = 0.0; F4,1
=
0.068 (diffracted intensities are normalized such as ꢀh,k Ih,k = 1.0, so that
the map shows the fluctuation of the electron density around the average
value, set to 0).The density levels are equally spaced, with an increment
of 0.2: dashed and full lines represent positive and negative levels, re-
spectively.Axis dimensions are in ångstrçms.
Self-assembly of 8-oxoI 2b to give a liquid-crystalline phase:
To ascertain whether increasing the amphiphilicity of the ox-
3444
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2007, 13, 3441 – 3449

Helical Self-Assembly of an 8-Oxoinosine
FULL PAPER
Scheme 1.Synthetic route to 2a and 2b.
oinosine derivative would increase its ability to self-assem-
ble, we prepared compound 2b (see Scheme 1).In this de-
rivative, aromatic groups and longer alkyl chains are present
on each of the three tails exposed to the solvent as com-
pared to the substituents on 2a.This structural feature
should increase the difference in lyophilicity between the
core (the 8-oxohypoxanthine base) and the rest of the mole-
cule.Furthermore, there may be stabilizing interactions be-
tween the aromatic groups.In fact, it has recently been re-
ported that aromatic side chains considerably enhance the
stability of double-helical foldamers.[13] Compound 2b was
indeed found to exhibit lyotropic liquid-crystalline proper-
ties in hydrocarbon solvents.Optical microscopy revealed
the presence of a birefringent fluid phase at c > 2.3% and
3.5% in hexadecane and heptane, respectively (where c is
the ratio of the weight of 2b to the total weight of the
sample) (Figure 7).
high concentration.Thus, at concentrations higher than
15%, a series of diffraction peaks was seen in the X-ray pro-
file.The characteristic peak positions confirmed the liquid-
crystalline nature of the sample (Figure 8).The low-angle
X-ray diffraction experiments confirmed the existence of
a liquid-crystalline order that is only clearly detectable at
Figure 8.X-ray diffraction profile obtained from a 30% (w/w) solution of
2b in heptane.
X-ray diffraction region is dominated by a strong peak, the
position of which depends on the solvent concentration,
while higher order diffraction peaks can only be detected
after very long exposure times.Under optimal conditions,
two or three further peaks were detected, which helped in
the assessment of the lattice symmetry.[9,10] The observed
p
p p p
peak reciprocal spacings were in the ratio 1: 3: 4: 7: 9,
indicating a two-dimensional hexagonal packing of the struc-
tural elements.
A characteristic peak, centered at about 3.74 Š (Q =
1.68 Šꢀ1), was observed in the high-angle region, the posi-
tion of which was independent of the concentration.The
presence of this peak indicates the columnar nature of the
Figure 7.Polarized optical microscopy image of a 10% (w/w) solution of
2b in hexadecane.Magnification 100”.
Chem. Eur. J. 2007, 13, 3441 – 3449
ꢀ 2007 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
3445

G.Gottarelli, G.P.Spada et al.
[9,10]
aggregates of 2b
and suggests that the stacked bases
may be arranged in a helical fashion, as observed for deriva-
tive 1.[9] Assuming that the stacking distance between the 8-
oxoinosine residues is again the standard 3.4 Š, the ob-
served 3.74 Š peak suggests that the angle formed by the
bases in relation to a plane perpendicular to the column axis
is around 238.As in the case of derivative 1, the high-angle
region is also characterized by a large band at Q = 1.4 Šꢀ1,
which is related to the disordered conformation of the lipo-
philic substituents in the hydrocarbon region.
Two points were then considered: first, the dependence of
the inter-helix distance on the sample composition, and
second, the reconstruction of the electron density distribu-
tion.Figure 9 shows the unit cell dimension measured for
Figure 10.Calculated electron density distribution for the mesophase of
2b (c = 0.3). The normalized Fourier coefficients are as follows: F1,0
0.319; F2,1 = 0.192; F2,0 = 0.092; F3,1 = 0.067; F3,0 = ꢀ0.102 (diffracted
intensities are normalized such as ꢀh,k Ih,k = 1.0, so that the map shows
the fluctuation of the electron density around the average value, set to
0).The density levels are all equally spaced, with an increment of 02. :
dashed and full lines represent positive and negative levels, respectively.
Axis dimensions are in ångstrçms.
=
Figure 9.Dependence on concentration of the 2D hexagonal unit cell di-
Table 1.Structural data of the mesophases formed by 2b in heptane at
different weight concentrations (c). R is the cross-sectional radius of the
helix calculated from unit cell dimensions and sample concentration
([Eq.(2)]); R’ is the outer radius of the 8-oxohypoxanthine shell directly
measured from the electron density maps.
mension for 2b.
the hexagonal phase as a function of the concentration: in-
terestingly, the concentration dependence of the unit cell di-
c
cv,G
a [Š]
R
R’
ꢀ1/2
mension is cv,G
and consequently the swelling behavior
0.2
0.3
0.5
0.7
0.055
0.082
0.136
0.189
72.6
60.5
47.4
38.2
8.9
9.1
9.2
8.7
10.5
10.5
10.5
10.5
appears to be well described by a model composed of rod-
like elements of infinite length that move apart as dilution
proceeds (2D swelling, see reference [9]).By fitting the data
by using Equation (2), a radius of 9 Š was derived for the
cross-section of the 8-oxoinosine helices (only the aromatic
core 8-oxohypoxanthine was considered).Electron density
distributions have been calculated by using Equation (3)
and the sign of the Fourier coefficients Fh,k was obtained by
considering the helical geometric model described above.
The electron density map for the 8-oxoI derivative 2b at c
= 0.3 is presented in Figure 10: comparison with the map
obtained for sample 1 clearly illustrates the lower experi-
mental resolution (for example, no details concerning the
central cavity are visible) and the swelling determined by
the solvent.Nevertheless, the similar shape and size of the
cross-sections of the two helices can be readily appreciated
(see Table 1).
solutions, the latter being a strongly competing solvent for
hydrogen bonding.
1
The H NMR spectrum of 2b in [D6]DMSO at room tem-
perature (258C) shows a single series of peaks (see Support-
ing Information), the assignments for which are presented in
Table 2.
Each proton gives rise to a peak of roughly the same
width, with the exception of the NH(1) and NH(7) protons,
the peaks due to which are approximately seven times
broader.The 1H NMR spectrum of 2b in CDCl3 at room
temperature also shows a single series of peaks.Notably,
however, the signals are significantly broader than those ob-
served in [D6]DMSO: in particular, the signals due to H(2),
the sugar residue protons, and the NH protons are 10, 20,
and 40–50 times broader, respectively.The existence of ag-
gregates large enough to give rise to broad signals was ruled
out by means of an external reference.
Self-assembly of 8-oxoI 2b to give an isotropic solution: To
obtain information on the structure of the supramolecular
aggregates arising from 2b in isotropic solution, a compara-
tive NMR study was carried out in chloroform and DMSO
3446
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2007, 13, 3441 – 3449

Helical Self-Assembly of an 8-Oxoinosine
FULL PAPER
Table 2. 1H NMR (400 MHz) chemical shifts for a 13 mm solution of 2b (258C).Assignments were made on
the basis of COSY, NOESY, and ROESY spectra.
carbon bond, as well as cross-
peaks between protons H(1’)
and H(4’) and between protons
Solvent
NH(1)
NH(7)
H(2)
H(1’)
H(2’)
H(3’)
H(4’)
H
G
Ph protons
CDCl3
[D6]DMSO
12.39
12.68
11.20
11.59
7.81
7.92
6.35
6.07
6.53
6.30
6.16
6.04
4.70
4.65
4.77, 4.67
4.59, 4.48
8.07–7.84, 6.82
7.88–6.88
H(3’) and H(5’)/H
corresponding spectrum ob-
tained from solution in
A
[a] Diastereotopic protons have not been assigned.
a
CDCl3 showed substantially
the same cross-peak pattern,
In principle, proof of the aggregation of 2b in CDCl3
could be obtained by comparing the sign of the NOESY
cross-peaks with respect to the diagonal in [D6]DMSO and
CDCl3.However, this was not possible in our case as 2b dis-
played NOESY cross-peaks with the same sign as the diago-
nal in both solvents (i.e., wtc > 1.12 even in DMSO), due
to its high molecular mass (Mw = 1149.50). Indirect proof
of aggregation has nevertheless been obtained from analysis
of ROESY spectra, which were preferred to NOESY spec-
tra because of the possibility of distinguishing between a
cross-peak generated by a direct through-space interaction
and one due to relayed ROE or chemical exchange, regard-
less of tc.[14]
In principle, two modes of assembly may be adopted in
solution: one based on hydrogen bonding between the
NH(7)-CO(8) amide group of one molecule and the NH(1)-
CO(6) amide group of another molecule, the other based on
hydrogen bonding between two NH(7)-CO(8) amide groups
of two molecules and between two NH(1)-CO(6) amide
groups of two molecules.The first possibility leads to G-
quartet-like or helical ribbon superstructures, while the
latter gives rise to a linear self-assembled ribbon (see
Figure 3).
but with one salient difference, namely a positive, medium-
intensity cross-peak between H(1’) and both the H(5’) and
H(5’’) protons.The presence of this cross-peak seems to rule
G
out the linear ribbon structure (and supports a helical or a
stacked G-quartet-like structure).An intramolecular ROE
seems to be excluded due to the excessive distance of three
ꢀ
C C bonds; a spin diffusion effect also seems unlikely.Spin
diffusion through H(4’) would give a negative cross-peak be-
tween H(1’) and H(5’)/H(5’’), while the observed cross-peak
E
is positive.A doubly relayed ROE would give a cross-peak
with the correct sign, but considering that the sample was
not degassed and that the intensity of cross-peaks between
ꢀ
protons separated by single C C bonds is weak, this possi-
bility seems remote.Finally, an intermolecular ROE seems
to be excluded in the case of a linear assembly, due to the
clearly excessive distance.
We therefore conclude that the above-mentioned NOE
coupling arises from hydrogens belonging to nucleosides in
a partially stacked configuration (see Figure 11), as is the
To ascertain whether the supramolecular arrangement at
high concentration inferred from the X-ray diffraction ex-
periments persists in isotropic solution, we realized that, ac-
cording to models, such an arrangement implies an intermo-
lecular distance between NH(7) of an inosine moiety and
H(2) of another inosine located at a different level of the
supramolecular helix that is shorter than the intramolecular
distance of 5.9 Š. Thus, if a helical structure is present, a
ROESY cross-peak should be detected between the NH(7)
and H(2) signals in chloroform (but not in DMSO), provid-
ed that the spin lock is kept sufficiently short to avoid signif-
icant NH(7) magnetization decay with T11.ROESY spectra
were thus run with a very short mixing time of 15 ms, simi-
larly to experiments reported in the literature in the pres-
ence of rapidly relaxing protons.[15] Our ROESY results are
consistent with the existence of a supramolecular structure
in CDCl3.The required cross-peak between NH(7) and
H(2) is indeed only observed in CDCl3 solution (see Sup-
porting Information), its positive sign (negative diagonal)
confirming that it is due to a genuine ROE.Taken alone,
this finding does not prove that a helical structure is adopt-
ed in solution, but it definitely rules out the possibility of a
ribbon architecture in chloroform.
Figure 11.A molecular sketch of a fragment of the supramolecular helix
of 2b. ROE contacts H(1’)–H(5’/5’’) and H(2)–NH(7) are indicated by
A
lines.
case in a continuous, helical ribbon or a stacked G-quartet-
like structure.It should be noted that in previous examples
of stacked G-quartets in the presence of ions,[16] two sets of
proton signals were observed.
CD spectra of 2b in different solvents are presented in
Figure 12.This compound exhibits a first intense absorption
band at around 250–260 nm (with a shoulder at ca.275 nm)
and a second band at around 215 nm.In methanol (in which
H-bonded guanine architectures are generally absent), the
CD spectrum features weak signals in both wavelength re-
A ROESY spectrum obtained from a solution of 2b in
DMSO using a longer mixing time (200 ms) showed positive
cross-peaks between protons separated by one carbon–
Chem. Eur. J. 2007, 13, 3441 – 3449
ꢀ 2007 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
3447

G.Gottarelli, G.P.Spada et al.
templated G-quartets show multiple sets[16] (and this should
also apply to possible columns without metal ions), we be-
lieve that a continuous helical architecture is most likely
adopted by 8-oxoI 2b (Figure 3d).The situation is analogous
to that reported for 8-oxoG, the assembled structure of
which has been confirmed in this work by SAXS measure-
ments on a film.
Experimental Section
General: CD spectra were recorded on a JASCO J-710 spectropolarime-
ter using cells of the appropriate pathlength.NMR spectra were recorded
on Varian Mercury instruments at 300 or 400 MHz.X-ray diffraction ex-
periments were performed using a Philips PW1830 X-ray generator
equipped with a Guinier-type focusing camera operating in a vacuum; a
bent quartz crystal monochromator was used to select CuKa1 radiation (l
= 1.54 Š). The investigated Q range (Q = (4p sinq)/l, where 2q is the
full scattering angle) was between 0.068 and 2.3 Šꢀ1.Diffraction patterns
were recorded on a stack of two Kodak DEF-392 films; film densities
were measured by using a digital scanner and scattered intensities were
corrected for electronic noise and sample holder signals.For data evalua-
tion purposes, Bragg peaks were fitted by Lorentzian distributions and
peak intensities Ih,k (h,k are the reflection indices) were obtained by mul-
tiplying each peak area by its corresponding wave vector (Q/2p)2 (for a
discussion, see reference [17]).
Figure 12.CD (top) and absorption (bottom) spectra of a 1-m m solution
of 2b in methanol (258C, dotted line) and in cyclohexane at 258C
(dashed line) and 708C (solid line).In the inset: the melting profile in cy-
clohexane monitored at 203 nm (dashed line) and 220 nm (solid line).
2’,3’,5’-O-Tridecanoyl-8-oxoguanosine (1) was prepared according to ref-
erence [9].2 ’,3’,5’-O-Tridecanoyl-8-oxoinosine (2a) and 2’,3’,5’-O-tris(4-
dodecyloxybenzoyl)-8-oxoinosine (2b) were prepared starting from com-
mercial 3 (see Scheme 1) according to the procedures outlined below.
Derivatives 4, 5, and 6 were synthesized as reported in references [18],
[19], and [20], respectively.4-(Dodecyloxy)benzoyl chloride was pre-
pared according to standard procedures.
gions.In contrast, the CD spectrum in hydrocarbon solvents
shows a much stronger signal in the region 230–280 nm and
a very intense (DDe ꢂ 170; De/e ꢂ2”10ꢀ3), exciton-like, bi-
signate couplet centered at 215 nm.This spectrum does not
change with concentration in the range 0.01–1 mm, indicat-
ing that the structure of the species in solution is not affect-
ed by concentration.On increasing the temperature, the
shape of the CD spectrum of a 1 mm solution changes dra-
matically and at 708C the exciton couplet at 215 nm has
almost completely vanished.As no assignment of the elec-
tronic transitions of the 8-oxohypoxanthine chromophore is
available and, furthermore, in derivative 2b the benzoyl
chromophore also contributes to the CD, detailed analysis
of the spectrum is not possible.However, a few issues may
be considered: 1) that different CD spectra are obtained in
cyclohexane and methanol resembles the behavior of 8-
oxoG 1;[9] 2) the intensity of the CD signal and the g-factor
(De/e) in cyclohexane are only compatible with a “highly”
chiral structure, much more so than the chiral structures
composed of stacked G-quartets.
8-Oxoinosine (7): Na (0.35 g, 15 mmol) was dissolved in anhydrous meth-
anol (15 mL) and the resulting methoxide solution was combined with
compound 6 (1.50 g, 3.80 mmol). The product precipitated immediately,
and after 2 h the solid was isolated by filtration to give 7 (0.54 g; 50%
yield). 1H NMR (300 MHz, [D6]DMSO): d = 3.40–3.70 (m, 2H; H5’/
H5’’), 3.84 (m, 1H; H4’), 4.10 (m, 1H; H3’), 4.82 (m, 1H; H2’), 5.64 (dd,
1H; H1’), 7.89 ppm (s, 1H; ArH, H2).
2’,3’,5’-O-Tridecanoyl-8-oxoinosine (2a): 8-Oxoinosine (0.3 g, 1.0 mmol)
was dried over P2O5 in vacuo for 2 h at 508C and then suspended in an-
hydrous acetonitrile (8 mL).Redistilled Et 3N (0.5 mL, 3.6 mmol), dec-
anoic anhydride (1.2 mL, 3.3 mmol), and a catalytic amount of DMAP
were added, and the resulting mixture was stirred overnight.After evapo-
ration of the solvent in vacuo, the crude material was applied to a
column of silica gel.After washing with dichloromethane/acetone (95:5)
to remove the decanoic acid, 2’,3’,5’-O-tridecanoyl-8-oxoinosine was
eluted with dichloromethane/methanol (96:4).The solvents were evapo-
rated in vacuo to afford 2a (0.56 g; 76% yield). 1H NMR (300 MHz,
[D6]DMSO): d = 0.84 (m, 9H; CH3), 1.23 (m, 36H; CH2), 1.49 (m, 6H;
OC-CH2CH2), 2.30 (m, 6H; OC-CH2), 4.12–4.22 (m, 2H; H5’/H5’’), 4.33
(m, 1H; H4’), 5.58 (m, 1H; H3’), 5.81 (dd, 1H; H1’), 6.01 (m, 1H; H2’),
8.00 (s, 1H; ArH, H2), 11.61 (brs, 1H; NH), 12.60 ppm (brs, 1H; NH);
13C NMR (75 MHz, CDCl3): d = 14.29 (CH3), 14.44 (CH3), 22.93 (CH2),
25.05 (CH2), 25.11 (CH2), 25.83 (CH2), 29.55 (CH2), 29.69 (CH2), 32.13
(CH2), 34.15 (CH2), 34.20 (CH2), 34.28 (CH2), 35.94 (CH2), 63.70 (CH2),
70.79 (CH), 71.80 (CH), 79.94 (CH), 84.76 (CH), 109.62 (C), 144.30 (C),
144.51 (C), 145.66 (CH), 152.73 (C), 172.63 (C), 172.72 (C), 173.78 ppm
(C); elemental analysis calcd (%) for C40H66N4O9: C 64.32, H 8.91, N
7.50; found: C 64.46, H 9.29, N 7.59.
Conclusions
Considering that 1) the G-quartets are unable to stack in the
absence of metal ions,[5] 2) the anisotropy factor, g, is much
higher than those of chiral structures composed of stacked
2’,3’,5’-O-Tris(p-dodecyloxybenzoyl)-8-oxoinosine (2b): 8-Oxoinosine
(0.20 g, 0.7 mmol) was dried over P2O5 in vacuo for 2 h at 508C and then
suspended in anhydrous acetonitrile (6 mL).Redistilled Et 3N (0.35 mL,
2.51 mmol), 4-(dodecyloxy)benzoyl chloride (0.75 g, 2.31 mmol), and a
catalytic amount of DMAP were added.The mixture was stirred over-
1
G-quartets,[9] and 3) the H NMR spectra show only a single
set of signals whereas columnar structures of stacked, metal-
3448
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
Chem. Eur. J. 2007, 13, 3441 – 3449

Helical Self-Assembly of an 8-Oxoinosine
FULL PAPER
night.The reaction was stopped by adding methanol (2 mL) and, after
evaporation of the solvent in vacuo, the crude material was crystallized
from ethanol/chloroform to afford 2b as a white solid (0.54 g; 68%
yield). 1H NMR (300 MHz, [D6]DMSO): d = 0.84 (m, 9H; CH3), 1.23
(m, 48H; CH2), 1.38 (m, 6H; O-CH2CH2CH2), 1.69 (m, 6H; O-CH2CH2),
4.01 (m, 6H; O-CH2), 4.46–4.70 (m, 3H; H4’, H5’, and H5’’), 6.06 (m,
1H; H3’), 6.09 (d, 1H; H1’), 6.32 (dd, 1H; H2’), 6.88–7.02 (m, 6H; ArH),
7.74–7.95 (m, 7H; ArH), 11.61 (s, 1H; NH), 12.70 ppm (s, 1H; NH); 13C
G.Gottarelli, GP. .Spada, A.Garbesi, in
Comprehensive Supra-
molecular Chemistry, Vol. 9, Templating, Self-Assembly and Self-Or-
ganization (Eds.: J-.M.Lehn, J.L.Atwood, D.D.MacNicol, J.A.D.
Davies, F.Vçgtle, J-.P.Sauvage, M.W.Hosseini), Pergamon, Oxford,
1996, pp.483–506; A. Ghoussoub, J-.M. Lehn,
2005, 46, 5763; N.Sreenivasachary, J-.M.Lehn,
Sci. USA 2005, 102, 5938; A.B.Kotlyar, N.Borovok, T.Molotsky,
H.Cohen, E.Shapir, D.Porath, Adv. Mater. 2005, 17, 1901.
Chem. Commun.
Proc. Natl. Acad.
NMR (100 MHz, [D6]DMSO, 508C): d
=
14.09 (CH3), 22.30 (CH2),
[5] J.L.Sessler, M.Sathiosatham, K.Doerr, V.Lynch, K.A.Abboud,
Angew. Chem. 2000, 112, 1356; Angew. Chem. Int. Ed. 2000, 39,
1300; ; F.W.Kotch, V.Sidorov, Y-.F.Lam, K.J.Kayser, H.Li, M.S.
Kaucher, J.T. Davis, J. Am. Chem. Soc. 2003, 125, 15140; T. N.
25.67 (CH2), 28.75 (CH2), 28.80 (CH2), 28.93 (CH2), 29.20 (CH2), 29.24
(CH2), 31.53 (CH2), 63.28 (CH2), 68.22 (CH2), 70.86 (CH), 72.35 (CH),
78.66 (CH), 84.49 (CH), 109.14 (C), 114.49 (CH), 114.60 (CH), 114.76
(CH), 120.77 (C), 120.95 (C), 121.65 (C), 131.55 (CH), 131.67 (CH),
131.83 (CH), 143.93 (C), 145.33 (CH), 151.10 (C), 151.48 (C), 163.03 (C),
163.23 (C), 163.37 (C), 164.51 (C), 165.34 ppm (C); elemental analysis
calcd (%) for C67H96N4O12: C 70.01, H 8.42, N 4.87; found: C 70.09, H
8.18, N 4.70.
Pham, S.Masiero, G.Gottarelli, SP. .Brown,
J. Am. Chem. Soc.
2005, 127, 16018.
[6] K.Araki, R.Takasawa, I.Yoshikawa,
Chem. Commun. 2001, 1826;
J. Mater.
T.Sato, M.Seko, R.Takasawa, I.Yoshikawa, K.Araki,
Chem. 2001, 11, 3018; K.Araki, M.Abe, A.Ishizaki, T.Ohya,
Chem. Lett. 1995, 359; G.Gottarelli, S.Masiero, E.Mezzina, S.Pier-
accini, JP. .Rabe, P.Samorꢁ, GP. .Spada,
Chem. Eur. J. 2000, 6,
3242; T.Giorgi, F.Grepioni, I.Manet, P.Mariani, S.Masiero, E.
Acknowledgement
Mezzina, S.Pieraccini, L.Saturni, G.P.Spada, G.Gottarelli,
Eur. J. 2002, 8, 2143.
Chem.
This work has been supported by the MUR/PRIN project Modellazione
caratterizzazione di cristalli liquidi per strutture nano-organizzate
(2005035119), the MUR/FIRB project NOMADE (RNNE01YSR8_004),
and the University of Bologna.
[7] S.L.Forman, J.C.Fettinger, S.Pieraccini, G.Gottarelli, J.T.Davis,
J. Am. Chem. Soc. 2000, 122, 4060; M.S.Kaucher, W.A.Harrell, Jr.,
J.T.Davis, J. Am. Chem. Soc. 2006, 128, 38.
[8] G.Maruccio, P.Visconti, V.Arima, S.DꢂAmico, A.Biasco, E.
DꢂAmone, R.Cingolani, R.Rinaldi, S.Masiero, T.Giorgi, G.Got-
tarelli, Nano Lett. 2003, 3, 479; R.Rinaldi, E.Branca, R.Cingolani,
e
[1] J.-M. Lehn, Science 2002, 295, 2400; V.Balzani, A.Credi, FM. .
Raymo, J.F. Stoddart, Angew. Chem. 2000, 112, 3484; Angew.
Chem. Int. Ed. 2000, 39, 3348; S.I.Stupp, V.LeBonheur, K.Walker,
S.Masiero, G.P.Spada, G.Gottarelli,
3541; R.Rinaldi, G.Maruccio, A.Biasco, V.Arima, R.Cingolani, T.
Giorgi, S.Masiero, G.P.Spada, G.Gottarelli, Nanotechnology 2002,
13, 398; S.DꢂAmico, G.Maruccio, P.Visconti, E.DꢂAmone, R.Cin-
Appl. Phys. Lett. 2001, 78,
L.S.Li, K.E.Huggins, M.Keser, A.Amstutz,
384.
Science 1997, 276,
golani, R.Rinaldi, S.Masiero, G.P.Spada, G.Gottarelli,
tron. J. 2003, 34, 961.
Microelec-
[2] A.E. Rowan, R.J.M. Nolte,
Angew. Chem. 1998, 110, 65; Angew.
Chem. Int. Ed. 1998, 37, 63; J.C. Nelson, J.G. Saven, J.S. Moore,
P.G. Wolynes, Science 1997, 277, 1793; V.Berl, I.Huc, RG. .
Khoury, M.J.Krische, J-.M.Lehn, Nature 2000, 407, 720; L.J.Prins,
[9] T.Giorgi, S.Lena, P.Mariani, M.A.Cremonini, S.Masiero, S.Pier-
accini, J.P.Rabe, P.Samorꢁ, G.P.Spada, G.Gottarelli,
J. Am. Chem.
Soc. 2003, 125, 14741.
T.Timmerman, DN. .Reinhoudt,
10153; D.T.Bong, T.D.Clark, J.R.Granja, M.R.Ghadiri,
Chem. 2001, 113, 1016; Angew. Chem. Int. Ed. 2001, 40, 988; ;
J. Am. Chem. Soc. 2001, 123,
[10] G.Gottarelli, G.P.Spada, P.Mariani, in
Crystallography of Supra-
Angew.
molecular Compounds (Eds:. G.Tsoucaris, J.L.Atwood, J.Lipkow-
ski), Kluwer, Dordrecht, 1996, p.307.
[11] V.Luzzati, in Biological Membranes (Ed.: D. Chapman), Academic
Press, London, 1968, p.71.
K.E.S.Phillips, T.J.Katz, S.Jockusch, A.J.Lovinger, N.J.Turro,
Am. Chem. Soc. 2001, 123, 11899; H.Fenniri, P.Mathivanan, K.L.
Vidale, D.M. Sherman, K. Hallenga, K.V. Wood, J.G. Stowell,
Am. Chem. Soc. 2001, 123, 3854; R.Oda, I.Huc, M.Schutz, S.J.
Candau, F.C. MacKintosh, Nature 1999, 399, 566; G.W. Orr, L.J.
Barbour, J.L. Atwood, Science 1999, 285, 1049; V.Percec, M.
Glodde, T.K. Bera, Y. Miura, I. Shiyanovskaya, K.D. Singer,
V.S.K. Balagurusamy, P.A. Heiney, I. Schnell, A. Rapp, H-.W.
J.
J.
[12] F.Federiconi, P.Ausili, G.Fragneto, C.Ferrero, P.Mariani,
J. Phys.
Chem. B 2005, 109, 11037.
[13] D.Haldar, H.Jiang, J-.M.Lꢃger, I.Huc,
Angew. Chem. 2006, 118,
5609; Angew. Chem. Int. Ed. 2006, 45, 5483.
[14] D.Neuhaus, MP. .Williamson,
Structural and Conformational Analysis, VCH, Weinheim, 1989.
[15] U. Simonis, J.L. Dallas, F.A. Walker, Inorg. Chem. 1992, 31, 5349;
The Nuclear Overhauser Effect in
Spiess, S.D.Hudson, H.Duan, Nature 2002, 419, 384; M.L.Bushey,
A.Hwang, P.W.Stephens, C.Nuckolls, Angew. Chem. 2002, 114,
T.K. Shokhireva, M.J.M. Nesset, F.A. Walker,
Inorg. Chim. Acta
2952; Angew. Chem. Int. Ed. 2002, 41, 2828; ; A.P.H.J.Schenning,
A.F.M.Kilbinger, F.Biscarini, M.Cavallini, H.J.Cooper, P.J.Der-
rick, W.J.Feast, R.Lazzaroni, P.Leclere, L.A.McDonell, E.W.
Meijer, S.C.J.Meskers, J. Am. Chem. Soc. 2002, 124, 1269; H.En-
1998, 272, 204.
[16] E.Mezzina, P.Mariani, R.Itri, S.Masiero, S.Pieraccini, G.P.Spada,
F.Spinozzi, J.T.Davis, G.Gottarelli, Chem. Eur. J. 2001, 7, 388;
A.L. Marlow, E. Mezzina, G.P. Spada, S. Masiero, J.T. Davis, G.
Gottarelli, J. Org. Chem. 1999, 64, 5116; V.Gubala, J.E.Betancourt,
J.M.Rivera, Org. Lett. 2004, 6, 4735.
gelkamp, S.Middelbeek, RJ.M. .Nolte,
Science 1999, 284, 785;
J.H.K.K. Hirschberg, L. Brunsveld, A. Ramzi, J.A.J.M. Veke-
mans, R.P.Sijbesma, E.W.Meijer, Nature 2000, 407, 167.
[17] M.Rappolt, A.Hickel, F.Bringezu, K.Lohner,
Biophys. J. 2003, 84,
[3] T.Shimizu, R.Iwaura, M.Masuda, T.Hanada, K.Yase,
J. Am.
3111.
Chem. Soc. 2001, 123, 5947; Z.Li, C.A.Mirkin, J. Am. Chem. Soc.
[18] T.S. Lin, J.C. Cheng, K. Ishiguro, A.C. Sartorelli,
J. Med. Chem.
2005, 127, 11568; S.Sivakova, S.J.Rowan,
Chem. Soc. Rev. 2005,
1985, 28, 1481.
34, 9; L.Brunsveld, BJ.B. .Folmer, EW. .Meijer, RP..Sijbesa,
Chem. Rev. 2001, 101, 4071; S.Sivakova, J.Wu, C.J.Campo, P.T.
Mather, S.J.Rowan, Chem. Eur. J. 2006, 12, 446; K.Araki, I.Yoshi-
kawa, Top. Curr. Chem. 2005, 256, 133.
[19] R.E.Holmes, R.K.Robins, J. Am. Chem. Soc. 1964, 86, 1242.
[20] M.Ikehara, T.Maruyama, Chem. Pharm. Bull. 1976, 24, 565.
[4] J.T. Davis, Angew. Chem. 2004, 116, 684; Angew. Chem. Int. Ed.
2004, 43, 668; J.T. Davis, G.P. Spada, Chem. Soc. Rev. 2007, DOI:
Received: August 3, 2006
Revised: November 14, 2006
101. 039/b600282j; G.Gottarelli, G.P.Spada,
Chem. Rec. 2004, 4, 39;
Published online: January 16, 2007
Chem. Eur. J. 2007, 13, 3441 – 3449
ꢀ 2007 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
3449