Electrochemical characterisation of 6-iodomaltose, 6′-iodomaltose and 6-iodomaltotriose on a silver cathode and their one-pot electrochemical dimerisation to new mixed O/C maltotetraose and maltohexaose mimics

Article abstract of DOI:10.1002/chem.200900825

The electrochemical reduction on silver of peracetylated 6-iodo6-deoxy-β-maltose (2), 6-iodo-6-deoxy-β-maltotriose (3) and 6′-iodo-6′-deoxy-β-maltose (4) has been investigated by cyclic voltammetry and performed on a preparative scale, according to the st

Full text of DOI:10.1002/chem.200900825

FULL PAPER
DOI: 10.1002/chem.200900825
Electrochemical Characterisation of 6-Iodomaltose, 6’-Iodomaltose and
6-Iodomaltotriose on a Silver Cathode and Their One-Pot Electrochemical
Dimerisation to New Mixed O/C Maltotetraose and Maltohexaose Mimics
Angelo Alberti,[b] Dante Macciantelli,[b] Annamaria Naggi,[c] Elena Urso,[c]
Giangiacomo Torri,[c] and Elena Vismara*[a]
Dedicated to the Centenary of the Italian Chemical Society
ꢀ
C
Abstract: The electrochemical reduc-
tion on silver of peracetylated 6-iodo-
6-deoxy-b-maltose (2), 6-iodo-6-deoxy-
b-maltotriose (3) and 6’-iodo-6’-deoxy-
b-maltose (4) has been investigated by
cyclic voltammetry and performed on a
CH2 , primary radicals at C-6, which
(double bonds C-6’/C-5’ and C-5’/C-4’).
In turn, 4 afforded only the 6’,6’-dimer
maltotetraose mimic 17 in 60% yield,
accompanied by the reduced maltose
18 in 20% yield, in which the starting
CH2I became CH3. Compounds 7, 8,
12, 13 and 17 belong to a class of
mixed O/C malto-mimic oligosacchar-
could also abstract H-5’, becoming CH3
and generating the C-5’ quaternary
radicals that dimerised in 8 and 13, re-
spectively. These products were accom-
panied by the maltose derivatives 9, 10
and 11a/b in 42% overall yield and by
the maltotriose derivatives 14, 15 and
16 in 48% overall yield, respectively.
Compounds 9, 14 and 10, 15 came from
ꢀ
preparative scale, according to the stoi-
ꢀ
ꢀ
CH2I
ꢀ
C
chiometry,
G
Iꢀ. In agreement with the preparative
electrolysis results, cyclic voltammetry
showed different profiles for the reduc-
ing terminal-iodinated 2 and 3 and for
the non-reducing terminal-iodinated 4.
Compounds 2 and 3 partly dimerised
to maltotetraoses mimics 7 (6,6-dimer)
and 8 (5’,5’-dimer) in 38% overall yield
and to maltohexaose mimics 12 (6,6-
dimer) and 13 (5’,5’-dimer) in 30%
overall yield, respectively. Compounds
7 and 12 came from the dimerisation of
ides wherein an unnatural C C bond
ꢀ
between two saccharide units increases
metabolic stability compared to their
O-analogues and modulates the sugar
chain conformational flexibility, a fun-
damental parameter in determining
protein–carbohydrate binding. Direct
and spin-trapping EPR studies substan-
tiated the radical-based nature of the
dimerisation processes and allowed the
identification of some of the paramag-
netic species involved.
C
CH2 disproportionation to CH3 and
CH2=C, respectively (exocyclic double
bond C-6/C-5). Compounds 11a/b and
16 came from C-5’ radical reduction,
followed by acetate anion elimination
Keywords:
electrochemistry
radicals · silver
carbohydrates
maltomimics
·
·
·
Introduction
The electroreduction of halosaccharides on a silver cathode
is a well-established technique providing a mild, clean and
one-pot method for the preparation of double sugar units
[a] Prof. Dr. E. Vismara
ꢀ
through the formation of stable interglycosidic C C bonds.
Dipartimento di Chimica
More precisely, it is a synthetic procedure whereby dimeri-
sation of a carbon-centred radical follows the loss of a
halide anion from an electrochemically reduced halo sugar
and affords C-disaccharide mimics (see Scheme 1),[1–3] mixed
Materiali et Ingegneria Chimica “G. Natta“
Politecnico di Milano (Italy)
Fax : (+39)0223993180/3080
E-mail: elena.vismara@polimi.it
[b] Dr. A. Alberti, D. Macciantelli
ISOF-CNR Area di Ricerca
O/C-maltotetraose
mimics
from
acetobromomaltose
(ABM)[4] and mixed O/C-maltohexaose mimics from aceto-
via Pietro Gobetti 101, 40129 Bologna (Italy)
bromomaltotriose (ABMT;[5] see Scheme 2).
[c] Dr. A. Naggi, Dr. E. Urso, Dr. G. Torri
Actually, it should be emphasised that the real bioactive
moieties of natural polysaccharides are specific or unspecific
Istituto di Ricerche Chimiche e Biochimiche “G. Ronzoni”
via Giuseppe Colombo 81, 20133 Milano (Italy)
Chem. Eur. J. 2009, 15, 8005 – 8014
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
8005

majority of oligosaccharides are flexible, which allows them
to assume proper conformations to adapt themselves to the
binding site of a specific protein (favourable enthalpic con-
tribution). However, the binding brings along an unfavoura-
ble entropy contribution, because it reduces the flexibility of
the oligosaccharide. Thus, although high flexibility of oligo-
saccharides helps the binding, too much flexibility may
hamper the carbohydrate–protein complex formation. To
gain an understanding of structure/activity relationships of
natural and synthetic carbohydrate-based structures, many
studies were done in various directions: these include mod-
elling protein recognition of carbohydrates,[10] simulating
and modelling flexible sugar rings,[11] synthesising conforma-
tionally constrained oligosaccharides,[12] comparing C-glyco-
side and O-glycoside conformation[13] and more recently de-
veloping multidisciplinary approaches to the study of pro-
tein–carbohydrate complexes, made possible by the techni-
cal developments in NMR spectroscopy and X-ray diffrac-
tion.[14] Nevertheless, despite all these activities, this
understanding seems a long way from being achieved. The
most important property of the sulfated forms of the O/C-
maltohexaose mimics reported in Scheme 2 is that, in addi-
tion to the above features, they are biologically active and
have a real chance of being developed as new carbohydrate-
based drugs having anti-tumour and anti-metastatic activi-
ties.[15]
Scheme 1. Halo-monosaccharide electroreduction derivatives.
These positive results prompted us to prepare different
“malto-mimics” by strategically choosing the position of the
halogen atom in the starting halo sugar. Thus, we have en-
deavoured in the preparation of new structures derived
from the coupling of iodomaltose and iodomaltotriose in
which the iodine atom is bound to the exocyclic carbon of
either the reducing or non-reducing-sugar end unit.
Scheme 2. ABM and ABMT electroreduction derivatives.
oligosaccharides of more than four sugar units.[6] Due to
their small molecular size, oligosaccharides show a larger
bioavailability than their polymeric precursors. As a result,
large efforts have been addressed to preparing synthetic oli-
gosaccharides that may mimic the natural ones.
We report here on the outcome of the electrochemical re-
duction of the model compound 1,2:3,4-di-O-isopropyli-
dene-6-deoxy-6-iodo-a-d-galactose (1),[3,16] of 1,2,3,2’,3’,4’,6’-
hepta-O-acetyl-6-deoxy-6-iodo-b-maltose
(2),[17]
In recent years, we have begun focussing our attention on
linear malto-oligosaccharides, especially on maltohexaose,
whose sulfated forms have been found to be efficient inhibi-
tors of tumour growth and metastasis.[7] We then moved to
building up mixed O/C-malto-oligosaccharides and checking
whether or not their sulfated forms would mimic natural
malto-oligosaccharide. The sugar chains of these mimics are
1,2,3,2’,3’,6’,2’’,3’’,4’’,6’’-deca-O-acetyl-6-deoxy-6-iodo-b-mal-
totriose (3)[18] and 1,2,3,6,2’,3’,4’-hepta-O-acetyl-6’-deoxy-6’-
iodo-b-maltose (4).[17]
ꢀ
characterised by the presence of an interglycosidic C C
bond that is expected to induce chemical and metabolic sta-
bility with respect to malto-oligosaccharides, similarly to
what has been observed for C-glycosides that are less vul-
nerable to metabolic processing than their O-analogues and
for this reason have been studied as drug candidates and in-
hibitors of carbohydrate-processing enzymes.[8] The presence
ꢀ
of the interglycosidic C C bond is an important feature be-
cause, as shown by a molecular modelling conformational
analysis,[5] it modifies the geometry of the sugar chains and
makes their conformation rigid. In fact, the conformational
flexibility of oligosaccharides is critical in determining their
binding to protein and consequently their bioactivity.[9] The
8006
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 8005 – 8014

Characterisation of 6- and 6’-Iodomaltose and 6-Iodomaltotriose
FULL PAPER
Results
Cyclic voltammetry: The cyclic voltammetry on a silver
cathode of compounds 1–4 in acetonitrile was explored at
different sweep rates prior to carrying out the preparative
electrolysis.
Complex and irreversible voltammograms consisting of
several waves were observed in all cases, their relative inten-
sities being dependent on the sweep rate.
Compounds 1 and 4 exhibit similar trends. At scan rates
ꢁ200 mVsꢀ1 a first reduction wave at approximately ꢀ1.15
to ꢀ1.25 V and a broad second wave at higher potentials
(ca. ꢀ1.78 V for 1 and between ꢀ1.5 and ꢀ1.85 V for 4) are
observed, with the former wave corresponding to the reduc-
tion of the carbon–iodine bond. At 500 mVsꢀ1, 4 exhibits a
unique second wave between ꢀ1.3 and ꢀ1.9 V, whereas for
1 the two waves begin to be superimposed although still dis-
tinguishable. At 800 mVsꢀ1 only one broad band is visible
for both 1 and 4. In the voltammograms of 2 and 3, the re-
duction wave of the carbon–iodine bond around ꢀ1.2 V is
accompanied by a second wave very close to the first and a
third wave at more negative potentials. The second wave is
very weak, in fact hardly visible at scan rates ꢁ50 mVsꢀ1,
but becomes more and more relevant as the scan rate grows
faster (from 50 to 500 mVsꢀ1) while the first wave loses in-
tensity, shifting slightly to more negative potentials. In the
same scan range, the third wave shifts from 1.56 to 1.99 V.
At 800 mVsꢀ1 only the second wave remains visible in the
interval between 0 and ꢀ2.3 V, the third wave having proba-
bly shifted beyond ꢀ2.5 V.
This is evidenced in Figure 1, in which the forward reduc-
tion voltammograms are shown for compounds 1, 2 and 4
(the voltammogram observed for 3 is virtually identical to
that observed for 2).
After the first scan, small waves are also detected for all
sugars in the interval between 0 and ꢀ0.7 V. These are due
to the reduction of the iodine released as iodide anion from
the sugar after the first reduction.
Preparative electrolysis experiments: Potentiostatic electrol-
yses were carried out in acetonitrile in a two-compartment
cell, with the cathode and anode both being silver plates.
Electrolysis details, workup and product isolations have
been detailed in the Experimental Section. The isolated
products can be divided in two classes: dimers, that is, com-
pounds that have twice the number of sugar units of the
starting compound, and thus twice the molecular weight of
the intermediate radicals resulting from the loss of the
iodine anion (see Scheme 3), and compounds that have the
same number of sugar units as the starting compound and
thus a molecular weight close to that of the intermediate
radical (see Scheme 4).
Figure 1. Forward cyclovoltammetric reduction traces for compounds
top) 1, middel) 2 and bottom) 4. Scan rates are 50 (black), 100 (red), 200
(yellow), 500 (blue) and 800 (purple) mVsꢀ1
.
18.[17,19] In turn, 2 and 3 partly dimerised to maltotetraoses
mimics 7 (6,6-dimer) and 8 (5’,5’-dimer) and to maltohexa-
ose mimics 12 (6,6-dimer) and 13 (5’,5’-dimer), respectively.
These products were accompanied by the maltose deriva-
tives 9,[19] 10[20] and 11a/b, and by the maltotriose derivatives
14, 15[18] and 16, respectively.
The maintenance of the C5’ d-glucose configuration and
conformation in 8 and 13 was supported by optical rotation
and NMR spectroscopic characterisation, as detailed in the
Experimental Section. The alternative l-ido configuration is
In Table 1 we report their quantitative yields. Compounds
1 and 4 exhibit similar trends. Compound 1 was reduced to
the C-disaccharide mimic 5,[3] and to the galactose derivative
6.[16] Compound 4 afforded only the 6’,6’-dimer maltote-
traose mimic 17 accompanied by the maltose derivative
Chem. Eur. J. 2009, 15, 8005 – 8014
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8007

E. Vismara et al.
expected to contribute with negative optical rotation (a)
values.[21] Preliminary semi-empirical (AM1) calculations on
13 were also performed and elucidated the C2 symmetry
(see Figure 2).
Scheme 3. Dimerisation derivatives isolated in the preparative electroly-
sis of compounds 2–4.
Figure 2. Molecular modelling geometry of compound 13.
EPR spectroscopy: Although the first step of the electrore-
duction of the investigated iodo-sugars is conceivably the
formation of the corresponding radical anions that immedi-
ately undergo loss of an iodide anion with formation of the
corresponding 6-yl primary radicals, no EPR spectra could
be observed upon room-temperature electrochemical reduc-
tion of sugars 1–4 inside the cavity of the EPR spectrome-
ter.
Scheme 4. Reduction, disproportionation and rearrangement derivatives
isolated upon preparative electrolysis of compounds 1–4.
This was attributed to a very low persistence of the result-
ing radicals. Indeed, the persistence of these paramagnetic
species should increase at low temperature, but our equip-
ment setup did not allow low-temperature electrochemical
EPR experiments. Therefore, the generation of the 6-yl pri-
mary radicals from 1–4 at low temperature was attempted
by iodine abstraction (see Experimental Section).
Table 1. Product distribution in the preparative electrolysis of sugars 1–4.
Yield[a]
Yield[a]
Yield[a]
Yield[a]
Yield[a]
1[b]
2
3
–
71 (5)
9 (8)
12 (13)
60 (17)
15 (6)
–
–
–
–
29 (7)
18 (12)
–
12 (9,10)
24 (14,15)
–
30 (11a/b)
24 (16)
–
4
20 (18)
Figure 3a shows the EPR spectrum observed by irradiat-
ing with UV light at T=ꢀ708C a thoroughly degassed solu-
[a] % yield of the isolated compound that is given in parentheses.
[b] Data from ref. [3].
8008
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 8005 – 8014

Characterisation of 6- and 6’-Iodomaltose and 6-Iodomaltotriose
FULL PAPER
Scheme 5. Generation of 1a, MNP trapping (1c), dimerisation (5) and re-
duction (6).
for which a spectrum similar to that exhibited by 1a was ex-
pected. We believe this spectrum to be due to a rearranged
radical and, due to the nature of the isolated products, we
tentatively identify this new species as radical 2e, in which
the large doublet would be due to the hydrogen in position
4’ and the smaller one to either of the two hydrogen atoms
of the 6’-methylenic group. As was the case for 1, increasing
the temperature led to the disappearance of the EPR signal.
In the available temperature interval, an increase of the
smaller splitting and a decrease of the larger one were ob-
served when increasing the temperature.
Despite the fast rearrangement process undergone by rad-
ical 2a, the photolysis of 2 in the presence of MNP led to a
strong signal from 2c, a nitroxide structurally related to that
observed with sugar 1 (see Table 2 for spectral parameters).
The only difference here was the presence of a line-width al-
ternation effect due to hindered rotation around the nitro-
gen–C6 bond that causes an exchange of the two methylenic
hydrogen splittings. Through a careful simulation of the
spectral pattern, a value of approximately 1ꢁ108 sꢀ1 for the
exchange rate constant could be estimated.
Figure 3. EPR spectra observed upon photolysis of a degassed solution of
1 and hexabutylditin in cyclopropane (top) and a degassed solution of 1,
hexabutylditin and MNP in cyclopropane (bottom). The lines marked
with * are due to di-tert-butyl nitroxide. Experimental traces are in blue,
computer-reproduced traces in red.
tion of sugar 1 in cyclopropane containing a small amount
of hexabutylditin.
On the basis of the spectral parameters (see Table 2),[22]
the spectrum can be safely attributed to radical 1a, with the
two methylenic hydrogen atoms being responsible for the
2.414 mT triplet and the hydrogen in position 5 for the
2.711 mT doublet. The additional 0.076 mT quartet splitting
is due to three of the four remaining ring hydrogen atoms.
Increasing the temperature did not result in significant
changes of the spectral pattern but for a remarkable pro-
gressive decrease of the intensity, the EPR signal being
eventually undetectable above ꢀ308C.
When
a small amount of 2-methyl-2-nitrosopropane
(MNP) was added to the system, the spectrum shown in Fig-
ure 3b was instead observed. This is attributed[23,24] to nitro-
xide 1c resulting from the trapping of 1a by MNP (see
Scheme 5).
The low-temperature (T=ꢀ708C) photolysis of a solution
of 2 in cyclopropane/toluene containing hexabutylditin led
to the observation of an EPR spectrum consisting of a dou-
blet of doublets that could not be attributed to radical 2a,
Sugar 3 behaved similarly to sugar 2 in all respects. Its
photoreaction with hexabutylditin did not lead to the obser-
vation of radical 3a, but to a spectrum consisting again of a
doublet of doublets, attributed to the rearranged radical 3e
(Scheme 6), that exhibited a temperature dependence simi-
lar to that observed for 2e. Nitroxide 3c was similarly
formed when the photoreaction was carried out in the pres-
ence of MNP, but in this case the exchange rate constant
was faster (ca. 4ꢁ108 sꢀ1). Fur-
thermore, 3c was accompanied
Table 2. EPR spectral parameters for the radicals observed in the photoreactions of sugars 1–4 with hexabu-
tylditin in the absence or in the presence of MNP.
by another nitroxide (3 f); al-
though the trapping of the rear-
ranged tertiary radical 3e might
have been expected, the spec-
tral parameters of 3 f suggest
Radical
Hyperfine splitting constants [mT]
g factor
T [8C]
1a
1c
2e
2c
3e
3c
3 f
4a
4c
2.414 (2H), 2.711 (1H), 0.076 (3H)
0.071 (1H), 0.734 (1H), 1.256 (1H), 1.501 (1N)
0.683 (1H), 3.515 (1H)
0.0550 (1H), 0.499 (1H), 1.529 (1H), 1.510 (1N)
0.580 (1H), 3.835 (1H)
0.423 (1H), 1.297 (1H), 1.486 (1N)
0.251 (1H), 1.401 (1N)
0.112 (1H), 2.253 (2H), 2.399 (1H)
0.055 (1H), 0.499 (1H), 1.529 (1H), 1.510 (1N)
2.0025(9)
2.0060(7)
2.0030(6)
2.0062(0)
2.0031(4)
2.0061(2)
2.0061(0)
2.0024(3)
2.0061(7)
ꢀ70
23
ꢀ70
the trapping of
a secondary
23
carbon-centred radical.
ꢀ70
23
23
Like 1, and at variance with 2
and 3, when sugar 4 was photo-
reacted with hexabutylditin at
low temperature it led to the
ꢀ70
23
Chem. Eur. J. 2009, 15, 8005 – 8014
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8009

E. Vismara et al.
The electroreduction of 2 and 3 afforded the 5’,5’-dimeric
compounds 8 and 13, respectively, along with the dispropor-
tionation products 9, 10 and 14, 15 and of the 6,6-dimers 7
and 12, respectively (see Schemes 9 and 10). The isolation of
Scheme 6. Rearranged radicals 2e and 3e.
detection of radical 4a (see Table 2). As was the case for 1,
no rearranged radical could be detected. Also the spin-trap-
ping behaviour of 4 paralleled that of 1, affording nitroxide
4c, the spectrum of which exhibited exchange between the
methylenic hydrogen atoms (kꢂ8.5ꢁ106 sꢀ1).
Discussion
Scheme 9. Mechanism of formation of 7–11 following electroreduction
of 2.
Relevant mechanistic considerations based on the cyclic vol-
tammetry and EPR studies rationalise the formation of all
the known and unknown products involved in the prepara-
tive electrolysis of 1–4.
The electroreduction of 1 and 4 afforded the 6,6-dimers 5
and 17, along with the reduction products 6 and 18, respec-
tively (see Schemes 7 and 8). The first reduction wave falls
at a similar potential for the two derivatives, and corre-
sponds to the formation of their radical anions, which quick-
ly undergo loss of an iodide anion, thereby affording the
corresponding primary radicals 1a and 4a. As already ob-
served by other authors, the reduction potential is slightly
shifted towards more negative values, as the potential scan
rate is increased.[25] In the case of 1 and 4, the second wave
is attributed to reduction of radicals 1a and 4a to the corre-
sponding carbanions 1b and 4b. Thus, on a qualitative basis,
the low intensity of the second waves at lower scan rates ob-
served for 1 and 4 is attributed to rapid dimerisation of 1a
and 4a to form dimers 5 and 17.
Scheme 10. Mechanism of formation of 12–16 following electroreduction
of 3.
these 5’,5’-dimers clearly indicates the occurrence of a rear-
rangement of the initially formed primary radicals 2a and
3a to the corresponding isomers 2e and 3e. The mainte-
nance of the C5’ d-glucose configuration in 8 and 13 indi-
cates that 2e and 3e remain in an sp3 orbital, thus affording
only one product with their dimerisation. In addition, the
isolation of 11a and 11b confirms the rearrangement of 2a
to 2e, further reduced to the carbanion 2d, followed by ace-
tate anion elimination. Similarly 3e, reduced to 3d, elimi-
nates acetate anion to afford 16.
Scheme 7. Mechanism of formation of 5 and 6 following electroreduction
of 1.
The different chemical behaviour of 2 and 3 with respect
to that of 1 and 4, was also paralleled by different voltam-
metric behaviour (see Figure 1). The first wave is still to be
associated with the formation of the radical anions of the
starting compounds, whereas the second wave, almost invisi-
ble at low scan rate, is due to the reduction of radicals 2a
and 3a to the carbanions 2b and 3b. The third wave at
Scheme 8. Mechanism of formation of 17 and 18 following electroreduc-
tion of 4.
8010
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 8005 – 8014

Characterisation of 6- and 6’-Iodomaltose and 6-Iodomaltotriose
FULL PAPER
more negative potential is instead believed to reflect the fur-
ther reduction of the rearranged radicals 2e or 3e to the car-
banions 2d or 3d. Of course, as the scan rate grows, so does
the intensity of the second wave, with the reduction of 2a or
3a competing more favourably with the rearrangement to
2e or 3e.
Radical 1a could be observed and characterised (see
Table 2) by means of EPR spectroscopy upon photolysis at
low temperature of a solution of iodosugar 1 in the presence
of hexabutylditin. As the temperature was raised, the spec-
tral intensity decreased until complete disappearance of the
signal above ꢀ308C.
The voltammetric behaviour of 4 parallels that of 1, and
does not deserve special comment. Also, the EPR spectrum
observed when photolysing at low temperature a solution of
4 containing Bu6Sn2 was formally similar to that of 1a, and
consistent with the formation of the corresponding primary
radical 4a. Besides, although photolysis of 4 at room tem-
perature did not lead to the observation of any EPR spec-
troscopic signal, as it was the case for 1, the spectrum of the
nitroxide 4c (see Table 2) resulting from the trapping of 4a
was observed in the presence of MNP.
The failure to observe the EPR spectra of radicals 2a and
3a when photolysing 2 and 3 at low temperature in the pres-
ence of Bu6Sn2 seems consistent with a rapid rearrangement
of these species taking place and is in line with the afore-
mentioned mechanism (Schemes 9 and 10) suggested to ac-
count for the isolation of 5,5-dimers from both sugars. Al-
though the observed spectra have been attributed to the re-
arranged radicals 2e and 3e, the tentative assignment is
mainly based on the nature of the isolated products, the
spectral parameters being also consistent with other radicals
of different structure. Despite their possibly fast rearrange-
ment, the primary radicals 2a and 3a are nevertheless trap-
ped by MNP to give nitroxides 2c and 3c that are structural-
ly similar to those afforded by halo sugars 1 and 4. It is also
worth noting that MNP does not apparently trap the rear-
ranged radicals 2e and 3e, although the spectra of the re-
sulting di-tert-alkyl nitroxides might in principle be hidden
by the intense signal of di-tert-butyl nitroxide (see Fig-
ure 2b), a species always present in the photoreactions in-
volving MNP.
Again, we attribute this to the dimerisation process be-
coming fast enough to lower the steady-state concentration
of 1a below the instrumental detection threshold. If, on the
other hand, an efficient spin trap (for example, MNP) is
added to the solution, radical 1a is scavenged before being
able to dimerise even at room temperature and nitroxide 1c
is readily observed. It should be noted that the rate con-
stants for the trapping of most carbon-centred radicals by
MNP fall in the range between 106 and 107 mꢀ1 sꢀ1.[26] The
magnetic non-equivalence of the two methylenic hydrogen
atoms in nitroxide 1c might be explained in two ways. The
first is that the two hydrogen atoms, being adjacent to a
chiral centre, are diastereotopic and hence magnetically dif-
ferent under all conditions. As an alternative, the difference
of the two methylenic hydrogen atoms might be explained
admitting a hindered, actually almost frozen-out rotation of
the tert-butylnitroxyl fragment about the newly formed
carbon–nitrogen bond. In the case of nitroxide 1c, we do
not see any way to discriminate between these two possibili-
ties. The situation is somehow different for nitroxides 2c, 3c
and 4c in the spectra in which line-width alternation effects
are clearly evident. For two diastereotopic hydrogen atoms
to give rise to line-width alternation, their molecule must
exchange between two conformations corresponding to two
energy minima, each corresponding to different values of
the hydrogen coupling constants; each hydrogen would then
exchange with itself and the line-width broadening would
result from this difference. The fact that the first and last
line of the spectra do not broaden suggests that the varia-
tion of the couplings of one hydrogen is exactly counterbal-
anced by that of the other hydrogen, and in a limiting situa-
tion two isomers should be detected. It seems to us more
likely that the line-width alternation observed in the spectra
of 2c, 3c and 4c is due to the above-mentioned rotation of
Conclusion
Electroreduction of halo sugars on silver cathodes provided
synthetic mixed O/C-malto-oligosaccharides: for example,
three maltotetraose and two maltohexaose mimics coming
from CH2I reducing and non-reducing end-unit precursors,
characterised by a central interglycosydic carbon–carbon
bond instead of the usual O-glycosidic linkage present in
natural and a large part of synthetic oligosaccharides.
Experimental Section
ꢀ
the tert-butylnitroxyl fragment about the N C6 bond.
Materials: All commercially available reagents, including dry solvents,
were used as received. Organic extracts were dried over sodium sulfate,
filtered and concentrated under vacuum by using a rotary evaporator.
All compounds were dried under high vacuum. Reactions were moni-
tored by thin-layer chromatography on precoated Merck silica gel 60
F254 plates and visualised by staining with 10% H2SO4 in water and with
a solution of cerium sulfate (1 g) and ammonium heptamolybdate tetra-
hydrate (27 g) in water (469 mL) and concentrated sulfuric acid (31 mL).
Flash chromatography was performed on Fluka silica gel 60.
Indeed, it seems unsurprising to us that such rotation is
blocked in 1c, in which the sugar fragment is a three-fused-
ring rigid moiety that may cause severe hindrance to rota-
tion and conformational change. As for 2c, 3c and 4c, the
conformational freedom might facilitate the rotation despite
the larger size of the sugar moiety, and it would appear that
the location of the methylenic group, that is, on the reducing
end or non-reducing end of the sugar, may significantly
affect this rotation.
Analytical equipment and characterisation methods: Melting points were
measured with a 10ꢁ magnification Reichert microscope equipped with
a heating plate and are not corrected. Optical rotations were measured
using a Dr. Kernichen Propol digital automatic polarimeter.
Chem. Eur. J. 2009, 15, 8005 – 8014
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8011

E. Vismara et al.
1H and 13C NMR spectra were recorded at 303 K using Bruker Avance
500 and 600 spectrometers equipped with a 5 mm 1H/X inverse probe
and with a TCI cryoprobe, respectively. Assignments were made through
HMQC, COSY and TOCSY experiments.
63.7, 68.7, 69.2, 70.0, 70.1, 71.0, 71.7, 72.4, 73.1, 75.5, 91.5, 96.3, 96.4,
169.6–171.2 ppm (CH3CO).
Electrochemical equipment and general procedures: Cyclic voltammetry
and preparative electrolysis experiments were carried out using an Amel
2053 potentiostat coupled with an Amel 7800 interface (electrochemical
measurements managing software Junior Assist). Supporting electrolyte:
tetraethylammonium tetrafluoborate (TeaTfb). Solvent: commercial an-
hydrous acetonitrile. Electrodes: reference saturated calomel electrode
(SCE), silver cathode (2 mm diameter) and platinum wire (cyclic voltam-
metry), cathode and anode silver plates 1 cmꢁ3 cm (preparative electrol-
ysis). Cells: 50 mL homemade flask (cyclic voltammetry), sintered glass
diaphragm two-compartment cell (preparative electrolysis).
Analysis was made by means of ESI-Q-TOF mass spectrometry. The ac-
quisitions were performed by direct infusion of the sample solution (pre-
pared at the appropriate concentration in CH3CN) sprayed at 4 mLminꢀ1
in the ionisation source of an ESI-Q-TOF mass spectrometer (Bruker
Daltonics) operating in positive mode on the mass range from m/z 50 to
2000 (instrumental conditions: capillary voltage 4500 V, nebuliser gas
0.4 bar, dry gas 4.0 Lminꢀ1, dry temperature 1808C). Mass calibration
was performed using sodium formate clusters.
Analytical HPLC were performed using a Hypersil Column BDS C18
(250ꢁ4.6 mm). Rheodine valve volume: 20 mL. Eluent: 55:45 acetoni-
trile/water. Flow rate: 1.5 mLminꢀ1. UV detection at 210 nm. Samples
were dissolved in acetonitrile at the concentration of 1 mgmLꢀ1; 20 mL
were injected. Semi-preparative HPLC was performed under the same
conditions using an Hypersil BDS C18 250ꢁ10 mm column from Thermo
Cyclic voltammetry experiments: Cyclic voltammetry experiments were
run at 208C in a thermostat-controlled cell with 5 mm iodosugar. Tetrae-
thylammonium tetrafluoborate (TeaTfb; 0.1 mm) was used as supporting
electrolyte, in anhydrous acetonitrile with silver cathode, platinum anode
and SCE (saturated calomel electrode) as reference electrode.
Preparative electrolysis experiments: All the electrolyses were run in a
two-compartment cell, the anode and cathode being divided by a glass
frit. Cathode and anode were both silver plates, and the anode and cath-
ode electrolyte solutions consisted of TeaTfb (0.1m) in anhydrous aceto-
nitrile, and were briefly pre-electrolysed between ꢀ1.3 and ꢀ2.2 V until
the obtainment of a low and constant current intensity. After pre-elec-
trolysis, iodosugar was added to the cathode solution and extensively
electrolysed under N2 in potentiostatic conditions on the base of voltam-
metric measurements.
Hypersil, a Rheodine valve of 100 mL and a flow rate of 5 mLminꢀ1
.
Samples were dissolved in acetonitrile at the concentration of 50–
100 mgmLꢀ1; 70–100 mL were injected.
EPR equipment and methods: EPR spectra were recorded by means of
an upgraded ER200D-ESP300 Bruker X-band EPR spectrometer operat-
ing at a frequency of approximately 9.3 GHz. The spectrometer was
equipped with a standard variable temperature accessory, a dedicated
computer for the storage and manipulation of data, an NMR spectro-
scopic gaussmeter for the calibration of the magnetic field and a frequen-
cy counter for the determination of g factors that were corrected with re-
spect to the value of perylene radical cation in concentrated sulfuric acid
(2.00258). In a typical experiment, a thoroughly degassed solution of the
halo sugar and of hexabutylditin in cyclopropane (THF or toluene) was
cooled at the desired temperature inside the cavity of the spectrometer
and irradiated with the unfiltered light of a 1 kW high-pressure mercury
lamp. For spin-trapping experiments, a very small amount of MNP was
also added to the sample. Typical instrumental settings were as follows:
modulation amplitude 0.025 mT, receiver gain 1ꢁ105, power 1 mW, scan
width 10 mT. The experimental spectra were computer simulated using
custom-made software based on a Monte Carlo minimisation proce-
dure.[27]
Electrolysis of 2: After pre-electrolysis, 2 (1.76 g, 2.46 mmol) was added
to the cathode solution (70 mL) and exhaustively electrolysed under N2
at ꢀ1.35 V (5 h) and at ꢀ1.45 V (1 h). The solution at the cathode com-
partment was concentrated and precipitated with AcOEt. The solid sup-
porting electrolyte was filtered off. The filtrate was evaporated at re-
duced pressure and the residue was subjected to flash chromatography
on silica gel with a gradient of hexane/ethyl acetate. A ratio of 4:6
hexane/acetate is sufficient to elute and separate 9, 10, 11a and 11b,
whereas the elution of pure 7 and 8 needs a 3:7 ratio.
Electrolysis of 3: After pre-electrolysis, 3 (300 mg, 0.29 mmol) was added
to the cathode solution (10 mL) and exhaustively electrolysed under N2
at ꢀ1.30 V (3 h). The solution at the cathode compartment was concen-
trated and precipitated with AcOEt. The solid supporting electrolyte was
filtered off. The filtrate was evaporated at reduced pressure and the resi-
due was subjected to flash chromatography on silica gel with a gradient
of hexane/ethyl acetate. A gradient from 1:1 to 3:7 eluted 14, 15 and 16.
A further gradient from 2:8 to 1:9 was necessary to elute 12 and 13 in
mixture. The mixture of 12 and 13 was separated by semi-preparative
HPLC.
Synthesis of starting iodosugars 2–4: Compounds 2 and 4 were prepared
according to literature procedures and had melting points and spectral
data in agreement with those reported in the literature. Compound 3 was
characterised by 1H and 13C NMR spectra as they are not present in the
literature.
Acetobromomaltose and acetobromomaltotriose were prepared from
commercial maltose monohydrate and maltotriose.[28]
They were converted to the corresponding peracetylated 1,6-anydromal-
tose and 1,6-anydromaltotriose, precursors of the peracetylated 6-formyl-
maltose and 6-formyl-maltotriose, respectively.[19] Iodination of the pera-
cetylated 6-hydroxyl-maltose and 6-hydroxylmaltotriose, coming from the
formyl group hydrolysis, afforded 2 and 3, respectively.[29]
Electrolysis of 4: After the pre-electrolysis, 4 (200 mg, 0.27 mmol) was
added to the cathode solution (10 mL) and was exhaustively electrolysed
under N2 at ꢀ1.30 V (1 h) and at ꢀ1.35 V (2 h). The solution at the cath-
ode compartment was concentrated and precipitated with AcOEt. The
solid supporting electrolyte was filtered off. The filtrate was evaporated
at reduced pressure and the residue was subjected to flash chromatogra-
phy on silica gel (3:7 hexane/acetate) to separate the more polar 17 from
the less polar 18.
Maltose was first protected as 4’,6’-benzylidene maltose and was then
fully acetylated. Debenzylation gave peracetylated 4’,6’-hydroxyl maltose,
which was protected at the 6’-position with the trityl group and acetylat-
ed at the 4’-position.[30] After detritylation, 6’-hydroxyl peracetylated mal-
tose was iodinated to 4.[29]
Compound 3: 1H NMR (500 MHz, CDCl3; the three rings are termed A,
B and C for convenience, with ring C bearing I): d=2.14–1.98 (s;
CH3CO), 3.37 (dt, J1 =3.4 Hz, J2 =9.2 Hz, 1H), 3.47 (dd, J6aC,5C =4.0 Hz,
1,2,3,2’,3’,4’,6’-Hepta-O-acetyl-6-deoxy-a-maltos-6-yl dimer (7): M.p. 165–
1668C; [a]D =+98.2 (c=1 in CH3Cl); ESI-Q-TOF MS: m/z calcd for
C52H70O34Na1: 1261.3641; found: 1261.3602 [M+Na]+.
1,2,3,2’,3’,4’,6’-Hepta-O-acetyl-6-deoxy-6-methyl-a-maltos-5’-yl dimer (8):
M.p. 172–1738C; [a]D =+112.6 (c=1 in CHCl3); ESI-Q-TOF MS: m/z
calcd for C52H70O34Na1: 1261.3641; found: 1261.3631 [M+Na]+.
J
6aC,6bC =11.3 Hz; H-6aC), 3.61 (dd, J6bC,5C =3.2, J6bC,6aC =11.3 Hz; H-6bC),
3.89–3.98 (m, 4H), 4.02 (dd, J1 =2.6 Hz, J2 =12.5 Hz, 1H), 4.21 (dd, J1 =
3.5 Hz, J2 =12.3 Hz, 1H), 4.23 (dd, J1 =3.0 Hz, J2 =12.3 Hz, 1H), 4.57
1,2,3,2’,3’,6’,2’’,3’’,4’’,6’’-Deca-O-acetyl-6-deoxy-a-maltotrios-6-yl
dimer
(12): ESI-Q-TOF MS: m/z calcd for C76H102O50Na1: 1837.5331; found:
1837.5390 [M+Na]+.
(dd, J1 =1.9 Hz, J2 =12.3 Hz, 1H), 4.73 (dd,
10.3 Hz; H-2B), 4.84 (dd, J2A,1A =4.0 Hz, J2A,3A =10.6 Hz; H-2A), 4.94
(dd, J2C,1C =8.0 Hz, J2C,3C =8.9 Hz; H-2C), 5.04 (t, J4A,3A =9.8 Hz, J4A,5A
9.8 Hz; H-4A), 5.27–5.39 (m, 5H), 5.80 ppm (d, 1C,2C =8.0; H-1C);
13C NMR (125 MHz, CDCl3): d=7.2 (CH2I), 21.2–21.6 (CH3CO), 62.1,
J2B,1B =4.0 Hz, J2B,3B =
1,2,3,2’,3’,6’,2’’,3’’,4’’,6’’-Deca-O-acetyl-6-deoxy-6-methyl-a-maltotrios-5’-yl
dimer (13): M.p. 148–1498C; [a]D =+124.3 (c=1 in CH3Cl); ESI-Q-TOF
MS: m/z calcd for C76H102O50Na1: 1837.5331; found: 1837.5401 [M+Na]+.
=
J
8012
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 8005 – 8014

Characterisation of 6- and 6’-Iodomaltose and 6-Iodomaltotriose
FULL PAPER
Table 3. NMR spectroscopic characterisation of new mixed O/C-maltotetraose and -maltohexaose mimics 7, 8, 12, 13 and 17.[a]
7
8
12
13
17
1H
13C
1H
13C
1H
13C
1H
13C
1H
13C
d
J
d
d
J
d
d
J
d
d
J
d
d
J
d
H/C-1A
H/C-2A
H/C-3A
H/C-4A
H/C-5A
5.64
4.1
10.6
9.7
10.1
2.2
98.0
73.5
72.2
71.3
71.7
5.42
5.41
6.01
6.04
–
3.5
8.7
9.8
–
100.4
73.3
71.6
72.2
76.8
5.663
5.15
5.99
5.43
4.51
3.9
10.7
9.7
9.7
2.4
99.0
73.4
72.4
71.6
71.87
5.58
5.32
5.71
5.46
4.53
4.3
9.5
9.5
9.6
2.5
100.2
72.8
73.6
70.6
72.5
5.312
4.77
5.277
4.82
–
3.8
10.5
9.8
9.8
–
95.6
70.6
69.6
70.5
69.0
5.10
5.93
5.49
4.25
–
3.5
3.4
2.8
H/C-6A
4.32
4.73
5.73
5.13
5.45
3.61
2.74
12.0
64.7
4.30
5.22
5.93
5.35
5.48
3.38
3.54
12.6
66.4
4.43
4.48
5.64
4.90
5.93
4.18
4.28
12.2
64.5
4.43
4.50
5.69
5.21
5.81
5.08
–
12.7
63.8
1.61
1.67
5.78
4.98
5.302
3.94
3.86
n.d.
23.4
H/C-1 B
H/C-2 B
H/C-3 B
H/C-4 B
H/C-5 B
8.5
9.4
9.2
9.2
n.d.
94.7
74.7
78.7
77.9
76.9
8.3
9.3
9.0
10.3
5.0
94.7
74.3
76.4
84.9
75.2
4.1
10.5
9.0
9.7
2.4
97.4
73.9
75.4
75.0
71.81
3.5
8.8
8.6
–
100.3
73.4
75.8
79.4
76.9
8.1
8.1
9.1
9.1
2.4
91.4
71.1
75.0
73.1
73.2
–
4.1
H/C-6 B
1.63
1.67
n.d.
30.6
1.59
–
20.3
4.61
4.82
5.89
5.24
5.665
3.73
3.07
12.3
66.3
4.60
5.49
5.99
5.38
5.52
3.62
3.60
12.0
66.4
4.14
4.43
12.2
63.2
H/C-1 C
H/C-2 C
H/C-3 C
H/C-4 C
H/C-5 C
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
9.5
9.5
9.2
9.2
95.3
74.8
78.8
77.8
78.0
8.4
9.4
9.2
–
94.7
74.5
77.4
84.3
74.7
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
>10.0
2.8
–
H/C-6 C
–
–
–
–
–
–
–
–
1.72
1.87
1.6–1.9
n.d.
23.2
1.63
–
–
20.6
–
–
–
–
A/B/C
1.7–2.1
22–24
1.6–2.3
22–24
–
22–24
1.6–2.4
22–24
1.8–2.2
22–24
MeOꢀacetyl
[a] Chemical shifts (d) in ppm; coupling constants (J) in Hz; n.d.=not determined.
1,2,3,6,2’,3’,4’-Hepta-O-acetyl-a-maltos-6’-yl dimer (17): M.p. 121–1228C;
[a]D =+158.2 (c=1 in CH3Cl); ESI-Q-TOF MS: m/z calcd for
C52H70O34Na1: 1261.3641; found: 1261.3638 [M+Na]+.
(d, 1H; H-1A), 5.44 (d, 1H; H-3A), 5.62 ppm (d, J1B,2B =8.4 Hz, 1H; H-
1B); 13C NMR (125 MHz, CDCl3): d=17.3 (C-6B), 19.4–21.8 (CH3CO),
62.0 (C-6A), 69.3 (C-2A), 71.2 (C-5B, C-2B), 75.0 (C-3B), 76.3 (C-4B),
93.6 (C-1B), 95.4 (C-1A), 99.3 (C-4A), 169.1–170.4 ppm (CH3CO); ESI-
Q-TOF MS: m/z calcd for C24H32O15Na1: 583.1633; found: 583.1690
[M+Na]+.
Table 3 reports full details of the NMR spectroscopic characterisation of
7, 8, 12, 13 and 17 in C6D6. As concerns compounds 8 and 13, the NMR
3
3
spectroscopy J2B,3B and J3B,4B coupling constant values of the ring in-
1
4
Compound 16: H NMR (500 MHz, CDCl3, the three rings are termed A,
volved in C5’–C5’ bonds indicate the maintenance of the original C1 ring
B and C for convenience): d=1.36 (d, J6C,5C =6.1 Hz; CH3, H-6C), 1.90–
conformation. In addition, the [a]D values support a maintenance of the
C5’ d-glucose configuration, instead of the possible epimerisation of this
residue to the l-ido configuration. In fact, [a]D values of the tetramers in-
crease from +63.0, the original value of b-octaacetylmaltose, to +98.2
for 7, due to the effect of chain elongation, and up to +112.6 for tetra-8.
Whereas for the hexasaccharides, the [a]D value of +86 for b-undecaace-
tylmaltotriose increases to +124.3 for 13.
2.20 (m; CH3CO), 3.664 (m, J6C,5C =6.1 Hz; H-5C), 3.757 (dd, J4C,5C
4C,3C =9.2 Hz; H-4C), 4.090 (m; H-5A), 4.109 (m, J6bA,5A =2.4 Hz; H-
6bA), 4.154 (dd, J6aA,5A =4.9 Hz, J6aA,6bA =12.6 Hz; H-6aA), 4.280 (dt,
4B,3B =9.4 Hz, J4B,6B =1.5 Hz; H-4B), 4.767 (dd, J2B, 1B =3.8 Hz, J2B,3B
=
J
J
=
10.4 Hz; H-2B), 4.788 (dd, J2A,1A =3.8 Hz, J2A, 3A =10.8 Hz; H-2A), 4.840
(t, J6aB,4B =1.5 Hz; H-6aB), 4.946 (dd, J2C,3C =9.3 Hz, J2C,1C =8.4 Hz; H-
2C), 5.012 (t, J4A,5A =J4A,3A =9.8 Hz; H-4A), 5.027 (t; H-6bB), 5.214 (dd,
1
Compound 11a: H NMR (500 MHz, CDCl3; the two rings are termed A
J
2C,3C =J3C,4C =9.3 Hz; H-3C), 5.363 (t, J2B,3B =J3B,4B =9.8 Hz; H-3B), 5.378
(d, J1A,2A =3.7 Hz; H-1A), 5.417 (d, J1B,2B =3.6 Hz; H-1B), 5.448 (dd,
2A,3A =J3A,4A =9.8 Hz; H-3A), 5.69 ppm (d, 2C,1C =8.4 Hz; H-1C);
and B for convenience, exocyclic double bond between 5’ and 6’ in A):
d=1.28 (d, J6B,5B =6.1 Hz; CH3, H-6B), 1.98–2.11 (s; CH3CO), 3.60 (m;
H-5B), 3.71 (t, J4B,3B =J4B,5B =9.1 Hz; H-4B), 4.50 (t, J6aA,6bA =2.0 Hz,
J
J
13C NMR (125 MHz, CDCl3): d=19.3 (C-6C; CH3), 20.0–21.5 (CH3CO),
61.47 (C-6A), 68.03 (C-4A), 68.47 (C-5A), 69.24 (C-3A), 70.08 (C-3B),
70.158 (C-2B), 70.359 (C-2A), 71.26 (C-2C), 71.7 (C-5C), 71.99 (C-4B),
75.34 (C-3C), 75.68 (C-4C), 91.09 (C-1C), 95.30 (C-1A), 91.50 (C-1B),
95.7 (C-1C’), 99.97 (C-6B, C=CH2), 150.99 (quaternary C-5B, C=CH2),
169.0–170.6 ppm (CH3CO); ESI-Q-TOF MS: m/z calcd for C36H48O23Na1:
871.2479; found 871.2497 [M+Na]+. Quaternary C-5B shows in the
HMBC spectrum correlation peaks with H-3B, H-1B, H-6aB, H-6bB and
H-4B. In the HMBC spectrum, H-1C shows a correlation peak with C-O,
H-1B shows with C-4C, H-4B with C-1A.
J
6aA,4A =1.8 Hz; H-6aA), 4.67 (t, J6bA,6aA =2.0 Hz, J6bA,4A =1.8 Hz, H-6bA),
4.83 (dd, 1A,2A =3.8 Hz, 3A,2A =10.0; H-2A), 4.89 (t, 1B,2B =J3B,2B
J
J
J
=
9.2 Hz; H-2B), 5.18 (t; H-3B), 5.33 (t, J3A,4A =10.0 Hz; H-3A), 5.42 (dd;
H-4A), 5.42 (d; H-1A), 5.64 ppm (d, J1B,2B =8.2 Hz; H-1B); 13C NMR
(125 MHz, CDCl3): d=18.9 (C-6B), 19.4–21.8 (CH3CO), 68.2 (C-3A),
68.6 (C-4A), 69.8 (C-2A), 71.2 (C-2B), 71.6 (C-5B), 75.5 (C-3B, C-4B),
91.2 (C-1B), 95.2 (C-1A), 97.9 (C-6A), 150.8 (C-5A), 169.1–170.4 ppm
(CH3CO); ESI-Q-TOF MS: m/z calcd for C24H32O15Na1: 583.1633; found:
583.1689 [M+Na]+.
1
Compound 11b: H NMR (500 MHz, CDCl3, the two rings are termed A
and B for convenience, endocyclic double bond between 4’ and 5’ in A):
d=1.22 (J6B,5B =6.1 Hz; CH3, H-6B), 1.98–2.11 (s; CH3CO), 3.55 (m, 1H;
H-5B), 3.64 (t, J4B,3B =J4B,5B =9.3 Hz; H-4B), 4.41 (d, J6aA,6bA =13.3 Hz,
Acknowledgements
1H; H-6aA), 4.37 (d, J6bA,6aA =13.3 Hz, 1H; H-6bA), 4.87 (dd, J1A,2A
=
3.0 Hz, J3A,2A =8.6 Hz, 1H; H-2A), 4.89 (t, J1B,2B =J3B,2B =9.2 Hz, 1H; H-
2B), 4.95 (d, J=2.2 Hz, 1H; H-4A), 5.22 (t, J=9.3 Hz, 1H; H-3B), 5.38
The authors would like to acknowledge partial financial support from
Fifth Framework Programme (FP5)/Quality of Life and Management of
Chem. Eur. J. 2009, 15, 8005 – 8014
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8013

E. Vismara et al.
Living Resources, contract nos. QLK3-CT-2002-02049 (“Heparanase In-
hibitors in Antiangiogenic and Antimetastatic Cancer Therapy”) and
MIUR-FIRB2001 (“Glycosaminoglycans Elastase and Selectins Antago-
nist as Anti-Inflammatory Drugs”). We also thank Dr. Roberto Santar-
siero, Dr. Marco Guerrini and Dr. Sara Guglieri of the Ronzoni and
NMR Centre staff, as well as Dr. Stefano Elli and Dr. Daniel Pizzolato
of “G. Natta” for their contribution to the Experimental Section.
Canales, M. Jimꢂnez, J. L. Roldan, A. Garcꢃa-Herrero, L. Iturrino,
J. L. Asensio, F. J. CaÇada, A. Romero, H.-C. Siebart, S. Andrꢄ, D.
Solꢃs, H.-J. Gabius, J. Jimꢂnez-Barbero, Biochim. Biophys. Acta Gen.
Subj. 2001, 1568, 225–236; c) A. Canales, J. Angulo, R. Ojeda, M.
Bruix, R. Fayos, R. Lozano, G. Gimenez-Gallego, M. Martin-Lomas,
P. M. Nieto, J. Jimꢂnez-Barbero, J. Am. Chem. Soc. 2005, 127, 5778–
5779; d) A. Almond, B. O. Petersen, J. Ø. Duus, Biochemistry 2004,
43, 5853–5863.
[15] E. Vismara, G. Torri, A. Naggi, I. Vlodavsky, Patent Application
Number EP08160464, 2008.
[16] E. Vismara, A. Donna, F. Minisci, A. Naggi, N. Pastori, G. Torri, J.
Org. Chem. 1993, 58, 959–963.
[17] H. P. Wessel, M. Trumtel, R. Minder, J. Carbohydr. Chem. 1996, 15,
523–548.
[18] K. Takeo, T. Kuge, Carbohydr. Res. 1976, 48, 282–289.
[19] S. Cottaz, C. Apparu, H. Driguez, J. Chem. Soc. Perkin Trans. 1
1991, 2235–2241.
[1] A. Alberti, M. A. Della Bona, D. Macciantelli, F. Pelizzoni, G. Sello,
G. Torri, E. Vismara, Tetrahedron 1996, 52, 10241–10248.
[2] M. Guerrini, G. Torri, P. R. Mussini, S. Rondinini, E. Vismara,
Chem. Commun. 1998, 1575–1576.
[3] S. Rondinini, P. R. Mussini, G. Sello, E. Vismara, J. Electrochem.
Soc. 1998, 145, 1108–1112.
[4] A. Alberti, S. Bertini, M. Comoli, M. Guerrini, A. Mele, E. Vismara,
Tetrahedron 2000, 56, 6291–6297.
[20] R. Blattner, R. J. Ferrier, P. Prasit, Chem. Commun. 1980, 944–945.
[21] J. F. Stoddard in Stereochemistry of carbohydrates. Wiley-Inter-
science, New York, 1971.
[22] F. A. Neugebauer in Landolt-Bçrnstein Magnetic properties of free
radicals, New Series, Group II, Vol. 17, Part b (Ed.: H. Fischer),
Springer, Berlin, 1987.
[23] A. R. Forrester in Landolt-Bçrnstein Magnetic properties of free rad-
icals. New Series, Group II, Vol. 9, Part c (Eds.: H. Fischer, K.-H.
Hellwege), Springer, Berlin, 1979.
[24] A. Alberti in Landolt-Bçrnstein Magnetic properties of free radicals.
New Series, Group II, Vol. 26, Subvolume d (Ed.: H. Fischer),
Springer, Berlin, 2005.
[5] M. Guerrini, S. Guglieri, R. Santarsiero, E. Vismara, Tetrahedron:
Asymmetry 2005, 16, 243–253.
[6] A. Imberty, S. Perꢂz, Chem. Rev. 2000, 100, 4567–4588, and referen-
ces therein.
[7] a) C. R. Parish, C. Freeman, K. J. Brown, D. J. Francis, W. B.
Cowden, Cancer Res. 1999, 59, 3433–3441; b) C. R. Parish, W. B.
Cowden, Patent US006143730A.
[8] H. Gong, M. R. Gagne, J. Am. Chem. Soc. 2008, 130, 12177–12183.
[9] a) J. P. Carver, Pure Appl. Chem. 1993, 65, 763–770; b) P. K. Qasba,
Carbohydr. Polym. 2000, 41, 293–309.
[10] A. Laederach, P. J. Reilly, Proteins Struct. Funct. Bioinf. 2005, 60,
591–597, and references therein.
[25] C. P. Andrieux, I. Gallado, J.-M. Savꢂant, J. Am. Chem. Soc. 1989,
111, 1620–1626.
[26] T. J. Kemp, Prog. React. Kinet. Mech. 1999, 24, 287–358.
[27] M. Lucarini, B. Luppi, G. F. Pedulli, B. P. Roberts, Chem. Eur. J.
1999, 5, 2048–2054.
[28] I. Damager, C. E. Olsen, B. L. Moller, M. S. Motawia, Synthesis
2002, 418–426.
[29] B. Classon, Z. Liu, B. Samuelsson, J. Org. Chem. 1988, 53, 6126.
[30] K. Takeo, K. Shinmitsu, Carbohydr. Res. 1984, 133, 135–145.
[11] M. Seo, N. Castillo, R. Ganzynkowicz, C. R. Daniels, R. J. Woods,
T. L. Lowary, P.-N. Roy, J. Chem. Theory Comput. 2008, 4, 184–191.
[12] a) R. W. Armstrong, B. R. Teegarden, J. Org. Chem. 1992, 57, 915–
922; b) N. Navarre, A. H. van Oijen, G. J. Boons, Tetrahedron Lett.
1997, 38, 2023–2026; c) R. Alibꢂs, D. R. Bundle, J. Org. Chem. 1998,
63, 6288–6301; d) M. C. Galan, A. P. Venot, J. Glushka, A. Imberty,
G.-J. Boons, J. Am. Chem. Soc. 2002, 124, 5964–5973.
[13] J.-F. Espinosa, M. Martꢃn-Pastor, J. L. Asensio, H. Deitrich, M.
Martꢃn-Lomas, R. Schmidt, J. Jimꢂnez-Barbero, J. Am. Chem. Soc.
1996, 118, 10862–10871.
[14] a) J. Jimenꢂz-Barbero, J. L. Asensio, F. J. CaÇada, A. Proveda, Curr.
Opin. Struct. Biol. 1999, 9, 549–555; b) J. M. Alonso-Plaza, M. A.
Received: March 31, 2009
Published online: July 14, 2009
8014
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 8005 – 8014