Mol Cell Biochem
DOI 10.1007/s11010-017-3108-8
Aegle marmelos differentially affects hepatic markers of glycolysis,
insulin signalling pathway, hypoxia, and inflammation in HepG2
cells grown in fructose versus glucose-rich environment
H. Aggarwal1 • J. Nair1 • P. Sharma1 • R. Sehgal1 • U. Naeem1 • P. Rajora1
R. Mathur1
Received: 7 March 2017 / Accepted: 1 July 2017
Ó Springer Science+Business Media, LLC 2017
Abstract Fructose consumption is responsible for the
onset of insulin resistance (IR), and metabolic syndrome. It
possesses no functional utility in body and its detrimental
effects on hepatic metabolic milieu are beyond those pro-
duced by glucose. The need of the hour is to identify
fructose-induced IR as an unique pathological state to be
managed differentially. The effect of aqueous leaf extract
of Aegle marmelos (AM) on hepatic markers of insulin
resistance using HepG2 cells cultured in either fructose or
glucose-rich environment is investigated. Human hepato-
cellular carcinoma cells (HepG2) were grown under stan-
dard conditions in either—DMEM without glucose (NC),
DMEM with high glucose 25 mM (Glu), DMEM-glu-
cose?0.55 mM fructose (FC1), DMEM-glucose?1 mM
fructose (FC2) or DMEM-glucose?1 mM fruc-
tose?0.1 lM insulin (FC3). The cells were treated with
either AM, rutin, quercetin, metformin or pioglitazone and
assessed for levels of hexokinase, phosphofructokinase
(PFK), aldehyde dehydrogenase, phosphatidylinositol
kinase (PI3K), signal transducer and activator of tran-
scription-3 (STAT-3), mitochondrial target of rapamycin
(mTOR), hypoxia-induced factor (HIF-1a), vascular
endothelial growth factor (VEGF) and tumour necrosis
factor (TNF-a). Summarily, when results from fructose-
& R. Mathur
mathurajani@gmail.com
1 vine fruits, berries, honey and root vegetables. However,
Department of Pharmacology, Delhi Institute of
Pharmaceutical Sciences and Research (DIPSAR), Pushp
Vihar, Sector-3, MB Road, New Delhi 110017, India
•
and glucose-rich environment were compared, then (1) IR
was more pronounced in former; (2) AM performed better
in former; (3) metformin and pioglitazone were equivocal
in either; (4) rutin and quercetin showed deviant effects
from AM; and lastly (5) effects of rutin were closer to AM
than quercetin. We hypothesize that AM ameliorates
fructose-induced IR through a mechanism which is distinct
from standard drugs and not shared by individual phyto-
constituents in toto.
Keywords Fructose
Glucose
Insulin resistance

HepG2
Aegle marmelos
Rutin
Quercetin
Introduction
Insulin resistance (IR) is the pre-diabetic state with com-
plex phenotype involving altered glucose metabolism,
dyslipidemia, impaired glucose tolerance, pro-inflamma-
tory and hypoxic state. This silent scrooge often preludes
type-II diabetes, cardiovascular and metabolic disorders
with increased risk of morbidity and mortality [1]. Glob-
ally, a quarter of adult population is said to have
unknowingly fallen prey to IR [2].
Increased prevalence and early incidence of IR are
mainly attributable to lack of physical activity and con-
sumption of fructose-rich diet [3]. Fructose or fruit sugar, is
a simple ketonic monosaccharide, and has been identified
as the most harmful of all the sugars, especially with regard
to the pathogenesis of type 2 diabetes and obesity [4]. For
thousands of years, the fructose consumption by humans
was quite stable (@16–20 g/day) and largely derived from
due to westernization of diets, the fructose consumption
has now escalated to about 85–100 g/day [5]. Most of the
123

consumed fructose is in the form of caramel, syrups and
molasses that are rampantly used in the preparation of
beverages, energizing drinks and bars, breakfast cereals,
baked goods etc. [6]. Fructose is highly lipogenic and
reportedly contributes to tissue insulin insensitivity, meta-
bolic defects, hepatic inflammation and development of a
pre-diabetic or IR state [7, 8].
Following oral ingestion, fructose is absorbed by facil-
itated transport from the jejunum into the portal vein where
the insulin facilitates its hepatic influx via glucose trans-
porters (GLUT). In the pathogenesis of IR due to chronic
ingestion of excess fructose, the following events are
reported to occur (i) paradoxical insulin independent
fructose absorption with no quantitative defects in insulin
receptor, (ii) altered hepatic metabolic milieu in favour of
lipogenesis and glycogenesis and against glycolysis and
gluconeogenesis, (iii) diminished hepatic insulin down-
stream signalling cascade (phosphatidylinositol-4,5-bis-
phosphate 3-kinase/mechanistic target of rapamycin, PI3K/
mTOR), (iv) raised oxidative stress (tumour necrosis fac-
tor-a, TNF-a), and (v) raised hypoxic state (Hypoxia
Inducible Factor-1a, HIF-1a and vascular endothelium
growth factor (VEGF)) [9, 10]. These markers are identi-
fied as hallmark characteristics of fructose-induced hepatic
IR and are widely documented and studied as therapeutic
targets [11–14].
Aegle marmelos (Rutaceae), commonly known as bael,
is a medicinal plant, the leaves of which are an important
drug in the ancient system of medicine and attributed with
anti-inflammatory, anti-microbial, anti-lipidemic and hep-
atoprotective activities [12, 15]. However, the effect of
bael leaf extract in curtailing the metabolic and hormonal
downstream alterations induced by fructose, leading to the
onset of IR remains unclear, and the present study evalu-
ates the same.
Several potentially active compounds in leaves of A.
marmelos have been identified: c-sitosterol, aegelin, lupeol,
rutin, quercetin, marmesinin, b-sitosterol, flavone, gly-
coside, O-isopentenyl halfordiol, marmeline and pheny-
lethyl cinnamamide [12]. The pharmacological activity of a
plant extract (in this case, aqueous leaf extract of A. mar-
melos) is the combined effect of the cocktail of phyto-
chemicals and cannot be pinned down on a single unique
phytoconstituent. In accordance with international guide-
lines that have been unanimously set by global agencies
like the World Health Organization (WHO), United
Nations Industrial Development Organization (UNIDO),
and International Certification Services (ICS), the plant
extract under investigation should be authenticated using a
phytochemical reference standard (PRS), which may either
be a therapeutically active compound or any compound
unique to the plant or its major phytochemical constituent,
123
Mol Cell Biochem
and the present study uses rutin to authenticate the aqueous
leaf extract of AM [16].
Materials and methods
Collection and identification of plant material
Fresh leaves of A. marmelos (Linn.) (Family-Rutaceae)
were collected locally. The taxonomy of the plant was
authenticated by Principal Scientist—National Bureau of
Plant Genetic Resources (Indian Council of Agricultural
Research) Pusa campus, New Delhi, India and the voucher
specimen is deposited at the NBPGR herbarium with ref.
NHCP/NBPGR/2014-16.
Extract preparation
The collected leaves were weighed and washed with tap
water, shade dried, powdered and mixed with distilled water
in 1:10 w/v ratio. The mixture was shaken for 2 h on a
mechanical shaker (Scientech Technologies, India) and
filtered. The filtrate was frozen at -20 °C and lyophilized at
-80 °C (Allied Frost, India) to obtain aqueous leaf extract
of A. marmelos (AM) and was stored in a desiccator.
Extract standardization: quantification of rutin
in AM
Rutin, a flavonoid reported to be present in leaves of A.
marmelos was quantitatively analysed by reversed phase
high performance liquid chromatography (RP-HPLC) using
Prominence UFLC system (Shimadzu Corporation, Kyoto,
Japan) connected by Prominence communication bus mod-
ule (CPM-20, Shimadzu Corporation, Kyoto, Japan) to data
processing software, LC solutions (Shimadzu Corporation,
Kyoto, Japan) installed in windows 7 operating system. The
chromatographic separation was performed using isocratic
elution (methanol:water::50:50, pH 2.8 with ortho-phos-
phoric acid), pumped by Prominence liquid chromatograph
(LC-20AD, Shimadzu Corporation, Kyoto, Japan) at a flow
rate of 1.0 ml/min in a reverse phase analytical enable C-18
G column (250 9 4.6 mm ID, S/N WL06-082). The system
was maintained at room temperature which yielded a column
back pressure of 2200–2400 psi. Prior to injecting samples,
the column was equilibrated for at least 30 min with the
mobile phase. The samples were injected by Hamilton syr-
inge (Microlitre TM #702, 0.0025 ml) into the 20 ll loop
Rheodyne manual sample injector with switch (Rheo-
dyneÒ Injector P/N 7725i), for run time of 10 min and the
effluent was monitored at 256 nm by Prominence diode-ar-
ray detector (SPD-M20A, Shimadzu Corporation, Kyoto,
Mol Cell Biochem
Japan). All chromatographic data were recorded and pro-
cessed using LC solution software.
Preparation of standard stock and sample solution
of rutin and AM
The standard stock solution of rutin (1 mg/ml) was diluted
with the mobile phase to give a working concentration range
of 5–80 lg/ml. The AM was dissolved in prefiltered distilled
water to get the final concentration of 1 mg/ml, filtered and
ultrasonicated. The stock (1 mg/ml) was further diluted to
500, 200, 100 lg/ml with mobile phase.
Calibration curves
Calibration curve was constructed by analysing the con-
centration of standard rutin in the range of 5–80 lg/ml. The
AM was spiked with of standard rutin (40 lg/ml) in the ratio
of 1:1 and an aliquot of 20 ll of the solution was injected.
The recovery of rutin from AM was calculated to measure
accuracy of the developed method. Further, the developed
method was validated for parameters such as linearity, limit
of detection, limit of quantification, accuracy, and precision
in accordance with ICH guidelines.
Cell culture
The human hepatocellular carcinoma cell line (HepG2)
was obtained from National Centre for Cell Sciences
(NCCS), Pune, India, and grown in culture medium con-
sisting of Dulbecco’s Modified Eagle Medium (DMEM),
supplemented with heat inactivated fetal bovine serum
(FBS, 12%) and Penicillin–Streptomycin Antibiotic (1%)
(HiMedia Laboratories, India) and cultured in a humidified
5% CO2 incubator (Shel lab, USA) at 37 °C. The culture
medium was replaced every 48 h and cells were passaged
at 80–90% confluence.
Experimental design: IR induction by insulin
and fructose treatment
Confluent flask of HepG2 cells was trypsinized (Trypsin,
0.25% ? EDTA, 0.02%), centrifuged (800 rpm 9 8 min),
re-suspended, counted, and seeded (1 9 105 cells/well/
2 ml). The cells were divided into different groups and
grown for 48 h either in—DMEM without glucose (NC),
DMEM with high glucose 25 mM (Glu), DMEM-glu-
cose?0.55 mM fructose(FC1), DMEM-glucose?1 mM
fructose (FC2) or DMEM-glucose?1 mM fruc-
tose?0.1 lM Insulin (FC3).
Cell treatment and sample collection
The cells in the different groups were treated either with
DMSO (0.1% v/v, VC Glu-VC3), AM (75 lg/ml, AM Glu-
AM3), rutin (3 lM, RUT Glu-RUT3), quercetin (7 lM,
QUER Glu-QUER3), metformin (5 mM, MET Glu-
MET3), pioglitazone (15 lM, PIO Glu-PIO3) for 48 h
under standard conditions (37 °C, 5% CO2). The super-
natant and cell lysates were collected and preserved at
-80 °C for further analysis.
Hexokinase estimation
The hexokinase activity was estimated through a coupled
reaction with glucose-6-phosphate dehydrogenase as per
the method described earlier [17]. Different concentrations
of the standard hexokinase enzyme (Merck, Germany)
were added to the samples and increase in absorbance due
to reduced NAD was recorded spectrophotometrically at
340 nm for 4 min. Change in absorbance per minute was
plotted against concentration of standard hexokinase
enzyme to calculate hexokinase activity in each sample.
Aldehyde dehydrogenase (ALDH) estimation
The aldehyde dehydrogenase (ALDH) activity was esti-
mated based on the reduction of NAD as per the method
described earlier [18]. Different concentrations of standard
ALDH enzyme (Merck, Germany) were added to samples
and increase in absorbance was recorded spectrophotomet-
rically at 340 nm for 5 min. Change in absorbance per
minute was plotted against concentration of standard ALDH
enzyme to calculate ALDH activity of in each sample.
Biochemical estimations
Biochemical estimation of Phosphofructokinase (PFK)
(Kinesis DX), PI3K (Bioassay Laboratory Technology),
p-tyr-STAT-3 (Ray Biotech), mTOR (Ray Biotech) and
HIF-1a (Ray Biotech) were conducted in the cell lysate.
Supernatant was used to estimate VEGF (Krishgen
Biosystems) and TNF-a (Krishgen Biosystems) using
respective human ELISA kits, in accordance with the
protocol detailed by the manufacturer.
Stastical analysis
All data are expressed as mean ± SD (n = 3). To evaluate
the significance of differences between the groups, one-
way ANOVA, followed by post hoc Tukey’s multiple
123

comparison test, was performed utilizing Graph pad prism
5.0.1 installed in Windows 8.1 operating system. p \ 0.05
was considered statistically significant.
Results
Quantification of rutin in AM
A sensitive isocratic RP-HPLC method was developed and
standardized with precise interday and intraday values to
give well-separated peaks of AM with LOD and LOQ as
0.18 and 0.55 lg/ml, respectively. Under these conditions,
rutin was detected at Rt = 7.12 ± 0.04 min, with linearity
between 5 and 80 lg/ml (r2 = 0.998). The concentration of
Fig. 1 RP-HPLC Chromatogram of aqueous extract of leaves of
Aegle marmelos (500 lg/ml) (a), standard rutin (40 lg/ml) (b), and
AM spiked with rutin. Well-separated, finely resolved, sharp peak of
123
Mol Cell Biochem
rutin in AM was calculated to be 8.52 lg/mg of AM
(Fig. 1).
Effect of AM on carbohydrate metabolizing enzymes
Effect of AM on hexokinase
Maximal increase in hexokinase activity was recorded in
Glu (p \ 0.05) as compared to NC (Fig. 2a). The hexoki-
nase activity was reduced in AM-Glu, RUT-Glu and
QUER-Glu but significantly increased (p \ 0.05) in MET-
Glu and PIO-Glu (Fig. 2b). AM1, RUT1, MET1 and PIO1
but not QUER1 significantly increased (p \ 0.05) hexoki-
nase activity as compared to FC1 (Fig. 2c). Hexokinase
activity was significantly increased (p \ 0.05) in AM2–3,
rutin was detected at 256 nm in extract and standard eluting at
Rt = 7.12 ± 0.04 min using methanol:water (1:1), pH 2.8, at flow
rate of 1 ml/min
Mol Cell Biochem
Fig. 2 Effect of AM (75 lg/ml) on hexokinase activity (IU/ml), a
key glycolysis enzyme. The activity of hexokinase was recorded
highest in hepatocytes grown in high glucose medium (Glu) and
lowest in high fructose medium (FC2) (a). Metformin (5 mM) and
Pioglitazone (15 lM) but not AM, rutin (3 lM) or quercetin (7 lM)
augment hexokinase activity in Glu (b). Rutin and AM augment
hexokinase activity, specifically in fructose-rich medium (FC1) and
out-perform metformin or pioglitazone (c). In fructose-induced-
insulin resistant state (FC2), hexokinase activity was significantly
elevated by AM and rutin that was greater than that by metformin or
pioglitazone (d). When insulin was externally provided to fructose-
induced insulin-resistant state (FC3), AM performed highest to
augment hexokinase activity towards sugar utilization. Each value is
the mean ± SD of n = 3. *p \ 0.05 vs. control using one-way
ANOVA, followed by post hoc Tukey’s multiple comparison test
123

RUT2–3, MET2–3 and PIO2–3 but not QUER2–3 as
compared to FC2–3 (Fig. 2d, e).
Effect of AM on phosphofructokinase (PFK)
The PFK activity was increased significantly (p \ 0.05) in
FC2 and FC3 as compared to NC (Fig. 3a). Significant
reduction (p \ 0.05) in PFK activity was recorded in AM-
Glu, RUT-Glu, QUER-Glu, MET-Glu and PIO-Glu as
compared to Glu (Fig. 3b). PFK was significantly reduced
(p \ 0.05) in AM1–3, RUT1–3 and QUER1–3 but raised
in MET1–3 and PIO1–3 as compared to FC1–3, respec-
tively (Fig. 3c–e).
Effect of AM on aldehyde dehydrogenase (ALDH)
The activity of ALDH was increased significantly
(p \ 0.05) in Glu, FC1–3 as compared to NC (Fig. 4a).
Significant elevation (p \ 0.05) in ALDH activity was
recorded in AM-Glu, RUT-Glu, QUER-Glu and PIO-Glu
as compared to Glu (Fig. 4 B). In AM1–3, RUT1–3,
QUER1–3, MET1–3 and PIO1–3, significant reduction in
ALDH activity was observed as compared to FC1–3,
respectively (Fig. 4c–e).
Effect of AM on secondary messengers of insulin
signalling
Effect of AM on phosphatidylinositol-4,5-bisphosphate
3-kinase (PI3K)
The level of PI3K was increased significantly (p \ 0.05) in
Glu and FC1 as compared to NC (Fig. 5 A). As compared
to NC, significantly reduced level of PI3K was recorded
FC2 that was elevated in FC3 (Fig. 5a). The level of PI3K
in AM-Glu, RUT-Glu, QUER-Glu, MET-Glu and PIO-Glu
remained significantly lowered as compared to Glu
(Fig. 5b). Significantly elevated level of PI3K was recor-
ded in AM1, RUT1 and PIO1 but not QUER1 and MET1,
as compared to FC1 (Fig. 5c). Level of PI3K in AM2–3,
RUT2–3, QUER2–3, MET2–3 and PIO2–3 was increased
significantly (p \ 0.05) as compared to FC2–3, respec-
tively (Fig. 5d–e).
Effect of AM on p-try-STAT-3
The level of STAT-3 was increased significantly
(p \ 0.05) in FC1–3 as compared to NC (Fig. 6a). STAT-3
remained significantly (p \ 0.05) elevated in AM-Glu,
RUT-Glu, QUER-Glu, MET-Glu, PIO-Glu as compared to
Glu (Fig. 6b). AM1–3, RUT1–3, QUER1–3 showed sig-
nificant reduction in level of STAT-3 as compared to
FC1–3, respectively (Fig. 6c–e). MET1 and PIO1 recorded
123
Mol Cell Biochem
significantly reduced level of STAT-3 as compared to FC1,
but not MET2 and PIO2 as compared FC2 (Fig. 6c, d).
Effect of AM on mTOR
The level of mTOR was increased significantly (p \ 0.05)
in Glu, as well as FC1–3, as compared to NC (Fig. 7a).
AM-Glu and PIO-Glu recorded significantly (p \ 0.05)
reduced level of mTOR as compared to Glu (Fig. 7b). The
mTOR level was significantly (p \ 0.05) lowered in
AM1–3, RUT1–3, QUER1–3, MET1–3, PIO1–3 as com-
pared to FC1–3, respectively (Fig. 7c–e).
Effect of AM on hypoxic and inflammatory
mediators
Effect of AM on VEGF
The level of VEGF was increased significantly (p \ 0.05)
in Glu, FC1–3 as compared to NC (Fig. 8a). Significant
reduction in the level of VEGF in AM Glu-AM3, RUT
Glu-RUT3, QUER Glu-QUER3, MET Glu-MET3, PIO
Glu-PIO3 was recorded as compared to Glu, FC1–3,
respectively (Fig. 8b–e).
Effect of AM on HIF-1a
The level of HIF-1a was increased significantly (p \ 0.05)
in Glu and FC2 as compared to NC (Fig. 9a). Significant
(p \ 0.05) reduction in level of HIF-1a was achieved in
AM-Glu and RUT-Glu as compared to Glu (Fig. 9b).
AM1–3, RUT1–3, QUER1–3 and PIO1–3 but not MET1–3
recorded significant lowering of HIF-1a level, as compared
to FC1–3 (Fig. 9c–e).
Effect of AM on TNFa
The levels of TNFa were increased significantly (p \ 0.05)
in Glu, as well as, FC1–3 as compared to NC (Fig. 10a).
The level of TNFa was significantly (p \ 0.05) reduced in
AM Glu-AM3, RUT Glu-RUT3, QUER Glu-QUER3,
MET Glu-MET3 and PIO Glu-PIO3 as compared to Glu-
FC3, respectively (Fig. 10b–e).
Discussion
Insulin resistance is a lifestyle-associated pathological state
that is assuming epidemic proportions, largely due to shift
in dietary preferences towards fructose-based products.
The metabolic effects of fructose as compared to eucaloric
glucose are more deleterious and exacerbated on energy
homeostasis [19].
Mol Cell Biochem
Fig. 3 Effect of AM (75 lg/ml) on concentration of phosphofruc-
tokinase (ng/ml), a major irreversible regulator of glycolysis. PFK
levels in HepG2 cells were significantly high in FC2 vs. NC. Insulin-
stimulated three-fold increase in PFK levels in FC3 vs. FC2 (a). PFK
concentration was significantly lowered in Glu following treatment
with AM, rutin (3 lM), quercetin (7 lM), as well as metformin
(5 mM) and pioglitazone (15 lM) (b). In fructose-rich environment,
FC1, PFK levels were lowered by AM, rutin and quercetin but not by
In order to study the metabolic milieu in hepatocytes
when provided with an environment that is either (i) glu-
cose deficient, (ii) high glucose (25 mM), (iii) high
metformin and pioglitazone (c). With progressive rise in fructose
concentration (FC2), concomitant fall in PFK activity is recorded in
AM-, rutin- and quercetin-treated HepG2 cells (d). In FC3, AM, rutin
and quercetin further reduced PFK levels but not metformin or
pioglitazone (e). Each value is the mean ± SD of n = 3. *p \ 0.05,
$
p \ 0.05, #p \ 0.05 vs. control using one-way ANOVA, followed by
post hoc Tukey’s multiple comparison test
fructose, (iv) higher fructose and lastly, (v) higher fruc-
tose ? insulin, HepG2 cells were grown in media as in NC
(-glucose), Glu (?25 Mm glucose), FC1 (?0.55 mM
123

Fig. 4 Effect of AM (75 lg/ml) on level of aldehyde dehydrogenase
(IU/ml), a glycogenesis enzyme. Relative level of ALDH was
significantly elevated in the order Glu \ FC2 \ FC1 = fc3 (a). Rutin
(3 lM), quercetin (7 lM), metformin (5 mM), AM and pioglitazone
(15 lM) significantly elevated ALDH levels in glucose-rich medium
fructose), FC2 (?1 mM fructose) and FC3 (?1 mM fruc-
tose?0.1 lM Insulin), respectively. The choice of provid-
ing fructose in the environment at the concentrations of
0.55 or 1 mM was made to simulate the physiological
concentration of fructose available in portal vein after a
fructose-rich meal [20]. Further, protective mechanism of
123
Mol Cell Biochem
(b). In FC1–3, ALDH levels were significantly lowered by AM, rutin,
quercetin, metformin and pioglitazone (c–e). Each value is the
mean ± SD of n = 3. *p \ 0.05 vs. control using one-way ANOVA,
followed by post hoc Tukey’s multiple comparison test
action of AM leaf extract (75 lg/ml) under the different
conditions was studied and compared with standard com-
pounds rutin (3 lM), metformin (5 mM) and pioglitazone
(15 lM).
The highlights of the present results are, firstly, it is
established that excess fructose (FC2) and excess Glu
Mol Cell Biochem
Fig. 5 Effect of AM (75 lg/ml) on the activity of PI3K (ng/ml) in
cultured HepG2 cells. Significantly elevated PI3K activity was
recorded in Glu and FC1 as compared to NC. Significantly diminished
PI3K activity, indicative of defective insulin signalling pathway, was
recorded in FC2 and slightly elevated with insulin in FC3 (a). In Glu,
the PI3K activity was significantly lower in AM, rutin (3 lM),
quercetin (7 lM), as well as metformin (5 mM) and pioglitazone
initiate divergent metabolic processes in hepatocytes and
IR-state is more definite in former than latter. Secondly, the
molecular responses elicited by AM were cohesive as
compared to, isolated phytochemicals or standard drugs, an
observation supported elsewhere [21, 22].
(15 lM) (b). Rutin, AM and pioglitazone but not metformin
improved insulin signalling by significantly increasing the activity
of PI3K in FC1 (c). Highly significant elevation of PI3K activity was
recorded with AM, rutin, quercetin, metformin and pioglitazone in
FC2 and FC3 (d, e). Each value is the mean ± SD of n = 3.
*p \ 0.05 vs. control using one-way ANOVA, followed by post hoc
Tukey’s multiple comparison test
One of the crucial rate-limiting enzymes of glycolysis is
hexokinase, that was recorded highest in the group Glu
indicating that in hepatocytes, hyperglycaemic state is first
managed by elevating mechanisms of glucose utilization.
Even though FC1 is also a hyperglycaemic environment,
123

Fig. 6 Effect of AM (75 lg/ml) on the activity (mAU) of signal
transducer and activator of transcription-3 (STAT-3), a metabolic
marker responsible for inducing insulin resistance. Glu and FC1–3
exhibited significantly increased levels of STAT-3, in cultured HepG2
cells (a). Rutin (3 lM), AM, quercetin (7 lM), metformin (5 mM)
the hexokinase activity was lower than that in Glu which
may be explained based on lower affinity constant (Km) of
hexokinase for fructose than for glucose. However, FC2
recorded the lowest hexokinase activity, suggesting that the
mechanisms that control energy homeostasis are not
effective against onslaught of fructose and the pathogenesis
of IR may have been initiated. Insulin marginally alleviates
123
Mol Cell Biochem
and pioglitazone (15 lM) significantly raised STAT-3 activity in Glu
(b). Rutin, quercetin, metformin, AM and pioglitazone inactivated
signalling molecule STAT-3 in FC1–3 (c–e). Each value is the
mean ± SD of n = 3. *p \ 0.05 vs. control using one-way ANOVA,
followed by post hoc Tukey’s multiple comparison test
fructose-induced-IR state as shown by hexokinase activity
in FC3. The standard drugs metformin and pioglitazone
augmented sugar oxidation via raised levels of hexokinase,
irrespective of sugar-type. However, AM and rutin were
more effective in reinforcement of hexokinase-mediated
glycolytic pathway to mitigate the fructose-induced IR
state.
Mol Cell Biochem
Fig. 7 Effect of AM (75 lg/ml) on the activity (mAU) of mam-
malian target of rapamycin mTOR, a marker of cellular processes
leading to insulin resistance. Glu and FC1–3 exhibited increased
mTOR activity (a). Pioglitazone (15 lM) and AM, but not rutin
(3 lM), quercetin (7 lM) or metformin (5 mM) reduced mTOR
Another key enzyme that possesses regulatory as well as
catalytic role in glycolysis is PFK. Its irreversible catalytic
action makes it the control site that sets the pace of
mammalian glycolytic pathway. Interestingly, high levels
of ATP allosterically inhibit the enzyme in the liver, thus
lowering its affinity for the substrate, fructose 6-phosphate.
Thus, the activity of PFK increases when the ATP/AMP
ratio is lowered or in other words, glycolysis is stimulated
activity in Glu (b). Fructose environment provoked the rise in mTOR
activity in FC1–3 was significantly reversed by AM, pioglitazone,
rutin, quercetin and metformin (c–e). Each value is the mean ± SD of
n = 3. *p \ 0.05, $p \ 0.05 vs. control using one-way ANOVA,
followed by post hoc Tukey’s multiple comparison test
as the energy charge falls [23]. In this light, the present
results evidence that of the three-excess sugar environment
(Glu, FC1, FC2), the energy homeostasis is skewed max-
imally in FC2. Despite high substrate availability (fruc-
tose), the PFK levels are highest, suggesting that ATP
levels may be lowest in FC2 as it may be the most con-
ducive environment for mitochondrial uncoupling and
phosphorylation inefficiency. Further studies that
123

Fig. 8 Effect of AM (75 lg/ml) on the concentration (pg/ml) of
vascular endothelium growth factor, VEGF, a marker of vascular
inflammation and insulin resistance. Glu and FC1–3 exhibited
significantly raised the concentration of VEGF (a). Rutin (3 lM),
quercetin (7 lM), metformin (5 mM), AM and pioglitazone (15 lM),
reduced VEGF concentration in Glu (b). Fructose environment
corroborate this hypothesis are required and are being
planned in our lab. As insulin is a known stimulator of PFK
[24], a three-fold rise in FC3 was expected. Significantly,
low PFK levels were observed in Glu following treatment
with AM, rutin, quercetin, as well as metformin and
pioglitazone, indicating that they may have been effective
in restoring coupling of substrate oxidation to ATP
123
Mol Cell Biochem
provoked rise in VEGF concentration in FC1–3 was significantly
reversed by AM, rutin, quercetin, metformin and pioglitazone (c–e).
Each value is the mean ± SD of n = 3. *p \ 0.05, vs. control using
one way ANOVA, followed by post hoc Tukey’s multiple comparison
test
formation. But, in fructose-rich environment (FC1), PFK
levels were lowered by AM, rutin and quercetin but not by
metformin and pioglitazone, suggesting former can possi-
bly mitigate fructose-induced mitochondrial uncoupling, an
important pharmacological property lacking in latter. With
progressive rise in fructose concentration (FC2), con-
comitant fall in PFK activity is recorded in AM-, rutin- and
Mol Cell Biochem
Fig. 9 Effect of AM (75 lg/ml) on the concentration (ng/ml) of
hypoxia-induced factor, HIF-1a, a marker of hypoxia and insulin
resistance. Glu and FC1–3 exhibited significantly raised the concen-
tration of HIF-1a (a). Rutin (3 lM), AM, quercetin (7 lM),
metformin (5 mM) and pioglitazone (15 lM), reduced HIF-1a
concentration in Glu (b). Fructose environment provoked rise in
quercetin-treated hepatocytes, further strengthening the
hypothesis. In insulin supplemented fructose-rich environ-
ment (FC3), AM, rutin and quercetin decrease PFK levels
effectively but not metformin or pioglitazone. In conclu-
sion, PFK results show that between fructose- and glucose-
induced IR state, AM, rutin and quercetin combat former
HIF-1a concentration in FC1–3 was significantly reversed by AM,
rutin, quercetin, and pioglitazone but not by metformin (c–e). Each
value is the mean ± SD of n = 3. *p \ 0.05, $p \ 0.05 vs. control
using one-way ANOVA, followed by post hoc Tukey’s multiple
comparison test
more effectively than latter, and, metformin and pioglita-
zone may act in latter but fail miserably in former.
Interestingly, the activity of PFK is also regulated by
metabolites that are formed under anaerobic conditions in
liver. It is known that excess fructose initiates anaerobic
metabolism to maintain energy homeostasis [25, 26] and
123

Fig. 10 Effect of AM (75 lg/ml) on the concentration (pg/ml) of
tumour necrosis factor, TNF-a, a marker of inflammation and insulin
resistance. Glu and FC1–3 showed significant rise in concentration of
TNF-a that was higher in glucose than in fructose-rich environment
(a). Rutin (3 lM), AM, quercetin (7 lM), metformin (5 mM) and
pioglitazone (15 lM), significantly reduced TNF-a concentration in
corroborated in present study by elevated levels of HIF-1a
and VEGF. Aegle marmelos, rutin, metformin and piogli-
tazone annulled HIF-1a and VEGF activity to mitigate
hypoxia and inflammatory responses. Further, significant
123
Mol Cell Biochem
Glu (b). Fructose environment provoked rise in TNF-a concentration
in FC1–3 was significantly reversed by AM, rutin, quercetin,
metformin and pioglitazone (c–e). Each value is the mean ± SD of
n = 3. *p \ 0.05, vs. control using one-way ANOVA, followed by
post hoc Tukey’s multiple comparison test
increase in levels of pro-inflammatory marker, TNFa, was
observed in FC groups. TNF-a is a very important cytokine
that is responsible for inflammatory assault in pathogenesis
of metabolic disorders by inhibiting insulin signalling
Mol Cell Biochem
pathways, impairing expression of glucose transporters and
increasing triacylglycerol concentration [27]. The potential
of AM, rutin, metformin and pioglitazone in reversing
inflammatory assault was demonstrated.
Aldehyde dehydrogenase belongs to a large gene family
that irreversibly catalyses the oxidation of a variety of
biological aldehydes. In the hepatic metabolism of fruc-
tose, ALDH plays an important role as it increases glyc-
erate levels that raises de novo lipogenesis and
hypertriglyceridemia. A two-and-a half-fold rise in ALDH
activity was recorded in FC1 as compared to Glu, which
highlights that fructose has greater potential for lipogenesis
and hypertriglyceridemia than glucose. Despite higher
concentration of fructose in FC2, a fall in ALDH activity
shows that functional proteins of glycolysis are affected,
but activity was improved in the presence of insulin. In
present study, AM, rutin, quercetin, metformin and
pioglitazone inhibited fructose-induced rise in ALDH but
not by glucose.
Here we present very interesting results in terms of
downstream signalling molecule, PI3K, as it clearly evi-
dences the IR state of the hepatocytes. The levels of PI3K
were recorded highest in Glu and least in FC2. The
diminishing levels of PI3K in hepatocytes with concomi-
tant rise in fructose availability, clearly points towards IR-
state in FC2 that could be marginally improved with insulin
in FC3. It was demonstrated that AM, rutin and pioglita-
zone are instrumental in activation of PI3K pathway to
ameliorate IR and increase glucose uptake along with
carbohydrate metabolism that may point towards their
insulinomimetic action.
Another downstream molecule of insulin signalling cas-
cade is STAT-3 that is linked with hypertriglyceridemic
state. In our study, AM significantly lowered STAT-3 levels
suggesting restoration of the JAK-STAT signalling pathway
to reduce the accumulation of triglycerides in the presence of
fructose but not glucose-rich environment. Rutin, metformin
and pioglitazone also reduced STAT-3 levels in fructose-rich
media. Fructose is reported to stimulate mTOR signalling
culminating to lipogenesis [28]. Present study shows that
when placed in high glucose or fructose environment, the
hepatocytes adopt the mTOR pathway to shunt excess sugar
towards lipogenesis. Aegle marmelos, rutin, metformin and
pioglitazone significantly lowered mTOR levels and reduced
the lipogenesis induced by fructose.
Taken together, this study evidences that excess fructose
subverts highly regulated glycolysis cascade and down-
regulates insulin downstream signal leading to IR in
HepG2 cells. Further, AM acts by activating PI3K,
inhibiting p-tyr-STAT-3 and hypoxic and pro-inflamma-
tory markers, and restores the levels of key enzymes in
favour of increased glycolysis and decreased lipogenesis to
reverse the fructose-induced hepatic metabolic milieu.
References
1. Kahn BB, Flier JS (2000) Obesity and insulin resistance. J Clin
Invest 106:473–481
2. Meigs JB (2003) Epidemiology of the insulin resistance syn-
drome. Curr Diab Rep 3:73–79
3. Goran MI, Dumke K, Bouret SG, Kayser B, Walker RW,
Blumberg B (2013) The obesogenic effect of high fructose
exposure during early development. Nat Rev Endocrinol 9:494–
500
4. Lustig RH (2010) Fructose: metabolic, hedonic, and societal
parallels with ethanol. J Am Diet Assoc 110(9):1307–1321.
doi:10.1016/j.jada.2010.06.008
5. Basciano H, Federico L, Adeli K (2005) Fructose, insulin resis-
tance, and metabolic dyslipidemia. Nutr Metab (Lond) 2:5
6. Douard V, Ferraris RP (2013) The role of fructose transporters in
diseases linked to excessive fructose intake. J Physiol 591:401–
414
7. Miller A, Adeli K (2008) Dietary fructose and the metabolic
syndrome. Curr Opin Gastroenterol. 2:204–209. doi:10.1097/
MOG.0b013e3282f3f4c4
8. Zavaroni I, Sander S, Scott S, Reaven GM (1980) Effect of
fructose feeding on insulin secretion and insulin action in the rat.
Metabolism 29:970–973
9. Rebollo A, Roglans N, Alegret M, Laguna JC (2012) Way back
for fructose and liver metabolism: bench side to molecular
insights. World J Gastroenterol 18(45):6552–6559. doi:10.3748/
wjg.v18.i45.6552
10. de Moura LP, Souza Pauli LS, Cintra DE, de Souza CT, da Silva
AS, Marinho R, de Melo MA, Ropelle ER, Pauli JR (2013) Acute
exercise decreases PTP-1B protein level and improves insulin
signaling in the liver of old rats. Immun Ageing 10(1):8. doi:10.
1186/1742-4933-10-8
11. Zheng X, Ke Y, Feng A, Yuan P, Zhou J, Yu Y, Wang X, Feng W
(2016) The mechanism by which amentoflavone improves insulin
resistance in HepG2 cells. Molecules 21(5):624. doi:10.3390/
molecules21050624
12. Maity P, Hansda D, Bandyopadhyay U, Mishra DK (2009) Bio-
logical activities of crude extracts and chemical constituents of
Bael, Aegle marmelos (L.) Corr. Indian J Exp Biol 47(11):849–
861
13. Wolever TM (2000) Dietary carbohydrates and insulin action in
humans. Br J Nutr. 83(1):S97–S102
14. Ueno M, Bezerra RM, Silva MS, Tavares DQ, Carvalho CR, Saad
MJ (2000) A high-fructose dietinduces changes in pp185 phos-
phorylation in muscle and liver of rats. Braz J Med Biol Res
33:1421–1427
15. Ansari P, Afroz N, Jalil S, Azad SB, Mustakim MG, Anwar S,
Haque SM, Hossain SM, Tony RR, Hannan JM (2017) Anti-
hyperglycemic activity of Aegle marmelos (L.) corr. is partly
mediated by increased insulin secretion, a-amylase inhibition,
and retardation of glucose absorption. J Pediatr Endocrinol Metab
30(1):37–47. doi:10.1515/jpem-2016-0160
16. Handa SS, Kapoor VK, Goel AK, Tandon N (eds) (2010) Phy-
tochemical reference standards of selected Indian medicinal
plants, vol 1. Medicinal PlantsUnit, Indian Council of Medical
Research, New Delhi
17. Brandstrup N, Kirk JE, Bruni C (1957) The hexokinase and
phosphoglucoisomerase activities of aortic and pulmonary artery
tissue in individuals of various ages. J Gerontol 12:166–171
18. Kraemer RJ (1968) Isolation and characterization of human liver
aldehyde dehvdrogenase. J Biol Chem 243:6402–6408
19. Elliott SS, Keim NK, Stern JS, Teff K, Havel PJ (2002) Fructose,
weight gain, and the insulin resistance syndrome. Am J Clin Nutr
76:911–922
123

20. Mayes PA (1993) Intermediary metabolism of fructose. Am J
Clin Nutr 58:754S–765S
21. Chen Q, Wang T, Li J, Wang S, Qiu F, Yu H, Zhang Y, Wang T
(2017) Effects of natural products on fructose-induced nonalco-
holic fatty liver disease (NAFLD). Nutrients 9:96
22. Tiwari AK, Rao JM (2002) Diabetes mellitus and multiple ther-
apeutic approaches of phytochemicals: present status and future
prospects. Curr Sci 83(1):30–38
23. Berg JM, Tymoczko JL, Stryer L (2002) Biochemistry, 5th edn.
Freeman WH, New York
24. Hamer MJ, Dickson AJ (1990) Control of glycolysis in cultured
chick embryo hepatocytes. Fructose 2, 6-bisphosphate content
and phosphofructokinase-1 activity are stimulated by insulin and
epidermal growth factor. Biochem J 269(3):685–690
123
Mol Cell Biochem
25. Hellerstein MK, Schwarz JM, Neese RA (1996) Regulation of
hepatic de novo lipogenesis in humans. Annu Rev Nutr 16:523–
557
26. Carroll VA, Ashcroft M (2006) Role of hypoxia-inducible factor
(HIF)-1alpha versus HIF-2alpha in the regulation of HIF target
genes in response to hypoxia, insulin-like growth factor-I, or loss
of von Hippel-Lindau function: implications for targeting the HIF
pathway. Cancer Res 66:6264–6270
27. Hyun B, Shin S, Lee A, Lee S, Song Y, Ha NJ, Cho KH, Kim K
(2013) Metformin down-regulates TNF-a secretion via suppres-
sion of scavenger receptors in macrophages. Immune Netw
13(4):123–132. doi:10.4110/in.2013.13.4.123
28. Kim J, Song G, Wu G, Bazer FW (2012) PNAS Plus: functional
roles of fructose. Proc Natl Acad Sci 109:E1619–E1628