Content uploaded by Dirk Jacobus Brand
Author content
All content in this area was uploaded by Dirk Jacobus Brand on Jul 26, 2018
Content may be subject to copyright.
Benzophenone C- and O‑Glucosides from Cyclopia genistoides
(Honeybush) Inhibit Mammalian α‑Glucosidase
Theresa Beelders,*
,†,‡
D. Jacobus Brand,
§
Dalene de Beer,
‡
Christiaan J. Malherbe,
‡
Sithandiwe E. Mazibuko,
⊥
Christo J. F. Muller,
⊥
and Elizabeth Joubert
†,‡
†
Department of Food Science, Stellenbosch University, Private Bag X1, Matieland 7602, South Africa
‡
Post-Harvest and Wine Technology Division, Agricultural Research Council (ARC) Infruitec-Nietvoorbij, Private Bag X5026,
Stellenbosch 7599, South Africa
§
Central Analytical Facility (CAF), Nuclear Magnetic Resonance Unit, Private Bag X1, Matieland 7602, South Africa
⊥
Diabetes Discovery Platform, South African Medical Research Council, P.O. Box 19070, Tygerberg 7505, South Africa
*
SSupporting Information
ABSTRACT: An enriched fraction of an aqueous extract prepared from the aerial parts of Cyclopia genistoides Vent. yielded a
new benzophenone di-C,O-glucoside, 3-C-β-D-glucopyranosyl-4-O-β-D-glucopyranosyliriflophenone (1), together with small
quantities of a known benzophenone C-glucoside, 3-C-β-D-glucopyranosylmaclurin (2). The isolated compounds showed α-
glucosidase inhibitory activity against an enzyme mixture extracted from rat intestinal acetone powder. Compound 2exhibited
significantly (p< 0.05) higher inhibitory activity (54%) than 1(43%) at 200 μM. In vitro tests in several cell models showed that
1and its 3-C-monoglucosylated derivative (3-C-β-D-glucopyranosyliriflophenone) were marginally effective (p≥0.05) in
increasing glucose uptake.
The global prevalence of diabetes is increasing at alarming
rates, leading to estimates of 439 million diabetics by
2030.
1
In Africa, South Africa tips the scale with a national
prevalence of over 7%.
2
It is projected that diabetes will rank as
the ninth leading cause of death in low-income countries in the
next few decades.
3
A sedentary lifestyle together with an
unhealthy diet, high in refined carbohydrates accompanied by a
low intake of fruits and vegetables, are considered contributing
factors. The search for natural and synthetic α-glucosidase
inhibitors
4,5
that delay the breakdown and absorption of
carbohydrates in the gut, thus mimicking the protective effect of
the drug acarbose, but without its side-effects,
6
is escalating.
Much prominence is given to polyphenols, with promising
results in vitro.
4
Synergistic effects between acarbose and
polyphenols suggest benefits in terms of dose reduction of the
drug.
7
Subclasses of phenolic compounds that show promise as
α-glucosidase inhibitors include the polyhydroxyxanthones and
polyhydroxybenzophenones, together with their glycosylated
derivatives.
8−11
Extracts of Cyclopia genistoides Vent. (honey-
bush) are principally renowned for high levels of the
tetrahydroxyxanthone C-glucoside mangiferin,
12
a bioactive
compound
13
with potent α-glucosidase inhibitory activity.
9
Recent investigation of the phenolic constituents of C.
genistoides revealed the presence of several benzophenone C-
glucosides, including 3-C-β-D-glucopyranosyliriflophenone and
3-C-β-D-glucopyranosylmaclurin.
14−16
The major benzophe-
none glycoside present in hot water extracts of C. genistoides
has been tentatively identified as an iriflophenone di-O,C-
hexoside thus far.
16
A compound with similar MS properties
has also been observed in extracts of another Cyclopia species,
C. subternata,
17
albeit at lower concentrations. In the present
paper the isolation and structural elucidation of this
benzophenone glucoside, identified as 3-C-β-D-glucopyrano-
syl-4-O-β-D-glucopyranosyliriflophenone (1), is described.
During the isolation process, small quantities of the known
benzophenone C-glucoside 3-C-β-D-glucopyranosylmaclurin
(2) was also obtained. These compounds were isolated from
the aerial parts of C. genistoides using solid-phase extraction
(SPE) for extract enrichment, followed by purification by
semipreparative liquid chromatography (LC). The ability of 1
and 2to inhibit mammalian α-glucosidase was assessed and
Received: September 16, 2014
Article
pubs.acs.org/jnp
© XXXX American Chemical Society and
American Society of Pharmacognosy Adx.doi.org/10.1021/np5007247 |J. Nat. Prod. XXXX, XXX, XXX−XXX
compared to that of 3-C-β-D-glucopyranosyliriflophenone.
Included in the α-glucosidase assay was the benzophenone
aglucone maclurin, with acarbose employed as positive control.
To date, the α-glucosidase inhibitory activities of 1and 2as C-
and O-glucosylated polyhydroxybenzophenones have not been
assessed. In vitro tests in several cell models were conducted to
investigate the ability of 1and 3-C-β-D-glucopyranosyliriflo-
phenone to increase glucose uptake in vitro.
■RESULTS AND DISCUSSION
Compound 1was obtained as a white, amorphous powder. Its
molecular formula was established as C25H30O15 from the
analyses of its HRESIMS and 13C NMR data. Initial LC-
ESIMSMS analyses of this compound in a C. genistoides sample
matrix
16
had pointed to an iriflophenone di-O,C-hexoside. The
elucidation of the structure of 1and, in particular, the
identification of the hexoside moieties and their point of
attachment to the aglycone, were done via 1D and 2D NMR
experiments. The relative configurations of the glycoside units
can normally be determined by a 1D/2D NOESY NMR
experiment. However, NMR spectroscopic data for 1proved
inconclusive for unambiguous identification of the individual
glycoside units, due to overlapping proton resonances. Since
the NMR spectra of 1closely resembled those of 3-C-β-D-
glucopyranosyliriflophenone, it was essential to first selectively
hydrolyze and identify the terminal O-linked sugar moiety,
followed by comparison of the resulting iriflophenone C-
hexoside to that of a commercial 3-C-β-D-glucopyranosyl-
iriflophenone standard, also present in C. genistoides.
18
The terminal O-linked hexoside moiety was identified as D-
glucose by enzymatic hydrolysis and GC-MS analysis. The
resulting iriflophenone C-hexoside was purified, and its 1H and
13C NMR data matched those of 3-C-β-D-glucopyranosyl-
iriflophenone. Having identified the β-D-glucopyranosyl con-
stituent units, the rest of the molecule and the point of
attachment of the O-β-D-glucopyranosyl unit were defined by
additional 1D NOESY, COSY, HSQC, and HMBC NMR
experiments.
The aromatic region between 6.0 and 8.5 ppm of the 1H
NMR spectra showed two o-coupled doublets and a singlet, δ
7.61 (2H, d, J= 8.6 Hz), 6.78 (2H, d, J= 8.6 Hz), 6.24 (1H, s),
reminiscent of an AA′BB′aromatic spin system with an
uncoupled singlet on a separate aromatic ring and supported by
the presence of aromatic carbons in the 13C NMR spectra of 1.
The 1H NMR spectra also presented the resonances of two β-
coupled (∼8 Hz anomeric coupling constant) glucoside units in
the characteristic 2.5−5.0 ppm region.
19
The presence of a
single carbon at δ193.1 in addition to the 12 aromatic carbons
in the 13C NMR spectrum was reminiscent of a benzophenone
aglycone unit.
19
The anomeric protons of the two glucopyr-
anosyl moieties at δ4.73 (1H, d, J= 7.8 Hz) and 4.66 (1H, d, J
= 9.8 Hz) were irradiated selectively utilizing a 1D NOESY
experiment to determine their position of attachment and other
proton resonances of the individual O-and C-linked
glucopyranosyl units. The anomeric proton of the glucopyr-
anosyl residue at δ4.73 (1H, d, J= 7.8 Hz, H-1‴) displayed a
strong NOE association with the H-5 singlet at δ6.24. This
assigned the O-linked glucosyl unit to C-4 of the
benzophenone A-ring. A weak NOE association of H-1‴with
H-4″(δ3.24) of the neighboring glucosyl moiety indicated the
close proximity of the glucosyl entities.
An NOE association was also observed between the three
axial H-1‴, H-3‴, and H-5‴protons (Figure 1), affirming the
glucose configuration of the O-linked sugar unit. The nature of
theC-linkedsugartotheaglyconewasconfirmed by
comparison of the hydrolytic products to the NMR spectra
of the commercial 3-C-β-D-glucopyranosyliriflophenone stand-
ard.
A COSY experiment confirmed the AA′BB′aromatic spin
system of the B-ring and the connectivity of some of the
protons on the individual glucosyl moieties. This includes only
the protons that are sufficiently resolved to be assigned by the
COSY experiment and did not overlap with the glucosyl proton
resonances.
The HSQC NMR experiment was used to assign all the
directly bonded protons (1JHC) to their respective carbons. All
the carbons were resolved successfully by an HMBC experi-
ment showing the long-range (2JHC,3JHC) connectivity between
protons and carbons [up to four bonds (4JHC) in some cases].
Those structurally significant HMBC correlations are shown in
Figure 2, enabling the assignment of the unprotonated carbons
in the molecule. Compound 1was thus identified as 3-C-β-D-
glucopyranosyl-4-O-β-D-glucopyranosyliriflophenone.
Compound 2was obtained as a light yellow, amorphous
powder. Accurate measurement of the pseudomolecular ions in
the positive and negative HRESIMS data, in conjunction with
13C NMR data, allowed a molecular formula of C19H20O11 to be
assigned to 2. The structure of 2was confirmed as 3-C-β-D-
Figure 1. Relevant NOE associations in 1.
Figure 2. Relevant long-range HMBC correlations in 1.
Journal of Natural Products Article
dx.doi.org/10.1021/np5007247 |J. Nat. Prod. XXXX, XXX, XXX−XXXB
glucopyranosylmaclurin by comparison of its observed and
reported 1H NMR and 13C NMR spectroscopic data.
14,20
Both benzophenone derivatives 1and 2, together with the
reference compounds, showed inhibitory activity against
mammalian α-glucosidase at various concentrations with all
values significantly differing (p< 0.05) from all other values.
The observed inhibitory activities were both concentration- and
compound-specific (Figure 3), showing a clear dose−response.
The activities of the benzophenones could thus be compared at
equimolar concentration levels, which gave insight into possible
structure−activity relations. Compound 2was the most active
inhibitor, followed by 3-C-β-D-glucopyranosyliriflophenone and
1, while maclurin showed the weakest inhibitory activity. It
appears that C-monoglucosides tend to be more effective
inhibitors than their corresponding aglucones; that is, 2is
significantly more active than maclurin (p< 0.05). This has also
been observed previously for 3-C-β-D-glucopyranosyl-
iriflophenone and its aglucone, iriflophenone.
9
Furthermore,
the higher activity of 2compared to that of 3-C-β-D-
glucopyranosyliriflophenone was attributed to the additional
3′-hydroxy group of 2. Previous studies on hydroxybenzophe-
nones
10
and hydroxyxanthones
8
have shown that increases in
the number of phenolic hydroxy groups on the basic diphenyl
ketone and dibenzo-γ-pyrone structures lead to significant
increases in α-glucosidase inhibitory activities. Moreover, for
hydroxyflavones it has been shown that a 3′-hydroxy group in
particular leads to increased inhibitory activity.
21
The presence
of an additional O-glucopyranosyl moiety at C-4 on the
diphenyl ketone structure of 1significantly (p< 0.05) lowered
α-glucosidase inhibitory activity compared to 3-C-β-D-
glucopyranosyliriflophenone. A similar decrease in activity
due to an additional C-glucopyranosyl moiety has also been
reported for 3-C-β-D-glucopyranosyl-5-C-β-D-glucopyranosyl-
iriflophenone compared to 3-C-β-D-glucopyranosyl-
iriflophenone.
9
The effect of 1and 3-C-β-D-glucopyranosyliriflophenone on
in vitro glucose uptake in L6 myocytes demonstrated marginal,
but not significant, increases in glucose uptake relative to the
vehicle control at concentrations of 10 μMof1(ca. 22%
increase; Figure 4A) and 100 μMof3-C-β-D-glucopyranosyl-
iriflophenone (ca. 27% increase; Figure 4B). Similar marginal
effects on glucose uptake were observed in 3T3-L1 adipocytes
(Figure 4C and D) and in C3A hepatocytes (data not shown).
It was anticipated that the isolated benzophenone glucoside (1)
and 3-C-β-D-glucopyranosyliriflophenone could increase in
vitro glucose uptake activity by activating the cellular energy
regulator AMPK, as demonstrated for the latter compound
22,23
in mature 3T3-L1 adipocytes and diabetic KK-Aymice. The
assessed compounds were, however, less effective at increasing
cellular glucose uptake than the reference pharmacological
agent metformin, mechanistically known to be an activator of
AMPK.
24
■EXPERIMENTAL SECTION
General Experimental Procedures. NMR spectra were recorded
on a Varian Unity Inova 600 NMR spectrometer with a 1H frequency
of 600 MHz and a 13C frequency of 150 MHz using a 5 mm inverse
detection PFG probe. The chemical shift frequencies are indicated on
the δscale. 1H and 13C NMR spectra were referenced to the residual
DMSO-d6peak at δ2.5 and 39.5, respectively. The spectra in acetone-
d6have the residual acetone peaks referenced at δ2.05 for 1H NMR
and at δ29.84 for the 13C NMR spectra. The spectra were recorded
using the standard VnmrJ instrument software and processed and
expanded further using the Mestrenova 9.0 software package.
HRESIMS analyses were conducted on an Acquity UPLC system,
fitted with a photodiode-array detector and coupled to a Synapt G2 Q-
TOF equipped with an electrospray ionization source (Waters,
Milford, MA, USA). UV spectra were recorded online. Chromato-
graphic conditions and MS parameters were as described by
Beelders.
16
Semipreparative LC was performed on a LaChrom
HPLC system (Merck Hitachi, Hitachi High Technologies, Japan)
comprising a quaternary pump, autosampler, variable-wavelength
detector, and diode-array detector and fitted with a Phenomenex
Gemini C18 column (5 μm, 110 Å, 150 ×10 mm) (Phenomenex,
Torrance, CA, USA). The column was protected by a guard column of
the same stationary phase (10 ×10 mm) and a high-pressure
semipreparative in-line filter (IDEX Health & Science, Oak Harbor,
WA, USA). Column temperature was maintained at 30 °C using an
external HPLC column oven (LKB Bromma, Sweden). Gas
chromatography was performed on an Agilent GC-MS instrument
(Agilent 6890N GC and Agilent 5975 MSD, Agilent Technologies,
Palo Alto, CA, USA) fitted with a 30 m Zebron ZB-SemiVolatiles
column with 0.25 mm inner diameter and 0.25 μmfilm thickness. The
GC chromatograms and mass spectra were evaluated using Agilent
MSD ChemStation (version D.02.00.237) software.
Plant Material. The leaves and fine stems from a selection of
Cyclopia genistoides bushes (Overberg type) were harvested from a
commercial plantation situated near Pearly Beach (Western Cape,
South Africa; GPS coor. −34.702, 19.618). Plant material was dried
without delay in a cross-flow drying tunnel at 40 °C for 16 h to a
moisture content of less than 10% and pulverized using a Retsch rotary
mill (1.4 mm sieve; Retsch GmbH, Haan, Germany).
Extraction and Isolation. Plant material (420 g) was extracted
with hot water (4.2 L, 93 °C) for 30 min using a ratio of 1:10 (w/v).
The crude extract was filtered through Whatman #4filter paper,
frozen, and freeze-dried using a VirTis Advantage Plus freeze-drier (SP
Scientific, Warminster, PA, USA). The freeze-dried hot water extract
(126.5 g) comprised 6.31% of 1and 0.53% of 2.
Enrichment of the crude extract in terms of 1and 2was performed
using C18 solid-phase extraction (Discovery DSC-18; 10 g/60 mL;
Sigma-Aldrich). The cartridge was conditioned sequentially with
MeOH and deionized water (60 mL). A solution of freeze-dried
extract (300 mg reconstituted in 50 mL of deionized water) was
applied to the cartridge, followed by flushing with deionized water
(100 mL). The target analytes were eluted with 5% aqueous MeOH
(300 mL). This process was repeated 48 times with new cartridges
(14.4 g of freeze-dried extract loaded), and the eluants were pooled,
vacuum-evaporated, and freeze-dried, yielding ca. 1.2 g of enriched
Figure 3. Percentage activity of rat α-glucosidase challenged with
various concentrations of 1,3-C-β-D-glucopyranosyliriflophenone, 2,
and maclurin.
Journal of Natural Products Article
dx.doi.org/10.1021/np5007247 |J. Nat. Prod. XXXX, XXX, XXX−XXXC
fraction. The enriched fraction comprised 56.5% of 1and 4.8% of 2.
The average recovery of target analytes after SPE was 75%.
The enriched fraction was reconstituted in deionized water (ca. 6
mg/mL), filtered, and subjected to semipreparative LC (MeCN−2%
HOAc (aq), 4:96, v/v) using a flow rate of 4.8 mL/min. Aliquots (400
μL) of the reconstituted, enriched fraction were injected repeatedly,
equaling 880 mg. The fractions containing 1and 2were collected
based on retention times using a Gilson FC203B fraction collector
(Gilson, Middleton, WI, USA) and pooled, and the organic solvent
was evaporated under vacuum. The remaining aqueous solutions were
filtered, frozen, and freeze-dried, yielding 370 mg of 1(purity >99% by
HRESIMS; total yield of 74% from the enriched extract) and 30 mg of
2(purity 95% by HRESIMS; total yield of 71% from the enriched
extract).
3-C-β-D-Glucopyranosyl-4-O-β-D-glucopyranosyl-
iriflophenone (1): white, amorphous powder; UV λmax online 234,
294 nm; 1H NMR 600 MHz (DMSO-d6, 298 K) δ7.61 (2H, d, J= 8.6
Hz, H-2′,6′), 6.78 (2H, d, J= 8.6 Hz, H-3′,5′), 6.24 (1H, s, H-5), 4.73
(1H, d, J= 7.8 Hz, H-1‴), 4.66 (1H, d, J= 9.8 Hz, H-1″), 3.6 (3H, m,
H-6″,2×H-6‴), 3.46 (2H, m, H-2″, H-6″), 3.25 (3H, m, H-3″, H-4″,
H-5″), 3.18 (2H, m, H-3‴, H-5‴), 3.08 (1H, t, J= 9.2 Hz, H-4‴), 2.88
(1H, t, J= 8.4 Hz, H-2″); 13C NMR 150 MHz (DMSO-d6, 298 K) δ
193.1 (C, CO), 161.8 (C, C-1′), 157.7 (C, C-1), 155.2 (C, C-2),
155.2 (C, C-4), 131.8 (CH, C-3′,5′), 130.2 (CH, C-4′), 114.8 (CH,
C-2′,6′), 110.0 (C, C-6), 106.2 (C, C-3), 100.5 (CH, C-1‴), 94.8
(CH, C-5), 81.0 (CH, C-3″), 78.0 (CH, C-5″), 77.1 (CH, C-5‴), 76.6
(CH, C-3‴), 74.9 (CH, C-1″), 73.2 (CH, C-2‴), 72.3 (CH, C-2″),
69.3 (CH, C-4″), 69.3 (CH, C-4‴), 60.5 (CH2, C-6‴), 60.1 (CH2,C-
6″); HRESIMS m/z569.1503 [M −H]−(calcd for C25H29O15,
596.1506); ESIMS [m/z(%)] 569 (100) [M −H]−, 1139 (20) [2M −
H]−; HRESIMS m/z571.1661 [M + H]+(calcd for C25H31O15,
571.1663); ESIMS [m/z(%)] 1141 (10) [2M + H]+, 571 (60) [M +
H]+, 409 (100) [M + H −162]+, 391 (60) [M + H −162 −H2O]+,
373 (10) [M + H −162 −2×H2O]+, 355 (10) [M + H −162 −3×
H2O]+, 313 (10) [M + H −258]+, 289 (15) [M + H −162 −120]+.
Acid Hydrolysis of 1. Aliquots of a solution of 1(105 mg in 70
mL of 1.1 M HCl) were heated in 5 mL glass Reacti-vials at 60 °Cina
Stuart heating block (Bibby Scientific Limited, Stone, UK) for 24 h.
The hydrolysis reaction was monitored using HPLC-DAD (86%
degradation at t= 24 h). The hydrolyzed mixture was cooled to room
temperature and adjusted to pH ∼4 using 1.1 M NaHCO3.
One half of this hydrolyzed mixture was vacuum-evaporated and
subjected to semipreparative LC (MeCN−2% HOAc (aq), 8:92, v/v)
to obtain intact 1and iriflophenone C-hexoside. The fraction
containing iriflophenone C-hexoside was collected based on retention
time, pooled, vacuum-evaporated, and freeze-dried, followed by NMR
analysis.
1H NMR and 13C NMR Data for 3-C-β-D-glucopyranosyl-
iriflophenone (hydrolyzed product of 1): 1H NMR 600 MHz
(acetone-d6, 298 K) δ7.64 (2H, d, J= 8.6 Hz, H-2′,6′), 6.85 (2H, d, J
= 8.6 Hz, H-3′,5′), 5.99 (1H, s, H-5), 4.92 (1H, d, J= 9.8 Hz, H-1″),
3.85 (2H, d, J= 3.1 Hz, 2 ×H-6″), 3.69 (H, t, J= 9.2 Hz, H-4″), 3.63
(H, t, J= 9.2 Hz, H-2″), 3.57 (H, t, J= 8.9 Hz, H-3″), 3.49 (H, m, H-
5″); 13C NMR 150 MHz (acetone-d6, 298 K) δ197.8 (CO), 162.6
(C-4′), 161.9 (C-4), 161.4 (C-2), 160.5 (C-6), 133.3 (C-1″), 132.5
(C-2′,6′), 115.3 (C-3′,5′), 106.6 (C-1), 104.7 (C-3), 96.9 (C-5), 82.1
(C-5″), 79.3 (C-3″), 76.6 (C-1″), 74.6 (C-4″), 70.6 (C-2″), 61.7 (C-
6″).
Identification of the O-Linked Hexose Residue in the
Hydrolyzed Product. The other half of the hydrolyzed product
was partially evaporated under vacuum, and a small volume of the
concentrated liquid was removed for an enzyme robot assay. This
assay was conducted on a Thermo Scientific Arena 20XT random
access chemistry analyzer (Thermo Fisher Scientific, Oy, Finland) with
the use of an Enytec fluid D-glucose no. 5140 enzyme reagent kit
Figure 4. 3H-2-DOG uptake in L6 myocytes exposed to (A) 1and (B) 3-C-β-D-glucopyranosyliriflophenone and 3T3-L1 adipocytes exposed to (C)
1and (D) 3-C-β-D-glucopyranosyliriflophenone. Insulin (positive control) and metformin (reference drug control) were included at concentrations
of 1 μM. Mean activity is expressed relative to the vehicle control at 100% ±standard error of the means. Three independent experiments were
performed each with three technical repeats. **p< 0.01; ***p< 0.001.
Journal of Natural Products Article
dx.doi.org/10.1021/np5007247 |J. Nat. Prod. XXXX, XXX, XXX−XXXD
(AEC-Amersham, Kayalami, South Africa) for the identification of
glucose according to the manufacturer’s instructions.
The remaining sample volume was freeze-dried, derivatized with
bis(trimethylsilyl)trifluoroacetamide (BSTFA) using the procedure
described by Roessner,
25
and subjected to GC-MS analysis. D-Glucose
and D-galactose standards (Sigma-Aldrich) were also derivatized prior
to analysis. Sample volumes of 1 μL were injected with a split ratio of
1:10. The injection temperature was 280 °C, the interface set to 280
°C, and the ion source adjusted to 240 °C. The carrier gas was helium
set at a constant flow rate of 1.0 mL/min. The temperature program
comprised a 6 °C oven temperature ramp from 70 to 76 °C within 1
min, followed by a 32 °C/min ramp to 300 °C, and a final 5 min
heating at 300 °C. Mass spectra were recorded over an m/zscanning
range of 40 to 650. Electron energy was 70 eV and solvent delay 8 min.
3-C-β-D-Glucopyranosylmaclurin (2): light yellow, amorphous
powder; UV λmax online 236, 290 (sh), 318 nm; 1H NMR 600 MHz
(DMSO-d6, 298 K) δ7.15 (1H, d, J= 1.93 Hz, H-2′), 7.06 (1H, dd, J
= 1.93, 8.24 Hz, H-6′), 6.74 (1H, d, J= 8.24 Hz, H-5′), 5.94 (1H, s, H-
5), 4.60 (1H, d, J= 9.77 Hz, H-1″), 3.61 (1H, d, 10.96 Hz, H-6a), 3.51
(2H, m, H-2″,6″), 3.20 (3H, H-3″,4″,5″); 13C NMR 150 MHz
(DMSO-d6, 298 K) δ194.6 (C, CO), 158.4 (C, C-4), 156.7 (C, C-
6), 156.3 (C, C-2), 150.0 (C, C-4′), 144.6 (C, C-3′), 131.0 (C, C-1′),
122.3 (CH, C-6′), 116.2 (CH, C-2′), 114.7 (CH, C-5′), 107.6 (C, C-
3), 103.6 (C, C-1), 94.8 (CH, C-5), 81.1 (CH, C-5″), 78.3 (CH, C-
3″), 74.8 (CH, C-1″), 72.1 (CH, C-4″), 69.6 (CH, C-2″), 60.4 (CH2,
C-6″); HRESIMS m/z423.0933 [M −H]−(calcd for C19H19O11,
423.0927); ESIMS [m/z(%)] 1271 (5) [3M −H]−, 847 (70) [2M −
H]−, 423 (100) [M −H]−, 303 (20) [M −H−120]−; HRESIMS m/
z425.1075 [M + H]+(calcd for C19H21O11, 425.1084); ESIMS [m/z
(%)] 425 (80) [M + H]+, 407 (100) [M + H −H2O]+, 389 (10) [M +
H−2×H2O]+, 371 (10) [M + H −3×H2O]+, 341 (10) [M + H −
84]+, 329 (15) [M + H −96]+, 305 (20) [M + H −120]+, 287 (5) [M
+H−120 −H2O]+.
α-Glucosidase Inhibitory Activity. A method for the determi-
nation of α-glucosidase inhibitory effects
26
was adapted for use on a
BioTek SynergyHT microplate reader with dispenser (BioTek
Instruments, Winooski, VT, USA). A mixture containing α-glucosidase
was extracted from rat intestinal acetone powder (Sigma-Aldrich) by
suspending ca. 350 mg of powder in 10 mL of cold KH2PO4buffer
(200 mM KH2PO4, pH 6.8 with KOH) followed by repeated
sonication on ice (sonication sequence repeated 12 times: 30 s
sonication with 25% amplitude, 1 min rest) using a model VCX600
high-intensity ultrasonic processor with a 3 mm stepped microtip
(Sonics & Materials Inc., Newton, CT, USA). The crude mixture was
centrifuged at 10000gfor 30 min at 4 °C in a Hettich Universal 320R
refrigerated centrifuge (Hettich Holding GmbH & Co. oHG,
Kirchlengern, Germany), and the supernatant was retrieved and
filtered using 0.45 μm pore size, 33 mm Millex HV PVDF filter
membranes (Merck Millipore). The supernatant was used as an
enzyme mixture after dilution to the standardized concentration based
on activity testing.
Activity determination of the enzyme mixture was performed daily
prior to each set of experiments, using the same procedure as for the
inhibition assays, but with H2O as sample controls and varying
dilutions of the enzyme mixture. Fluorescence measurements were
used to determine the correct concentration for optimal enzyme
activity, ca. 15 to 20 mg/mL of the original powder estimated as an
FL-value of 50 000 (λEX: 360 nm; λEM: 460 nm), 20 min after addition
of the substrate.
The inhibitory activities of 1,2,3-C-β-D-glucopyranosyl-
iriflophenone (Fluka, Sigma-Aldrich, St. Louis, MO, USA), and
maclurin (Sigma-Aldrich) were assessed at three concentration levels
ranging between 50 and 400 μM. Acarbose (Sigma-Aldrich), a known
inhibitor of mammalian α-glucosidase, was used as positive control at
65 μM. The following test procedure was employed: 20 μL of the
assay control (H2O), positive control, or target analyte at selected
concentration was added to 125 μL of a 200 mM KH2PO4buffer (pH
6.8) and 65 μL of the chosen enzyme dilution in 96-well black, flat-
bottom microplates with a clear bottom (Greiner Bio-One GmbH,
Rainbach im Muhlkreis, Austria). The mixture was preincubated at 37
°C for 15 min, followed by the addition of 40 μL of the substrate, 1.2
mM 7-O-α-D-glucopyranosyl-4-methylumbelliferone, by dispenser.
Fluorescence (λEX: 360 nm; λEM: 460 nm) was monitored over 30
min, and the net fluorescence (net FL) and percent enzyme activity
were calculated using the following formulas:
=−Net FL Fluorescence Fluorescence
30min 0mi
n
=×
%
Enzyme Activity 100 Net FL
Net FL
sample
assay control
Statistical analysis was performed with GraphPad Prism version 5.00
for Windows (GraphPad Software, San Diego, CA, USA, www.
graphpad.com) using one-way ANOVA with Tukey’smultiple
comparison posthoc test to determine significant differences between
values at the 95% confidence level (p< 0.05).
2-Deoxy-[3H]-D-glucose (3H-2-DOG) Uptake. To estimate in
vitro glucose uptake activity, cellular 3H-2-DOG uptake was assessed
by liquid scintillation counting, using the method described by
Mazibuko.
27
Briefly, L6 myoblasts (2.5 ×104cells/mL) and 3T3-L1
fibroblasts (2.0 ×104cells/mL) were seeded into 24-well plates in
Dulbecco’s modified Eagle’s medium (DMEM) supplemented with
10% fetal or normal calf serum, respectively. C3A hepatocytes were
seeded at 5.5 ×104cells/mL in Eagle’s minimal essential medium
(EMEM) supplemented with 10% fetal calf serum. All cells were
cultured at 37 °C in humidified air with 5% CO2. Cell culture media
DMEM, EMEM, and fetal and normal calf serum were obtained from
Lonza (Walkersville, MD, USA). The L6 myoblasts and 3T3-L1
fibroblasts were differentiated into myotubule-forming myocytes and
adipocytes, respectively, while the C3A hepatocytes were used as
semiconfluent cultures. For the 3H-2-DOG uptake experiments, cells
(1 h for L6 myocytes, and 3 h for 3T3-L1 adipocytes and C3A
hepatocytes) were exposed to 1and 3-C-β-D-glucopyranosyl-
iriflophenone at concentrations ranging from 0.001 to 100 μM.
Compounds were dissolved in DMSO and diluted with Krebs-Ringer
bicarbonate HEPES buffer (KRBH) containing 8 mM glucose (final
DMSO concentrations <0.004%). For glucose uptake determination,
cells were pulse-labeled for 15 min with 0.5 μCi/mL 3H-2-DOG
(American Radiolabeled Chemicals, Inc., St. Louis, MO, USA) in
glucose- and serum-free KRBH containing 1and 3-C-β-D-
glucopyranosyliriflophenone at the relevant concentrations. Insulin
and metformin (1,1-dimethylbiguanide hydrochloride) (Sigma-Al-
drich) both at 1 μM were included as positive and drug reference
controls, respectively. The amount of 3H-2-DOG taken up by cells was
measured using a liquid scintillation counter (2200CA Tricarb Series,
Packard Instrument Company, Downers Grove, IL, USA), and the
activity calculated as fmol (3H-2-DOG)/mg protein. Statistical analysis
was performed with GraphPad Prism version 5.00 for Windows using
one-way ANOVA with Dunnett’s multiple comparison post hoc test to
determine significant differences between values at the 95% confidence
level (p< 0.05).
■ASSOCIATED CONTENT
*
SSupporting Information
S1 (1H), S2 (13C), S3 (COSY), S4 (1D NOESY), S5 (HSQC),
and S6 (HMBC) NMR spectra of 3-C-β-D-glucopyranosyl-4-O-
β-D-glucopyranosyliriflophenone (1). S7: 1H NMR spectra for
3-C-β-D-glucopyranosyliriflophenone (obtained by acid hydrol-
ysis of 1), with an overlay of corresponding spectra for the
commercial reference standard. S8: The unreacted product of 1
is also compared to the starting material of 1to assess the
stability of the sugar moiety during the acid hydrolysis reaction.
This material is available free of charge via the Internet at
http://pubs.acs.org.
Journal of Natural Products Article
dx.doi.org/10.1021/np5007247 |J. Nat. Prod. XXXX, XXX, XXX−XXXE
■AUTHOR INFORMATION
Corresponding Author
*E-mail: beelderst1@arc.agric.za (T. Beelders). Phone: +27-21-
809-3441. Fax: +27 21-809-3430.
Notes
The authors declare no competing financial interest.
■ACKNOWLEDGMENTS
This work is based on the research supported in part by the
South African Department of Science and Technology (DST/
CON 0133/2012). Other financial support was received
through a grant from the Economic Competitive Support
Package for Agroprocessing to the ARC by the South African
Government. NRF grant holders (E.J. and C.J.M.) acknowledge
that opinions, findings, and conclusions or recommendations
expressed in any publication generated by the NRF-supported
research (IFRR Grant 85277 and Thuthuka Grant 87849) are
those of the authors and that the NRF accepts no liability
whatsoever in this regard. An NRF-DST Professional Develop-
ment Program Doctoral Scholarship to T.B. is acknowledged.
Mr. F. Joubert (Koksrivier) is acknowledged for supplying plant
material. Ms. W. Kuhn (Water Analysis Division, CAF,
Stellenbosch University) is thanked for the enzyme robot
assay, and Mr. F. Hiten (Mass Spectrometry Unit, CAF,
Stellenbosch University) for GC-MS analyses.
■REFERENCES
(1) Shaw, J. E.; Sicree, R. A.; Zimmet, P. Z. Diabetes Res. Clin. Pract.
2010,87,4−14.
(2) International Diabetes Federation. IDF Diabetes Atlas, 6th ed.;
International Diabetes Federation: Brussels, Belgium, 2013 (accessible
at http://www.idf.org/diabetesatlas).
(3) Mathers, C. D.; Loncar, D. PLoS Med. 2006,3, e442.
(4) Xiao, J.; Kai, G.; Yamamoto, K.; Chen, X. Crit. Rev. Food Sci. Nutr.
2013,53, 818−836.
(5) Park, H.; Hwang, K. Y.; Oh, K. H.; Kim, Y. H.; Lee, J. Y.; Kim, K.
Bioorg. Med. Chem. 2008,16, 284−292.
(6) Hollander, P. Diabetes Spectrum 2007,20, 159−165.
(7) Boath, A. S.; Stewart, D.; McDougall, G. J. Food Chem. 2012,135,
929−936.
(8) Liu, Y.; Zou, L.; Ma, L.; Chen, W.-H.; Wang, B.; Xu, Z.-L. Bioorg.
Med. Chem. 2006,14, 5683−5690.
(9) Feng, J.; Yang, X.-W.; Wang, R.-F. Phytochemistry 2011,72, 242−
247.
(10) Hu, X.; Xiao, Y.; Wu, J.; Ma, L. Arch. Pharm. Chem. Life Sci.
2011,344,71−77.
(11) Liu, Q.; Guo, T.; Li, W.; Li, D.; Feng, Z. Arch. Pharm. Chem. Life
Sci. 2012,345, 771−783.
(12) Joubert, E.; Joubert, M. E.; Bester, C.; De Beer, D.; De Lange, J.
H. S. Afr. J. Bot. 2011,77, 887−907.
(13) Vyas, A.; Syeda, K.; Ahmad, A.; Padhye, S.; Sarkar, F. H. Mini-
Rev. Med. Chem. 2012,12, 412−425.
(14) Kokotkiewicz, A.; Luczkiewicz, M.; Pawlowska, J.; Luczkiewicz,
P.; Sowinski, P.; Witkowski, J.; Bryl, E.; Bucinski, A. Fitoterapia 2013,
90, 199−208.
(15) Malherbe, C. J.; Willenburg, E.; De Beer, D.; Bonnet, S. L.; Van
der Westhuizen, J. H.; Joubert, E. J. Chromatogr. B 2014,951−952C,
164−171.
(16) Beelders, T.; De Beer, D.; Stander, M. A.; Joubert, L. Molecules
2014,19, 11760−11790.
(17) De Beer, D.; Schulze, A. E.; Joubert, E.; De Villiers, A.;
Malherbe, C. J.; Stander, M. A. Molecules 2012,17, 14602−14624.
(18) Kokotkiewicz, A.; Luczkiewicz, M.; Sowinski, P.; Glod, D.;
Gorynski, K.; Bucinski, A. Food Chem. 2012,133, 1373−1382.
(19) Jay, M. In The Flavonoids: Advances in Research since 1986;
Harborne, J. B., Ed.; Chapman and Hall: London, 1993; p 85.
(20) Tanaka, T.; Sueyasu, T.; Nonaka, G.-I.; Nishioka, I. Chem.
Pharm. Bull. 1984,32, 2676−2686.
(21) Li, H.; Song, F.; Xing, J.; Tsao, R.; Liu, Z.; Liua, S. J. Am. Soc.
Mass. Spectrom. 2009,20, 1496−1503.
(22) Zhang, Y.; Qian, Q.; Ge, D.; Li, Y.; Wang, X.; Chen, Q.; Gao, X.;
Wang, T. J. Agric. Food Chem. 2011,59, 11526−11533.
(23) Zhang, Y.; Liu, X.; Han, L.; Gao, X.; Liu, E.; Wang, T. Food
Chem. 2013,141, 2896−2905.
(24) Zhou, G.; Myers, R.; Li, Y.; Chen, Y.; Shen, X.; Fenyk-Melody,
J.; Wu, M.; Ventre, J.; Doebber, T.; Fujii, N.; Musi, N.; Hirshman, M.
F.; Goodyear, L. J.; Moller, D. E. J. Clin. Invest. 2001,108, 1167−1174.
(25) Roessner, U.; Wagner, C.; Kopka, J.; Trethewey, R. N.;
Willmitzer, L. Plant J. 2000,23, 131−142.
(26) Azuma, T.; Kayano, S.; Matsumura, Y.; Konishi, Y.; Tanaka, Y.;
Kikuzaki, H. Food Chem. 2011,125, 471−475.
(27) Mazibuko, M. E. In Vitro and in Vivo Effect of Aspalathus
linearis and Its Major Polyphenols on Carbohydrate and Lipid
Metabolism in Insulin Resistant Models. Ph.D. Thesis, University of
Zululand, KwaDlangezwa, South Africa, 2013.
Journal of Natural Products Article
dx.doi.org/10.1021/np5007247 |J. Nat. Prod. XXXX, XXX, XXX−XXXF