AMPK in the Small Intestine in Normal and Pathophysiological

GENERAL
ENDOCRINOLOGY
AMPK in the Small Intestine in Normal and
Pathophysiological Conditions
Elodie Harmel, Emilie Grenier, Ali Bendjoudi Ouadda, Mounib El Chebly,
Ehud Ziv, Jean François Beaulieu, Alain Sané, Schohraya Spahis, Martine Laville,
and Emile Levy
CD
R
Research Center (E.H., E.G., A.B.O., M.E.C., A.S., S.S., E.L.), Sainte-Justine MUHC, Montreal, Quebec,
Canada, H3T 1C5; Department of Nutrition (E.H., E.G., S.S., E.L.) and Department of Biochemistry
(M.E.C.), Université de Montréal, Montreal, Quebec, Canada, H3T 1C5; Diabetes Unit (E.Z.), Division of
Internal Medicine, Hadassah Ein Kerem Hospital, 120 Jerusalem, Israel-91; Canadian Institutes for Health
Research Team on the Digestive Epithelium (J.F.B., E.L.), Department of Anatomy and Cellular Biology,
Faculty of Medicine and Health Sciences, Université de Sherbrooke, Sherbrooke, Quebec, Canada, J1H
5N4; and CRNH Rhône-Alpes (E.H., M.L.), Université Lyon 1, Institut National de la Sante´ et de la
Recherche Me´dicale Unite´ Mixte de Recherche 1060, CENS, Centre Hospitalier Lyon-Sud, F-69310 Pierre
Bénite, France
Co
pi
aa
ut
or
iza
da
p
or
The role of AMPK in regulating energy storage and depletion remains unexplored in the intestine.
This study will to define its status, composition, regulation and lipid function, as well as to examine
the impact of insulin resistance and type 2 diabetes on intestinal AMPK activation, insulin sensitivity, and lipid metabolism. Caco-2/15 cells and Psammomys obesus (P. obesus) animal models were
experimented. We showed the predominance of AMPK␣1 and the prevalence of ␣1/␤2/␥1 heterotrimer in Caco-2/15 cells. The activation of AMPK by 5-aminoimidazole-4-carboxamide ribonucleoside and metformin resulted in increased phospho(p)-ACC. However, the down-regulation of
p-AMPK by compound C and high glucose lowered p-ACC without affecting 3-hydroxy-3-methylglutaryl-coenzyme A reductase. Administration of metformin to P. obesus with insulin resistance
and type 2 diabetes led to 1) an up-regulation of intestinal AMPK signaling pathway typified by
ascending p-AMPK␣-Thr172; 2) a reduction in ACC activity; 3) an elevation of carnitine palmitoyltransferase 1; 4) a trend of increase in insulin sensitivity portrayed by augmentation of p-Akt and
phospho-glycogen synthetase kinase 3␤; 5) a reduced phosphorylation of p38-MAPK and ERK1/2;
and 6) a decrease in diabetic dyslipidemia following lowering of intracellular events that govern
lipoprotein assembly. These data suggest that AMPK fulfills key functions in metabolic processes
in the small intestine. (Endocrinology 155: 873– 888, 2014)
M
ammalian 5⬘-AMP-activated protein kinase (AMPK)
is a serine/threonine protein kinase that acts as a sensor
of cellular energy homeostasis. It is expressed as a heterotrimer consisting of one catalytic ␣-subunit and 2 regulatory
subunits (␤ and ␥). The subunits ␣ and ␤ are each encoded by
2 genes (␣1/␣2; ␤1/␤2) whereas the ␥-subunit is encoded by
3 genes (␥1/␥2/␥3), yielding 12 possible heterotrimeric com-
plexes, which provide a molecular basis for the multiple roles
of the highly conserved AMPK signaling system in nutrient
regulation and utilization in mammalian cells (1). The different complexes confer tissue specificity. Surprisingly, little
information is available about the specific AMPK complex
composition in the small intestine, an organ highly involved
in nutrient transport and metabolism.
ISSN Print 0013-7227 ISSN Online 1945-7170
Printed in U.S.A.
Copyright © 2014 by the Endocrine Society
Received August 7, 2013. Accepted December 30, 2013.
First Published Online January 7, 2014
Abbreviations: ACC, acetyl-CoA carboxylase; AICAR, 5-aminoimidazole-4-carboxamide
ribonucleoside; AMPK, 5⬘-AMP-activated protein kinase; apo, apolipoprotein; CPT1, carnitine palmitoyltransferase 1; DGAT, diacylglycerol acyltransferase; FA, fatty acid; GSK,
glycogen synthetase kinase; HMG-CoA-R, 3-hydroxy-3-methylglutaryl-coenzyme A reductase; IR, insulin resistance; LPO, lipid peroxidation; MGAT, monoacylglycerol acyltransferase; MTP, microsomal triglyceride transfer protein; MW, molecular weight; OxS, oxidative stress; PGE2, prostaglandin E2; p-GSK, phospho-GSK; PI3K, phosphatidylinositol
3-kinase; qPCR, quantitative real-time PCR; shRNA, short hairpin RNA; T2D, type 2 diabetes; TG, triglyceride.
doi: 10.1210/en.2013-1750
Endocrinology, March 2014, 155(3):873– 888
endo.endojournals.org
03/06/2014
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 May 2014. at 06:51 For personal use only. No other uses without permission. . All rights reserved.
873
AMPK in Intestinal Absorptive Cells
Endocrinology, March 2014, 155(3):873– 888
metabolic role of AMPK, experiments were performed to
test how ablation of this enzyme by genetic manipulation
in this intestinal cell model impacted on intracellular insulin signaling and lipid homeostasis. Finally, to figure out
the status of intestinal AMPK in insulin resistance (IR) and
T2D, we employed the Psammomys obesus (P. obesus)
sand rat, a unique model of metabolic syndrome, and appraised the effects of metformin.
Materials and Methods
Cell culture
CD
R
Caco-2/15 cells were grown at 37°C with 5% CO2 in Dulbecco’s modified Eagle’s Medium (Wisent Bicocenter) containing 1% penicillin-streptomycin and 1% MEM nonessential
amino acids (Wisent Bicocenter) 1% sodium pyruvate (SigmaAldrich), and 1% glutamax (GIBCO BRL), and supplemented
with 10% decomplemented fetal bovine serum (Flow Cytometry
Standards Corp.). For experiments, cells were plated at a density
of 1 ⫻ 106 cells per well on 6-well polycarbonate Transwell filter
inserts plates (Costar), permitting separate access to the upper
and lower compartments of the monolayer. Cells were cultured
for 21 days to fully differentiate for optimal lipid synthesis (17,
18), unless mentioned otherwise. The medium was refreshed
every 48 hours.
Co
pi
aa
ut
or
iza
The control of AMPK activity is complex, and the classic view is that AMPK is allosterically activated by an
increase in the intracellular AMP/ATP ratio and/or by the
phosphorylated (p)-AMPK-Thr172 within the ␣-subunit.
The enzyme is activated in response to intracellular
stresses, leading to increased AMP/ATP ratios, and hormones that regulate whole-body energy metabolism (2, 3).
AMPK is also activated by powerful glucose-lowering
drugs such as 5-aminoimidazole-4-carboxamide ribonucleoside (AICAR) and metformin (4, 5). Activation of
AMPK leads to phosphorylation of multiple downstream
targets that restore energy imbalances. In particular,
AMPK attenuates lipogenesis in organs such as the liver
and muscle, through phosphorylation and inhibition of
acetyl-CoA carboxylases (ACC), resulting in lower malonyl-CoA levels and, thus, promoting mitochondrial
␤-oxidation while simultaneously suppressing fatty acid
(FA) synthesis (6). The cholesterol de novo synthesis is
controlled by AMPK as well through phosphorylation and
inhibition of 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMG-CoA-R) (6). Currently, the AMPK subunit
profile, as well as the AMPK phosphorylation degree, activity, and role in the small intestine in normal and pathophysiological conditions, is unresolved.
The huge interest in AMPK has been generated due to
both acute and chronic effects on glucose and lipid metabolism through phosphorylation of key protein substrates, but above all because of the associations between
the metabolic syndrome and the pathways that are regulated by AMPK at the cellular level. Previous studies have
suggested that AMPK signaling may be suppressed with
obesity and that therapeutic activation of AMPK may
therefore be beneficial (7–9). At present, AMPK system
appears as a major player in the development and/or treatment of type 2 diabetes (T2D), which is considerably increased in modern industrialized societies. Importantly,
AMPK activation has been shown to improve skeletal
muscle and liver insulin sensitivity through distinct mechanisms. The central question is how this explosion in
knowledge applies to the intestinal system (10, 11). Even
if AMPK’s importance in controlling lipid metabolism in
skeletal muscle and liver has been well established (12–
14), its role in regulating energy storage and depletion in
enterocytes remains unexplored. In fact, little attention
has been paid to the gastrointestinal tract, the first system
to face nutrients, and particularly to the small intestine, an
organ closely associated with diabetic postprandial dyslipidemia (15, 16).
Because little is known to date about the status of
AMPK in the intestine, we devoted multiple efforts towards defining its structure, activation, and regulation using intestinal Caco-2/15 cells. Furthermore, to clarify the
or
Harmel et al
da
p
874
Disruption of AMPK␣1 gene expression by short
hairpin RNA (shRNA)
Exponentially growing Caco-2/15 cells were transfected either with the pGIPZ nonsilencing (Mock) lentiviral shRNAmir
vector (V2LHS_57699) or the pGIPZ vector harboring shRNAmir
expression cassette against AMPK-␣1 (V3LHS_348528) (Open
Biosystems). Because AMPK knockdown was very difficult to carry
out in fully differentiated Caco-2/15 cells, we transfected cells at
postconfluence states (4 – 6 days), and we left them achieve differentiation for additional postconfluence days (10 days) before the
different measurements. Although these cells do not have 21 days of
postconfluence, differentiation was achieved as evidenced by various biomarkers such as ohmic resistance and sucrase activity.
Knockdown of AMPK-␣1 was assessed by RT-PCR and Western
blot. Forward primer AGGAAGAATCCTGTGACAAGCAC and
reverse primer CCGATCTCTGTGGAGTAGCAGT were used for
gene modification.
Animals
P. obesus gerbils (males) from the Hebrew University colonies
were obtained from Harlan Laboratories Ltd (Jerusalem, Israel)
and traited as previously described (19 –23). Briefly, upon weaning at 3 weeks of age, the animals were maintained on a lowenergy diet containing 2.38 kcal/g (Koffolk) until the beginning
of the experiments. They were then switched to a high-energy
diet (2.93 kcal/g; Weizmann Institute of Science, Rehovot, Israel)
for 2 weeks, which differed only in lipid composition. The animals were housed individually in polypropylene cages in a constant temperature of 22–23°C and humidity-controlled animal
facility with a 12-hour light, 12-hour dark cycle as described
previously (18, 20 –23). Water and food were supplied ad libi-
03/06/2014
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 May 2014. at 06:51 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/en.2013-1750
endo.endojournals.org
tum. Tail blood was drawn for glucose and insulin determinations,
which allowed classification into 3 groups: normoglycemia/normoinsulinemia (control), normoglycemia/hyperinsulinemia (insulinresistant), and hyperglycemia/hyperinsulinemia (diabetic) according to our previous studies (19, 20). For the metformin treatment,
the drug was daily dissolved in drinking water (300 mg/kg body
weight) and administered orally for 3 weeks. Metformin concentrations in water were readjusted twice a week after measurement
of daily water intake. Some animals were daily injected with compound C (50 nM), a powerful inhibitor of AMPK, for 3 days before
being humanely destroyed. The untreated animals received drinking water without metformin ad libitum. All experimental procedures performed in the study were authorized by the Institutional
Animal Care Committee.
875
of 2 mL of 2% intralipid was orally administered by gavage in
12-hour fasted animals as described previously (22). One hour
later, Triton WR-1339 (400 mg/kg body weight) was injected,
blood samples were collected, chylomicrons were isolated by
ultracentrifugation, and TGs and apo B-48 moieties were determined as described previously (24, 25).
Microsomal TG transfer protein (MTP) assays
Intestinal microsomes were used as the source of MTP activity
that was determined by the transfer of radiolabeled triacylglycerol from donor small unilamellar vesicles as described previously (22).
Monoacylglycerol acyltransferase (MGAT) and
diacylglycerol acyltransferase (DGAT) activities
The activities of MGAT and DGAT were also determined in
microsomes as reported previously (22).
CD
R
RNA extraction and quantitative real-time PCR
(qPCR)
qPCR were performed using Quantitect SYBR Green kit (Life
Technologies) in an ABI Prism 7500 Sequence Detection System.
The analyses were performed for each gene and for glyceraldehydes 3-phosphate dehydrogenase (as a housekeeping gene) in
the same plate in triplicate for each sample. The relative mRNA
fold-changes between the 3 animal groups were calculated using
the 2-⌬⌬Ct method (29). The primers used have been listed in
Supplemental Table 1 published on The Endocrine Society’s
Journals Online web site at http://endo.endojournals.org.
iza
da
p
Briefly, the jejunum from P. obesus was washed and cut into
explants, which were transferred onto lens paper, with the mucosal side facing up in each organ culture dish according to our
previously described techniques (24, 25). After a 30-minute stabilization period, the medium was replaced with a fresh one
containing a final amount of 1.0 ␮mol/mL unlabeled oleic acid
with 0.5 ␮Ci of [14C]oleic acid in a micellar mixture (6.6 mM
sodium taurocholate, 1 mM oleic acid, 0.5 mM monoolein, 0.1
mM cholesterol, and 0.6 mM phosphatidylcholine) (26). Intestinal explants from P. obesus were cultured for 3 hours. After this
incubation period, tissue integrity was confirmed by morphologic (lighted electron microscopy [EM]) and biochemical (sucrase activity) studies.
or
Intestinal organ culture
or
Reagents and antibodies
Co
pi
aa
ut
Compound C was purchased from CALBIOCHEM. AICAR
and antibodies against phospho-AMPK␣ (p-AMPK-Thr172),
phospho-acetyl CoA carboxylase (p-ACC-Ser79), phospho-Akt/
PKB (p-Akt/PKB-Ser473), phospho-glycogen synthetase kinase
(GSK)3␤ (p-GSK3␤-Ser9), phospho-p38 MAPK (p-p38 MAPKThr180/Tyr182), p38 MAPK, ERK1/2, p-ERK1/2, ACC, and
AMPK␥1 (Cell Signaling Technology). Anticarnitine palmitoyltransferase 1 (CPT1), AMPK␤1, AMPK␤2, and apolipoprotein
(apo)-B48 were from Santa Cruz Biotechnology; AMPK␣1 and
AMPK␣2 were from Bethyl Laboratories, and antiphospho (pHMG-CoA-R) was from EMD Millipore. Anti-␤-actin antibody
and all other reagents, unless stated otherwise, were from Sigma.
Carnitine palmitoyltransferase-1 (CPT1) and ACC
activities
CPT1 (EC.2.3.1.21) and ACC activities were assessed as well
described in detail in our previous studies (27, 28).
Analytic procedures
Plasma glucose was determined by the glucose oxidase
method and insulin levels were assessed by RIA (Phadesph; Kabi
Pharmacia Diagnostics,). Plasma triglyceride (TG) and cholesterol levels were measured colorimetrically (Roche,) as described
in detail in our previous studies (19, 22).
In vivo intestinal fat absorption
To examine whether higher chylomicron secretion contributed to enhanced in vivo lipemia in diabetic P. obesus, a volume
Western blots
The Bradford assay (Bio-Rad Laboratories) was used to estimate protein concentration. Proteins were denatured in sample
buffer containing sodium dodecyl sulfate and ␤-mercaptoethanol, separated on a 4%–20% gradient SDS-PAGE, and electroblotted onto nitrocellulose membranes. Nonspecific binding
sites of the membranes were blocked with defatted milk proteins
followed by the addition of primary antibodies directed to various targeted proteins. The relative amount of primary antibody
was detected with species-specific horseradish peroxidase-conjugated secondary antibody. Even if identical protein amounts of
tissue homogenates were applied, the ␤-actin protein was used to
confirm equal loading on SDS-PAGE. Blots were developed and
the protein mass was quantitated using an HP Scanjet scanner
equipped with a transparency adapter and software (UN-SCANIT-GEL 6.1). Of note, in several experiments, we employed
Western blot detection of multiple proteins displaying either different molecular weight (MW) or similar MW (such as phosphorylated proteins) using the stripping technique. We first analyzed the expression of distinct MW proteins that
electrophoretically migrated to different places in the same blot.
Thereafter, we stripped the immunoblots and reprobed with a
second antibody using the same blot to concomitantly detect the
phosphorylated forms. Altogether, the distinct proteins and their
phosphorylated forms share the ␤-actin (43 kDa) on the same
Western blot.
03/06/2014
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 May 2014. at 06:51 For personal use only. No other uses without permission. . All rights reserved.
876
Harmel et al
AMPK in Intestinal Absorptive Cells
Endocrinology, March 2014, 155(3):873– 888
Estimation of lipid peroxidation (LPO)
LPO in the intestinal tissue content of LPOs was quantified
with a commercial kit assay (Cayman Chemical), which measures lipid peroxides directly utilizing the redox reactions with
ferrous ions.
Total plasma antioxidant capacity
The total plasma antioxidant capacity was determined as described previously (30). This method detects hydrosoluble
and/or liposoluble antioxidants by measuring the chemiluminescence inhibition time induced by 2,2-azobis (2-amidinopropane). The system was calibrated with the vitamin E analog
Trolox.
Inflammatory biomarkers
TNF␣ and prostaglandin E2 (PGE2) were measured by ELISA
kit according to the manufacturer’s (R&D Systems) protocol.
Statistical analysis
da
p
or
CD
R
Statistical analyses of data were performed with Prism 5.0
software (GraphPad Software). All values were expressed as the
mean ⫾ SEM. The data were evaluated by Student’s two tailed
t test and ANOVA, when appropriate. A P value ⬍ 0.05 was
considered statistically significant.
Pharmacologic modulation of AMPK
AMPK stimulation appears to be a fundamental process of the response of many cell types to various stresses.
Given the paucity of information in the small intestine, we
evaluated the potential of AICAR to act on AMPK in
Caco-2/15 cells. As shown in Figure 1A, treatment of
Caco-2/15 cells with AICAR induced a dose-dependent
p-AMPK-Thr172. Because no cytotoxicity was noted with
high concentrations of AICAR using 3-[4,5-diméthylthiazol-2yl]-2,5-diphényltétrazolium bromide test and cell
differentiation markers (data not shown), we selected the
concentration of 8 mM to ensure maximal effects on pAMPK-Thr172. With this concentration, the magnitude of
the p-AMPK-Thr172/AMPK ratio was maximal at 3 hours
post-AICAR addition to Caco-2/15 cells before returning
to baseline values at 8 hours (Figure 1B). Under the same
conditions, we could observe a 2-fold rise in p-ACC-Ser79
(Figure 1C) but not in p-HMG-CoA-R (Figure 1D), two
known substrates of AMPK.
Thereafter, we turned to metformin, the most used antidiabetic agent worldwide. When a dose-dependent curve
from 0.1 mM to 4 mM during 3 hours was assessed, a
10-fold increase was observed in p-AMPK in the range of
0.1– 0.5 mM (data not shown), as exemplified by the concentration of 0.5 mM metformin (Figure 1E), which induced a strong stimulation of p-ACC-Ser79 (5-fold) (Figure
1F) with no significant effect on p-HMG CoA-R (Figure
1G). On the other hand, as reported by available literature
in different types of cells, compound C at a 40 ␮M caused
an inhibition of p-AMPK-Thr172 in a time course experiment (data not shown). The inhibition effect (50%) of
compound C on p-AMPK-Thr172 was detectable as early as
the first hour of cell incubation (data not shown) and deepened at 3 hours (90%) (Figure 2A), whereas this level was
maintained for 24 hours (data not shown). As a consequence, a reduction of p-ACC-Ser79 (Figure 2B) was noted,
but once again p-HMG-CoA-R was unaffected (Figure
2C).
iza
Results
Co
pi
aa
ut
or
Identification of the subunits of the major AMPK
heterotrimeric complex
Because AMPK exists in most of the tissues as a heterotrimeric complex that confers different properties and
shows relative tissue specificity, we first identified the
composition of the major AMPK isoforms (␣, ␤, ␥) in
intestinal Caco-2/15 cells. Although qPCR analyses revealed the presence of the 7 isoforms (␣1, ␣2, ␤1, ␤2, ␥1,
␥2 and ␥3), the predominant subunits were ␣1, ␤2, and ␥1
(Supplemental Figure 1A). The data were validated by
Western blot (Supplemental Figure 1B). In order to determine the direct interactions of the catalytic subunits (␣)
with the 2 regulatory subunits (␤ and ␥), the coimmunoprecipitation technique was employed. This experimental
approach allowed us to show a more intense cooperation
between ␣1, ␤2, and ␥1 whereas ␣2 showed more affinity
for ␤1 and ␥3 (Supplemental Figure 1C). Interestingly,
differences were noted between differentiated and nondifferentiated Caco-2/15 cells: the gene expression of AMPK
isoforms ␣1, ␤2, and ␥1 was higher in differentiated than
counterparts in nondifferentiated Caco-2/15 cells (data
not shown), suggesting a distinct regulation of AMPK in
mature and immature cells. Importantly, immunofluorescent studies were also performed using human intestinal
tissue. They confirmed the presence of AMPK and illustrated its location in the apical part of the absorptive cells
(Supplemental Figure 2).
Regulation of AMPK by glucose
Regulation of AMPK by physiological and high glucose
concentrations was assessed. A significant reduction in
p-AMPK-Thr172 resulted following treatment of the intestinal Caco-2/15 cells with 5 and 25 mM of glucose compared with medium without glucose (Figure 2D) at 4 hours
because this period of time was amply sufficient to initiate
AMPK activation with 25 mM of glucose concentration
(Figure 2E). Similarly, a substantial decrease was noticed
in p-ACC-Ser79 (Figure 2F) with the addition of glucose to
the medium, but p-HMG-CoA-R remained without significant changes at 5 mM and was lowered only at 25 mM
glucose (Figure 2G).
03/06/2014
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 May 2014. at 06:51 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/en.2013-1750
endo.endojournals.org
A
B
AICAR
Control 1 mM
2 mM
877
4 mM
6 mM
8 mM
p-AMPK
AMPK
β-actin
7
**
6
*
5
4
3
2
1
5
4
3
2
1
Control 1 mM 2 mM 4 mM 6 mM 8 mM
1h
D
AICAR
8h
3h
Control
AICAR
p-HMG-COA-R
HMG-COA-R
or
ACC
3.5
3.0
**
iza
2.5
2.0
or
1.5
ut
1.0
2.0
1.5
1.0
0.5
0.0
AICAR
Control
AICAR
E
Control
Metformin
p-AMPK
AMPK
β-actin
Control
2
***
6
5
4
3
2
1
1
0
Control
Metformin
0
Metformin
HMG-COA-R
7
p-ACC/ACC
3
Control
p-HMG-COA-R
β-actin
*
G
Metformin
ACC
4
p-AMPK/AMPK
F
p-ACC
Control
Metformin
p-HMG-COA-R/HMG-COA-R
Co
pi
Control
aa
0.5
β-actin
da
p
β-actin
0.0
Control
R
Control
0
CD
C
p-ACC
p-ACC/ACC
***
6
p-HMG-COA-R/HMG-COA-R
0
p-AMPK/AMPK
p-AMPK/AMPK
7
β-actin
1.5
1.0
0.5
0.0
Control
Metformin
Figure 1. Effects of AICAR and metformin on AMPK phosphorylation and impact on its substrates in differentiated Caco-2/15 cells. AMPK
activation was estimated as a function of concentrations of AICAR (A) and incubation time (B). With the best conditions obtained, we evaluated
the effects of AICAR (8 mM and 3 hours) and metformin (0.5 mM and 3 hours) on AMPK (E) ACC (C and F) and HMG-CoA-R (D and G),
respectively. Protein mass and phosphorylation were examined by SDS-PAGE and Western blot. As noted in Materials and Methods, we employed
Western blot detection of multiple proteins with different MW or with similar MW (such as phosphorylated proteins) using stripping methods.
Values are expressed as means ⫾ SEM for 3 independent experiments performed in triplicate. *, P ⬍ .05; **, P ⬍ .001; ***, P ⬍ .0001 vs
controls.
03/06/2014
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 May 2014. at 06:51 For personal use only. No other uses without permission. . All rights reserved.
Harmel et al
AMPK in Intestinal Absorptive Cells
A
Control
B
Compound C
C
Control
p-AMPK
Compound C
Control
p-ACC
AMPK
ACC
β-actin
β-actin
1.5
1.0
0.5
HMG-COA-R
1.0
**
0.5
***
0.0
0.0
Control
Control
Compound C
D
5 mM
Control
Compound C
R
2.5
***
1.0
***
0.0
5 mM
25 mM
aa
0 mM
ut
0.5
1.5
1.0
iza
1.5
da
p
2.0
p-AMPK/AMPK
2.0
or
p-AMPK/AMPK
0.5
or
β-actin
***
0.5
0.0
***
0h
0,5h
5 mM
Co
0 mM
pi
GLUCOSE
F
1h
G
***
4h
8h
GLUCOSE
0 mM
25 mM
5 mM
25 mM
p-HMG-COA-R
HMG-COA-R
p-HMG-COA-R/HMG-COA-R
β-actin
10
p-ACC/ACC
1.0
CD
AMPK
ACC
1.5
E
25 mM
p-AMPK
p-ACC
β-actin
0.0
Compound C
GLUCOSE
0 mM
Compound C
p-HMG-COA-R
1.5
p-ACC/ACC
p-AMPK/AMPK
Endocrinology, March 2014, 155(3):873– 888
p-HMG-COA-R/HMG-COA-R
878
8
6
4
2
*
**
5 mM
25 mM
0
0 mM
β-actin
1.5
1.0
**
0.5
0.0
0 mM
5 mM
25 mM
Figure 2. Effects of compound C and glucose on AMPK and substrate phosphorylation in differentiated Caco-2/15 cells. The optimal incubation
period and concentration were determined for compound C (40 ␮M and 1 hour). Under these conditions, we determined the effects of
compound C on AMPK (A), ACC (B), and HMG-CoA-R (C). Furthermore, activation of AMPK was estimated as a function of physiological (5 mM)
and supraphysiological (25 mM) concentrations of glucose (D). To evaluate the incubation periods, we used the supraphysiological (25 mM)
concentrations of glucose (E). Under these conditions (4 hour incubation), we determined the effects of physiological (5 mM) and
supraphysiological (25 mM) glucose concentration on ACC (F) and HMG-CoA-R (G). Protein mass and phosphorylation were examined by SDSPAGE and Western blot. As noted in Materials and Methods, we employed Western blot detection of multiple proteins with different MW or with
similar MW (such as phosphorylated proteins) using stripping methods.Values are expressed as means ⫾ SEM for 3 independent experiments
performed in triplicate. *, P ⬍ .05; **, P ⬍ .001; ***, P ⬍ .0001 vs control cells or 0 mM glucose concentration.
03/06/2014
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 May 2014. at 06:51 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/en.2013-1750
endo.endojournals.org
or
CD
R
same trend of decline in the 2 experimental P. obesus
groups compared with controls. When the ratio of
p-AMPK␣ to total AMPK␣ was calculated, it was significantly declined in the jejunum from animals with IR with
only a trend of decrease in diabetic animals (Figure 5C).
We next determined how the decrease in phosphorylation and protein content in AMPK impacts on intestinal
ACC, its principal substrate, which catalyzes the carboxylation of acetyl-CoA to form malonyl-CoA, and represents the key enzyme in the control of intracellular FA
metabolism (31, 32). Our data showed a significant decrease in the p-ACC-Ser79 without any substantial changes
in the protein content in the jejunal segment of the 2 experimental P. obesus groups compared with controls (Figure 5, D and E). Therefore, when the ratio of p-ACC-Ser79
to total ACC was calculated, it was significantly reduced
in IR with a trend of decrease in T2D compared with
control animals, indicating enhanced ACC activity (Figure
5F). Accordingly, the FA synthase protein expression increases in intestinal diabetic animal groups compared with
controls (data not shown). Because lessening of AMPK
activity may also result in a decline in FA catabolism, we
investigated CPT1, the rate-limiting enzyme for ␤-oxidation of FA. A substantial decrease was noted in intestinal
CPT1 mRNA abundance (Figure 5G) and protein mass
(Figure 5H) in P. obesus animals with IR and T2D.
Given the positive relationship between AMPK and insulin sensitivity, we reasoned that the attenuation in
AMPK activation in IR and T2D conditions may be linked
to derangements in insulin signaling at the level of the
jejunum. Therefore, we determined whether the intestine
from animals with IR and T2D differed from control P.
obesus with regard to their abundance and phosphorylation of the instrumental proteins Akt, p38-MAPK, and
ERK1/2. Our results showed a trend of decrease in p-AktSer473
without significant changes in Akt protein mass in
the jejunum of IR and T2D animals compared with control
group (Figure 6A). On the other hand, our findings revealed a rise in the degree of p-p38 MAPK-Thr180/Tyr182 and
p-ERK1/2 in IR and T2D animals even though their protein mass remained unchanged (data not shown), thereby
resulting in higher ratios of p38-MAPK-Thr180/Tyr182/p38MAPK protein mass (Figure 6B) and p-ERK1/2/ERK1/2
protein expression (Figure 6C). Overall, our data indicate
an insulin signal dysfunction in link with unaccented
AMPK phosphorylation.
We also examined the impact of the subtle decrease in
p-Akt on activity of GSK3, an important Akt substrate in
the insulin-signaling pathway. In fact, mammals express 2
isoforms, GSK3␣ and GSK3␤, which share similar kinase
domains but differ considerably in their termini. Inactivation of GSK3␤ appears to be the major route by which
da
p
Lentiviral knockdown of AMPK␣1 and impact on
AMPK target proteins in Caco-2/15 cells
To define the specific role of AMPK, stable Caco-2/15
cells with AMPK␣1 knockdown were established using
shRNA constructs targeting AMPK␣1 mRNA. AMPK␣1
expression was measured in the selected clones at 10-day
differentiation in comparison with noncoding shRNA
vector control cells. A significant reduction in gene expression (⬃80%) was noted in AMPK␣1 shRNA-infected
cells compared with control cells (Mock) infected with
empty vector as measured by RT-PCR (Figure 3A). The
inhibition of AMPK␣1 protein expression (by transduction with shRNA) was confirmed by Western blot analysis
(Figure 3B). In these conditions, there was a decline in
p-AMPK-Thr172 (Figure 3C). Furthermore, AMPK␣1
knockdown led to p-ACC-Ser79 reduction (60%) (Figure
3D), suggesting an enhancement of ACC activity, but it did
not result in alterations in p-HMG-CoA-R (Figure 3E).
Interestingly, the residual AMPK responded to activation
by AICAR and to inhibition by glucose and compound C
(Figure 3F).
Co
pi
aa
ut
or
iza
Body weight and plasma biochemical parameters
in P. obesus
To get insight into the in vivo role of AMPK␣ in the
small intestine, we employed the P. obesus, a unique model
of nutritionally induced T2D (19 –23). As illustrated in
Supplemental Figure 3, a moderate weight gain was recorded in IR and T2D groups that also developed hyperinsulinemia. However, the hyperglycemic and hyperinsulinemic group acquired diabetes because the high insulin
levels failed to contain the progression of hyperglycemia
(Supplemental Figure 3).
Intestinal AMPK␣/ACC signaling pathway in P.
obesus
We subsequently focused on the influence of IR and
T2D on AMPK␣ activation status in the small intestine in
association with the main cellular insulin-signaling pathways, which are responsible for the uptake, utilization,
and secretion of several metabolic products. We first assessed the effects of IR and T2D on the gene expression of
AMPK␣1 and AMPK␣2 using qPCR. A predominance
was noted in AMPK␣1 relatively to AMPK␣2 in the jejunum (data not shown), but both isoforms were significantly reduced by the IR and diabetic conditions compared with the healthy P. obesus animals (Figure 4). Then,
we examined the protein expression of the catalytic
AMPK␣ by Western blotting and found it markedly decreased in the jejunum of animals with IR and TD2 (Figure
5A). Evaluation of the p-AMPK␣ was then undertaken. As
shown in Figure 5B, p-AMPK␣ at Thr-172 followed the
879
03/06/2014
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 May 2014. at 06:51 For personal use only. No other uses without permission. . All rights reserved.
Harmel et al
AMPK in Intestinal Absorptive Cells
A Mock
Endocrinology, March 2014, 155(3):873– 888
B Mock
AMPKα1-/-
AMPKα1
1.4
1.4
1.2
1.2
AMPKα1-/-/β-actin
β-actin
AMPKα1-/-/GAPDH
GAPDH
1.0
0.8
0.6
0.4
***
Mock
D
0.5
p-HMG-CoA-R
β-actin
da
p
1.5
1.0
iza
1.0
*
1.5
1.0
0.5
or
0.5
ut
***
0.0
Mock
AMPKα1-/-
0.001
pi
F
0.0
AMPKα1-/-
Mock
aa
AMPKα1-/-
0.0001
0.0001
Co
3.5
AMPKα1-/-
Mock
R
2.0
AMPKα1-/-
CD
1.5
P-ACC/β-actin
P-ACC
β-actin
E
AMPKα1-/-
Mock
or
AMPKα1-/-
β-actin
P-AMPK/β-actin
***
AMPKα1-/-
P-AMPK
Mock
0.6
0.0
Mock
0.0
0.8
0.2
0.0
Mock
1.0
0.4
0.2
C
AMPKα1-/-
AMPKα1
P-HMG-COA-R/β-actin
880
3.0
P-AMPK/β-actin
2.5
2.0
0.0001
1.5
0.0001
0.05
1.0
0.05
0.5
0.0
Mock
AMPKα1-/-
Mock
+ AICAR
AMPKα1-/+ ACAR
Mock
+HG
AMPKα1-/+HG
Mock
+ CC
AMPKα1-/+ CC
Figure 3. Effects of pharmacologic agents and knockdown on AMPK activation in differentiated Caco-2/15 cells. Exponentially growing Caco-2/
15 cells were transfected either with the pGIPZ nonsilencing (scramble) lentiviral shRNAmir vector (V2LHS_57699) or the pGIPZ vector harboring
shRNAmir expression cassette against AMPK-␣1 (V3LHS_348528) as described in Materials and Methods. Knockdown of AMPK-␣1 was measured
by RT-PCR (A) and Western blot (B). Protein mass and phosphorylation of AMPK at Thr172 (p-AMPK) (C), ACC at Ser79 (p-ACC) (D), and HMGCoA-R at Ser872 (p-HMG-COA-R) (E) were examined by SDS-PAGE and Western blot in normal (Mock) and genetically modified Caco-2/15 cells
for AMPK␣1 (AMPK␣1⫺/⫺). In a second step, phosphorylation of AMPK was evaluated in the presence of AICAR (8 mM), high glucose (HG, 25
mM), and compound C (CC, 40 ␮M) in Caco-2/15 cells depleted of AMPK␣1 (F). As noted in Materials and Methods, we employed Western blot
detection of multiple proteins with different MW with similar MW (such as phosphorylated proteins) using stripping methods. Values are expressed
as means ⫾ SEM for 3 independent experiments performed in triplicate. *, P ⬍ .05; ***, P ⬍ .0001 vs mock cells. GAPDH, glyceraldehyde 3phosphate dehydrogenase.
03/06/2014
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 May 2014. at 06:51 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/en.2013-1750
endo.endojournals.org
881
TGs, phospholipids, and cholesteryl
esters in cells (Supplemental Figure
140
5). As to media, metformin was ca140
pable of lessening these classes of lip120
120
ids (Figure 8A) and TG-rich lipopro100
100
tein production (Figure 8B) via
80
down-regulation of apo B-48 bio80
genesis (Figure 8C) and Sar1B GT60
60
*
Pase (Figure 8D), as well as enzy**
**
40
40
matic activities of MTP (Figure 8E),
20
MGAT (Figure 8F), and DGAT (Fig20
ure 8G) in T2D P. obesus animals. In
0
0
order to confirm the in vitro overFigure 4. AMPK␣1 and AMPK␣2 gene expression in intestinal P. obesus animals following 6production of lipoprotein assembly,
hour fasting. cDNA was synthesized from total RNA extracted from jejunal specimens. qPCR were
we examined in vivo lipid absorpperformed using Quantitect SYBR Green kit and primer targeting gene sequences of AMPK␣1,
AMPK␣2, and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (housekeeping gene). All
tion. To this end, Intralipid was
results were normalized to GAPDH, and fold changes were calculated using 2-⌬⌬Ct method. Data
orally administered by gavage to 12are expressed as means ⫾ SEM for control (CTR; n ⫽ 11), insulin resistant (IR; n ⫽ 8/group) and
hour fasted animals followed 1 hour
diabetic (T2D; n ⫽ 8) animals. *, P ⬍ .05; **, P ⬍ .001 vs controls.
later by Triton WR-1339 injection.
insulin activates glycogen synthesis (33, 34). We, there- Diabetic animals exhibited a rise in chylomicron-TG and
fore, quantified GSK3␤ (Figure 6D) as well as its phos- chylomicron-Apo B-48, but metformin treatment markphorylation extent (Figure 6E) in the intestine of P. obesus. edly reduced diabetes-induced hyperchylomicronemia
p-GSK3␤-Ser9 was decreased, which resulted in a reduced (Figure 8, H and I).
p-GSK3␤-Ser9/GSK3␤ ratio in the jejunum of IR and T2D
Taken together, these findings indicate that metformin
animals compared with controls (Figure 6F).
attenuates IR-induced harmful effects by activating
In order to gain more insight into the specific contri- AMPK and controlling lipid metabolism. To investigate
bution of AMPK to insulin signaling and lipid homeosta- the beneficial involvement of AMPK in the metforminsis, we treated diabetic P. obesus with metformin, one of mediated lipid changes, P. obesus animals on metformin
the best AMPK-activating drugs. We noted an improve- treatment were daily injected with compound C (50 nM)
ment in systemic glycemia, hyperinsulinemia, triglyceri- during 3 days before being humanely destroyed. As docdemia, and cholesterolemia (Supplemental Table 2). At umented in supplemental data (Supplemental Table 3), the
the jejunum level, metformin treatment improved some- favorable effects of metformin on AMPK, lipids, apo B-48,
what p-AMPK-Thr172 (Figure 7A) and insulin sensitivity, MTP, and Sar1B GTPase were lightened by compound C
as indicated by the increased p-Akt (Figure 7B), and de- administration.
creased p-p38-MAPK-Thr180/Tyr182 (Figure 7C) and
ERK1/2 (Figure 7D) in T2D groups compared with control animals. Because excessive oxidative stress (OxS) and Discussion
inflammation promote IR in a feed-forward cycle, we evaluated their biomarkers in the intestine of P. obesus with Native AMPK is a ␣/␤/␥ heterotrimer with multiple suband without metformin treatments. Whereas the diabetic unit isoforms encoded by 7 genes (␣1, ␣2, ␤1, ␤2, ␥1, ␥2,
condition triggered OxS (documented by high lipid per- ␥3) allowing for up to 12 possible heterotrimeric combioxide levels and weak total plasma antioxidant capacity) nations, which provide a molecular basis for the multiple
and inflammation (acknowledged by elevated TNF-␣ and roles of the highly conserved AMPK-signaling system in
PGE2), the administration of metformin led to a regular- nutrient regulation and utilization in mammalian cells (1).
For example, in human skeletal muscle, the ␣2/␤2/␥3 hetizing trend (Figure 7, E–H).
Because the main biological effect of AMPK is to down- erotrimer has been identified as the AMPK complex that
regulate FA lipogenesis, we examined the ACC enzyme, is primarily activated by exercise, whereas the activities of
the activity of which was inhibited by metformin (Supple- complexes containing the ␣1 catalytic subunit remain unmental Figure 4A). Conversely, the stimulation of CPT1 changed (35). On the other hand, in the liver, the ␣1 and
activity by metformin was recorded (Supplemental Figure ␣2 isoforms each account for half of the enzyme activity
4B). In the same line, metformin treatment was able to and are predominantly associated with the ␤1 subunit
lower lipid esterification of the major lipid classes, eg, (36). The ␣/␤/␥ combinations are worth investigating, beCTR
IR
T2D
pi
aa
ut
or
iza
da
p
or
CD
R
AMPKα2/GAPDH
B
Co
AMPKα1/GAPDH
A
03/06/2014
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 May 2014. at 06:51 For personal use only. No other uses without permission. . All rights reserved.
Harmel et al
AMPK in Intestinal Absorptive Cells
Endocrinology, March 2014, 155(3):873– 888
Co
pi
aa
ut
or
iza
da
p
or
CD
R
882
Figure 5. AMPK␣, ACC, and CPT1 expressions in P. obesus animals following 6-hour fasting. Jejunal specimens were homogenized and proteins
from the 3 animal groups were run on SDS-PAGE and immunoblotted against AMPK␣ (A) and p-AMPK␣-Thr172 (B), followed by the calculation of
the ratio of phosphorylated AMPK␣ to total AMPK␣ (C), ACC (D), and p-ACC-Ser79 (E), followed by the the calculation of the ratio of p-ACC to
total ACC (F), and determination of CPT1 (H) and ␤-actin. As noted in Materials and Methods, we employed Western blot detection of multiple
proteins with different MW or with similar MW (such as phosphorylated proteins) using stripping methods. On the other hand, cDNA was
synthesized from total RNA extracted from jejunal specimens. qPCRs were performed using Quantitect SYBR Green kit and primer targeting
sequences of CPT1 and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (housekeeping gene) (G). Data are expressed as means ⫾ SEM for
control (CTR; n ⫽ 11), IR (n ⫽ 8), and diabetic (T2D) animals (n ⫽ 8/group). *, P ⬍ .05; **, P ⬍ .001; ***, P ⬍ .0001 vs controls; #, P ⬍ .05 vs IR
animals.
03/06/2014
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 May 2014. at 06:51 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/en.2013-1750
endo.endojournals.org
883
CTR
IR
T2D
CTR
IR
T2D
P-p38 MAPK
Akt
P38 MAPK
β-actin
β-actin
P-p38 MAPK/p38 MAPK
120
100
P-Akt/Akt
80
60
40
B
CTR
IR
C
T2D
300
300
**
***
250
200
P-ERK1/2/ERK1/2
A
P-Akt Ser473
**
150
100
200
100
50
20
0
0
E
T2D
P-GSK3β
β-actin
β-actin
P-GSK3β/β-actin
80
20
0
80
60
ut
aa
40
*
pi
60
100
or
100
Co
GSK3β/β-actin
120
IR
iza
120
CTR
F
T2D
200
P-GSK3β/GSK3β
IR
or
CTR
da
p
D
GSK3β
CD
R
0
150
100
*
40
50
20
0
0
Figure 6. IR biomarkers and their phosphorylated protein forms in P. obesus animals. Data are expressed as means ⫾ SEM for control (CTR; n ⫽
11), IR (n ⫽ 8), and diabetic (T2D) animals (n ⫽ 8/group). Following homogenization of jejunal specimens, proteins from the 3 animal groups were
run on SDS-PAGE and immunoblotted against p-Akt-Ser473 (A), p-p38 MAPK-(Thr180/Tyr182) (B), p-ERK1/2 (C), GSK3␤ (D), and p-GSK3␤-Ser9 (E),
as well as against their corresponding proteins to calculate the ratio (F). As noted in Materials and Methods, we employed Western blot detection
of multiple proteins with different MW with similar MW (such as phosphorylated proteins) using stripping methods. Data are expressed as
means ⫾ SEM for control (CTR; n ⫽ 11), IR (n ⫽ 8), and diabetic (T2D; n ⫽ 8) animals. *, P ⬍ .05; **, P ⬍ .001;***, P ⬍ .0001 vs controls.
cause they confer different properties to the AMPK complexes and serve several roles within a given tissue. Our
experiments indicate that the prevailing catalytic subunit
of AMPK complex in the enterocyte is ␣1, and its preferential associated regulatory partners are ␤2 and ␥1. This
is the first demonstration for the “signature” of the AMPK
trimeric complex in the small intestine.
The ␣-subunit is encoded either by the PRKAA1 or
PRKAA2 gene whereas the ␤- and ␥-subunits are encoded
by the PRKAB1or PRKAB2 and the PRKAG1, PRKAG2,
or PRKAG3 genes, respectively (37). Because each isoform varies from its sibling forms, the complement of isoforms present in a complex can impact the role and re-
sponse of AMPK within the cell (38). At present, we do not
know the functional distinctions of the AMPK complex
containing the ␣1, ␤2, and ␥1 in the small intestine. Although the functions of the noncatalytic ␤- and ␥-subunits
remain unclear, different features have been reported for
the 2 catalytic ␣-subunits. Within the cell, the 2 isoforms
are differentially compartmentalized. The AMPK␣1 subunit is distributed evenly throughout the cytosol, whereas
the ␣2 is localized to the nucleus in times of energy demand
(39). This disparity in distribution suggests a specialized
role for different complexes in cellular regulation, with
␣1-containing complexes functioning to control signaling
pathways, whereas ␣2-containing complexes regulate
03/06/2014
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 May 2014. at 06:51 For personal use only. No other uses without permission. . All rights reserved.
884
Harmel et al
AMPK in Intestinal Absorptive Cells
Endocrinology, March 2014, 155(3):873– 888
To study the regulation of AMPK
in Caco-2/15 intestinal cells, several
activators or inhibitors were used.
150
150
AICAR was able to stimulate AMPK
A
B
whereas compound C demonstrated
*
100
100
*
its inhibitory actions. As expected
from these data, AMPK phosphorylated its ACC target in enterocytes
50
50
according to the chemical agent
stimulus but was curiously unable to
0
0
act on HMG-CoA-R. Although the
300
250
modulation of HMG-CoA-R by
C
D
**
**
200
phosphorylation/dephosphoryla200
#
tion has been well established in var#
150
ious tissues, divergent data were re100
100
ported for the intestine that has a
50
potent capability to synthesize cholesterol and plays an important role
0
0
in cholesterol and lipoprotein meF 1500
E 50
tabolism. In fact, Field et al (40)
***
40
failed to inactivate intestinal micro1000
###
somal HMG-CoA-R by Mg2⫹-ATP
30
**
whereas others showed that intesti***
20
500
*###
nal reductase might be sensitive to
10
regulation by the phosphorylation/
0
0
dephosphorylation mechanism (41).
Our data clearly indicate that HMGCoA-R in Caco-2/15 cells did not respond to AMPK.
G 15
H 60
Due to its role in maintaining en*
ergy balance, a dysfunction in
**
10
40
AMPK-signaling pathway may result in perturbations at the systemic
5
20
level that contribute to development
of metabolic disorders. In support,
0
0
there is a strong correlation between
low AMPK activation state, mainly
due to overnutrition and lack of exercise, and metabolic disorders such
as IR (42). However, little work has
Figure 7. Effect of metformin on phosphorylation of AMPK, IR markers, as well as on de novo
LPO, total plasma antioxidant capacity (TRP), and concentration of inflammatory biomarkers in
been done on the relationship bethe jejunum of diabetic P. obesus animals. Following metformin treatment for 3 weeks, treated
tween AMPK and insulin sensitivity
and nontreated animals were humanely destroyed, and the jejunum was removed. Aliquots were
via a scrutiny of insulin cascade sigincubated with insulin (100 nM for 20 minutes) before being analyzed for IR biomarkers by Western
blot. AMPK was analyzed by the same procedure, but was not incubated with insulin. LPO (E), total
naling. This aspect was specifically
plasma antioxidant capacity (TRAP) (F) as well as inflammatory markers (TNF␣ and PGE2) (G and H),
examined in the small intestine of P.
were determined using jejunal explants derived from animal groups as described in Materials and
obesus, which showed alterations in
Methods. Data are means ⫾ SEM for 4 animals/group. *, P ⬍ .05; **, P ⬍ .001; ***, P ⬍ .0001 vs
#
###
controls; , P ⬍ .05; , P ⬍ .0001 vs diabetic animals. CTR, control.
several components of the insulin receptor-signaling pathways. Indeed, a
transcription and gene expression. Additional future re- decrease in phosphatidylinositol 3-kinase (PI3K) (data not
search is requested in this critical underexplored area to shown), a trend of drop in p-Akt, and a rise in p38 and
determine the influence of ␣1, ␤2, and ␥1 on insulin sig- ERK1/2 characterized the jejunum of animals with IR and
naling and lipoprotein assembly.
T2D, suggesting a reduced insulin sensitivity in the jeju-
or
da
p
aa
PGE2
(ρg/mg protein/ml)
pi
Co
TNFα
(ρg/mg protein/ml)
ut
or
iza
LPO
(ρmol/mg protein)
TRAP
(ηM/mg protein)
CD
R
P-ERK1/2/ERK1/2
P-P38MAPK/P38MAPK
P-Akt/Akt
P-AMPK/AMPK
CTR
T2D
T2D + Metformin
03/06/2014
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 May 2014. at 06:51 For personal use only. No other uses without permission. . All rights reserved.
endo.endojournals.org
885
Co
pi
aa
ut
or
iza
da
p
or
CD
R
doi: 10.1210/en.2013-1750
Figure 8. Effect of metformin on de novo lipid synthesis, apo B biogenesis, TG-rich lipoprotein secretion, and key proteins in lipoprotein assembly
(MGAT, DGAT, Sar1B, MTP, GTPase), as well as on in vivo lipid absorption in the intestine of diabetic P. obesus. Jejunal explants from the different
groups were incubated with [14C]oleic acid substrate for 3 hours. Lipids (TG, PL, CE) in media (A) were then extracted with chloroform-methanol
(2:1, vol/vol) isolated by thin layer chromatography and quantitated. On the other hand, jejunal explants were incubated for 3 hours either with
[14C]oleic acid or methionine-free medium containing [35S]methionine to assess lipoprotein production and apo B-48 biogenesis, respectively (B
and C). Protein expression of Sar1B GTPase was run on SDS-PAGE and immunoblotted (D), whereas the activity of MTP (E), MGAT (F), and DGAT
(G) was enzymatically measured. Intralipid was orally administered by gavage to 12-hour fasted animals followed 1 hour later by Triton WR-1339
injection as described in Materials and Methods to TG (H) and apo B-48 (I) content in chylomicrons (CM). Data are means ⫾ SEM for 4 animals/
group. *, P ⬍ .05; **, P ⬍ .001; ***, P ⬍ .0001 vs control and diabetic animals. PL, phospholipids; CE, cholesteryl ester.
03/06/2014
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 May 2014. at 06:51 For personal use only. No other uses without permission. . All rights reserved.
AMPK in Intestinal Absorptive Cells
Endocrinology, March 2014, 155(3):873– 888
CD
R
denced by lipid peroxidation that probably surpasses the
capacity of the antioxidant defenses to detoxify them in
the small intestine of diabetic animals. Because an increase
in OxS-derived inflammation has been hypothesized to be
a major mechanism in the pathogenesis and progression of
IR, we examined inflammatory markers and detected a
rise in the concentrations of TNF␣ and PGE2 in the intestinal tissue of diabetic animals. These findings emphasize that the OxS machinery and inflammatory signaling
are interrelated in the small intestine with the ensuing positive impact on the production of chylomicrons. Importantly, activation of AMPK by metformin mediates an antiinflammatory and antioxidative phenotype in the gut
along with beneficial effects on the genesis of apo B48containing lipoproteins and postprandial lipids, which
may potentially improve the proatherosclerosis condition
in diabetes. The present work stresses that AMPK activation by metformin may counteract OxS and inflammation
in the small intestine of diabetic P. obesus. OxS and inflammation are major stress conditions present in the metabolic syndrome (48). Previous studies have reported that
AMPK induction by its metformin agonist was able to
reduce the systemic inflammation by decreasing the level
of inflammatory markers in metabolic syndrome (49) and
by lowering plasma macrophage migration-inhibitory factor concentrations in obese patients (50). Similarly, a few
investigations have revealed a close link between the reduced AMPK activity and inflammation in adipose tissue
and macrophages (51, 52). These experiments indicate
that the activity of AMPK regulates the oxidative and inflammatory responses, which may induce IR and impact
on lipoprotein production. Because metformin is not a
specific AMPK activator, various investigators have demonstrated the direct effect of AMPK on nuclear factor-␬B
that represents the principal pathway involved in the activation of both the innate and adaptive immune systems
(53, 54).
Although numerous studies have provided evidence directly linking metformin to AMPK-signaling pathway for
the various metabolic effects, mounting evidence suggests
that other AMPK-independent mechanisms are involved,
including the mitochondrial machinery. Therefore, to uncover the direct effects of AMPK, we treated the P. obesus
animals with compound C, and intestinal lipoprotein assembly was reevaluated. Our data clearly showed that
inhibition of AMPK by compound C attenuated the favorable effects of metformin, thereby suggesting AMPK
involvement. Therefore, conserving AMPK activity at
normal levels can protect against harmful effects on lipid
homeostasis.
Of note, few experiments (n ⫽ 3) were carried out in
control P. obesus animals in fasting (48 hour) and refeed-
Co
pi
aa
ut
or
iza
num of IR and T2D animals because defects in PI3K/Akt
signaling have been implicated in IR (43) and T2D (44).
The activation of this pathway is initiated at the plasma
membrane, where phosphatidylinositol (44, 45) trisphosphate, generated by PI3K recruits Akt to the membrane
for phosphorylation and activation, which allows it to
adopt an active conformation and proceeds to phosphorylate a variety of protein substrates involved in insulin
signaling and diverse cellular processes. Therefore, concomitantly to the reduced activation of jejunal AMPK in
P. obesus, a marked decrease was noticed in PI3K that
plays a pivotal role for the metabolic effects of insulin,
with a trend of drop on Akt, however sufficient to lessen
p-GSK3. Importantly, the coupling of Akt and GSK3 leads
to inverse changes in their activities: when Akt activity is
high, GSK3 is maintained in a serine-phosphorylated, inhibited state, whereas decreases in Akt activity lead to
dephosphorylation and activation of GSK3 (46). Herein
we report a decrease in p-Akt and p-GSK3␤, which may
represent an important signaling pathway in the intestine
of P. obesus with IR and T2D. Particularly, the decline in
p-GSK3␤ may impact on the intracellular lipid metabolism, because elevated p-GSK3 promotes proteasomal
degradation of sterol-regulatory element-binding proteins, a transcription factor that turns on the expression of
genes involved in lipid biosynthesis (47). In fact, the protein expression of sterol-regulatory element-binding proteins was higher in the small intestine of insulin-resistant
and diabetic P. obesus animals compared with controls
(data not shown) and coincided with abnormal lipid synthesis, apo B-48 biogenesis, and TG-rich lipoprotein
production.
Because metformin is a powerful activator of AMPK,
we anticipated that its administration to P. obesus may
restore the aforementioned intestinal abnormalities in insulin signaling, lipid homeostasis, and lipoprotein metabolism. Such an experimental approach can help us identify, for example, the cross talk between AMPK and
insulin signaling in the enterocyte. Our hypothesis proved
true in view of the beneficial effects of metformin on restoration of AMPK activation, improvement of systemic
IR, amelioration of enterocyte insulin sensitivity, and restitution of lipid/lipoprotein homeostasis.
An important question that was addressed in the present work was whether the small intestine of animals with
metabolic disorders is endowed with impaired insulin responsiveness/signaling. Moreover, it was crucial to determine whether the mechanisms include OxS and inflammation. Finally, efforts were necessary to explore whether
these pathologic conditions trigger apo B-48-containing
lipoprotein overproduction in the small intestine of diabetic animals with IR. In fact, significant OxS was evi-
or
Harmel et al
da
p
886
03/06/2014
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 May 2014. at 06:51 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/en.2013-1750
endo.endojournals.org
or
CD
R
2. Hardie DG. AMP-activated/SNF1 protein kinases: conserved guardians of cellular energy. Nat Rev Mol Cell Biol. 2007;8:774 –785.
3. Kemp BE, Stapleton D, Campbell DJ, et al. AMP-activated protein
kinase, super metabolic regulator. Biochem Soc Trans. 2003;31:
162–168.
4. Wojtaszewski JF, Jørgensen SB, Hellsten Y, Hardie DG, Richter EA.
Glycogen-dependent effects of 5-aminoimidazole-4-carboxamide
(AICA)-riboside on AMP-activated protein kinase and glycogen synthase activities in rat skeletal muscle. Diabetes. 2002;51:284 –292.
5. Hayashi T, Hirshman MF, Kurth EJ, Winder WW, Goodyear LJ.
Evidence for 5⬘ AMP-activated protein kinase mediation of the effect
of muscle contraction on glucose transport. Diabetes. 1998;47:
1369 –1373.
6. Carling D, Hardie DG. The substrate and sequence specificity of the
AMP-activated protein kinase. Phosphorylation of glycogen synthase and phosphorylase kinase. Biochim Biophys Acta. 1989;1012:
81– 86.
7. Brusq JM, Ancellin N, Grondin P, et al. Inhibition of lipid synthesis
through activation of AMP kinase: an additional mechanism for the
hypolipidemic effects of berberine. J Lipid Res. 2006;47:1281–
1288.
8. Chen MB, McAinch AJ, Macaulay SL, et al. Impaired activation of
AMP-kinase and fatty acid oxidation by globular adiponectin in
cultured human skeletal muscle of obese type 2 diabetics. J Clin
Endocrinol Metab. 2005;90:3665–3672.
9. Dolinsky VW, Dyck JR. Role of AMP-activated protein kinase in
healthy and diseased hearts. Am J Physiol Heart Circ Physiol. 2006;
291:H2557–H2569.
10. Zang M, Zuccollo A, Hou X, et al. AMP-activated protein kinase is
required for the lipid-lowering effect of metformin in insulin-resistant human HepG2 cells. J Biol Chem. 2004;279:47898 – 47905.
11. Zang M, Xu S, Maitland-Toolan KA, et al. Polyphenols stimulate
AMP-activated protein kinase, lower lipids, and inhibit accelerated
atherosclerosis in diabetic LDL receptor-deficient mice. Diabetes.
2006;55:2180 –2191.
12. Kahn BB, Alquier T, Carling D, Hardie DG. AMP-activated protein
kinase: ancient energy gauge provides clues to modern understanding of metabolism. Cell Metab. 2005;1:15–25.
13. Ruderman NB, Saha AK, Kraegen EW. Minireview: malonyl CoA,
AMP-activated protein kinase, and adiposity. Endocrinology. 2003;
144:5166 –5171.
14. Zong H, Ren JM, Young LH, et al. AMP kinase is required for
mitochondrial biogenesis in skeletal muscle in response to chronic
energy deprivation. Proc Natl Acad Sci USA. 2002;99:15983–
15987.
15. Hsieh J, Hayashi AA, Webb J, Adeli K. Postprandial dyslipidemia in
insulin resistance: mechanisms and role of intestinal insulin sensitivity 1. Atheroscler Suppl. 2008;9:7–13.
16. Adeli K, Lewis GF. Intestinal lipoprotein overproduction in insulinresistant states. Curr Opin Lipidol. 2008;19:221–228.
17. Grenier E, Maupas FS, Beaulieu JF, et al. Effect of retinoic acid on
cell proliferation and differentiation as well as on lipid synthesis,
lipoprotein secretion, and apolipoprotein biogenesis. Am J Physiol
Gastrointest Liver Physiol. 2007;293:G1178 –G1189.
18. Levy E, Stan S, Delvin E, et al. Localization of microsomal triglyceride transfer protein in the Golgi: possible role in the assembly of
chylomicrons. J Biol Chem. 2002;277:16470 –16477.
19. Levy E, Spahis S, Ziv E, et al. Overproduction of intestinal lipoprotein containing apolipoprotein B-48 in Psammomys obesus: impact
of dietary n-3 fatty acids. Diabetologia. 2006;49:1937–1945.
20. Levy E, Lalonde G, Delvin E, et al. Intestinal and hepatic cholesterol
carriers in diabetic Psammomys obesus. Endocrinology. 2010;151:
958 –970.
21. Zoltowska M, Ziv E, Delvin E, et al. Circulating lipoproteins and
hepatic sterol metabolism in Psammomys obesus prone to obesity,
hyperglycemia and hyperinsulinemia. Atherosclerosis. 2001;157:
85–96.
Acknowledgments
Co
pi
aa
ut
or
iza
da
p
ing. These states up-regulated (908 ⫾ 113) and downregulated (533 ⫾ 61) jejunal AMPK activity (nmol/g tissue/hr), respectively, compared with controls (675 ⫾ 87)
(P ⬍ .05), which confirms the role of AMPK as an enterocyte regulator and sensor of energy homeostasis. However, additional experiments are definitely needed to confirm the data and to extend the studies to other diabetic
tissues in P. obesus animals especially because fasting increased the AMP/ATP ratio and AMPK in adipose tissue,
but not in muscle or liver tissue of Wistar rats (55).
In conclusion, our experiments provide strong biochemical evidence that AMPK has its proper signature and
responds to established pharmacologic agents in the Caco2/15 cell line, a remarkable intestinal model. IR and T2D
alter AMPK phosphorylation in the small intestine, which
impacts on key target proteins controlling metabolic pathways of insulin signaling, FA oxidation, and lipid/lipoprotein synthesis in the P. obesus, a unique polygenic animal model for obesity and nutritionally induced T2D,
which has been studied extensively in our laboratory. Restoration of AMPK activity with metformin alleviated various aforementioned abnormalities, particularly those related to dysregulation of lipid homeostasis as highlighted
by compound C, a powerful AMPK inhibitor. Therefore,
our studies showed that the salutary effects of metformin
on the elevated lipids associated with IR and T2D states
depend on AMPK and start in the gastrointestinal tract,
the first system to face obesogenic nutrients and antidiabetic drugs, to be closely associated with diabetic postprandial lipemia and diet-induced intestinal OxS and inflammation, which represent significant mediators for the
development of obesity and IR.
We thank Dr Randy W Schekman (Howard Hugues Medical
institute Reagents of the University of California) for providing
Sar1b antibody.
Address all correspondence and requests for reprints to:
Dr. Emile Levy, Research Centre, Gastroenterology, hepatology and nutrition Unit, CHU-Sainte-Justine, 3175 Côte Ste
Catherine, Montréal, Québec, Canada H3T 1C5. E-mail:
[email protected]
This work was supported by grants from the Canadian Institutes of Health Research (MOP 10584), the Canadian Institutes of Health Research Team Grant (CTP-82942), and the JA
deSève Research Chair in Nutrition.
Disclosure Summary: The authors have nothing to disclose
References
1. Gao G, Fernandez CS, Stapleton D, et al. Non-catalytic ␤- and ␥-subunit isoforms of the 5⬘-AMP-activated protein kinase. J Biol Chem.
1996;271:8675– 8681.
887
03/06/2014
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 May 2014. at 06:51 For personal use only. No other uses without permission. . All rights reserved.
AMPK in Intestinal Absorptive Cells
Endocrinology, March 2014, 155(3):873– 888
41.
42.
43.
44.
45.
46.
CD
47.
48.
49.
ut
or
50.
51.
aa
pi
Co
complexes containing the alpha2 isoform. Biochem J. 1998;334 ( Pt
1):177–187.
Field FJ, Erickson SK, Shrewsbury MA, Cooper AD. 3-Hydroxy-3methylglutaryl coenzyme A reductase from rat intestine: subcellular
localization and in vitro regulation. J Lipid Res. 1982;23:105–113.
Oku H, Ide T, Sugano M. Reversible inactivation-reactivation of
3-hydroxy-3-methylglutaryl coenzyme A reductase of rat intestine.
J Lipid Res. 1984;25:254 –261.
Roduit R, Nolan C, Alarcon C, et al. A role for the malonyl-CoA/
long-chain acyl-CoA pathway of lipid signaling in the regulation of
insulin secretion in response to both fuel and nonfuel stimuli. Diabetes. 2004;53:1007–1019.
Zhu S, Sun F, Li W, et al. Apelin stimulates glucose uptake through
the PI3K/Akt pathway and improves insulin resistance in 3T3–L1
adipocytes. Mol Cell Biochem. 2011;353:305–313.
Farese RV, Sajan MP, Standaert ML. Insulin-sensitive protein kinases (atypical protein kinase C and protein kinase B/Akt): actions
and defects in obesity and type II diabetes. Exp Biol Med (Maywood). 2005;230:593– 605.
Bozulic L, Hemmings BA. PIKKing on PKB: regulation of PKB activity by phosphorylation. Curr Opin Cell Biol. 2009;21:256 –261.
White MF. Insulin signaling in health and disease. Science. 2003;
302:1710 –1711.
Sundqvist A, Bengoechea-Alonso MT, Ye X, et al. Control of lipid
metabolism by phosphorylation-dependent degradation of the
SREBP family of transcription factors by SCF(Fbw7). Cell Metab.
2005;1:379 –391.
Roberts CK, Sindhu KK. Oxidative stress and metabolic syndrome.
Life Sci. 2009;84:705–712.
Akbar DH. Effect of metformin and sulfonylurea on C-reactive protein level in well-controlled type 2 diabetics with metabolic syndrome. Endocrine. 2003;20:215–218.
Dandona P, Aljada A, Ghanim H, et al. Increased plasma concentration of macrophage migration inhibitory factor (MIF) and MIF
mRNA in mononuclear cells in the obese and the suppressive action
of metformin. J Clin Endocrinol Metab. 2004;89:5043–5047.
Gauthier MS, O’Brien EL, Bigornia S, et al. Decreased AMP-activated protein kinase activity is associated with increased inflammation in visceral adipose tissue and with whole-body insulin resistance
in morbidly obese humans. Biochem Biophys Res Commun. 2011;
404:382–387.
Yang Z, Kahn BB, Shi H, Xue BZ. Macrophage ␣1 AMP-activated
protein kinase (␣1AMPK) antagonizes fatty acid-induced inflammation through SIRT1. J Biol Chem. 2010;285:19051–19059.
Bai A, Ma AG, Yong M, et al. AMPK agonist downregulates innate
and adaptive immune responses in TNBS-induced murine acute and
relapsing colitis. Biochem Pharmacol. 2010;80:1708 –1717.
Wu X, Mahadev K, Fuchsel L, Ouedraogo R, Xu SQ, Goldstein BJ.
Adiponectin suppresses I␬B kinase activation induced by tumor necrosis factor-␣ or high glucose in endothelial cells: role of cAMP and
AMP kinase signaling. Am J Physiol Endocrinol Metab. 2007;293:
E1836 –E1844.
Kajita K, Mune T, Ikeda T, et al. Effect of fasting on PPAR␥ and
AMPK activity in adipocytes. Diabetes Res Clin Pract. 2008;81:
144 –149.
R
40.
iza
22. Zoltowska M, Ziv E, Delvin E, et al. Cellular aspects of intestinal
lipoprotein assembly in Psammomys obesus: a model of insulin resistance and type 2 diabetes. Diabetes. 2003;52:2539 –2545.
23. Zoltowska M, Ziv E, Delvin E, Lambert M, Seidman E, Levy E. Both
insulin resistance and diabetes in Psammomys obesus upregulate the
hepatic machinery involved in intracellular VLDL assembly. Arterioscler Thromb Vasc Biol. 2004;24:118 –123.
24. Levy E, Thibault L, Me´nard D. Intestinal lipids and lipoproteins in
the human fetus: modulation by epidermal growth factor. J Lipid
Res. 1992;33:1607–1617.
25. Levy E, Me´nard D, Delvin E, et al. The polymorphism at codon 54
of the FABP2 gene increases fat absorption in human intestinal explants. J Biol Chem. 2001;276:39679 –39684.
26. Levy E, Marcel Y, Deckelbaum RJ, et al. Intestinal apoB synthesis,
lipids, and lipoproteins in chylomicron retention disease. J Lipid
Res. 1987;28:1263–1274.
27. Montoudis A, Seidman E, Boudreau F, et al. Intestinal fatty acid
binding protein regulates mitochondrion ␤-oxidation and cholesterol uptake. J Lipid Res. 2008;49:961–972.
28. Ben Djoudi Ouadda A, Levy E, Ziv E, et al. Increased hepatic lipogenesis in insulin resistance and type 2 diabetes is associated with
AMPK signalling pathway up-regulation in Psammomys obesus.
Biosci Rep. 2009;29:283–292.
29. Livak KJ, Schmittgen TD. Analysis of relative gene expression data
using real-time quantitative PCR and the 2(-⌬⌬ C(T)) method.
Methods. 2001;25:402– 408.
30. Santini SA, Marra G, Giardina B, et al. Defective plasma antioxidant
defenses and enhanced susceptibility to lipid peroxidation in uncomplicated IDDM. Diabetes. 1997;46:1853–1858.
31. Kim YS, Kolattukudy PE. Malonyl-CoA decarboxylase from the
uropygial gland of waterfowl: purification, properties, immunological comparison, and role in regulating the synthesis of multimethylbranched fatty acids. Arch Biochem Biophys. 1978;190:585–597.
32. Ruderman NB, Saha AK, Vavvas D, Witters LA. Malonyl-CoA, fuel
sensing, and insulin resistance. Am J Physiol. 1999;276:E1–E18.
33. Mora A, Sakamoto K, McManus EJ, Alessi DR. Role of the PDK1PKB-GSK3 pathway in regulating glycogen synthase and glucose
uptake in the heart. FEBS Lett. 2005;579:3632–3638.
34. McManus EJ, Sakamoto K, Armit LJ, et al. Role that phosphorylation of GSK3 plays in insulin and Wnt signalling defined by
knockin analysis. EMBO J. 2005;24:1571–1583.
35. Birk JB, Wojtaszewski JF. Predominant ␣2/␤2/␥3 AMPK activation
during exercise in human skeletal muscle. J Physiol. 2006;577:
1021–1032.
36. Stapleton D, Woollatt E, Mitchelhill KI, et al. AMP-activated protein kinase isoenzyme family: subunit structure and chromosomal
location. FEBS Lett. 1997;409:452– 456.
37. Cheung PC, Salt IP, Davies SP, Hardie DG, Carling D. Characterization of AMP-activated protein kinase ␥-subunit isoforms and
their role in AMP binding. Biochem J. 2000;346(Pt. 3):659 – 669.
38. Moffat C, Ellen Harper M. Metabolic functions of AMPK: aspects
of structure and of natural mutations in the regulatory ␥ subunits.
IUBMB Life. 2010;62:739 –745.
39. Salt I, Celler JW, Hawley SA, et al. AMP-activated protein kinase:
greater AMP dependence, and preferential nuclear localization, of
or
Harmel et al
da
p
888
52.
53.
54.
55.
03/06/2014
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 30 May 2014. at 06:51 For personal use only. No other uses without permission. . All rights reserved.

Download Report