Exploiting the therapeutic potential of 8-β- d -glucopyranosylgenistein: Synthesis, antidiabetic activity, and molecular interaction with islet amyloid polypeptide and amyloid β-peptide (1-42)

Article abstract of DOI:10.1021/jm501069h

8-β-d-Glucopyranosylgenistein (1), the major component of Genista tenera, was synthesized and showed an extensive therapeutical impact in the treatment of STZ-induced diabetic rats, producing normalization of fasting hyperglycemia and amelioration of exce

Full text of DOI:10.1021/jm501069h

Article
pubs.acs.org/jmc
Exploiting the Therapeutic Potential of
8‑β‑D‑Glucopyranosylgenistein: Synthesis, Antidiabetic Activity, and
Molecular Interaction with Islet Amyloid Polypeptide and Amyloid
β‑Peptide (1−42)
Ana R. Jesus,† Catarina Dias,† Ana M. Matos,†,‡ Rodrigo F. M. de Almeida,† Ana S. Viana,†
Filipa Marcelo,∥ Rogerio T. Ribeiro,‡,§ Maria P. Macedo,‡,§ Cristina Airoldi,⊥ Francesco Nicotra,⊥
́
Alice Martins,† Eurico J. Cabrita,∥ Jesus Jimenez-Barbero,# and Amelia P. Rauter*,†
́
́
́
†Center of Chemistry and Biochemistry, Department of Chemistry and Biochemistry, Faculdade de Cien
Ed C8, Piso 5, Campo Grande, 1749−016 Lisboa, Portugal
‡CEDOC Chronic Diseases Center, Faculdade de Cien
cias Med
1150-082, Lisboa, Portugal
§APDP, Diabetes Portugal Education and Research Center, APDP-ERC, Rua do Salitre, 118-120, 1250-203 Lisboa, Portugal
∥REQUIMTE, CQFB, Department of Chemistry, Faculdade de Cien
cias e Tecnologias, Universidade Nova de Lisboa, Quinta da
Torre, 2829-516 Caparica, Portugal
⊥Department of Biotechnology and Biosciences, University Milano Bicocca, Piaza della Sciencia 2-4, 20126, Milano, Italy
#Centro de Investigaciones Biolog
icas, C.S.I.C., Ramiro de Maeztu 9, 28040 Madrid, Spain
̂
cias, Universidade de Lisboa,
̂
́ ̂
icas, Universidade Nova de Lisboa, Rua Cam
ara Pestana, 6, 6a,
̂
́
S
* Supporting Information
ABSTRACT: 8-β-D-Glucopyranosylgenistein (1), the major com-
ponent of Genista tenera, was synthesized and showed an extensive
therapeutical impact in the treatment of STZ-induced diabetic rats,
producing normalization of fasting hyperglycemia and amelioration
of excessive postprandial glucose excursions and and increasing β-
cell sensitivity, insulin secretion, and circulating insulin within 7 days
at a dose of 4 (mg/kg bw)/day. Suppression of islet amyloid
polypeptide (IAPP) fibril formation by compound 1 was
demonstrated by thioflavin T fluorescence and atomic force
microscopy. Molecular recognition studies with IAPP and Aβ1−42
employing saturation transfer difference (STD) confirmed the same
binding mode for both amyloid peptides as suggested by their
deduced epitope. Insights into the preferred conformation in the
bound state and conformers’ geometry resulting from interaction with Aβ1−42 were also given by STD, trNOESY, and MM
calculations. These studies strongly support 8-β-D-glucopyranosylgenistein as a promising molecular entity for intervention in
amyloid events of both diabetes and the frequently associated Alzheimer’s disease.
Amyloids from different diseases may share a common pathway
for fibril formation, since they have common structural
properties.2 In the case of type 2 DM and AD, studies have
also shown that both IAPP and Aβ small and soluble oligomers
are actually the most toxic aggregates to β-cells and neurons,
respectively, when compared to larger amyloid fibrils.3,4 Thus,
targeting these prefibrillar amyloidogenic proteins seems to be a
rational approach.
INTRODUCTION
■
Because of population growth, aging, lifestyle alterations, and
increasing prevalence of obesity, the past 2 decades have seen
an explosive increase in people diagnosed with Diabetes mellitus
(DM). Therefore, research on new medicines for the
prevention and treatment of this chronic disease remains
mandatory. Type 2 DM is caused by a combination of
resistance to insulin and impaired insulin secretion, a
dysfunction associated with islet amyloid deposits derived
from islet amyloid polypeptide (IAPP), a protein coexpressed
and secreted with insulin by pancreatic β-cells. This
accumulation of amyloid fibrils, also encountered in many
other age related degenerative diseases, namely, in Alzheimer’s
disease (AD), occurs as an outcome of protein misfolding.1
Polyphenolic molecular entities have demonstrated the
ability to inhibit the formation of β-amyloid fibrils in vitro
and their associated cytotoxicity.5 In particular, genistein has
Received: July 16, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jm501069h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
the ability to prevent fibrillization and oligomerization and has
fibril-destabilizing effects on Aβ1−40 and Aβ1−42 in vitro.6 In
addition, (−)-epigallocatechin 3-gallate (EGCG)7,8 and
1,2,3,4,6-penta-O-galloyl-β-D-glucopyranose (PGG)9,10 were
reported to inhibit in vitro amyloid formation of both IAPP
and Aβ peptides. A dietary supplementation of EGCG for 5
weeks resulted in reduction of blood glucose levels11 and has
proven to delay the onset of type I diabetes,12 while PGG was
found to ameliorate high-fat-diet-induced diabetes in C57BL/6
mice.13
Scheme 1. Synthesis of 8-β-D-Glucopyranosylgenistein (1)
The plant kingdom is indeed a valuable source of compounds
effective against amyloid disorders, in particular diabetes.14,15
Our previous results revealed that Genista tenera,16−21 including
its ethyl acetate extract, has a promising antidiabetic activity on
an STZ-induced diabetic animal model. In vitro toxicity studies
showed no evidence for extract acute cytotoxicity or
genotoxicity.21 The O-glycosylated flavonoid components
lowered blood glucose levels of STZ induced diabetic Wistar
rats,20 although their activity was not comparable to that of the
extract.21 Its major component was identified as 8-β-D-
glucopyranosylgenistein (1)17 which is not commercially
available. Hence, synthesis of this glucosylisoflavone is
presented here and evaluated for its antidiabetic activity and
toxicity.
a
(a) (1) NaH, DMF, 0 °C, 30 min; (2) BnBr, rt, 24 h, 87%; (b)
AcOH, H2SO4 1 M, 90−95 °C, 24 h, 81%; (c) Ac2O, Py, rt, 1 h, 96%;
(d) (1) K2CO3, DMF, 0 °C, 30 min; (2) BnBr, rt, 1 h, 69%; (e) R2 =
Ac, R3 = Bn, Sc(OTf)3, DCE, Drierite, −30 °C, 30 min, then rt, 5 h,
49%; (f) R2 = R3 = H; (1) TMSOTf, CH2Cl2−CH3CN (1:1), Drierite,
−40 °C, 30 min, then rt, 3 h, 56%; (2) K2CO3, DMF, 0 °C, 30 min,
then BnBr, rt, 1 h, 74%; (g) (1) 1,4-dioxane, aq NaOH 50% (w/v), 18
h, reflux; (2) Ac2O, Py, DMAP, 1 h, rt, 60%; (h) (1) TTN(III),
(MeO)3CH, MeOH, 24 h, 40 °C; (2) THF, MeOH, aq NaOH 50%
(w/v), 4 h, rt, 63%; (i) MeOH, EtOAc, Pd/C, H2, 6 h, rt, 96%.
RESULTS AND DISCUSSION
■
Chemistry. Preparation of 8-β-D-glucopyranosylgenistein
(1) started from 2,4-dibenzyloxy-6-hydroxyacetophenone,
whose synthesis was successfully accomplished in one single
step by dibenzylation of acetophloroglucinol in 69% yield
(Scheme 1) as opposed to the four-step procedure previously
reported by Kumazawa et al.22 Its C-glycosylation succeeded in
the presence of catalytic Sc(OTf)3 with perbenzylglucosyl
acetate. When compared to the previously described conditions
based on reaction of the glucosyl fluoride catalyzed by boron
trifluoride etherate,22 our methodology resulted in a reprodu-
cible, less expensive, and cleaner reaction, giving compound 8
through a Fries-type rearrangement in reasonable yield (49%).
In the alternative, access to 8 succeeded by C-glucosylation of
acetophloroglucinol with perbenzylglucosyl acetate catalyzed by
TMSOTf in 56% yield, followed by benzylation. Reaction of 8
with 4-benzyloxybenzaldehyde in aqueous NaOH 50% (w/v)
followed by acetylation afforded chalcone 9 in 60% overall
yield. These conditions have proven reproducible, while those
previously reported, based on NaOMe in MeOH solution, did
not result in the expected yield.23
The oxidative rearrangement of 9 with thallium(III) nitrate
(TTN) was performed in basic conditions with aqueous
solution of NaOH 50%, leading to the glucosylisoflavone 10 in
63% isolated yield. Debenzylation by catalytic hydrogenation
with Pd/C gave 8-β-D-glucopyranosylgenistein (1) in 96%
yield.
Biological Evaluation. Effect of 8-β-D-Glucopyranosylge-
nistein (1) Treatment on Diabetic Animals. Compound 1
showed an extensive therapeutical impact on an animal model
of diabetes characterized by β-cell failure, in which the
streptozotocin (STZ) induced diabetic Wistar rats were treated
with intraperitoneal (ip) injections of 1 (4 (mg/kg bw)/day)
for 7 days. Normalization of fasting hyperglycemia and a major
amelioration of excessive postprandial glucose excursions to
values close to those observed on healthy normal rats were
found, as assessed by an intragastric glucose tolerance test
(Figure 1). When compared to the known polyphenols that
inhibit IAPP fibrillization, namely, EGCG and PGG, compound
1 is a far more promising candidate. Indeed, a dosage of 100
(mg/kg bw)/day of EGCG for 2 weeks could only slightly
improve oral glucose tolerance in high-fat-induced diabetic
mice.11 However, at concentrations of 60−90 (mg/kg bw)/day,
this compound was capable of reducing the risk of type I
diabetes, which results from the autoimmune-mediated
destruction of pancreatic β-cells and deficient insulin
production.12 Additionally, PGG lowered blood glucose level
of high-fat-diet-induced diabetic mice only at a dose of 50 (mg/
kg bw)/day and after 21 days of treatment.13 On the seventh
day with the same dose only 50% reduction of diabetic mice
blood glucose levels were reached, corresponding to 37%
increased circulating glucose when compared to the normal
control, in contrast to the complete normalization achieved by
compound 1 at a dose of 4 (mg/kg bw)/day.
Treatment with compound 1 also increased glucose-induced
insulin secretion as estimated through C-peptide insulin
parameters, an effect also reported for EGCG.11 Also,
circulating insulin was enhanced (Figures S1 and S2 in
Supporting Information) when compared to the values
observed after administration of STZ resulting from pancreatic
β-cells destruction. Hence, β-cell sensitivity was clearly
ameliorated by compound administration (Figure S3).
ThT Fluorescence and AFM Monitoring of IAPP Fibril
Formation in the Presence 8-β-D-Glucopyranosylgenistein
(1). Rat IAPP amino acid sequence differs from that of humans
by the substitution of proline residues in the amyloidogenic
B
dx.doi.org/10.1021/jm501069h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 1. (A) Glycemic curves showing basal fasting and postload (2 mg glucose/kg) values for intragastric glucose tolerance test. (B) Area under
the curve (AUC) for the glycemic curves for all groups. 8G = 8-β-D-glucopyranosylgenistein. (∗∗) Differences were considered statistically significant
when p < 0.05.
Figure 2. (A) Effect of 1 on IAPP fibrillization, monitored by AFM (z = 10 nm): (top row) control, IAPP (25 μM) incubated for 0, 2, and 6 h at 37
°C without 1; (bottom row) IAPP (25 μM) incubated for 0, 2, and 6 h at 37 °C, in the presence of compound 1 (50 μM). (B) Inhibition of hIAPP
amyloid formation by 8-β-D-glucopyranosylgenistein (1), monitored by ThT fluorescence intensity. Thioflavin-T monitored kinetic experiments of
hIAPP are shown in the absence (full circles) or presence of 8-β-D-glucopyranosylgenistein (open circles). Arrows indicate the times at which the
AFM images were acquired (0, 2, and 6 h).
Figure 3. Section analysis of the fibrils and spherical aggregates observed in the control sample, by AFM.
region of the peptide.24 As a result, it does not form amyloid
fibrils or is toxic to β-cells, meaning that, interestingly,
compound 1 exerts its antidiabetic effect in rats by a mechanism
that does not relate to the inhibition of IAPP amyloid
aggregation. Still, considering that EGCG and PGG are
known to possess this ability, we were interested in
C
dx.doi.org/10.1021/jm501069h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 4. STD result for the 8-β-D-glucopyranosylgenistein (1)/IAPP mixture at 298 K along with the corresponding controls. The STD of the IAPP
oligomers at the irradiation point (δ −0.5 ppm) (a) was performed to discard the presence of any contaminant, which could interfere in the STD
spectrum. The STD of IAPP alone appears completely clean (a). No NMR signals were detected in the STD control of the compound 1 at the same
irradiation point (b). The STD experiment was performed on the 8-β-D-glucopyranosylgenistein/IAPP system with 25:1 molar ratio under identical
experimental conditions (c). Clear STD signals of 8-β-D-glucopyranosylgenistein (1) appeared on the spectrum (c, d).
investigating whether 8-β-D-glucopyranosylgenistein (1) can
indeed act as an antidiabetic in humans by a multitarget
mechanism of action, including the prevention of IAPP amyloid
oligomerization. Thus, the binding of compound 1 to IAPP
oligomers was investigated.
hallmark of AD. We then investigated the interaction of 1 with
amyloid oligomers. The molecular recognition studies of
compound 1 with IAPP and Aβ1−42 oligomers were conducted
employing saturation transfer difference (STD) NMR, a robust
method to detect the binding of ligands to a given receptor.27
Indeed, binding of chemically different molecules to Aβ1−42
oligomers has been characterized by STD.27a The proper
oligomeric state of IAPP and Aβ1−42 peptides to be employed
for STD experiments was generated using previously reported
conditions (see Supporting Information).28 Figure 4 shows the
STD for the 8-β-D-glucopyranosylgenistein/IAPP mixture along
with the corresponding controls, which appeared completely
clean (Figure 4a and Figure 4b). The STD experiment was then
performed on the 8-β-D-glucopyranosylgenistein/IAPP system
under identical experimental conditions (Figure 4c). Clear STD
signals of compound 1 appeared on the spectrum, in the
presence of IAPP (Figure 4c and Figure 4d) or Aβ1−42 (Figure
S5) oligomers. These results showed, in a nonambiguous way,
that 1 interacts with both amyloid peptides.
Analysis of the STD signals obtained for 1 allowed
determination of its binding epitope in the presence of IAPP
or Aβ1−42 peptide oligomers (Figure 5). STD intensities
(Figure 5) clearly indicated that the aromatic aglycone is
involved in the binding. Nevertheless, several sugar resonances
appeared, suggesting their participation in the molecular
recognition process. The deduced epitope is the same for
both amyloid peptides, indicating the same binding mode for
both amyloid aggregates. For comparison reasons also recorded
was a STD experiment of the genistein itself in the presence of
Aβ1−42 peptide oligomers under identical experimental
conditions of its glycosylated analog. Interestingly, the epitope
map of the genistein (Figure S6) is rather distinct when
glycosylated or not. Without the sugar all protons of the
aromatic molecule, genistein, receive now almost the same
percentage of saturation indicating that no particular proton is
in closest contact with the Aβ1−42 receptor. This result points
out that the introduction of glucose residue on genistein
structure induces a preferential binding mode of the aromatic
genistein to the Aβ1−42 oligomers.
Inhibition of IAPP fibril formation by 1 at a ratio of 1:2
(hIAPP/1) was assessed using atomic force microscopy (AFM)
and fluorescence of thioflavin T. The fluorescence results show
that at time zero there are no fibrils present, and after a lag
phase fibrils start to form (Figure 2B). As a negative control, rat
IAPP, which is known not to form fibrils, was used to confirm
that the protocol and experimental results are being correctly
interpreted, and indeed no changes in ThT fluorescence were
detected along the course of the entire experiment (Supporting
Information, Figure S4). Human IAPP predominantly
aggregates into fibrils with discrete heights of 2−3 nm,
although spherical species with heights ranging from 3 to 9
nm are also formed in the absence of compound 1 (Figures 2
and 3). These latter species, commonly referred as oligomers,
have been proposed to act as intermediates for fibril
formation.25 We have also detected the presence of some
elongated fibrils after 2 h that turned into complex multiple
fibrils within 6 h. This is consistent with the time dependence
of ThT fluorescence, which shows that, for this sample, large
fibrils started to form at ∼5 h. Conversely, after incubation of
the peptide with 1 for 6 h, IAPP fibrils were not detected
(Figure 2), and again AFM and ThT fluorescence results were
consistent. Although a minor amount of spherical aggregates
were still observed, their apparent density was significantly
reduced when compared to that imaged without compound 1
(control). Already at the initial stages, the addition of 1 resulted
in the decrease of the oligomers’ density prior to fibril
formation. These studies suggest that the binding of 1 results in
the inhibition of IAPP amyloid formation, even in their earlier
stages. These oligomers, as mentioned above, are thought to be
the most toxic type of aggregates.
NMR Binding Experiments with Aβ1−42 Oligomers and
hIAPP. Recent studies have pointed out multiple lines of
evidence linking the incidence of diabetes to the development
of Alzheimer’s disease.26 While type 2 DM is associated with
IAPP deposits, the accumulation of β-amyloid (Aβ) fibrils is a
In order to identify the bound conformation of 1 to the
amyloid oligomers, trNOESY experiments29 were carried out.
D
dx.doi.org/10.1021/jm501069h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
absence of NOEs between the glucose and aglycone moiety for
1 in solution suggests a conformational equilibrium of different
conformers, where no conformation is defined as the major
conformer in solution.
Insights on the geometry of these conformers were obtained
by using molecular mechanics (MM) calculations, with the
MM3* force field.30,31 Two conformations around the C1″−
C8 linkage were identified, which only differed in 2.6 kJ/mol.
The energy barrier for their interconversion was estimated in
only 9 kcal/mol, supporting their simultaneous presence in free
solution (see above). The global minimum (defined by an
antigeometry for the H1″−C1″−C8−C7 torsion angle) is
shown in violet in Figure 7, while the alternative local minimum
(with a syn-orientation) is displayed in yellow.
Figure 5. Epitope mapping obtained for 8-β-D-glucopyranosylgenistein
(1) with IAPP (black labels) and Aβ1−42 (red labels) peptide
oligomers. The atom numbering used in the NMR analysis is
displayed.
The NOESY spectrum of the free compound was first recorded
in the absence of amyloid oligomers as control (Figure 6a),
yielding positive NOE cross peaks. In contrast, strong and
negative NOEs were detected in the presence of Aβ1−42 (Figure
6b). This fact reflects an increase of the effective rotational
motion correlation time of the molecule in the presence of the
Aβ oligomers, further supporting the existence of binding
process of 1 to a large molecular entity,30 here represented by
the peptide oligomeric state. The orientation of the glucosyl
group with respect to the oligomer was deduced from
inspection of the trNOESY cross peaks of compound 1 in
the presence of Aβ1−42. Indeed, the trNOESY recorded at 310
K (Figure 6b) showed a cross peak between H2 of the aglycone
and H1″ (also H3″ and H5″, mediated by spin diffusion). This
cross peak was absent in the NOESY spectrum of the ligand in
the free state (Figure 6a). The existence of this cross peak
suggests that in the bound state the glucosyl group adopts a
preferential conformation whose α-face points toward H2 at
the fused bicyclic moiety of the aglycone. In contrast, the
Figure 7. Two conformations around the C1″−C8 linkage of 1. The
global minimum is shown in violet, while the local minimum is in
yellow. trNOE experimental contacts are highlighted showing that 8-β-
D-glucopyranosylgenistein (1) adopts the global minimum conforma-
tion in the presence of Aβ1−42 oligomers.
According to the trNOESY results, compound 1 preferen-
tially adopts the anti-orientation (structure in violet at Figure 7)
in the presence of Aβ1−42 oligomers. In contrast, it could be
assumed that both conformers (anti and syn) coexist in
solution.
Figure 6. (a) 400 MHz 2D-NOESY spectra of 1 (2 mM) with a mixing time of 0.8 s. (b) trNOESY of the mixture containing Aβ1−42 (80 μM) and 1
(2 mM), with a mixing time of 0.3 s. Both samples were dissolved in deuterated PBS, at pH 7.5 and 310 K. Positive cross peaks are in red, and
negative ones are in blue.
E
dx.doi.org/10.1021/jm501069h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
ferric chloride, and a 10% sulfuric acid in ethanol, followed by heating
at 120 °C, as visualization agents. Silica gel 60G (0,040−0,063 mm)
from Merck (Germany) was used for column chromatography (CC).
Nuclear magnetic resonance (NMR) spectra were recorded at room
temperature (rt), on a Bruker Advance 400 apparatus, using
tetramethylsilane (TMS) as an internal standard. Chemical shifts δ
are given in ppm, and NMR data are in full agreement with that given
in the literature.22,23 Purity of all compounds, >95%, was detected by
high resolution mass spectra, acquired on a Bruker Daltonics APEX
Ultra FT-ICR mass spectrometer (Billerica, MA, USA) equipped with
a 7.0 T actively shielded superconducting magnet from Magnex
Scientific. The ions were generated from an external Apollo II dual
ESI/MALDI source from Bruker Daltonics. Samples were introduced
by means of an infusion pump at a flow rate of 120 μL/h. The
nebulizer gas (N2) flow rate was set to 2.5 L/min, and the drying gas
(N2) flow rate was set to 4.0 L/min at a temperature of 220 °C. The
capillary voltage was set to 4100 V and the spray shield voltage was set
to 3700 V.
The sugar moiety is acknowledged to improve bioavailability
(e.g., by increasing water solubility) and targeting as well as
promoting blood−brain barrier (BBB) crossing.32,33 Thus,
given the structural features of glucosyl isoflavone (1), this
compound has potential to combine the biocompatible
properties of sugars and the well-known ability of polyphenols
to inhibit amyloid fibril formation by specific aromatic
interactions. Although genistein demonstrated fibril-destabiliz-
ing effects and the ability to prevent Aβ oligomerization,6 the
presence of the sugar moiety in compound 1 seems to be a
crucial structural feature for the complete aggregate suppres-
sion. A recent study showed that several polyphenol glycosides
are capable of dissociating Aβ oligomers into nontoxic soluble
peptides, in contrast to the respective aglycones, which convert
them into large off-pathway aggregates.34 Hence, this C−C
linked glucosylgenistein, which is not chemically or enzymati-
cally cleaved in vivo, is expected to play a similar role. In fact,
the STD experiments here presented demonstrated that the
presence of the sugar is able to tune the binding mode of the
interaction of genistein to the Aβ receptors. This seems to
occur in a synergistic combination of the aglycone and the
sugar moiety, where the aglycone initiates the remodeling
process by disrupting intermolecular contacts between Aβ
peptide through π-stacking interaction, whereas CH−π stacking
interactions driven by association of the CH sugar bonds with
π-electron cloud of the aromatic rings were suggested to cause
interaction of sugars with the newly exposed aromatic rings,
ultimately preventing Aβ residues from associating between
them. Both this study and our STD experimental results give
insights into the advantages of compound 1 when compared to
genistein in the suppression of amyloid fibril formation.
2,4-Dibenzyloxy-6-hydroxyacetophenone (7). To a solution
of acetophloroglucinol 6 (10.02 g, 59.6 mmol) in DMF was added
K2CO3 (2.2. equiv). After stirring for 10 min at 0 °C, BnBr (2.2. equiv)
was added and the mixture stirred for 1 h. HCl, 2 M, was added, and
the mixture was poured into water and extracted with EtOAc. Organic
layers were combined, washed with brine, dried over MgSO4, and
concentrated. Compound 7 was purified by CC (10:1 hexane/EtOAc)
in 69% yield (14.32 g, 41.12 mmol). Rf = 0.73 (4:1 petroleum ether/
EtOAc); mp 103.5−104.0 °C (lit. mp 108−109 °C);22 1H NMR
(CDCl3) δ = 14.22 (s, 1H; OH-8), 7.47−7.40 (m, 10H, CH-Ph), 6.22
3
3
(d, J(5,7) = 2.3 Hz, 1H; H-7), 6.15 (d, J(5,7) = 2.3 Hz, 1H; H-5),
5.09 (s, 2H; CH2Ph-4), 5.08 (s, 2H; CH2Ph-6), 2.61 (s, 3H; H-1); 13
C
NMR (CDCl3) δ 203.2 (C-2); 167.6 (C-6, C-8); 162.1 (C-4); 135.9
(Cq-4); 135.7 (Cq-6); 128.8, 128.8, 128.5, 128.4, 128.1, 127.7 (CH,
Ph); 106.3 (C-3); 94.8 (C-7); 92.4 (C-5); 71.2 (CH2Ph-6); 70.3
(CH2Ph-4); 33.4 (C-1). HRMS: calcd for [M + Na] 371.1254; found
371.1257.
2,4-Dibenzyloxy-6-hydroxy-5-(2,3,4,6-tetra-O-benzyl-β-D-
glucopyranosyl)acetophenone (8). Method A. 1-O-Acetyl-2,3,4,6-
tetra-O-benzyl-α-D-glucopyranose (5, 1.28 g, 2.20 mmol) and 7 (2.0
equiv) were dissolved in DCE (10 mL) in the presence of Drierite
(100 mg). The solution was stirred at −30 °C, and Sc(OTf)3 (0.25
equiv) was added. The stirring continued for 30 min at −30 °C and
then at room temperature for 5 h. The reaction was quenched with
water and filtered through Celite, extracted with DCM, and
concentrated. Compound 8 was purified by CC (10:1 petroleum
ether/EtOAc) and isolated in 49% yield as a syrup (0.94 g, 1.08
mmol).
Method B. To a solution of acetophenone (6, 5.96 mmol, 2.0 equiv)
dissolved in ACN (25 mL) was added a solution of compound 4 (1.61
g, 2.98 mmol) in DCM (25 mL). Drierite (100 mg) was added, and
the reaction was kept at −40 °C under N2 atmosphere. TMSOTf (0.5
equiv) was added to the reaction, and the cooling bath was kept for 30
min. The bath was removed and reaction reached room rt and was run
for 3 h. The reaction was quenched with NaHCO3, extracted with
DCM, washed with brine, dried over MgSO4, and concentrated in
vacuum. The desired product was purified by CC (7:1 hexane/ethyl
acetate) and isolated in 56% yield (1.15 g, 1.67 mmol). Rf = 0.21 (2:1
hexane/EtOAc); 1H NMR (CDCl3) δ = 7.34−6.98 (m, 20H; CH-Ph),
5.89 (s, 1H; H-7), 4.88 (d, 1H, J1′,2′ = 9.71 Hz, H-1′); 4.97−4.48 (m,
8H; CH2Ph), 3.81−3.60 (m, 6H; H-2′, H-3′, H-4′, H-5′, H-6a′, H-
6b′), 2.56 (s, 3H; H-1); 13C NMR (CDCl3) δ = 203.9 (C-2); 164.0
(C-6); 161.9 (C-4); 161.4 (C-8); 138.3, 137.7, 137.0, 136.6 (Cq, Ph);
128.6, 128.5, 128.3, 128.1, 127.7, 127.6 (CH, Ph); 105.7 (C-5);
102.4(C-3); 97.1 (C-7); 86.2 (C-2′); 81.7 (C-3′); 78.7 (C-4′); 77.4
(C-5′); 75.7 (C-1′); 75.9, 75.2, 74.8, 74.5 (CH2Ph); 68.1 (C-6′); 32.9
(C-1). HRMS: calcd for [M + Na] 713.2721; found 713.2735.
This C-glucosyl derivative was dissolved in DMF (23 mL). To this
solution K2CO3 (2.2 equiv) was added at 0 °C and stirred for 10 min
at 0 °C. Then BnBr (2.2 equiv) was added and the reaction was
allowed to reach rt and run for 2 h. After completion, the reaction was
neutralized with HCl, 2M, extracted with DCM, washed with brine,
The acute toxicity of 1 in eukaryotic cells was assessed by the
MTT cell viability assay. Compound 1 exhibited an IC50 value
of 1.25 mg/mL, almost 10 times higher than that of the
commercial drug chloramphenicol (0.143 mg/mL). The
potential genotoxic activity of 1 was also evaluated by scoring
chromosomal aberrations on metaphase spread, and no
evidence for genotoxic risk was found.21
CONCLUSION
■
8-β-D-Glucopyranosylgenistein (1) has been successfully
accessed by synthesis, produces normalization of fasting
hyperglycemia in STZ induced diabetic Wistar rats, and
interferes in glucose excursions increasing glucose-induced
insulin secretion and insulin sensitivity. This nontoxic
glucosylflavonoid is the most potent antidiabetic molecular
entity described so far that interacts with both toxic Aβ1−42 and
IAPP amyloid oligomers, preventing amyloid fibrillization
characteristic of diabetic patients. These properties encourage
its further exploitation toward innovative therapeutics to
circumvent the tendency of diabetic patients to develop AD.
Its use as a functional food ingredient will provide specific
health benefits for the prevention of functional and cognitive
decline, a major issue to succeed in promoting a healthy life and
an active aging.
EXPERIMENTAL SECTION
■
Chemistry. Reagents and solvents were purchased from Sigma-
Aldrich (Barcelona, Spain). All nonaqueous reactions were carried out
under an atmosphere of N2 using freshly distilled and dry solvents,
unless otherwise noted. Reactions were monitored by thin layer
chromatography (TLC), which was carried out on 0.25 mm silica gel
F254 plates (Merck, Germany) using UV light, a 5% ethanol solution of
F
dx.doi.org/10.1021/jm501069h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
dried over MgSO4, and concentrated. The product was isolated by CC
(6:1 hexane/EtOAc) in 74% yield (1.08 g, 1.24 mmol). Rf = 0.33 (4:1
petroleum ether/EtOAc); 1H NMR (CDCl3) δ = 14.11(s, 1H; OH-4),
7.43−6.89 (m, 30H; CH-Ph), 6.36 (s, 1H; H-7), 4.96−4.45 (m, 13H;
CH2Ph, H-1′), 3.74−3.37 (m, 6H; H-2′, H-3′, H-4′, H-5′, H-6a′, H-
6b′), 2.52 (s, 3H; H-1); 13C NMR (CDCl3) δ = 203.1 (C-2); 164.7
(C-4); 163.5 (C-6); 160.7 (C-8); 138.7, 138.6, 138.2, 138.1, 137.9,
137.4, 137.5, 136.4, 135.8; (Cq. Ph); 128.5, 128.5, 128.4, 128.3, 128.2,
128.1, 128.0, 128.0, 127.9, 127.8, 127.7, 127.6, 127.6, 127.6, 127.3,
126.2 (CH. Ph); 111.5 (C-5); 105.9 (C-3); 90.0 (C-7); 82.1 (C-3′);
78.8 (C-5′); 77.6 (C-4′); 77.4, 77.1, 76.7, 75.7, 75.7, 75.7, 75.3, 75.3,
75.0, 73.5, 73.4, 73.2 (CH2Ph); 72.8 (C-2′); 70.0 (C-1′); 68.4 (C-6′);
33.5(C-1). HRMS: calcd for [M + Na] 893.3660; found 893.3694.
(2E)-1-[2-Acetoxy-4,6-dibenzyloxy-3-(2,3,4,6-tetra-O-benzyl-
β-D-glucopyranosyl)]phenyl-3-(4-benzyloxyphenyl)pro-2-en-1-
one (9). To a solution of 8 (2.44 g, 2.80 mmol) and p-
benzyloxybenzaldehyde (1.5 equiv) in 1,4-dioxane (27.9 mL) was
added an aqueous solution of NaOH 50% (27.9 mL). The reaction
mixture was stirred in reflux for 24 h. After this time HCl, 2 M, was
added and the mixture was extracted with DCM, washed with brine,
dried over MgSO4, and concentrated. The residue was dissolved in
pyridine (10 mL/g residue) and acetic anhydride (2.0 equiv/OH).
The mixture was stirred for 30 min, and then pyridine was removed.
Compound 9 was isolated by CC (5:1 petroleum ether/EtOAc) as a
syrup in 60% overall yield (1.86 g, 1.68 mmol). Rf = 0.50 (3:1
petroleum ether/EtOAc); 1H NMR (CDCl3) δ = 7.49−7.21 (m, 38H;
CH-Ph, H-3, H-3′, H-5′), 6.96 (d, 3J(2′,3′) = 8,6 Hz, 1H; H-6′), 6.91
mL), which was stirred at room temperature for 2 h under hydrogen
atmosphere. Catalyst was filtered off under Celite and washed with
MeOH. The filtrate was concentrated and purified by CC (6:1
petroleum ether/EtOAc) to give compound 1 as a syrup in 96% yield
1
(0.039g, 0.09 mmol). Rf = 0.35 (1:1 petroleum ether/EtOAc); H
NMR (acetone-d6) δ = 13.05 (s, 1H; OH-5), 8.06 (s, 1H; H-2), 7.31
(d, 3J(2′,3′) = 8.9 Hz, 2H; H-2′, H-6′), 6.76 (d, 2H; H-3′, H-5′), 6.12
3
(s, 1H; H-6), 4.91 (d, J(1″,2″) = 9.9 Hz, 1H; H-1″), 3.74−3.70 (m,
3H; H-2″, H-6a″, H-6b″); 3.56 (t, 3J(2′,3′)=9.93 Hz, 1H; H-4″); 3.48
3
3
3
(t, 1H, J(2″,3″) = 9.9 Hz, H-3″); 3.38 (td, J(5″,6a″) = J(5″,6b″) =
2.9 Hz, 1H; H-5″); 13C NMR (acetone-d6) δ 182.1 (C-4); 164.0 (C-
5); 163.3 (C-7); 158.5 (C-4′); 156.7 (C-8a);154.4 (C-2); 131.3 (C-2′,
C-6′); 123.8 (C-3); 123.0 (C-1′); 116.2 (C-3′, C-5′); 106.1 (C-4a);
104.1 (C-8); 100.8 (C-6); 82.1 (C-5″); 79.4 (C-3″); 75.8 (C-1″); 73.7
(C-2″); 70.9 (C-4″); 61.9 (C-6″). HRMS: calcd for [M + H]
433.1129; found 433.1132.
Biological Studies. Animals. Tests were conducted using male
Wistar rats, with an average weight of 250 g. Animals were maintained
under stable conditions of temperature (25 °C), light−dark periods
(12 h), and feeding (maintenance rat chow). Both food and water
were available ad libitum. Food was removed 24 h before testing to
ensure that animals were on the fasting state. Free access to water was
maintained during this period. The animals were bred and utilized at
the Animal House of the Faculty of Medical Sciences, Lisbon,
Portugal. This facility has SPF (specific pathogen free) standards, as
supervised by the DGVPortuguese General Directorate of
Veterinary.
Induction of Diabetes. A state of hyperglycemia adequate to the
diagnosis of diabetes was induced experimentally through chemical
intervention. A sole intraperitoneal (ip) injection of streptozotocin
(STZ), previously dissolved in saline, was administered at a dose of 40
mg/kg bw. Hyperglycemia was checked 2 days after STZ
administration by quantifying glucose on a blood sample collected
by tail puncture. The animals of the normal control group received
instead an injection of the same volume of saline, with glycemia
checked also after 2 days.
Experimental Animal Groups. Animals were randomly divided into
three groups. Group I (control) was given one saline injection, and 2
days after started a 7-day treatment with saline + 5% ethanol. This
group represents normoglycemic control. Group II (STZ) was first
treated with streptozotocin (40 mg/kg bw, ip) and then for 7 days
with saline + 5% ethanol. This group represents the diabetic condition.
Group III (STZ + 8G) was first given STZ (40 mg/kg bw, ip) and
then a 7-day treatment with synthesized 8-β-D-glucopyranosyliso-
flavone (1, 4 mg/kg/day in saline + 5% ethanol, ip).
Glucose Tolerance Curve and Associated Insulin Parameters.
Animals were anesthetized with sodium pentobarbital (65 mg/kg bw)
after a 24 h fasting period. Immediately after, they were placed on a
homeothermic apparatus. Body temperature was maintained at 37 °C
to avoid metabolic changes induced by hypoglycemia. An exterior loop
was surgically placed between the femoral vein and artery, and a
catheter was placed on the stomach. After surgery completion,
recovery before testing was allowed for a minimum period of 30 min.
Anesthesia was maintained throughout the experiment with a constant
sodium pentobarbital perfusion on the femoral vein. Glucose tolerance
testing was done first by monitoring blood glycemia at the fasting state
for 20 min, after which 2 mL of a glucose solution (2 mg glucose/kg
bw) was administered through the gastric catheter directly into the
stomach. Blood glycemia was thus measured at regular intervals by a
bench glucose analyzer, both on baseline fasting (from −20 to 0 min)
and on the postload period (from 0 to 180 min). Blood samples were
also collected for insulin and C-peptide quantification, both at baseline
and at postload time points (−20, −10, 0, 5, 15, 30, 45, 60, 90, 120,
180 min). These samples were quickly centrifuged, and serum was
stored at −80 °C for RIA analysis.
3
(d, J(2,3) = 16,15 Hz, 1H; H-2), 6.43 (s, 1H; H-5′), 5.14 (s, 2H,
CH2Ph-7), 5.04−4.14 (m, 4H; CH2Ph, H-1‴, H-2‴), 3.91−3.48 (m,
5H; H-3‴, H-4‴, H-5‴; H-6a‴, H-6b‴); 2.09 (s, 3H; OCH3-Ac); 13
C
NMR (CDCl3) δ = 191.9 (C-1); 169.4 (CO-Ac); 160.7 (C-6′); 159.6
(C-1′, C-4′); 157.5 (C-4″); 149.5 (C-2″); 144.8 (C-3); 138.9, 138.8,
138.6, 138.1, 136.7, 136.6 (Cq, Ph); 136.2 (Cq, Ph-7); 128.8, 128.7,
128.6, 128.5, 128.4, 128.3, 128.2, 128.1, 128.0, 127.9, 127.8, 127.7,
127.6, 127.5, 127.4, 127.3, 127.2(CH, Ph); 127.1 (C-3′, C-5′); 126.3
(C-2); 118.9 (C-1″); 115.3 (C-2′, C-6′); 113.4 (C-3″); 96.7 (C-5″);
87.3 (C-3‴); 81.5 (C-2‴); 79.4 (C-5‴); 78.1 (C-4‴); 75.2; 74.9; 74.2;
73.1; 71.4; 70.8; 70.2 (CH2Ph); 73.7 (C-1‴); 69.7 (CH2Ph-7); 69.2
(C-6‴); 29.9 (OCH3). HRMS: calcd for [M + Na] 1129.4497; found
1129.4512.
4′,5,7-Tri-O-benzyl-8-(2,3,4,6-tetra-O-benzyl-β-D-glucocopy-
ranosyl)genistein (10). TTN (2.0 equiv) was added to a solution of
9 (1.85 g, 1.67 mmol) in (MeO)3CH (45 mL) and MeOH (45 mL).
The reaction mixture stirred for 24 h at 40 °C and then sodium
bissulfite was added to promote the reduction of Tl(III) to Tl(I). Solid
was removed by filtration. Water was added, and the mixture was
extracted with DCM. The combined extracts were dried over MgSO4,
filtered off, and concentrated. The yellow residue was dissolved in
THF (21 mL) and MeOH (21 mL), and then aqueous NaOH 50%
(8.6 mL) was added and the reaction stirred for 4 h at room
temperature. After reaction completed, HCl, 2 M, was added and the
mixture was extracted with DCM, dried over MgSO4, filtered off, and
concentrated. The residue was separated by CC (petroleum ether/
EtOAc, 5:1) to give 10 as a syrup in 63% overall yield (1.12g, 1.05
mmol). Rf = 0.45 (petroleum ether/EtOAc, 3:1); 1H NMR (CDCl3) δ
= 7.77 (s, 1H; H-2), 7.60−6.95 (m, 35H; CH-Ph), 6.84 (d, 3J(2′,3′) =
J(5′,6′) = 7.2 Hz, 2H; H-2′, H-6′), 6.75 (d, 2H; H-3′, H-5′), 6.40 (s,
1H; H-6)*; 5.25−4.05 (m, 14H; CH2Ph, H-1″); 3.95−3.53 (m, 6H;
H-2″, H-3″, H-4″, H-5″, H-6a″, H-6b″); 13C NMR (CDCl3) δ = 180.9
(C-4), 163.2 (C-7), 163.0 (C-5), 155.9 (C-8a), 152.5 (C-2), 138.7,
138.6, 138.5, 138.3, 138.2, 137.8, 137.1 (Cq, Ph), 129.9, 130.1 (C-2′,
C-6′); 128.7, 128.6, 128.5, 128.4, 128.3, 128.2, 128, 1, 128.0, 127.9,
127.8, 127.6, 127.5, 127.3, 127.2, 127.1 (CH, Ph), 115.0, 115.1 (C-3′,
C-5′),123.0 (C-3), 122.9 (C-1′), 106.4 (C-4a), 104.8 (C-8), 96.6 (C-
6), 87.9 (C-2″), 79.6 (C-5″), 78.5 (C-4″), 75.9, 75.6, 75.2, 74.8, 74.2,
73.5, 73.2 (CH2Ph), 74.5 (C-3″), 68.9 (C-6″). *Duplication of the
peaks was observed, corresponding to rotamers. HRMS: calcd for [M
+ H] 1063.4416; found 1063.4465.
Data Analysis. Data is shown as the mean standard error. Mean
relates to n observations, in which n represents the number of animals
tested (between five and eight, depending on the group). Mean values
between groups were compared using one-way analysis of variance
8-β-D-Glucopyranosylgenistein (1). Pd/C (25 mg) was added to
a solution of 10 (0.1 g, 0.094 mmol) in MeOH (3 mL) and EtOAc (1
G
dx.doi.org/10.1021/jm501069h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
1
(ANOVA), followed by a Tukey post-test. Differences were
considered statistically significant when p < 0.05.
was added to the final concentration of 0.4 mM. A H spectrum was
acquired with a recycle delay of 60 s to achieve the complete relaxation
of all the resonances at each scan. The quantification was performed by
comparing the DSS methyl resonance integral with the 8-β-D-
glucopyranosylgenistein (1) aromatic resonance integrals.
Kinetics of IAPP Fibrillization by ThT Fluorescence and AFM
Morphological Studies. hIAPP (human islet amyloid polypeptide,
sequence K1CNTATCATQRLANFLVHSSNNFGAILSSTN-
VGSNTY37-NH2) and rIAPP (rat islet amyloid polypeptide, sequence
K1CNTATCATQRLANFLVRSSNNLGPVLPPTNVGSNTY37-NH2)
were purchased from Peptide 2.0 (Chantilly, USA) as lyophilized
powder. 1,1,1,3,3,3-Hexafluoropropan-2-ol (HFIP) and thioflavin T
(ThT) were purchased from Sigma-Aldrich (Barcelona, Spain). hIAPP
was dissolved in HFIP followed by brief vortexing to mix and dissolve.
The stock solution was sealed with Parafilm and stored at 4 °C until
required at a concentration of 100 μM. The HFIP treatment maintains
the amyloidogenic peptide in a monomeric soluble state and allows for
experimental reproducibility.35 Prior to experimental use, HFIP was
evaporated under a gentle N2 flow, followed by vacuum pump. The
residue was then resuspended in phosphate buffer (pH 7.4, 100 mM
sodium phosphate, 100 mM NaCl). Each fibril formation reaction was
performed at the final concentrations of 25 μM IAPP and 5 μM ThT,
with 2% DMSO. Experiments were carried out in the absence
(control) and in the presence of 8-β-D-glucopyranosylgenistein (1) at
the concentration of 50 μM in a 96-well plate in duplicate.
Fluorescence was monitored at 25 °C, with excitation at 440 nm
and emission at 480 nm, using a SpectraMax Gemini microplate reader
with slit width of 5 nm and a cutoff filter at 455 nm and stirring before
each reading. The emission intensity was measured over a time course
of 22 h until a plateau for the respective intensity levels was reached.
As a negative control, rat IAPP, which is known not to form fibrils, was
used at a final concentration of 25 μM to confirm that the protocol
and experimental results are being correctly interpreted (Figure S4).
For the morphological characterization, Nanoscope IIIa multimode
atomic force microscope produced by Digital Instruments (Veeco,
Santa Barbara, CA) was used. All measurements were carried out by
tapping mode AFM using etched silicon tips with a resonance
frequency of ∼300 kHz at a scan rate of ∼1.5 Hz. The images were
acquired in ambient conditions (20 °C) by placing a drop of the
solution onto freshly cleaved mica for 15 min, rinsing with water and
drying with pure N2. The AFM experiments were performed with
aliquots of the same initial sample that was used for ThT fluorescence
assays to allow direct comparison of the data.
Molecular Mechanics (MM) Calculations. Molecular mechanics
were conducted with MacroModel 9.6.20736 as implemented in
version 9.1.207 of the Maestro suite,37 using the MM3* force field.31,38
A systematic variation of the torsional degrees of freedom of the
molecules permitted generation of different starting structures that
were further minimized to provide the corresponding local minima.
Only the same two minima were always found for each molecule (O-
and C-glucosides). The continuum GB/SA solvent model39 was
employed, and the general PRCG (Polak−Ribiere conjugate gradient)
method for energy minimization was used. An extended cutoff was
applied.
ASSOCIATED CONTENT
* Supporting Information
■
S
Antidiabetic activity (C-peptide, insulin, and β-cell sensitivity
graphs), ThT kinetics of IAPP fibrillization (rIAPP negative
control graph; hIAPP and rIAPP sequence), and NMR binding
studies with islet amyloid peptide IAPP and Aβ1−42 oligomers
(STD NMR, 2D-NOESY, and trNOESY supporting data). This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
■
*Phone: +351217500952 or +351964408824. E-mail:
aprauter@fc.ul.pt.
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
Notes
The authors declare no competing financial interest.
NMR Binding Studies with Islet Amyloid Peptide IAPP and Aβ1−42
Oligomers. NMR experiments were recorded on a Bruker Avance 600
MHz spectrometer equipped with a triple channel cryoprobe head
(IAPP) or on a Varian 400 MHz Mercury instrument (Aβ1−42).
Immediately before use, lyophilized Aβ1−42 and IAPP were dissolved in
10 mM NaOD in D2O at a concentration of 160 μM, then diluted 1:1
with 10 mM phosphate buffer saline, pH 7.4, containing 100 mM
NaCl (PBS). To these samples were added the compound 8-β-D-
glucopyranosylgenistein to a final concentration of 2 mM and ∼1 mg
in the case of ethyl acetate extract of G. tenera. The pH of each sample
was verified with a microelectrode (Mettler Toledo) for 5 mm NMR
tubes and adjusted with NaOD and/or DCl. All pH values were
corrected for isotope effect. Selective saturation of the protein
resonances (on resonance spectrum) was performed by irradiating at
−0.5 ppm (IAPP) or at −1.0 ppm (Aβ1−42) using a series of Gaussian-
shaped or Eburp2.1000-shaped pulses (50 ms) for a total saturation
time of 2.0 s. For the reference spectrum (off resonance), the samples
were irradiated at 100 ppm. At the end, STD experiments were
recorded at two temperatures 298 and 310 K with a ligand/amyloid
oligomers molar ratio of 25:1. For NOESY and trNOESY experiments
the basic Varian and Bruker sequences were employed. The trNOESY
analysis is based on the well-known fact that given the appropriate
kinetic conditions for the ligand−receptor association process, NOE
cross peaks for small ligands, when free in solution, display a different
sign (positive cross peaks, opposite signal to the diagonal peaks) than
those recorded for their bound states, in the presence of the receptor
(negative cross peaks, same signal of diagonal peaks).27 A mixture of
ligand/amyloid oligomers 25:1 was employed to record trNOESY
experiments. For the quantitative NMR experiment, 1.5 mg of ethyl
acetate extract of G. tenera was dissolved in 550 μL of D2O and DSS
ACKNOWLEDGMENTS
■
The authors thank Fundaca
̧
o para a Cien
̂
cia e a Tecnologia
̃
(FCT) for financial support of the Projects PTDC/QUI/
67165/2006, Pest-OE/QUI/UI0612/2013, and RECI/BBB-
BQB/0230/2012, as well as for the Ph.D. grant of Ana M.
Matos (Grant SFRH/BD/93170/2013), the postdoctoral grant
of Filipa Marcelo (Grant SFRH/BPD/65462/2009), and IF
2012/2013 initiatives (P.O.P.H., F.S.E.). FCT is also acknowl-
edged for the Project REDE/1501/REM/2005, and Dr. Carlos
Cordeiro is acknowledged for providing the HRMS spectrum
̂
with the FTICR-MS at the Faculdade de Ciencias da
Universidade de Lisboa, Portugal. The European Commission
is acknowledged for the approval of INOVAFUNAGEING
Commitment. The European Community’s Seventh Frame-
work Programme (FP7/2007−2013) under Grant Agreements
610359 and 212043 is gratefully acknowledged. F.N. and C.A.
thank Fondo per la Promozione di Accordi Istituzionali,
Progetto No. 4779 “Network Enabled Drug Design (NEDD)”
from Lombardia. J.J.-B. thanks MICINN of Spain (Grant
CTQ2009-08536) and the Cariplo Foundation for funding.
ABBREVIATIONS USED
■
STZ, streptozotocin; STD, saturation transfer difference; NMR,
nuclear magnetic resonance; trNOESY, transferred nuclear
Overhauser effect spectrometry; IAPP, islet amyloid polypep-
tide; ThT, thioflavin T; ATM, atomic force microscopy; Aβ1−42
,
H
dx.doi.org/10.1021/jm501069h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
biological actions, and clinical potential. Curr. Med. Chem. 2013, 20,
899−907. (c) Qi, L.; Liu, E.; Chu, C.; Peng, Y. B.; Cai, H. X.; Li, P.
Anti-diabetic agents from natural productsan update from 2004 to
2009. Curr. Top. Med. Chem. 2010, 10, 434−457.
(15) (a) Stefani, M.; Rigacci, S. Protein folding and aggregation into
amyloid: the interference by natural phenolic compounds. Int. J. Mol.
Sci. 2013, 14, 12411−12457. (b) Ansari, N.; Khodagholi, F. Natural
products as promising drug candidates for the treatment of
Alzheimer’s disease: molecular mechanism Aspect. Curr. Neuro-
pharmacol. 2013, 11, 414−429.
amyloid β 1−42; bw, body weight; MM, molecular mechanics;
DM, diabetes mellitus; AD, Alzheimer’s disease; EGCG,
(−)-epigallocatechin 3-gallate; PGG, 1,2,3,4,6-penta-O-galloyl-
β-D-glucopyranose; TTN, thallium(III) nitrate; ip, intra-
peritoneal; AUC, area under the curve; hIAPP, human islet
amyloid polypeptide; rIAPP, rat islet amyloid polypeptide; Aβ,
β-amyloid; BBB, blood−brain barrier; TLC, thin layer
chromatography; CC, column chromatography; TMS, tetra-
methylsilane; rt, room temperature; mp, melting point; HFIP,
1,1,1,3,3,3-hexafluoropropan-2-ol
(16) Borges, C.; Martinho, P.; Martins, A.; Rauter, A. P.; Ferreira, M.
A. A. Characterization of flavonoids extracted of Genista tenera by FAB
MS/MS. Rapid Commun. Mass Spectrom. 2001, 15, 1760−1767.
(17) Rauter, A. P.; Martins, A.; Borges, C.; Ferreira, J.; Justino, J.;
Bronze, M. R.; Coelho, A. V.; Choi, Y. H.; Verpoorte, R. Liquid
chromatography-diode array detection-electrospray ionisation mass
spectrometry/nuclear magnetic resonance analyses of the anti-
hyperglycemic flavonoid extract of Genista tenera. Structure elucidation
of a flavonoid-C-glycoside. J. Chromatogr., A 2005, 1089, 59−64.
(18) Edwards, E. L.; Rodrigues, J. A.; Ferreira, J.; Goodall, D. M.;
Rauter, A. P.; Justino, J.; Thomas-Oates, J. Capillary electrophoresis-
mass spectrometry characterization of secondary metabolites from the
anti-hyperglycemic plant Genista tenera. Electrophoresis 2006, 27,
2164−2170.
REFERENCES
■
(1) Chiti, F.; Dobson, C. M. Protein misfolding, functional amyloid,
and human disease. Annu. Rev. Biochem. 2006, 75, 333−366.
(2) Glabe, C. Common mechanisms of amyloid oligomer patho-
genesis in degenerative diseases. Neurobiol. Aging 2006, 27, 570−575.
(3) Meier, J. J.; Kayed, R.; Lin, C.; Gurlo, T.; Haataja, L.; Jayasinghe,
S.; Langen, R.; Glabe, C. G.; Butler, P. C. Inhibition of human IAPP
fibril formation does not prevent β-cell death: evidence for distinct
actions of oligomers and fibrils of human IAPP. Am. J. Physiol.:
Endocrinol. Metab. 2006, 291, 1317−1324.
(4) Baglioni, S.; Casamenti, F.; Bucciantini, M.; Luheshi, L. M.;
Taddei, N.; Chiti, F.; Dobson, C. M.; Stefani, M. Prefibrillar amyloid
aggregates could be generic toxins in higher organisms. J. Neurobiol.
2006, 26, 8160−8167.
(5) Ladiwala, A. R.; Dordick, J. S.; Tessier, P. M. Aromatic small
molecules remodel toxic soluble oligomers of amyloid β through three
independent pathways. J. Biol. Chem. 2011, 286, 3209−3218.
(6) Hirohata, M.; Ono, K.; Takasaki, J.; Takahashi, R.; Ikeda, T.;
Morinaga, A.; Yamada, M. Anti-amyloidogenic effects of soybean
isoflavones in vitro: fluorescence spectroscopy demonstrating direct
binding to Aβ monomers, oligomers and fibrils. Biochim. Biophys. Acta
2012, 1822, 1316−1324.
(7) Cao, P.; Raleigh, D. P. Analysis of the inhibition and remodeling
of islet amyloid polypeptide amyloid fibers by flavanols. Biochemistry
2012, 51, 2670−2683.
(8) Meng, F.; Abedini, A.; Plesner, A.; Verchere, C. B.; Raleigh, D. P.
The flavanol (−)-epigallocatechin 3-gallate inhibits amyloid formation
by islet amyloid polypeptide, disaggregates amyloid fibrils and protects
cultured cells against IAPP-induced toxicity. Biochemistry 2010, 49,
8127−8133.
(9) Fujiwara, H.; Tabuchi, M.; Yamaguchi, T.; Iwasaki, K.; Furukawa,
K.; Sekiguchi, K.; Ikarashi, W.; Kudo, Y.; Higuchi, M.; Saido, T. C.;
Maeda, S.; Takashima, A.; Hara, M.; Yaegashi, N.; Kase, Y.; Arai, H. A
traditional medicinal herb Paeonia suffruticosa and its active
constituent 1,2,3,4,6-penta-O-galloyl-β-D-glucopyranose have potent
anti-aggregation effects on Alzheimer’s amyloid β proteins in vitro and
in vivo. J. Neurochem. 2009, 109, 1648−1657.
(10) Bruno, E.; Pereira, C.; Roman, K. P.; Takiguchi, M.; Kao, P.;
Nogaj, L. A.; Moffet, D. A. IAPP aggregation and cellular toxicity are
inhibited by 1,2,3,4,6-penta-O-galloyl-β-D-glucose. Amyloid 2013, 20,
34−38.
(11) Wolfram, S.; Raederstorff, D.; Preller, M.; Wang, Y.; Teixeira, S.
R.; Riegger, C.; Weber, P. Epigallocatechin gallate supplementation
alleviates diabetes in rodents. J. Nutr. 2006, 136, 2512−2518.
(12) Fu, Z.; Zhen, W.; Yuskavage, J.; Liu, D. Epigallocatechin gallate
delays the onset of type 1 diabetes in spontaneous non-obese diabetic
mice. Br. J. Nutr. 2011, 105, 1218−1225.
(13) Mohan, C. G.; Viswanatha, G. L.; Savinay, G.; Rajendra, C. E.;
Halemani, P. D. 1,2,3,4,6-Penta-O-galloyl-β-D-glucose, a bioactivity
guided isolated compound from Mangifera indica inhibits 11β-HSD-1
and ameliorates high fat diet-induced diabetes in C57BL/6 mice.
Phytomedicine 2013, 20, 417−426.
(14) (a) Salimifar, M.; Fatehi-Harranabad, Z.; Fatehi, M. A review on
natural products for controlling type 2 diabetes with an emphasis on
their mechanisms of actions. Curr. Diabetes Rev. 2013, 9, 402−411.
(b) Chang, C. L. T.; Chen, W. C.; Chen, H. M.; Yang, N. S.; Yang, W.
C. Natural cures for type 1 diabetes: a review of phytochemicals,
(19) Rauter, A. P.; Ferreira, J.; Martins, A.; Santos, R. G.; Serralheiro,
M. L.; Borges, C.; Araujo, M. E.; Silva, F.; Goulart, M.; Justino, J.;
́
Rodrigues, J.; Edwards, E.; Thomas-Oates, J. P.; Noronha, J. P.; Pinto,
R.; Mota-Filipe, H. Bioactivity studies and chemical profile of the
antidiabetic plant Genista tenera. J. Ethnopharmacol. 2009, 122, 384−
393.
(20) Rauter, A. P.; Martins, A.; Borges, C.; Mota-Filipe, H.; Pinto, R.;
Sepodes, B.; Justino, J. Antihyperglycaemic and protective effects of
flavonoids on streptozotocin-induced diabetic rats. Phytother. Res.
2010, 24, S133−S138.
(21) Rauter, A. P., Jesus, A. R., Martins, A., Dias, C., Ribeiro, R.,
Macedo, M.-P., Justino, J., Mota-Filipe, H., Pinto, R., Sepodes, B.,
Medeiros, M., Barbero, J. J., Airoldi, C., Nicotra, F. New C-
glycosylpolyphenol antidiabetic agents, effect on glucose tolerance
and interactionwith beta-amyloid. Therapeutic applications of the
synthesized agent(s) and of Genista tenera ethyl acetate extracts
containing some of those agents. WO 2013132470A2, 2013
(submitted Mar 9, 2012; publication date, Sep 12, 2013).
(22) Kumazawa, T.; Kimura, T.; Matsuba, S.; Sato, S.; Onodera, J.
Synthesis of 8-C-glucosylflavones. Carbohydr. Res. 2001, 334, 183−
193.
(23) Sato, S.; Hiroe, K.; Kumazawa, T.; Jun-Ichi, O. Environmentally
friendly C-glycosylation of phloroacetophenone with unprotected D-
glucose using scandium(III) trifluoromethanesulfonate in aquous
media: key compounds for the synthesis of mono- and di-C-
glycosylflavonoids. Carbohydr. Res. 2006, 341, 1091−1095.
(24) Westermark, P.; Engstrom, U.; Johnson, K. H.; Westermark, G.
̈
T.; Betsholtz, C. Islet amyloid polypeptide: pinpointing amino acid
residues linked to amyloid fibril formation. Proc. Natl. Acad. Sci. U.S.A.
1990, 87, 5036−5040.
(25) Wei, L.; Jiang, P.; Xu, W.; Li, H.; Zhang, H.; Yan, L.; Chan-Park,
M. B.; Liu, X.; Tang, K.; Mu, Y.; Pervushin, K. The molecular basis of
distinct aggregation pathways of islet amyloid polypeptide. J. Biol.
Chem. 2011, 286, 6291−6300.
(26) (a) Mehla, J.; Chauhan, B. C.; Chauhan, N. B. Experimental
induction of type 2 diabetes in aging-accelerated mice triggered
Alzheimer-like pathology and memory deficits. J. Alzheimer’s Dis. 2014,
39, 145−162. (b) Kim, B.; Backus, C.; Oh, S.; Feldman, E. L.
Hyperglycemia-induced tau cleavage in vitro and in vivo: a possible
link between diabetes and Alzheimer’s disease. J. Alzheimer’s Dis. 2013,
34, 729−739. (c) Morris, J. K.; Vidoni, E. D.; Honea, R. A.; Burns, J.
M. Impaired glycemia and Alzheimer’s disease. Neurobiol. Aging 2014,
35, 585−589.
(27) (a) Mayer, M.; Meyer, B. Characterization of ligand binding by
saturation transfer difference NMR spectroscopy. Angew. Chem., Int.
I
dx.doi.org/10.1021/jm501069h | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Ed. 1999, 38, 1784−1788. (b) Mayer, M.; Meyer, B. Group epitope
mapping by saturation transfer difference NMR to identify segments of
a ligand in direct contact with a protein receptor. J. Am. Chem. Soc.
2001, 123, 6108−6117. (c) Meyer, B.; Peters, T. NMR spectroscopy
techniques for screening and identifying ligand binding to protein
receptors. Angew. Chem., Int. Ed. 2003, 42, 864−890.
(28) (a) Airoldi, C.; Sironi, E.; Dias, C.; Marcelo, F.; Martins, A.;
Rauter, A. P.; Nicotra, F.; Jimenez-Barbero, J. Natural compounds
́
against Alzheimer’s disease: molecular recognition of Aβ1-42 peptide
by Salvia sclareoides extract and its major component, rosmarinic acid,
as investigated by NMR. Chem.Asian J. 2013, 8, 596−602.
(b) Airoldi, C.; Cardona, F.; Sironi, E.; Colombo, L.; Salmona, M.;
Cambianica, I.; Ornaghi, F.; Sancini, G.; Nicotra, F.; La Ferla, B.
Fluorescent amyloid β-peptide ligand derivatives as potential
diagnostic tools for Alzheimer’s disease. Pure Appl. Chem. 2013, 85,
1813−1823. (c) Airoldi, C.; Colombo, L.; Manzoni, C.; Sironi, E.;
Natalello, A.; Doglia, S. M.; Forloni, G.; Tagliavini, F.; Del Favero, E.;
Cantu, L.; Nicotra, F.; Salmona, M. Tetracycline prevents Aβ oligomer
toxicity through an atypical supramolecular interaction. Org. Biomol.
Chem. 2011, 9, 463−472. (d) Airoldi, C.; Zona, C.; Sironi, E.;
Colombo, L.; Messa, M.; Aurilia, D.; Gregori, M.; Masserini, M.;
Salmona, M.; Nicotra, F.; La Ferla, B. Curcumin derivatives as new
ligands of Aβ peptides. J. Biotechnol. 2011, 156, 317−324. (e) Airoldi,
C.; Cardona, F.; Sironi, E.; Colombo, L.; Salmona, M.; Silva, A.;
Nicotra, F.; La Ferla, B. cis-Glyco-fused benzopyran compounds as
new amyloid-β peptideligands. Chem. Commun. 2011, 47, 10266−
10268.
́
(29) NMR Spectroscopy of Glycoconjugates; Jimenez-Barbero, J.,
Peters, T., Eds.; Wiley-VCH: Weinheim, Germany, 2002.
(30) Ni, F.; Zhu, Y. Accounting for ligand−protein interactions in the
relaxation-matrix analysis of transferred nuclear Overhauser effects. J.
Magn. Reson. B 1994, 103, 180−184.
(31) Allinger, N. L.; Yuh, Y. H.; Lii, J. H. Molecular mechanics. The
MM3 force field for hydrocarbons. 1. J. Am. Chem. Soc. 1989, 111,
8551−8566.
(32) Qutub, A. A.; Hunt, C. A. Glucose transport to the brain: a
systems model. Brain Res. 2005, 49, 595−617.
(33) Storr, T.; Merkel, M.; Song-Zhao, G. X.; Scott, L. E.; Green, D.
E.; Bowen, M. L.; Thompson, K. H.; Patrick, B. O.; Schugar, H. J.;
Orvic, C. Synthesis, characterization, and metal coordinating ability of
multifunctional carbohydrate-containing compounds for Alzheimer’s
therapy. J. Am. Chem. Soc. 2007, 129, 7453−7463.
(34) Ladiwala, A. R. A.; Mora-Pale, M.; Lin, J. C.; Bale, S. S.;
Fishman, Z. S.; Dordick, J. S.; Tessier, P. M. Polyphenolic glycosides
and aglycones utilize opposing pathways to selectively remodel and
inactivate toxic oligomers of amyloid β. ChemBioChem 2011, 12,
1749−1758.
(35) Zagorski, M. G.; Yang, J.; Shao, H.; Ma, K.; Zeng, H.; Hong, A.
Methodological and chemical factors affecting amyloid beta peptide
amyloidogenicity. Methods Enzymol. 1999, 309, 189−203.
(36) MacroModel, version 9.6; Schrodinger, LLC: New York, 2008.
̈
(37) Maestro, version 9.1; Schrodinger, LLC: New York, 2010.
̈
́
(38) Martín-Pastor, M.; Espinosa, J. F.; Asensio, J. L.; Jimenez
Barbero, J. A comparison of the geometry and of the energy results
obtained by application of different molecular mechanics force fields to
methyl α-lactoside and the C-analogue of lactose. Carbohydr. Res.
1997, 298, 15−49.
(39) Still, W. C.; Tempczyk, A.; Hawley, R. C.; Hendrickson, T.
Semianalytical treatment of solvation for molecular mechanics and
dynamics. J. Am. Chem. Soc. 1990, 112, 6127−6129.
J
dx.doi.org/10.1021/jm501069h | J. Med. Chem. XXXX, XXX, XXX−XXX