Archives of Biochemistry and Biophysics 505 (2011) 231–241
Contents lists available at ScienceDirect
Archives of Biochemistry and Biophysics
journal homepage: www.elsevier.com/locate/yabbi
Supplement of TCA cycle intermediates protects against high
glucose/palmitate-induced INS-1 beta cell death
Sung-E Choi a,1, Youn-Jung Lee a,1, Geum-Sook Hwang b,1, Joo Hee Chung b, Soo-Jin Lee a, Ji-Hyun Lee a,
Seung Jin Han c, Hae Jin Kim c, Kwan-Woo Lee c, Youngsoo Kim d, Hee-Sook Jun e, Yup Kang a,⇑
a
Institute for Medical Sciences, Ajou University School of Medicine, Suwon, Republic of Korea
Korea Basic Science Institute, Seoul, Republic of Korea
Department of Endocrinology and Metabolism, Ajou University School of Medicine, Suwon, Republic of Korea
d
Cancer Research Institute, College of Medicine, Seoul National University, Seoul, Republic of Korea
e
Leegilya Cancer and Diabetes Institute, Gachon University of Medicine and Science, Inchon, Republic of Korea
b
c
a r t i c l e
i n f o
Article history:
Received 25 August 2010
and in revised form 12 October 2010
Available online 18 October 2010
Keywords:
Anaplerosis
Beta cell
Glucolipotoxicity
Mitochondria
TCA cycle
a b s t r a c t
The aim of this study is to investigate the effect of mitochondrial metabolism on high glucose/palmitate
(HG/PA)-induced INS-1 beta cell death. Long-term treatment of INS-1 cells with HG/PA impaired energyproducing metabolism accompanying with depletion of TCA cycle intermediates. Whereas an inhibitor of
carnitine palmitoyl transferase 1 augmented HG/PA-induced INS-1 cell death, stimulators of fatty acid
oxidation protected the cells against the HG/PA-induced death. Furthermore, whereas mitochondrial
pyruvate carboxylase inhibitor phenylacetic acid augmented HG/PA-induced INS-1 cell death, supplementation of TCA cycle metabolites including leucine/glutamine, methyl succinate/a-ketoisocaproic acid,
dimethyl malate, and valeric acid or treatment with a glutamate dehydrogenase activator, aminobicycloheptane-2-carboxylic acid (BCH), significantly protected the cells against the HG/PA-induced death. In
particular, the mitochondrial tricarboxylate carrier inhibitor, benzene tricarboxylate (BTA), also showed
a strong protective effect on the HG/PA-induced INS-1 cell death. Knockdown of glutamate dehydrogenase or tricarboxylate carrier augmented or reduced the HG/PA-induced INS-1 cell death, respectively.
Both BCH and BTA restored HG/PA-induced reduction of energy metabolism as well as depletion of
TCA intermediates. These data suggest that depletion of the TCA cycle intermediate pool and impaired
energy-producing metabolism may play a role in HG/PA-induced cytotoxicity to beta cells and thus,
HG/PA-induced beta cell glucolipotoxicity can be protected by nutritional or pharmacological maneuver
enhancing anaplerosis or reducing cataplerosis.
Ó 2010 Elsevier Inc. All rights reserved.
Introduction
Insulin deficiency caused by a loss of pancreatic beta cells and a
subsequently impaired compensation for insulin resistance
contributes toward the development of type 2 diabetes [1]. Increased levels of free fatty acids (FFAs)2 are believed to induce a
beta cell loss in type 2 diabetic patients, and this was termed ‘lipo-
⇑ Corresponding author. Address: Institute for Medical Science, Ajou University
School of Medicine, Wonchon-dong san 5, Yongtong-gu, Suwon, Gyeonggi-do 442749, Republic of Korea. Fax: +82 31 219 4482.
E-mail address: [email protected] (Y. Kang).
1
These authors contributed equally to this work.
2
Abbreviations used: CPT-1, carnitine palmitoyl transferase-1; ER, endoplasmic
reticulum; FFA, free fatty acid; FAO, fatty acid oxidation; GDH, glutamate dehydrogenase; LC-CoA, long-chain acyl-coenzyme A; PPAR-a, peroxisome proliferatoractivated receptors-alpha; ROS, reactive oxygen species; PARP, poly-(ADP-ribose)
polymerase; PC, pyruvate carboxylase; TC, tricarboxylate carrier; TCA, tricarboxylate
acid; TG, triacylglycerol.
0003-9861/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.abb.2010.10.011
toxicity’. Since FFAs in conjunction with hyperglycemia potentiates
the lipotoxicity, the high glucose/FFA-induced toxicity was an augmented form of lipotoxicity and termed ‘glucolipotoxicity’ [2]. In
fact, long-term exposure to FFA induced beta cell death in culture
and in isolated islets [3], and elevated levels of glucose augmented
this FFA-induced cell death [4]. This in vitro beta cell death by elevated FFA and high concentration of glucose is considered to represent in vivo glucolipotoxicity. Cell death was mainly apoptotic with
cytochrome c release, caspase 3 activation, and DNA fragmentation
[3]. Saturated fatty acids such as palmitic and stearic acids are
generally cytotoxic to beta cells, while unsaturated fatty acids like
linoleic, oleic, and palmitoleic acids, are not, and actually protect
cells from saturated FFA-induced death [5].
The molecular and cellular mechanisms involved in FFA-induced beta cell death as well as the potentiation of FFA-induced
death by glucose are not fully understood. Ceramide and the ceramide-induced up-regulation of nitric oxide (NO) have been suggested to be important mediators of FFA-induced beta cell death
232
S.-E. Choi et al. / Archives of Biochemistry and Biophysics 505 (2011) 231–241
[3,5–7]. The involvement of NO-mediated mitochondrial DNA
damage in FFA-induced INS-1 beta cell death was reported [8].
On the other hand, FFA-induced activation of novel protein kinase
Cs (PKCs), which included PKC-d, was suggested to contribute to
beta cell death [9,10]. Activation of NF-jB signal was implicated
in FFA-induced beta cell death since IjB kinase inhibitor was found
to reduce palmitate-induced INS-1 cell death [11]. It was also reported that elevated concentrations of FFAs or high concentration
of glucose produced reactive oxygen species (ROS) and the signals
generated from oxidative stress might contribute to induction of
beta cell glucolipotoxicity [11–13]. Ca2+-dependent mechanisms
were suggested to be involved in FFA-induced beta cell death
[14,15]. Down-regulation of the IRS/PI3 kinase/Akt signalling pathway was a critical modulator of FFA-induced beta cell death [16].
Recently, induction of endoplasmic reticular (ER) stress was reported to play a critical role in FFA-induced beta cell death [17,18].
Although several cytotoxic mediators for FFA-induced cytotoxicity have been reported, it is not determined how glucose/FFAmediated metabolic changes induce activation of such cytotoxic
signals. Elevated glucose levels were thought to potentiate FFA-induced beta cell toxicity through lipid partitioning [19]. FFAs are
transported into the mitochondria via carnitine palmitoyl transferase-1 (CPT-1) as acyl-CoA form and undergo b-oxidation when
FFAs are used as the energy source. However, FFA metabolism
can be shifted from the oxidation to esterification pathway when
the concentration of glucose is simultaneously elevated [19]. The
increased flux of glucose and enhanced glycolysis in beta cells promotes anaplerosis, which is the replenishment of tricarboxylate
acid (TCA) cycle intermediates, through the activation of pyruvate
carboxylase (PC). The enhanced anaplerosis results in elevation of
malate and citrate levels in the mitochondria and ultimately promotes cataplerosis, which accelerates the transport of TCA cycle
intermediates to the cytosol [20]. The increased cytosolic citrate
can be converted to malonyl-CoA through sequential enzymatic
reactions involving ATP citrate lyase and acetyl-CoA carboxylase
(ACC). Accumulation of malonyl-CoA inhibits CPT-1 activity, thereby switching fatty acid metabolism from oxidation (FAO) to lipid
synthesis (Supplemental Fig. 1) [21]. Ultimately, the accumulation
of long-chain acyl-CoA (LC-CoA) or lipid signalling molecules such
as phosphatidic acid, lysophosphatidic acid, diacylglycerol (DAG),
and ceramide were suggested to play a critical role in FFA-induced
lipotoxicity and glucose-induced augmentation [19]. In addition to
enhancing the malonyl-CoA pathway, high levels of glucose can
further promote the lipogenic pathway through induction of the
master lipogenic transcription factor, sterol regulatory elementbinding protein-1c (SREBP-1c) [22]. Recently, El-Assaad et al.
added that early promotion of lipid esterification was major contributors to high glucose/FFA-induced glucolipotoxicity [23].
Although lipid accumulation and subsequent activation of lipidmediated signals was associated with FFA-induced toxicity to beta
cells [24], triacylglycerol (TG) itself is not likely toxic because TG is
a biologically inert molecule [25]. In contrast, TG accumulation
could provide survival signals to FFA-treated beta cells by sequestrating toxic lipid intermediates [26,27]. Enhanced synthesis of TG
through over-expression of stearoyl-CoA desaturase 1 (SCD1) was
reported to reduce FFA-induced cytotoxicity [28]. Recently, the
TG/FFA cycling, which enhances lipid esterification in conjunction
with increased lipolysis, was suggested to act as the mechanism
preventing overnutrition-induced glucolipotoxicity [29]. Enhanced
TG/FFA cycling was postulated to divert toxic glucose/FFA metabolism to a useless energy exhaustion metabolism [30]. Collectively,
it is not certain whether the promotion of lipid synthetic pathways
by exposure to high glucose and FFA plays a critical role in glucolipotoxicity. Recently, a report that incomplete fatty acid oxidation
through depletion of TCA cycle intermediates contributed to FFAinduced insulin resistance suggest that mitochondrial dysfunction,
rather than activation of lipid-mediated signals, may be a direct
mediator for FFA-induced toxicity to beta cells [31]. Furthermore,
decrease of various mitochondrial enzymes such as pyruvate carboxylase, glutamate dehydrogenase, isocitrate dehydrogenase,
and ATP synthase in beta cells under hyperglycemic and hyperlipidemic conditions supports that mitochondrial dysfunction through
metabolic failure may be a critical contributor to high glucose/FFAinduced cytotoxicity to beta cells [23,32,33].
The present study was initiated to determine whether metabolic
impairment in mitochondria was involved in high glucose/FFA-induced glucolipotoxicity to beta cells and whether the maintenance
of TCA cycle intermediate pool could protect against the toxicity.
We initially tested whether mitochondrial energy-producing
metabolism was impaired in high glucose/palmitate (HG/PA)-treated INS-1 beta cells. We measured the level of intracellular ATP
and the oxidation rate of glucose and palmitate and determined
the levels of TCA cycle intermediates in the HG/PA-treated cells.
We next investigated the role of glucose/FFA metabolism in HG/
PA-induced cytotoxicity using modulators of anaplerosis and cataplerosis as well as modulators of fatty acid oxidation. The effect of reduced anaplerosis and cataplerosis on HG/PA-induced INS-1 cell
death was examined through gene knockdown of glutamate dehydrogenase (GDH) and mitochondrial tricarboxylate carrier (TC),
respectively. We finally investigated whether GDH activator and
TC inhibitor would be able to restore the HG/PA-induced reduction
of oxidation metabolism and depletion of TCA cycle intermediates.
Materials and methods
Reagents
All chemicals, including glucose, palmitate, 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazoilium bromide (MTT), bezafibrate,
etomoxir, phenylacetic acid (PAA), T0901317, fumonisin B1, myriosin, GF109203X, chelerythrine, N-acetylcysteine (NAC), reduced
glutathione (GSH), methyl pyruvate, leucine, glutamine, monomethyl succinate, a-ketoisocaproate, dimethyl malate, valeric acid,
2-aminobicyclo[2.2.1]heptan-2-carboxylic acid (BCH), 2,3-benzenetricarboxylate (BTA), and ATP were purchased from either Merck
Bioscience (Darmstadt, Germany) or Sigma–Aldrich (St. Louis, MO).
The chemicals were dissolved in either appropriate media solution
or dimethyl sulfoxide (DMSO) and then treated at the required
working dilution. All chemicals were handled in accordance with
the supplier’s recommendations. Anti-caspase 3 and PARP antibodies were purchased from Cell Signalling Technology (Beverly, MA).
Preparation of palmitate
Preparation of palmitate was slightly modified from Listenberg’s protocol [34]. Palmitate/BSA (bovine serum albumin) conjugates were prepared through soaping palmitate with sodium
hydroxide and mixing with BSA. Briefly, a 20 mM solution of palmitate in 0.01 M NaOH was incubated at 70 °C for 30 min and
the fatty acid soaps were then complexed with 5% fatty acid-free
BSA in phosphate-buffered saline (PBS) in 1:3 volume ratio. The
complexed fatty acids consisted of 5 mM palmitate and 3.75%
BSA. The palmitate/BSA conjugates were diluted in 10% FBS culture
medium (approximately 0.4% BSA) and administered to cultured
cells. Molar ratio of palmitate to BSA in 0.4 mM palmitate is
3.5:1 and the concentration of BSA is approximately 0.7%.
Cells
INS-1 rat insulinoma cells were grown in RPMI 1640 medium
supplemented with 10% fetal bovine serum (Invitrogen, Carlsbad,
S.-E. Choi et al. / Archives of Biochemistry and Biophysics 505 (2011) 231–241
CA), 100 U/mL penicillin, 100 g/mL streptomycin, and 10 mM
HEPES at 37 °C in a humidified atmosphere containing 95% air
and 5% CO2.
Viability assay
Briefly, the cells were treated with MTT (0.5 mg/ml) at 37 °C for
2 h. Supernatants were discarded and acidic isopropanol (0.04 N
HCl) was then added. After incubating at room temperature for
30 min, absorbance was measured at 570 nm using a microplate
reader (BIO-RAD, Hercules, CA).
Quantitation of cell death
Cell death was determined by measuring cytoplasmic histoneassociated DNA fragments using the Cell Death Detection ELISA
kit (Roche Applied Science, Mannheim, Germany). Briefly, the cells
that were treated with palmitate were lysed by adding incubation
buffer (supplied from kit). After centrifugation (20,000g, 10 min),
the supernatant was pipetted onto an anti-histone-coated microplate. After incubation for 90 min at 25 °C, the wells were rinsed
with washing solution (supplied from kit). Anti-DNA PODs were
then bound to the nucleosome complex by adding conjugation
solution (supplied from kit) and incubating at 25 °C for 90 min.
After washing the wells with washing solution, the colour was
developed by adding ABTS (2,20 -azino-di-[3-ethylbenzthiazoline
sulphonate]) substrate solution (1 mg/ml) and by incubating with
shaking at 250 rpm for 10–20 min. The amount of peroxidase retained in the nucleosome complex was determined by measuring
the absorbency at 405 nm.
Semi-quantitative reverse transcriptase-polymerase chain reaction
(RT-PCR)
Expression levels of GDH and mitochondrial TC mRNAs were
determined using the semi-quantitative RT-PCR technique. Semiquantitative RT-PCR was performed with the Takara RNA PCR kit
Version 3.0 (Takara, Shiga, Japan). Briefly, INS-1 cell cDNAs were
synthesised with avian myeloblastosis virus (AMV) reverse transcriptase and random 9-mers, and underwent PCR amplification
with primer sets of GDH (forward primer: 50 -CTCAGGTACCCGTCTGATCTTCTGTGCTCCCG, backward primer: 50 -CTCACTCGAGCGAGG
TGAAGAGGGTGGTAAAG) and TC (forward primer: 50 -GCTGGAGGC
ATCGAAATCTGCAT, backward primer: 50 -TAGCTGTGAGGCCTTGG
TATGTC). GDH DNAs were amplified under the following condition; denaturation at 95 °C for 5 min followed by 26 cycles of denaturation at 95 °C for 30 s, annealing at 60 °C for 30 s, and extension
at 72 °C for 1 min. On the other hand, TC DNA was amplified by 29
cycles under the same conditions. The reaction was terminated by
extension at 72 °C for 10 min. The amplified DNA was analysed by
1% agarose gel electrophoresis, and the relative DNA quantities
were measured by densitometric analysis on the basis of amplified
cyclophilin DNA.
Immunoblotting
The cells were suspended in RIPA buffer [150 mM NaCl, 1% NP40, 0.5% deoxycholate, 0.1% sodium dodecyl sulphate, 50 mM Tris–
Cl, (pH 7.5), and protease inhibitor cocktail (Roche Applied Science,
Mannheim Germany)] and then incubated on ice for 30 min. Whole
proteins were extracted by differential centrifugation (10,000g,
10 min) and protein concentrations in lysates were determined
using protein assay kits (Bio-Rad, Hercules, CA). An equal volume
of 2 sodium dodecyl sulphate (SDS) sample buffer [125 mM
Tris–Cl, (pH 6.8), 4% SDS, 4% 2-mercaptoethanol, 20% glycerol]
was added to cell lysates, and equivalent amounts of protein
(30 lg) were loaded onto 10–15% polyacrylamide gels, electropho-
233
resed, and then electrophoretically transferred onto Polyvinylidene
Fluoride (PVDF) membranes (Millipore, Bedford, MA). After blocking these membranes with 5% skimmed milk for 30 min, target
antigens were reacted with primary antibodies and subsequently
secondary antibodies (horseradish peroxidase-conjugated antimouse IgG or anti-rabbit IgG antibodies). Immunoreactive bands
were then developed using an enhanced chemiluminescence
detection system (Amersham Pharmacia Biotech, Arlington Height,
IL). Band intensity was determined by densitometric analysis using
a one-dimensional Quantity OneÒ 1D image analysis system. Maximum intensity of the band was converted to 100%, and relative
intensities were calculated on the basis of maximum.
Measurement of oxaloacetate, alpha-ketoglutarate and citrate
To determine the level of TCA cycle intermediates, metabolism
assay kits for oxaloacetate, alpha-ketoglutarate, and citrate (BioVision Inc., Mountain View, CA) were used. Briefly, cells were
scraped, washed with PBS, and homogenised in assay buffer provided by supplier. Insoluble particles were removed by differential
centrifugation (15,000g, 10 min). Protein concentrations in extracts
were determined using protein assay kits (Bio-Rad, Hercules, CA).
Residual enzymes in extracts were removed by perchloric acid/
KOH precipitation protocol (BioVision Inc., Cat. # K808-200). The
resulting supernatant was used for measurement of metabolites.
The supernatants were mixed with each assay buffer containing
enzyme mix, developer, and probe. After incubation at room temperature for 30 min, absorbance was measured at 570 nm. The
quantity of each metabolite was determined by normalisation to
intracellular protein content.
Measurement of oxidation rate of glucose and palmitate
Cells were incubated in 0.2% BSA/KRB (0.2 mM glucose, 24 mM
NaHCO2, 1.2 mM MgCl2, 1 mM Hepes, 129 mM NaCl, 4.8 mM KCl,
1.2 mM KH2PO4, 2.5 mM CaCl2) containing 2 mM glucose and
0.1 lCi/ml [U-14C] glucose for 1 h or in 0.2% BSA/KRB containing
2 mM glucose, 1 mM carnitine, 0.2 mM palmitate, 0.5 lCi/ml
[1-14C] palmitate for 2 h, respectively. The supernatant containing
labelled CO2 was transferred into a new tube and then the reaction
was stopped by adding 1/6 volume of cold 40% perchloric acid. The
CO2 was absorbed in 1 M KOH solution by sealing the bottle containing the reaction tube and KOH tube for 16 h. The amount of
labelled CO2 was measured by counting the radioactivity of absorbed CO2 in KOH solution with liquid scintillation. Normalisation
relative to intracellular protein content in live cells was performed.
The relative amount of labelled CO2 compared to CO2 produced
from non-treated INS-1 cells represented the relative oxidation
rate of glucose or palmitate in HG/PA-treated cells.
Measurement of ATP
Quantitation of intracellular ATP levels was determined using a
CellTiter-Glo™ luminescence kit (Promega Life Science, WI). INS-1
cells were mixed with the same volume of CellTiter-Glo reagent
and lysed by vortexing. After incubating at room temperature for
10 min, luminescence was measured using a TD-20/20 luminometer (Turner Designs, Sunnyvale, CA). The quantity of ATP was
calculated using an ATP standard curve. The relative level of ATP
compared to that from non-treated INS-1 cells represented the relative percent of ATP in HG/PA-treated cells.
Small interfering RNAs and transfection
The 21-nucleotide small interfering RNA duplexes were designed
and synthesised from Samchully Pharm Co. (Seoul, Korea). The sequences are as follows: GFP, 50 -GUUCAGCGUGUCCGGCGAGTT; rat
234
S.-E. Choi et al. / Archives of Biochemistry and Biophysics 505 (2011) 231–241
GDH (GenBank: X14223.1), 50 -GAUCAAUCCCAAGAACUAU; rat TC
(GenBank: NM_017307), 50 -GUGUCAUAGUGACUGUCAU. INS-1
cells were transfected with siRNA duplex using a pipet type electroporator (Digital biotechnology, Seoul, Korea) according to manufacturer’s instruction. Briefly, 2.5 lg each siRNA duplex with 3 lg
pCDNA in 100 ll R buffer (Digitalbiotechnology) were transfected
into 5 106 INS-1 cells under conditions of 1650 V and 10 ms width.
After transfection, 1.5 106 cells were seeded into each well of a 6well plate and incubated for 36 h.
Statistics
Data are represented as means ± SE from at least three independent experiments. Differences between the groups were determined by one way analysis of variance (ANOVA) using the SPSS
statistical analysis program version 7.5 (SPSS, Cray, NC, USA). Duncan’s multiple range test was performed for evaluation of differences between the groups. A value of p < 0.05 was considered to
indicate a significant difference between groups.
Results
Augmentation of palmitate-induced INS-1 cell death by high
concentration of glucose
Since saturated fatty acid was reported to be cytotoxic to beta
cells [3], we re-examined the palmitate-induced cytotoxicity to
INS-1 beta cells. INS-1 cells were exposed to different concentrations of palmitate in the presence of 11 mM glucose for 18 h. As
shown in Fig. 1A, a low dose of palmitate (0.1 mM) slightly increased viability of INS-1 cells but 0.3 mM and 0.4 mM of palmitate
(PA) significantly reduced the cell viability. Palmitate at a concentration of 0.4 mM reduced cell viability to 64%. Since FFA-induced
lipotoxicity was reported to be synergized with high concentration
of glucose [4], we examined the PA-induced viability reduction in
response to varying concentrations of glucose. As shown in
Fig. 1B, glucose dose-dependently potentiated the palmitate-induced reduction of INS-1 cell viability. Incubation of INS-1 cells
with 0.4 mM palmitate in the presence of 5 mM glucose for 18 h
reduced cell viability by around 13%. Incubation of the cells in
the presence of 25 mM glucose (high glucose: HG), however, reduced viability by around 56%. HG/PA-induced viability reduction
of INS-1 cells was time-dependent (Fig. 1C). Viability began to decline at 12 h after HG/PA treatment although that at 6 h was
slightly increased. The Cell Death Detection assay demonstrated
that fragmented DNAs were detected by palmitate treatment in
the presence of 5 mM glucose and that the palmitate-induced
DNA fragmentation was augmented by 25 mM glucose (Fig. 1D).
The DNA fragmentation by HG/PA treatment was also time-dependent (Fig. 1E). In particular, DNA fragmentation at 12 h was not significantly increased although cell viability at the same time was
significantly reduced (Fig. 1D and E), suggesting that viability
reduction by HG/PA was followed by HG/PA-induced DNA fragmentation. The level of the cleaved form of caspase 3 and poly
(ADP-ribose) polymerase (PARP), demonstrating caspase 3 activation, was gradually increased in palmitate-treated cells in proportion to increasing concentrations of glucose (Fig. 1E). These data
demonstrated that saturated fatty acid-induced apoptotic cell
death in INS-1 beta cells and the fatty acid-induced apoptotic
death was augmented by high concentration of glucose.
Reduction of ATP level, oxidation rate of glucose and palmitate, and
TCA cycle intermediates in HG/PA-treated INS-1 cells
Since mitochondrial dysfunction was suggested to be a mediator of HG/FFA-induced glucolipotoxicity, we examined the intra-
cellular level of ATP in HG/PA-treated INS-1 cells. As shown
Fig. 2A, the ATP level was slightly increased at 6 h after HG/PA
treatment, but the level was significantly reduced by around 35%
at 12 h, the time when viability reduction began to be observed.
Since ATP level was lower, oxidation rate of glucose and palmitate
was next investigated in HG/PA-treated cells. As shown in Fig. 2B
and C, pre-treatment with HG/PA for 12 h significantly reduced
oxidation rate of glucose and palmitate. On the other hand, the oxidation rate of glucose and palmitate was slightly increased at 6 h
after HG/PA treatment. These data also demonstrated that energy-producing capability in mitochondria was impaired with similar kinetics with viability reduction in HG/PA-treated INS-1 cells.
Whereas bezafibrate as a peroxisome proliferator-activated receptors-alpha (PPAR-a) agonist increased the oxidation rate of palmitate, etomoxir as an inhibitor of CPT-1 reduced the oxidation rate
of palmitate (Fig. 2C). Since impaired glucose metabolism and subsequent reduction of TCA cycle intermediate pool was suggest to
be a characteristic of lipid-induced mitochondrial stress [31], we
next examined the level of TCA cycle intermediates including oxaloacetate, citrate, and a-ketoglutarate using a metabolism assay
kit. As shown in Fig. 2D, the oxalacetate level began to decrease
at 3 h after HG/PA treatment and then continuously declined. On
the other hand, the level of citrate and a-ketoglutarate in HG/PAtreated INS-1 cells peaked at 6 h after HG/PA treatment and then
began to decrease at 9 h. These data demonstrated that levels of
all three TCA cycle intermediates in HG/PA-treated INS-1 cells
were significantly lower than those in untreated cells at the time
when the viability reduction was detected, but when HG/PA-induced DNA fragmentation was not observed. Reduction kinetics
of TCA cycle intermediates was in accord with that of intracellular
ATP and oxidation rate of glucose and palmitate (Fig. 2A–D). These
results suggest that long-term treatment of INS-1 cells with HG/PA
impaired the oxidative metabolism of glucose and FFA, partly
through depletion of TCA cycle intermediates.
Role of fatty acid oxidation in HG/PA-induced INS-1 cell death
Enhanced fatty acid oxidation (FAO) metabolism was supposed
to protects against palmitate-induced cell death since inhibition of
FAO promotes the palmitate-induced beta cell death [4]. We reexamined the effect of FAO on HG/PA-induced INS-1 cell death.
FAO was stimulated by treatment of the cells with the PPAR-a agonist bezafibrate or the AMP-activated protein kinase (AMPK) activator 5-aminoimidazole-4-carboxamide ribonucleoside (AICAR)
[35]. These two agents dose-dependently and significantly protected against HG/PA-induced DNA fragmentation (Fig. 3A and
B). The effective concentrations of AICAR or bezafibrate did not increase DNA fragmentation of INS-1 cell in the absence of palmitate.
In contrast, etomoxir, an inhibitor of carnitine palmitoyltransferase
1 (CPT-1) that catalyzes the rate-limiting step of the oxidation of
FFAs [4], dose-dependently and significantly augmented HG/PA-induced DNA fragmentation (Fig. 3C). These data demonstrated that
stimulation of FAO protected against HG/PA-induced INS-1 cell
death, but inhibition of FAO promoted the cell death.
Supplement of TCA cycle intermediates protects against HG/PAinduced INS-1 cell death
Pyruvate produced through glycolysis is converted to oxaloacetate through pyruvate carboxylase (PC), resulting in replenishment
of tricarboxylic acid (TCA) cycle intermediates (anaplerosis) and finally providing cytosolic citrate (cataplerosis) for fatty acid/lipid
synthesis (Supplemental Fig. 1). To determine whether depletion
of TCA cycle intermediates was involved in HG/PA-induced INS-1
cell death, the activity of PC was blocked by treatment with PC
inhibitor phenylacetic acid (PAA) and the effect of PAA on
S.-E. Choi et al. / Archives of Biochemistry and Biophysics 505 (2011) 231–241
235
Fig. 1. Augmentation of palmitate-induced INS-1 cell death by high concentration of glucose. (A) INS-1 cells were treated with different concentrations of palmitate in the
presence of 11 mM glucose (Glu) for 18 h. Cell viability was measured using an MTT assay. Data are represented as means ± SEM from three independent experiments.
*p < 0.05; **p < 0.01 vs. viability of cells with BSA (0.7%) and glucose (11 mM). (B) INS-1 cells were treated with 0.4 mM palmitate (PA) in the presence of different
concentrations of glucose for 18 h. *p < 0.05; **p < 0.01 vs. viability of cells with BSA (0.7%) and glucose (5 mM). (C) INS-1 cells were treated with 0.4 mM palmitate (PA) in the
presence of 25 mM glucose (HG) for the indicated times. *p < 0.05; **p < 0.01 vs. viability of untreated cells (0 h incubation). (D) INS-1 cells were treated with 0.4 mM
palmitate (PA) in the presence of different concentrations of glucose for 18 h. Apoptotic death was determined by an increase in DNA fragmentation (Cell Death Detection
ELISA). *p < 0.05; **p < 0.01 vs. absorbance from untreated INS-1 cells. (E) INS-1 cells were treated with 0.4 mM palmitate (PA) in the presence of 25 mM glucose for the
indicated times and DNA fragmentation was measured. **p < 0.01 vs. absorbance from untreated INS-1 cells. (F) INS-1 cells were treated with 0.4 mM palmitate (PA) in the
presence of different concentrations of glucose for 15 h. Cleaved caspase 3 and PARP was detected by immunoblotting with anti-cleaved caspase 3 and anti-PARP antibodies.
P-Cas 3, C-Cas 3, P-PARP, and C-PARP, represent pro-caspase 3, cleaved caspase 3, uncleaved PARP, and cleaved PARP, respectively.
Fig. 2. Decline in ATP levels, decreased oxidation rate of glucose and palmitate, and reduction of TCA cycle intermediates in HG/PA-treated INS-1 cells. (A) INS-1 cell were
treated with HG/PA for the indicated times. Quantitation of intracellular ATP was determined using the CellTiter-Glo™ luminescence kit. Data are represented as
means ± SEM from three independent experiments. *p < 0.05; **p < 0.01 vs. ATP in untreated INS-1 cells. Oxidation rates of glucose (B) or palmitate (C) were determined by
measurement of labelled CO2 produced from D-[U-14C] glucose or [1-14C] palmitate during a 1- or 2-h incubation in HG/PA-treated INS-1 cells, respectively. PPAR-a agonist
bezafibrate and carnitine palmitoyl transferase-1 (CPT-1) inhibitor etomoxir were used as the positive and negative controls of palmitate oxidation, respectively. Data are
represented as means ± SEM from three independent experiments. *p < 0.05; **p < 0.01 vs. glucose oxidation rate in untreated INS-1 cells. #p < 0.05; ##p < 0.01 vs. palmitate
oxidation rate in untreated INS-1 cells. (D) Levels of oxaloacetate, citrate, and a-ketoglutarate in cellular extracts were then determined using a metabolism assay kit
(Biovision). Relative levels of metabolites in HG/PA-treated and HG/PA-untreated cells were calculated. **p < 0.01 vs. oxaloactate in HG/PA-untreated INS-1 cells. #p < 0.05;
##
p < 0.01 vs. citrate in HG/PA-untreated INS-1 cells. $p < 0.05; $$p < 0.01 vs. a-ketoglutarate in HG/PA-untreated INS-1 cells.
HG/PA-induced death was then examined. PAA dose-dependently
augmented HG/PA-induced DNA fragmentation at the concentra-
tions that did not influence the death of INS-1 cells in the absence
of palmitate (Fig. 4A). This data strongly suggest that depletion of
236
S.-E. Choi et al. / Archives of Biochemistry and Biophysics 505 (2011) 231–241
Fig. 3. Effect of FAO modulators on HG/PA-induced INS-1 cell death. INS-1 cells were treated with 25 mM glucose (HG) and 0.4 mM palmitate (PA) in conjunction with
different concentrations of FAO modulators, such as bezafibrate (PPAR-a agonist) (A), AICAR (AMPK activator) (B), and etomoxir (CPT-1 inhibitor) (C). After 18 h incubation,
level of DNA fragmentation was measured using Cell Death Detection ELISA. *p < 0.05; **p < 0.01 vs. DNA fragmentation of HG/PA-treated INS-1 cells.
TCA cycle intermediates through reduced anaplerosis is involved in
HG/PA-induced INS-1 cell death. Since TCA cycle intermediates are
replenished through treatment with various metabolic fuels, the
effect of the metabolic fuels on HG/PA-induced INS-1 cell death
was tested (Supplemental Fig. 1). While treatment of the cells with
methyl pyruvate, a membrane transferable form of pyruvate, did
not influence HG/PA-induced DNA fragmentation (Fig. 4B), leucine
and glutamine, which replenish a-ketoglutarate, as well as monomethyl succinate and a-ketoisocaproic acid, which replenish succinic acid and acetyl-CoA, significantly and dose-dependently
reduced HG/PA-induced DNA fragmentation (Fig. 4C and D). Dimethyl malate, a membrane transferable TCA cycle intermediate
malate, also significantly reduced HG/PA-induced DNA fragmentation (Fig. 4E). In addition, valeric acid, a five carbon-chain FFA that
replenishes succinyl-CoA through propionyl-CoA significantly and
dose-dependently reduced HG/PA-induced DNA fragmentation
(Fig. 4F). In particular, BCH, an activator of glutamate dehydrogenase (GDH) that can replenish a-ketoglutarate in the TCA cycle
from glutamate, very effectively reduced HG/PA-induced DNA fragmentation and the protective effect was dose-dependent (Fig. 4G).
Since the TCA cycle intermediate pool could be expanded by blocking transportation of mitochondrial citrate to the cytosol, the effect
of BTA as an inhibitor of mitochondrial tricarboxylate carrier (TC)
on HG/PA-induced cytotoxicity was examined (Supplemental
Fig. 1). BTA also dose-dependently and significantly reduced the
HG/PA-induced DNA fragmentation (Fig. 3H). Since activator of
GDH or inhibitor of TC reduced HG/PA-induced DNA fragmentation, we confirmed the protective effect of BCH or BTA on HG/PA-
Fig. 4. Effect of anaplerotic and cataplerotic modulators on HG/PA-induced INS-1 cell death. INS-1 cells were treated with 25 mM glucose (HG) and 0.4 mM palmitate (PA) in
conjunction with different concentrations of TCA cycle modulators, such as PAA (phenylacetic acid; inhibitor of pyruvate carboxylase) (A), methyl pyruvate (B), Leu/Gln
(leucine/glutamine) (C), MMS/KIC (monomethyl succinate/a-ketoisocaproate) (D), DMM (dimethyl malate) (E), valeric acid (F), BCH (aminobicyclo-heptane-2-carboxylic
acid; activator of glutamate dehydrogenase) (G), and BTA (benzene tricarboxylate; inhibitor of the mitochondrial tricarboxylate carrier) (H). After 18 h incubation, level of
DNA fragmentation was measured using Cell Death Detection ELISA. *p < 0.05; **p < 0.01 vs. DNA fragmentation of HG/PA-treated INS-1.
S.-E. Choi et al. / Archives of Biochemistry and Biophysics 505 (2011) 231–241
induced cell death through reduced cleavage of caspase 3 and
PARP. As shown in Fig. 5, exposure of INS-1 cell to HG/PA for
15 h increased the level of cleaved caspase 3 in conjunction with
cleaved PARP. However, treatment with BCH or BTA significantly
attenuated HG/PA-induced cleavage of PARP and caspase 3. FAO
stimulator bezafibrate also showed similar attenuating effect on
HG/PA-induced caspase 3 cleavage. Collectively, all our attempts
for enhancing TCA cycle intermediate pool through enhanced anaplerosis or reduced cataplerosis showed evident protective effect
on HG/PA-induced INS-1 cell death.
Augmentation of HG/PA-induced INS-1 cell death by knockdown of
glutamate dehydrogenase (GDH) and attenuation of the death by
knockdown of mitochondrial tricarboxylate carrier (TC)
Since pharmacological agents enhancing anaplerosis or reducing cataplerosis showed a protective effect on HG/PA-induced
INS-1 cell death, we examined whether gene knockdown of GDH
or TC would demonstrate similar augmenting or attenuating effects on HG/PA-induced INS-1 cell death, respectively. While
knockdown of GDH is thought to inhibit supply of TCA cycle intermediate a-ketoglutalate, knockdown of TC is thought to increase
TCA cycle intermediate pool (Supplemental Fig. 1). Knockdown of
GDH or TC was determined by semi-quantitative RT-PCR. As shown
in Fig. 6A and B, transfection of GDH or TC siRNAs clearly reduced
mRNA level of GDH or TC, respectively. As shown in Fig. 6A, knockdown of GDH gene significantly augmented HG/PA-induced cleavage of caspase 3 and PARP. The augmenting effect by GDH
knockdown on HG/PA-induced caspase 3 activation was also demonstrated in INS-1 cell treated with BCH in the presence of HG/PA.
On the other hand, knockdown of mitochondrial TC significantly
reduced HG/PA-induced cleavage of PARP and caspase 3 (Fig. 6B).
The reducing effect by TC knockdown was also shown in INS-1 cell
treated with BTA in the presence of HG/PA. These data demonstrated that reduction of TC attenuated the HG/PA-induced INS-1
beta cell death, while reduction of GDH augmented the HG/PA-induced cell death. These results support the notion that depletion of
TCA cycle intermediate pool plays a role in HG/PA-induced beta
cell cytotoxicity.
Restoration of HG/PA-induced reduction of TCA cycle intermediates,
glucose/palmitate oxidation rate, and ATP level by BCH or BTA
treatment
Since the attempts to enhance TCA cycle intermediate pool
protected INS-1 cells against HG/PA-induced cell death, depletion in
237
the TCA cycle intermediates and subsequent reduction of energy-producing metabolism was postulated to be involved in HG/PA-induced
cytotoxicity. We tested whether treatment of the cells with BCH or
BTA would increase the level of TCA cycle intermediates and restore
ATP levels with enhanced oxidation rate in HG/PA-treated INS-1 cells.
Treatment of the INS-1 cells with BCH or BTA without palmitate
tended to increase the level of TCA cycle intermediates including oxalacetate, citrate and a-ketoglutarate (Fig. 7A–C). In particular, both
BCH and BTA significantly prevented HG/PA-induced reduction of
these intermediates (Fig. 7A–C). FAO stimulator bezafibrate also
showed similar preventive effect on HG/PA-induced reduction of
the intermediates. In accordance with an increase of TCA cycle intermediates, both BCH and BTA significantly prevented HG/PA-induced
reduction of glucose or palmitate oxidation (Fig. 7D and E). On the
other hand, bezafibrate significantly increased the oxidation rate of
palmitate (Fig. 7E), but decreased glucose oxidation in HG-treated
INS-1 cells (Fig. 7D). Bezafibrate also prevented HG/PA-induced
reduction of oxidation rate of glucose and palmitate in INS-1 cells
(Fig. 7D and E). In accordance with a restoration of oxidation rate,
treatment with bezafibrate, BCH or BTA significantly restored the
reduced ATP level in HG/PA-treated INS-1 cells (Fig. 7F). These data
suggested that HG/PA-induced inhibition of energy-producing
metabolism could be rescued by supplementation of TCA cycle intermediates through enhanced anaplerosis or reduced cataplerosis.
Discussion
High concentration of glucose amplified palmitate-induced INS1 cell death. Bezafibrate and AICAR, which is known to increase
FAO, protected against this HG/PA-induced cytotoxicity. However,
etomoxir, which inhibits FAO by blocking acyl-CoA transport into
mitochondria, potentiated this cytotoxicity. On the other hand,
most attempt to expand the TCA cycle intermediate pool, including
treatment of cells with leucine/glutamine, monomethyl succinate/
a-ketoisocaproic acid, dimethyl malate, valeric acid, BCH and BTA,
significantly protected against HG/PA-induced cytotoxicity. In fact,
the level of TCA cycle intermediates in HG/PA-treated INS-1 cells
was significantly low at the time when HG/PA-induced viability
reduction began to be observed. The oxidation rate of glucose
and palmitate was declined and the level of ATP was also reduced
in HG/PA-treated cells. The reductions of oxidation rate and ATP level in HG/PA-treated cells were restored after treatment of the cells
with bezafibrate, BCH or BTA. These results suggest that an insufficient supply of TCA cycle intermediates or fuel substrates into
mitochondria and subsequent impaired energy-producing metabolism may play a role in HG/PA-induced beta cell cytotoxicity.
Fig. 5. Protective effect of FAO stimulator, anaplerosis enhancer, and cataplerosis inhibitor on HG/PA-induced caspase 3 activation. INS-1 cells were treated with 25 mM
glucose (HG) and 0.4 mM palmitate (PA) in the presence of 0.5 mM bezafibrate (Bez), 10 mM BCH or 2.5 mM BTA for 15 h. Cleaved caspase 3 and cleaved PARP were detected
by immunoblotting (IB) with anti-cleaved caspase 3 and anti-PARP antibodies, respectively (A). P-Cas3, C-Cas3, P-PARP, and C-PARP, represent pro-caspase 3, cleaved caspase
3, uncleaved PARP, and cleaved PARP, respectively. **p < 0.01 vs. cleaved caspase 3 from HG/PA-treated INS-1 cells. ##p < 0.01 vs. cleaved PARP from HG/PA-treated cells.
238
S.-E. Choi et al. / Archives of Biochemistry and Biophysics 505 (2011) 231–241
Fig. 6. Knockdown effect of glutamate dehydrogenase and mitochondrial tricarboxylate carrier on HG/PA-induced INS-1 cell death. INS-1 cells were transfected with siRNA
duplex of green fluorescent protein (GF), glutamate dehydrogenase (GD) (B), or tricarboxylase carrier (TC) (C). After 36 h of incubation, INS-1 cells were treated with HG/PA,
HG/PA/BCH, or HG/PA/BTA for 15 h. RNA level of GDH and TC was determined by semi-quantitative RT-PCR (RT). Cleaved caspase 3 and cleaved PARP and were detected by
immunoblotting analysis. P-Cas3, C-Cas3, P-PARP, and C-PARP represent pro-caspase 3, cleaved caspase 3, uncleaved PARP, and cleaved PARP, respectively. **p < 0.01 vs.
cleaved caspase 3 from HG/PA-treated INS-1 cells. ##p < 0.01 vs. cleaved PARP from HG/PA-treated cells. $p < 0.05; $$p < 0.01 vs. cleaved caspase 3 from BCH- or BTA-treated
INS-1 cells. +p < 0.05; ++p < 0.01 vs. cleaved PARP from BCH- or BTA-treated cells.
Initially it was suggested that the accumulation of toxic lipid
intermediates such as ceramide, LC-CoA, DAG, or phospholipids
and subsequent activation of the lipid-mediated signals was linked
in beta cell glucolipotoxicity [19]. Reactive oxygen species (ROS)
produced during augmented lipogenesis was suggested to contribute to beta cell glucolipotoxicity [11,13]. However, several reports
opposed these suggestions. Ceramide was not believed to play a
role in FFA-induced lipotoxicity since a ceramide synthase inhibitor did not block FFA-induced cytotoxicity [26,33,36]. We also observed that fumonisin B1 and myriocin, inhibitors of sphingosine
N-acetyl transferase and serine palmitoyltransferase, respectively,
afforded little protection against HG/PA-induced cytotoxicity (Supplemental Fig. 2A). Although inhibiting PKC was weakly protective
(Supplemental Fig. 2B), Welters et al. reported that PKCd was not
required for palmitate-induced cytotoxicity to beta cells [37]. The
data that treatment of the INS-1 cells with antioxidants such as
NAC and GSH did not protect against HG/PA-induced cytotoxicity
suspected the role of ROS in HG/PA-induced cytotoxicity (Supplemental Fig. 2C). Recently, El-Assaad et al. reported that ROS were
not implicated in beta cell glucolipotoxicity [23]. In fact, it has been
reported that FFA-induced cytotoxicity and lipid accumulation are
inversely related [26], and that concomitant promotion of lipid
esterification and lipolysis could detoxify glucolipotoxicity [30].
Of particular note was that T0901317, an liver X receptor (LXR)
agonist that stimulated FFA/lipid synthesis, strongly protected
against HG/PA-induced cytotoxicity, weakening support for a potential role of glucose-induced lipogenesis in HG/PA-induced cytotoxicity (Supplemental Fig. 2D). Collectively, recent reports as well
as our data do not support the hypothesis that accumulation of a
specific lipid intermediate and subsequent activation of the lipidmediated signal may be the direct mediators in HG/PA-induced
beta cell glucolipotoxicity.
Stimulation or inhibition of FAO protected or augmented HG/
PA-induced beta cell death, respectively. These data are consistent
with a study published by El-Assaad et al., which reported that FAO
stimulator AICAR protected beta cells against HG/FFA-induced
glucolipotoxicity while FAO inhibitor etomoxir amplified the toxicity [4]. This data including ours suggested that FAO might be impaired in HG/PA-treated cells. When beta cells are simultaneous
exposed to high levels of glucose and FFA, a transient increase of
malonyl-CoA is believed to block acyl-CoA transport into the mitochondria and ultimately block FAO since malonyl-CoA inhibits
CPT-1 [19,21]. A report showing that over-expression of CPT-1
counteracted beta cell glucolipotoxicity also suggests that transportation of acyl-CoA to mitochondria may be impaired in HG/
PA-treated cells [38]. In addition to impaired transportation, oxida-
S.-E. Choi et al. / Archives of Biochemistry and Biophysics 505 (2011) 231–241
239
Fig. 7. Restoration of the HG/PA-induced reduction of mitochondrial metabolism by BCH and BTA. INS-1 cell were treated with HG/PA for 12 h in the absence or presence of
0.5 mM bezafibrate (Bez), 10 mM BCH, or 2.5 mM BTA The levels of oxaloacetate (A), citrate (B), and a-ketoglutarate (C) in cellular extracts were then determined using a
metabolism assay kit (Biovision). Data are represented as means ± SEM from three independent experiments. *p < 0.05; **p < 0.01 vs. oxaloacetate, citrate, and aketoglutarate in HG-treated INS-1 cells. #p < 0.05; ##p < 0.01 vs. oxaloacetate, citrate, and a-ketoglutarate in HG/PA-treated INS-1 cells. Oxidation rates of glucose or palmitate
were determined by measurement of labelled CO2 produced from D-[U-14C] glucose (D) or [1-14C] palmitate (E). *p < 0.05 vs. oxidation rate in HG-treated INS-1 cells.
##
p < 0.01 vs. oxidation rate in HG/PA-treated INS-1 cells. (F) ATP level was measured using the CellTiter-GloTM luminescence kit. #p < 0.05: ##p < 0.01 vs. ATP level in HG/PAtreated INS-1 cells.
tion of FFAs in mitochondria may be incomplete in beta cells since
incomplete FAO was reported to be a contributor to FFA-induced
insulin resistance [31]. On the other hand, Xu et al. reported that
high level palmitate as well as high concentration of glucose upregulated expression of pyruvate dehydrogenase kinase in beta
cells [39], which reduced pyruvate dehydrogenase activity and
subsequently inhibited supply of pyruvate to the TCA cycle. This
result suggests that the supply of both pyruvate and FFA as energy
substrates is not sufficient in mitochondria, and, subsequently, energy-producing metabolism to compensate the enhanced lipogenic
pathway can be impaired in HG/PA-treated cells. In fact, our experimental data showing that the oxidation rate of glucose as well as
palmitate declined and that intracellular ATP level was significantly low in HG/PA-treated INS-1 cells support that impaired energy-producing metabolism may be due to an insufficient supply of
pyruvate and FFA into mitochondria. Reports showing that the flux
of the TCA cycle was reduced in type 2 diabetes and that an enhanced mitochondrial metabolism accounted for adaptation of
beta cells to high-fat diet-induced insulin resistance [40,41] support that appropriate oxidation metabolism is essential for maintenance of cell viability in HG/PA-exposed beta cells. In addition to
an insufficient supply of fuel substrates, it was reported that the
ATP synthesis process was impaired in FFA-treated cells [42,43]
and expression of the molecules related to energy metabolism
were decreased in islets of diabetic subjects [44]. Collectively, energy-producing metabolism is impaired in HG/PA-treated cells,
which may explain the induction of beta cell glucolipotoxicity.
PAA, which blocked supply of TCA cycle intermediate oxaloacetate, potentiated HG/PA-induced INS-1 cell death. This result suggests that insufficient supply of the TCA cycle intermediates can
also be a critical mediator for HG/PA-induced toxicity. In fact, most
attempts to increase the TCA cycle intermediates protected the INS1 cells against HG/PA-induced cytotoxicity. A knockdown study of
GDH or TC augmenting or attenuating HG/PA-induced INS-1 cell
death, respectively, supports that maintenance of TCA cycle intermediate pool through enhanced anaplerosis and reduced cataplerosis is critical for maintaining cell viability in HG/PA-treated INS-1
cells. In fact, our preliminary experiment that treatment with both
PAA and etomoxir, which may impair supply of TCA cycle metabolites, induced similar mode of cytotoxicity induced by HG/PA. Treatment of INS-1 cells with PAA and etomoxir induced apoptotic death
and the PAA/etomoxir-induced death was prevented by treatment
with FAO stimulators or GDH activator (Data not shown). Since
FFA-induced insulin resistance was able to be caused by lack of
TCA intermediates, we measured cellular level of TCA cycle intermediates. Levels of oxaloactate, citrate, and a-ketoglutarate in
HG/PA-treated INS-1 cells were significantly lower than those in
untreated cells at the time when the viability reduction was detected. In particular, the cellular level of oxaloacetate was remarkably reduced at 3 h after HG/PA treatment, suggesting that HG/PA
treatment would decrease activity of PC in INS-1 cell. Busch et al.
reported that beta cells exposed to palmitate showed lower expression of PC in a gene array study [45] and MacDonald et al. demonstrated a decreased level of mitochondrial metabolic enzymes
including PC in pancreatic islets of patients with type 2 diabetes
[33]. In particular, Lu et al. provided evidence that mitochondrial
defects including lower level of PC and GDH played a critical role
in the pathology of insulin resistance-induced beta cell failure in
diabetic MKR mice [32]. Very recently, reduction of PC activity
was also suggested to be a cause for development of Type 2 diabetes
in Agouti-K mice [46]. All these data support that insufficient supply of TCA cycle intermediates through impaired PC and/or GDH
240
S.-E. Choi et al. / Archives of Biochemistry and Biophysics 505 (2011) 231–241
may plays a role in HG/PA-induced beta cell cytotoxicity. Our data
showing that methyl pyruvate did not afford protection against HG/
PA-induced cell cytotoxicity suggested the anaplerotic pathway
through PC is impaired in HG/PA-treated INS-1 cells. Recent report
stating that an LXR agonist could augment anaplerotic pathway
through enhanced activation of PC may explain our result that
T0901317 had strong protective effect on HG/PA-induced death
(Supplemental Fig. 2D), even though its enhanced lipogenic activity
[47]. Furthermore, a report that over-expression of PPAR up-regulated expression of anaplerotic pyruvate carboxylase [48] also suggests that FAO stimulation can also up-regulate the anaplerotic
process, supporting that FAO stimulators protect against HG/PA-induced glucolipotoxicity through a sufficient supply of TCA cycle
intermediates. In particular, our experiments showing that enhanced anaplerosis by BCH and reduced cataplerosis by BTA restored HG/PA-induced reduction of oxidation rate and ATP level
suggest that the impaired energy metabolism in HG/PA-treated
cells is partly due to insufficient supply of TCA cycle intermediates.
Interestingly, imbalanced mitochondrial energy production in HG/
PA-treated cells may explain beta cell dysfunctions observed in
type 2 diabetes, such as impaired glucose-stimulated insulin secretion and impaired expression of insulin gene [49–51].
Although metabolic failure through depletion of TCA cycle
intermediates was suggested to be a cause of HG/PA-induced
glucolipotoxicity, the intracellular mechanism how failure of energy-producing metabolism activates toxic signals in INS-1 beta
cells is not clearly defined. Earlier reports suggest that oxidative
stress, ER stress, Ca2+ signal, and impaired insulin signal were
intracellular mediators for HG/PA-induced cytotoxicity. In fact,
these cytotoxic mediators can be linked to mitochondrial dysfunction. Enhanced production of ROS and subsequent activation of
oxidative stress signal was a typical toxic mediator produced from
mitochondrial dysfunction and therefore, could be a possible mediator for HG/PA-induced cytotoxicity to INS-1 cells [23]. However,
since antioxidants such as NAC and GSH did not protect against
HG/PA-induced cytotoxicity, ROS and oxidative stress does not
seem to be a mediator for HG/PA-induced toxicity. Since coupling
mitochondrial dysfunction to ER stress responses was recently reported in hepatic cells [52], metabolic failure in mitochondrion
through depletion of TCA cycle intermediates may induce HG/PAinduced cytotoxicity through ER stress induction. It will be further
studied whether the ER stress can be a critical linker between
mitochondrial metabolic failure and HG/PA-induced INS-1 cell
death. Another mediator linking mitochondrial dysfunction to
glucolipotoxicity can be elevation of intracellular Ca2+. Since it
was reported that defective mitochondrial function caused the
cells to be vulnerable to Ca2+-mediated toxic signals [53], metabolic failure in beta cell mitochondria may induce HG/PA-induced
cytotoxicity through activation of Ca2+ signals. A recent report that
mitochondrial dysfunction inhibited insulin signalling pathway
[54] strongly suggests that mitochondrial dysfunction may induce
beta cell glucolipotoxicity through inhibition of insulin signalling
pathway.
To test whether HG/PA treatment induced cytotoxicity in primary beta cells and whether the HG/PA-induced toxicity was
reduced by BCH and BTA, we initially investigated the glucose-induced augmentation of PA-induced viability reduction of primary
islet cells. High concentration of glucose did not potentiate PA-induced viability reduction of primary islet cells isolated from nondiabetic rats, but potentiated the viability reduction of primary
islet cells isolated from diabetic animals, such as Otsuka Long
Evans Tokushima Fatty (OLETF) and Zucker Diabetic Fatty (ZDF)
(Supplemental Fig. 3). Protective effect through BCH and BTA on
HG/PA-induced cytotoxicity was also observed in primary islet
cells isolated from these diabetic rats (Supplemental Fig. 3). These
results suggest that metabolic failure in mitochondria can also be a
mediator for glucolipotoxicity to primary beta cells. Although our
data demonstrate that supplement of TCA cycle intermediates protects against HG/PA-induced toxicity to primary beta cells, further
research is needed to establish whether elevated FFA and high concentration of glucose induce beta cell loss in diabetic subjects
through metabolic failure in mitochondria. The concentrations of
glucose and palmitate used in this study are higher than the concentrations observed in normal human plasma. However, the concentrations would be similar to the concentrations experienced in
diabetic patients. Even in extreme cases, the postprandial blood
glucose concentration can easily be over 25 mM [55] and the level
of saturated fatty acid can also be over 0.4 mM in severely diabetic
patients [56]. At least, the concentrations we used can be relevant
concentrations for explaining beta cell loss in severely diabetic
patients.
In conclusion, metabolic failure in mitochondrion is thought to
play a critical role in HG/PA-induced cytotoxicity to beta cells. The
HG/PA-induced cytotoxicity could be protected by supplement of
the TCA cycle intermediates as well as stimulation of FAO. Thus,
insufficient capacity of mitochondrial energy metabolism in beta
cells may determine the shift from glucolipoadaptation to glucolipotoxicity in diabetic subjects. Nutritional or pharmacological approaches enhancing TCA cycle intermediate pool in mitochondria
can be novel targets to protect against beta cell loss characterised
by type 2 diabetes.
Acknowledgments
This study was supported by grants from Korea Research Foundation (KRF 2009-0072584) and Innovative Research Institute for
Cell Therapy (A062260).
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.abb.2010.10.011.
References
[1] M.Y. Donath, J. Storling, K. Maedler, T. Mandrup-Poulsen, J. Mol. Med. 81
(2003) 455–470.
[2] V. Poitout, R.P. Robertson, Endocr. Rev. 29 (2008) 351–366.
[3] R. Lupi, F. Dotta, L. Marselli, S. Del Guerra, M. Masini, C. Santangelo, G. Patane,
U. Boggi, S. Piro, M. Anello, E. Bergamini, F. Mosca, U. Di Mario, S. Del Prato, P.
Marchetti, Diabetes 51 (2002) 1437–1442.
[4] W. El-Assaad, J. Buteau, M.L. Peyot, C. Nolan, R. Roduit, S. Hardy, E. Joly, G.
Dbaibo, L. Rosenberg, M. Prentki, Endocrinology 144 (2003) 4154–4163.
[5] K. Maedler, J. Oberholzer, P. Bucher, G.A. Spinas, M.Y. Donath, Diabetes 52
(2003) 726–733.
[6] R. Okuyama, T. Fujiwara, J. Ohsumi, FEBS Lett. 545 (2003) 219–223.
[7] M. Shimabukuro, Y.T. Zhou, M. Levi, R.H. Unger, Proc. Nat. Acad. Sci. USA 95
(1998) 2498–2502.
[8] V. Grishko, L. Rachek, S. Musiyenko, S.P. Ledoux, G.L. Wilson, Free Radic. Biol.
Med. 38 (2005) 755–762.
[9] K. Eitel, H. Staiger, J. Rieger, H. Mischak, H. Brandhorst, M.D. Brendel, R.G.
Bretzel, H.U. Haring, M. Kellerer, Diabetes 52 (2003) 991–997.
[10] A.M. Hennige, F. Ranta, I. Heinzelmann, M. Düfer, D. Michael, H. Braumüller,
S.Z. Lutz, R. Lammers, F. Bosch, H.U. Häring, S. Ullrich, Diabetes 59 (2010) 119–
127.
[11] S. Piro, M. Anello, C. Di Pietro, M.N. Lizzio, G. Patanè, A.M. Rabuazzo, R. Vigneri,
M. Purrello, F. Purrello, Metabolism 51 (2002) 1340–1347.
[12] R. Robertson, H. Zhou, T. Zhang, J.S. Harmon, Cell. Biochem. Biophys. 48 (2007)
139–146.
[13] S.S. Choe, A.H. Choi, J.W. Lee, K.H. Kim, J.J. Chung, J. Park, K.M. Lee, K.G. Park, I.K.
Lee, J.B. Kim, Diabetes 56 (2007) 1534–1543.
[14] S.E. Choi, H.E. Kim, H.C. Shin, H.J. Jang, K.W. Lee, Y. Kim, S.S. Kang, J. Chun, Y.
Kang, Mol. Cell. Endocrinol. 272 (2007) 50–62.
[15] K.S. Gwiazda, T.L. Yang, Y. Lin, J.D. Johnson, Am. J. Physiol. Endocrinol. Metab.
296 (2009) E690–701.
[16] M.K. Lingohr, L.M. Dickson, C.E. Wrede, I. Briaud, J.F. McCuaig, M.G. Myers, C.J.
Rhodes, Mol. Cell. Endocrinol. 209 (2003) 17–31.
[17] E. Karaskov, C. Scott, L. Zhang, T. Teodoro, M. Ravazzola, A. Volchuk,
Endocrinology 147 (2006) 3398–3407.
S.-E. Choi et al. / Archives of Biochemistry and Biophysics 505 (2011) 231–241
[18] E. Lai, G. Bikopoulos, M.B. Wheeler, M. Rozakis-Adcock, A. Volchuk, Am. J.
Physiol. Endocrinol. Metab. 294 (2008) E540–550.
[19] M. Prentki, E. Joly, W. El-Assaad, R. Roduit, Diabetes 51 (Suppl. 3) (2002) s405–
s413.
[20] M.J. MacDonald, L.A. Fahien, L.J. Brown, N.M. Hasan, J.D. Buss, M.A. Kendrick,
Am. J. Physiol. Endocrinol. Metab. 288 (2005) E1–15.
[21] D. Saggerson, Annu. Rev. Nutr. 28 (2008) 253–272.
[22] M.B. Sandberg, J. Fridriksson, L. Madsen, V. Rishi, C. Vinson, H. Holmsen, R.K.
Berge, S. Mandrup, Mol. Cell. Endocrinol. 240 (2005) 94–106.
[23] W. El-Assaad, E. Joly, A. Barbeau, R. Sladek, J. Buteau, I. Maestre, E.
Pepin, S. Zhao, J. Iglesias, E. Roche, M. Prentki, Endocrinology 151
(2010) 3061–3073.
[24] I. Briaud, J.S. Harmon, C.L. Kelpe, V.B. Segu, V. Poitout, Diabetes 50 (2001) 315–
321.
[25] G.J. Wanten, A.H. Naber, Mini. Rev. Med. Chem. 4 (2004) 847–857.
[26] M. Cnop, J.C. Hannaert, A. Hoorens, D.L. Eizirik, D.G. Pipeleers, Diabetes 50
(2001) 1771–1777.
[27] L.L. Listenberger, X. Han, S.E. Lewis, S. Cases, R.V. Farese Jr., D.S. Ory, J.E.
Schaffer, Proc. Natl. Acad. Sci. USA 100 (2003) 3077–3082.
[28] A.K. Busch, E. Gurisik, D.V. Cordery, M. Sudlow, G.S. Denyer, D.R. Laybutt, W.E.
Hughes, T.J. Biden, Diabetes 54 (2005) 2917–2924.
[29] C.J. Nolan, M. Prentki, Trends Endocrinol. Metab. 19 (2008) 285–291.
[30] C.J. Nolan, J.L. Leahy, V. Delghingaro-Augusto, J. Moibi, K. Soni, M.L. Peyot, M.
Fortier, C. Guay, J. Lamontagne, A. Barbeau, E. Przybytkowski, E. Joly, P.
Masiello, S. Wang, G.A. Mitchell, M. Prentki, Diabetologia 49 (2006) 2120–
2130.
[31] T.R. Koves, J.R. Ussher, R.C. Noland, D. Slentz, M. Mosedale, O. Ilkayeva, J. Bain,
R. Stevens, J.R. Dyck, C.B. Newgard, G.D. Lopaschuk, D.M. Muoio, Cell Metab. 7
(2008) 45–56.
[32] H. Lu, V. Koshkin, E.M. Allister, A.V. Gyulkhandanyan, M.B. Wheeler, Diabetes
59 (2010) 448–459.
[33] M.J. MacDonald, M.J. Longacre, E.C. Langberg, A. Tibell, M.A. Kendrick, T. Fukao,
C.G. Ostenson, Diabetologia 52 (2009) 1087–1091.
[34] L.L. Listenberger, D.S. Ory, J.E. Schaffer, J. Biol. Chem. 276 (2001) 14890–14895.
[35] W.J. Lee, M. Kim, H. Park S, H.S. Kim, M.J. Jeon, K.S. Oh, E.H. Koh, J.C. Won, M.S.
Kim, G.T. Oh, M. Yoon, K.U. Lee, J.Y. Park, Biochem. Biophys. Res. Commun. 340
(2006) 291–295.
[36] N. Beeharry, J.A. Chambers, I.C. Green, Apoptosis 9 (2004) 599–607.
241
[37] H.J. Welters, S.A. Smith, M. Tadayyon, J.H. Scarpello, N.G. Morgan, J. Mol.
Endocrinol. 32 (2004) 227–235.
[38] E.M. Sol, E. Sargsyan, G. Akusjärvi, P. Bergsten, Biochem. Biophys. Res.
Commun. 375 (2008) 517–521.
[39] J. Xu, J. Han, P.N. Epstein, Y.Q. Liu, Biochem. Biophys. Res. Commun. 344 (2006)
827–833.
[40] M. Fex, M.D. Nitert, N. Wierup, F. Sundler, C. Ling, H. Mulder, Diabetologia 50
(2007) 74–83.
[41] P. Schrauwen, M.K. Hesselink, Diabetologia 51 (2008) 1694–1697.
[42] R. Köhnke, J. Mei, M. Park, D.A. York, C. Erlanson-Albertsson, Nutr. Neurosci. 10
(2007) 273–278.
[43] M.A. Abdul-Ghani, F.L. Muller, Y. Liu, A.O. Chavez, B. Balas, P. Zuo, Z. Chang, D.
Tripathy, R. Jani, M. Molina-Carrion, A. Monroy, F. Folli, H. Van Remmen, R.A.
DeFronzo, Am. J. Physiol. Endocrinol. Metab. 295 (2008) 678–685.
[44] H. Lu, Y. Yang, E.M. Allister, N. Wijesekara, M.B. Wheeler, Mol. Cell. Proteomics
7 (2008) 1434–1451.
[45] A.K. Busch, D. Cordery, G.S. Denyer, T.J. Biden, Diabetes 51 (2002) 977–987.
[46] J. Han, Y.Q. Liu, J. Endocrinol. 204 (2010) 143–152.
[47] T. Ogihara, C. Chung, G.L. Vestermark, J.C. Garmey, R.J. Ketchum, X. Huang, K.L.
Brayman, M.O. Thorner, J.J. Repa, R.G. Mirmira, C. Evans-Molina, J. Biol. Chem.
285 (2010) 5392–5404.
[48] F. Frigerio, T. Brun, C. Bartley, A. Usardi, D. Bosco, K. Ravnskjaer, S. Mandrup, P.
Maechler, Diabetologia 53 (2010) 331–340.
[49] A. Boucher, D. Lu, S.C. Burgess, S. Telemaque-Potts, M.V. Jensen, H. Mulder,
M.Y. Wang, R.H. Unger, A.D. Sherry, C.B. Newgard, J. Biol. Chem. 279 (2004)
27263–27271.
[50] N.M. Hasan, M.J. Longacre, S.W. Stoker, T. Boonsaen, S. Jitrapakdee, M.A.
Kendrick, J.C. Wallace, M.J. MacDonald, J. Biol. Chem. 283 (2008) 28048–
28059.
[51] A. Tzatsos, P.N. Tsichlis, J. Biol. Chem. 282 (2007) 18069–18082.
[52] J.H. Lim, H.J. Lee, M.H. Jung, J. Song, Cell. Signal. 21 (2009) 169–177.
[53] N. Brustovetsky, T. Brustovetsky, K.J. Purl, M. Capano, M. Crompton, J.M.
Dubinsky, J. Neurosci. 23 (2003) 4858–4867.
[54] J.A. Kim, J. Wei, J.R. Sowers, Circ. Res. 102 (2008) 401–414.
[55] A. Babu, A. Mehta, P. Guerrero, Z. Chen, P.M. Meyer, C.K. Koh, R. Roberts, J.
Schaider, L. Fogelfeld, Endocr. Pract. 15 (2009) 696–794.
[56] N. John, J.N. Clore, J. Allred, D. White, J. Li, J. Stillman, Metabolism 51 (2002)
1471–1477.