B

I

Dwvdocicmc

ptads

TfC

CTCr

N©

ASIC NUTRITIONAL INVESTIGATIONS

Role of Dietary Fatty Acids and AcuteHyperglycemia in Modulating Cardiac Cell Death

Sanjoy Ghosh, MSc, Ding An, MSc, Thomas Pulinilkunnil, MSc, Dake Qi, MSc,Howard C. S. Lau, BSc, Ashraf Abrahani, MSc, Sheila M. Innis, PhD, and

Brian Rodrigues, PhDFrom the Division of Pharmacology and Toxicology, Faculty of Pharmaceutical Sciences, and

the Department of Pediatrics, The University of British Columbia, Vancouver,British Columbia, Canada

OBJECTIVE: We examined the effect of dietary manipulation of palmitic acid (20% [w/w] palm oil [PO])on cardiomyocyte apoptosis in the rat heart under normoglycemic and hyperglycemic conditions in vivo.We used 20% (w/w) sunflower oil (SO; a diet rich in �-6 polyunsaturated fatty acids) as an isocaloriccontrol.METHODS: Adult male Wistar rats were fed experimental diets containing normal laboratory chow (5%corn oil) or a high fat diet (AIN-76A with PO or SO) for 4 wk. Subsequently, to induce diabetes, rats wereinjected with streptozotocin (55 mg/kg, intravenously). After 4 d of diabetes, hearts were tested forevidence of lipotoxicity and cell death, and the serum for its related markers.RESULTS: Feeding PO and SO magnified palmitic and linoleic acid contents within lipoproteins and heartsrespectively. Compared with SO, PO diabetic hearts demonstrated significantly higher levels of apoptosis,with an altered Bax:Bcl-2 ratio, augmented lipid peroxidation, and protein modification by formation ofnitrotyrosine. Interestingly, SO-fed diabetic animals demonstrated an increase in serum lactate dehydro-genase and myocardial necrotic changes.CONCLUSION: In marked contrast to results obtained in vitro, PO feeding led to only a minor fraction ofcardiomyocytes undergoing apoptosis and suggests that, in the intact heart, protective mechanisms couldbe triggered that dampen excessive apoptosis. Of greater clinical significance was the observation that“heart-friendly” vegetable oils such as SO, rich in �-6 polyunsaturated fatty acids, could precipitatecardiac necrosis, and questions its beneficial role in the cardiovascular system, especially followingdiabetes. Nutrition 2004;20:916–923. ©Elsevier Inc. 2004

KEY WORDS: apoptosis, necrosis, palmitic acid, linoleic acid, diabetes

NTRODUCTION

iabetes is a significant risk factor for cardiovascular diseases,ith most of these complications being attributed to coronaryascular pathology. However, in humans and animal models ofiabetes, an additional heart muscle-specific disease in the absencef any vascular pathology has also been described, termed diabeticardiomyopathy.1 Probable candidates to explain this heart diseasenclude autonomic abnormalities, metabolic disorders, abnormalontractile protein and enzyme function, interstitial fibrosis,1 and,ore recently, apoptosis, a regulated, energy-dependent, cell sui-

ide mechanism.2,3

An abnormal increase in plasma glucose has been suggested toredispose cardiomyocytes to death by apoptosis such that con-ractile function is ultimately altered. Changes in protein kinase Cctivity,4 generation of reactive oxygen species (ROS)5 mitochon-rial dysfunction,6 and activation of p53 and the renin-angiotensinystem3,7 have been put forward to explain this hyperglycemia-

he studies described in this paper were supported by operating grantsrom the Heart and Stroke Foundation of British Columbia and Yukon,anada.

orrespondence to: Brian Rodrigues, PhD, Division of Pharmacology andoxicology, Faculty of Pharmaceutical Sciences, The University of Britisholumbia, 2146 East Mall, Vancouver, BC, Canada V6T 1Z3. E-mail:

odrigue@unixg.ubc.ca

utrition 20:916–923, 2004Elsevier Inc., 2004. Printed in the United States. All rights reserved.

induced cell death. During diabetes, because glucose transport andoxidation are defective, cardiac energy production is almost ex-clusively by breakdown of fatty acids (FAs) that are supplied inexcess to the heart.8 Recently, cardiac apoptosis has been inde-pendently linked to an overload of FAs, specifically saturated FAssuch as palmitic acid.9,10 Palmitic acid in excess causes intracel-lular accumulation of ceramide and ROS11 and, in neonatal andadult cardiomyocytes, induces apoptosis in the absence or presenceof hyperglycemia.10,12 All of these experiments used free palmiticacid and in vitro model systems. Although this setting partlyresembles palmitic acid derived from plasma, it does not accountfor other FA sources (e.g., provision of palmitic acid via break-down of triacylglycerol [TG] within lipoproteins), nor does it takeinto account the effect of palmitic acid in a milieu of diverse FAsunder in vivo conditions. Moreover, in isolated cardiomyocytes,the metabolic fate of palmitic acid is likely to be altered, given thedecreased energy demand of non-beating, quiescent cells. The objec-tive of the present study was to determine the effect of dietarymanipulation of palmitic acid on cardiomyocyte apoptosis in the ratheart, under normoglycemic and hyperglycemic conditions, in vivo.

MATERIALS AND METHODS

Experimental Animals and High-Fat Diets

Animals were cared for in accordance with the principles promul-

gated by the Canadian Council of Animal Care and the University

0899-9007/04/$30.00doi:10.1016/j.nut.2004.06.013

ofmcI[T(hdwsimg[af

I

TaWwh

dpsctsme

S

BrfdauTdadlsgtScrwhed

C

AnsdttnfluqwNvttwninaSuAqe

SR

CRT

PCFT

I

CD

LCSCCSPMVCFFT

*pmfc†afH

Nutrition Volume 20, Number 10, 2004 917Dietary Fats and Cell Death in the STZ-Diabetic Heart

f British Columbia. Male Wistar rats (220 to 240 g) were obtainedrom the University of British Columbia Animal Care Unit andaintained under a 12-h light, 12-h dark (7:00 AM to 7:00 PM)

ycle. Animals were fed a high-fat (HF) diet (The Americannstitute of Nutrition [AIN] 76A diet supplemented with 20%w/w] palm oil; Research Diets Inc., New Brunswick, NJ, USA;able I) or a normal laboratory chow for 4 wk. Normal chow

Laboratory Rodent Diet 5001, PMI Feeds, Richmond, VA, USA)ad the following composition (kcal%): protein, 28; carbohy-rates, 60; fats, 12; total, 4.0 kcal/g. Energy contribution from fatsas 40% of the total energy intake in the HF groups. Previous

tudies have established that feeding of 20% (w/w) palm oil (PO)ncreases the palmitic acid content of circulating lipoproteins,13 aajor source of FAs to the heart. An additional isocaloric control

roup (AIN-76A supplemented with 20% [w/w] sunflower oilSO]), which increases polyunsaturated fatty acids (PUFAs), suchs linoleic acid, was included to compare the effects of high fateeding per se. Water was provided ad libitum.

nduction of Diabetes

o induce diabetes, halothane-anesthetized rats were injected withsingle intravenous injection of streptozotocin (STZ; 55 mg/kg).ith this dose of STZ, there is only partial destruction of �-cellsith an approximately 50% decrease in plasma insulin and stable

TABLE I.

COMPOSITION OF EXPERIMENTAL HIGH-FAT DIETS

Sunflower oil diet Palm oil diet

g% kcal% g% kcal%

rotein 24.1 21 24.1 21arbohydrate 44.9 39 44.9 39at 20.7 40 20.7 40otal (kcal/g) 4.62 100 4.62 100

ngredients g kcal g kcal

asein, 80 mesh 200 800 200 800L-methionine 3 12 3 12oDex 10 75 300 75 300orn starch 75 300 75 300ucrose 218.8 875 218.8 875ellulose, BW200 50 — 50 —orn oil 10 90 10 90unflower oil 165 1485 — —alm oil — — 165 1485ineral mix* 35 — 35itamin mix† 10 40 10 40holine bitartrate 2 — 2 —D&C dye no. 40 0.1 — — —D&C dye no. 1 — — 0.1 —otal 843.9 3902 843.9 3902

Mineral mix (mg/g): di-calcium phosphate 500; magnesium oxide 24;otassium citrate 220; potassium sulfate 52; sodium chloride 74; chro-ium KSO4 · 12H2O 0.55; cupric carbonate 0.3; potassium iodate 0.01;

erric citrate 6; manganous carbonate 3.5; sodium selenite 0.01; zincarbonate 1.6; sucrose 118.03Vitamin mix (mg/g): vitamin A 0.8; vitamin D3 1; vitamin E 10; men-

dione sodium bisulfite 0.08; biotin 1% 2; cyancocobalamin 0.1% 1;olic acid 0.2; nicotinic acid 3; calcium pantothenate 1.6; pyridoxine-Cl 0.7; riboflavin 0.6; thiamin HCl 0.6; sucrose 978.42.

yperglycemia. Some animals fed HF diets for 4 wk were made s

iabetic and kept for 4 d after STZ administration. An acute timeeriod was chosen: 1) at least in vitro, saturated FAs have beenhown to induce apoptosis within 24 h,10,12 and 2) maximumardiac apoptosis occurred on the third day after STZ administra-ion.3 HF diets were continued throughout the 4 d of diabetes;ubsequently, rats were anesthetized with sodium pentobarbital (65g/kg, intraperitoneally), blood was collected, and hearts were

xcised for morphologic and biochemical analyses.

eparation of Lipoproteins and Cardiac Free FAs

efore removal of the heart, blood was withdrawn from the infe-ior vena cava. Serum was obtained after centrifugation at 3000gor 20 min and then used for isolation of major lipoproteins byensity gradient ultracentrifugation.14 Briefly, ultracentrifugationt 288 000g at 15°C for 18 h was carried out in an L8-80 Beckmanltracentrifuge (Beckman Coulter, Inc., Fullerton, CA, USA).ubes were removed from the centrifuge with care so as not toisturb the gradient layers. Very low-density lipoprotein (VLDL)nd chylomicrons float on top (�1.006 g/mL), whereas low-ensity lipoprotein (LDL) separates at 1.019 g/mL, high-densityipoprotein (HDL) at 1.21 g/mL, and lipoprotein-deficient serumtays at the bottom. The lipoprotein layers were removed with alass Pasteur pipette, and the TG fraction of the isolated lipopro-eins (VLDL and chylomicrons) were analyzed for individual FAs.eparation of polar and non-polar lipids, including individuallasses, was achieved using high-performance liquid chromatog-aphy (Waters 2690 Alliance HPLC, Milford, MA, USA) equippedith an auto-sampler and column heater.15 Total lipids from theeart were extracted and solubilized in chloroform:methanol:ac-tone:hexane (4:6:1:1 v/v/v/v), and free FA (FFA) was analyzed asescribed above.

ardiac Apoptosis

fter termination, part of the left ventricle was fixed in 10%eutral buffered formalin and embedded in paraffin, and 5-�mections were prepared. Deoxynucleotidyl transferase (TdT) me-iated dUTP nick end labeling (TUNEL) assay was carried out onhese sections using a Fluorescein-Fragel DNA fragmentation de-ection kit (Oncogene, Manhasset, NY, USA). During apoptosis,uclear DNA fragmentation results in 3�-OH overhangs to whichuorescein isothiocyanate (FITC) labeled nucleotides were boundsing terminal TdT (with the emission of fluorescence). Subse-uent to TUNEL staining and to identify cardiomyocytes, sectionsere labeled with mouse monoclonal anti-desmin antibody (1:50;ovocastra Labs Ltd., Newcastle, UK) for 1 h as described pre-iously.16 Anti-mouse Alexa Fluor 549 conjugated secondary an-ibody (Molecular Probes, Eugene, OR, USA; 10 �g/mL) was usedo visualize cardiomyocytes. Sections were finally counterstainedith 4,6-diamidino-2-phenylindole (DAPI, to visualize normaluclei) and examined using confocal fluorescent microscopy. Pos-tive (DNAse 1; 10 min, 100 �L of a 1-mg/mL solution) andegative (omitting TdT during TUNEL) controls were also gener-ted using normal chow- and PO-fed heart sections, respectively.16

lides were visualized using FITC, DAPI, and Texas Red filtersnder a Biorad 600 Confocal Microscope at 400� magnification.t least six random sections of each heart were observed anduantified using the imaging software Northern Eclipse. Values arexpressed as the number of TdT-labeled nuclei per 106 nuclei.

emiquantitative Reverse Transcriptase Polymerase Chaineaction (RT-PCR)

ardiac gene expressions were measured with semiquantitativeT-PCR. Briefly, total RNA was extracted using Trizol (Lifeechnologies, Bethesda, MD, USA). Subsequently, reverse tran-

cription of total RNA was carried out using oligo (dT) primers

(rTABGUwsaaaim

O

OrosnwtpoTwafnUgpsFA

C

Cr

LMpawssrwrmmthSiw((sawf

S

SacCaL

S

Vft

BSSSF

*†‡§wfF

918 Ghosh et al. Nutrition Volume 20, Number 10, 2004

Invitrogen, La Jolla, CA, USA). PCR primers were designed fromat sequences available in GenBank and are as follows: BCl-2, 5�-ATGATAACCGGGAGATCGTG-3� (forward) and 5�-CAG-TGCCGGTTCAGGTACTC-3� (reverse; accession no. L14680);AX, 5�-GCGAATTGGAGATGAACTGG-3� (forward) and 5�-TGAGCGAGGCGGTGAGGAC-3� (reverse; accession no.49729); �-actin, 5�- CGTAAAGACCTCTATGCCAA-3� (for-ard) and 5�-AGCCATGCCAAATGTGTCAT-3� (reverse; acces-

ion no. J00691). Linearity of the PCR was found to be between 25nd 40 cycles. PCR products were electrophoresed on a 1.7%garose gel stained with ethidium bromide and densitometricallynalyzed. Levels of expression were indicated as the ratio of signalntensity for the mRNA of interest relative to that for �-actin

RNA.17

xidative Stress-Related Parameters

xidative stress was determined by measuring thiobarbituric acid-eactive substances (TBARS) such as malondialdehyde (MDA)18

r by the appearance of nitrotyrosine, a biomarker for nitrosativetress.2,5,19 For TBARS, frozen heart tissue was ground in liquiditrogen and homogenized. After acidification of the homogenateith 1% phosphoric acid and treatment with 0.6% TBA solution,

he mixture was heated in a boiling water bath for 1 h in theresence of 0.4% butylated hydroxy toluene to prevent furtherxidation of the tissue. After cooling, 1:2 adduct of MDA andBA was extracted into 4 mL of n-butanol, and the absorbanceas measured at 540 nm against 1,1,3,3-tetramethoxypropane used

s the standard. Nitrotyrosine immunohistochemistry was per-ormed on formalin-fixed sections. A rabbit polyclonal anti-itrotyrosine antibody (1:50; Cayman Chemicals, Ann Arbor, MI,SA) with Alexa Fluor 594 conjugated goat anti-rabbit immuno-lobulin G (IgG; Molecular Probes) were used to detect nitratedroteins in the rat heart. For cardiomyocyte-specific staining, theseections were cross stained for desmin and reacted with Alexaluor 350 conjugated goat anti-mouse IgG (Molecular Probes).16

ll nuclei were stained with SYTOX Green (Molecular Probes).

ardiac Necrosis

ardiac necrosis leads to a loss of cell membrane integrity with20

TA

GENERAL CHARACT

Normal control

ody weight (g) 381 � 9erum FFA (mM/L) 0.5 � 0.1erum glucose (mM/L) 8.6 � 0.2erum insulin (ng/mL) 2.8 � 0.4As in lipoprotein TG (% of total FA)Palmitic acid 20.9 � 0.9Oleic acid 26.0 � 1.3Linoleic acid 26.4 � 1.5

Significantly different from respective controls.Significantly different from normal chow.Significantly different from sunflower oil fed diabetic animals (P � 0.0Values are means � standard error (n � 4–6 animals per group). Animk before diabetes induction (55 mg/kg STZ). Diabetic animals were kep

rom fed animals.A, fatty acid; FFA, free fatty acid; STZ, streptozotocin; TG, triacylglyce

elease of lactate dehydrogenase (LDH) into the serum. Serum s

DH was estimated using an appropriate kit (Sigma, St. Louis,O, USA). Positive controls were generated by injection of iso-

roterenol (50 mg/kg, subcutaneously) into a normal chow-fed rat,nd the heart was removed after 48 h.21 Cardiac necrotic changesere also evaluated histochemically using hematoxylin and eosin

taining. Briefly, heart tissues were fixed in 10% formalin andectioned at 5 �M. Sections were deparaffinized in xylene andehydrated using various grades of ethanol. Mayer’s hematoxylinas applied for 15 min. Subsequently, the slides were washed in

unning tap water for 20 min and counterstained with eosin for 1in. Slides were dehydrated with alcohol, cleared in xylene, andounted. Cardiac necrosis was further substantiated immunohis-

ochemically using cardiac troponin T (cTnT) staining.22,23 Briefly,eart tissue was fixed in 10% formalin and sectioned at 5 �M.ections were deparaffinized in xylene and rehydrated using var-

ous grades of ethanol. Non-specific binding of immunoglobulinsas blocked by incubating the sections in 10% sheep serum

Sigma) for 20 min. A mouse monoclonal antibody against cTnTSerotec Ltd., Oxford, UK) was used as the primary antibody andections incubated overnight at 4°C. After washing the excessntibody, FITC-conjugated sheep anti-mouse IgG (Sigma; 1:50)as used as the secondary antibody. Negative controls were per-

ormed using normal mouse IgG.

erum Measurements

erum samples collected at termination were stored at �20°C untilssayed. Commercially available kits were used to measure glu-ose, TG, LDH (Sigma, St. Louis, MO, USA), and FFAs (Wakohemicals, Neuss, Germany). Serum insulin was measured usingdouble-antibody radioimmunoassay kit from Linco Research (St.ouis, MO, USA).

tatistical Analysis

alues are means � standard error. One-way analysis of varianceollowed by Bonferroni’s test or unpaired Student’s t test was usedo determine differences between group mean values. The level of

II.

TICS OF ANIMALS§

Sunflower oil Palm oil

trol Diabetic Control Diabetic

14† 393 � 11* 473 � 8† 389 � 11*0.1 1.8 � 0.8*† 0.4 � 0.0 1.1 � 0.2*†0.5 23.8 � 0.5*† 13.4 � 0.6† 23.1 � 0.3*†1.1† 1.5 � 0.6*† 5.1 � 1.8† 0.8 � 0.3*†

0.6† 12.7 � 1.2† 27.5 � 1.1 31.1 � 1.2†‡0.9 13.5 � 1.2† 45.5 � 1.0† 36.4 � 0.5†‡2.5† 40.8 � 3.3† 11.9 � 0.3† 16.8 � 0.7†‡

re maintained on normal chow or the two different high-fat diets for 4d before being killed. Values are those obtained before death and were

BLE

ERIS

Con

485 �0.4 �

11.1 �5.5 �

10.9 �20.1 �46.4 �

5).als wet for 4

rol

tatistical significance was set at P � 0.05.

R

FnIgwmsctlTfS

0�coFmaaipd1oci

ataiiwuadg

ppPSd

Ftocmecc

FdcEmfd0

Nutrition Volume 20, Number 10, 2004 919Dietary Fats and Cell Death in the STZ-Diabetic Heart

ESULTS

eeding with 20% (w/w) fat increased body weight compared withormal chow, with no difference between rats fed PO or SO (TableI). There was no difference in food intake between the two HFroups (data not shown). Although not the focus of this study,hen compared with serum insulin and glucose values from nor-al chow-fed rats, HF-fed animals were hyperinsulinemic, with

lightly elevated serum glucose (Table II). Diabetes for 4 d de-reased serum insulin and caused hyperglycemia. More impor-antly, induction of diabetes in HF-fed animals caused a substantialoss of body weight and marked elevation of serum FFA (Table II).able II (lower panel) lists the specific FAs in the lipoprotein TG

raction. PO feeding increased palmitic and oleic acids, whereas

ig. 1. Relative proportions of major free fatty acids in the heart subsequento HF feeding and diabetes. Animals were maintained on normal chow orne of the two HF diets. Cardiac free fatty acids were extracted withhloroform:methanol:acetone:hexane solvent, converted to their respectiveethyl esters, and separated by gas chromatography. *Significantly differ-

nt from respective control. †Significantly different from sunflower oil fedontrol. #Significantly different from sunflower oil fed diabetic. @Signifi-antly different from normal chow-fed control (P � 0.05). HF, high fat.

O feeding increased linoleic acid in lipoprotein TG. w

HF feeding induced an increase in total (normal control 0.2 �.03 versus palm control 1.4 � 0.1; sunflower control 1.6 � 0.1;g/mg protein; P � 0.001) and specific (Figure 1) cardiac FFAsompared with normal chow. In a pattern almost similar to thatbserved with lipoprotein TG, comparison of specific cardiacFAs between the HF groups showed that feeding PO successfullyagnified saturated fatty acid such as palmitic acid, whereas SO

ugmented cardiac �-6 polyunsaturated fatty acids (PUFAs) suchs linoleic acid (Figure 1). After diabetes, total cardiac FFA levelsn diabetic hearts from both HF groups remained unchanged com-ared with the respective controls (palm control 1.4 � 0.1; palmiabetic 1.4 � 0.1; sunflower control 1.6 � 0.1; sunflower diabetic.8 � 0.1; �g/mg protein). However, on induction of diabetes,leic and linoleic acids increased compared with their respectiveontrols, with oleic acid demonstrating the maximum amplificationn PO-fed diabetics (Figure 1).

To determine whether dietary manipulation of palmitic acidffects cardiac apoptosis during normoglycemia, but more impor-antly, during diabetes, cardiac apoptosis was evaluated in HF-fednimals by TUNEL assay by using desmin-specific counterstain todentify cardiomyocytes. In all HF groups, apoptotic cell deathncreased compared with normal chow-fed controls (Figure 2),ith maximum amplification observed in PO diabetic hearts (Fig-re 2). This augmented cellular demise occurred independent ofny further increase in total FFA or free palmitic acid in the POiabetic heart (Figure 1) and illustrates the synergistic effect oflucotoxicity in this process.

The apoptotic pathway in the heart is controlled by a number ofroteins of the Bcl-2 gene family that can promote (e.g., Bax) orrevent (e.g., Bcl-2) apoptosis.24,25 Bax mRNA expression inO-fed controls was significantly higher than in normal chow- orO-fed controls (Figure 3A). With the induction of diabetes, POiabetic hearts demonstrated the highest expression of Bax,

ig. 2. Effect of 4 d of diabetes on cardiac apoptosis in animals fed an HFiet. TUNEL assay was performed on rat heart sections, quantification ofardiac apoptosis was done by using the imaging software Northernclipse, and results are expressed TdT nuclei per 106 nuclei. Results are theeans � standard error of six rats in each group. *Significantly different

rom respective control. #Significantly different from sunflower oil-fediabetic. @Significantly different from normal chow-fed control. (P �.05). HF, high fat; TdT, deoxynucleotidyl transferase.

hereas Bcl-2 expression was highest in SO-fed diabetic rats

(pB(

rltaM(athiw

citmmi

d(ioSm

anhdddc

FapBPBdr#

hc

FaCsm#

tmSPdh

920 Ghosh et al. Nutrition Volume 20, Number 10, 2004

Figure 3B). Because a balance between pro- and anti-apoptoticroteins of the Bcl-2 family is key toward determining cell death,25

ax/Bcl-2 ratio showed a significant increase in PO diabetic ratsFigure 3C).

ROS such as superoxide (O�2) and hydroxyl radical (�OH) andeactive nitrogen species such as peroxynitrite (ONOO�) causeipid and protein modifications.26 Lipid peroxidation can be de-ected by measuring MDA. Although lipid peroxidation increasedcross all HF groups compared with normal control, levels ofDA were the highest in hearts isolated from PO-fed diabetic rats

Figure 4, top). Protein modifications can be detected by theppearance of nitrotyrosine, formed as a consequence of peroxyni-rite attack on tyrosine residues.2,5,19 Nitrotyrosine labeling was theighest in PO diabetic rat heart myocytes, with only sparse stainingn SO diabetic and PO control hearts (Figure 4A–D). Nitrotyrosineas undetectable in SO control hearts.

Myocardial cell death can propagate through apoptotic or ne-rotic pathways. Human diabetic cardiomyocytes demonstratedncreases in apoptosis and necrosis by 85-fold and 4-fold, respec-ively.2 Interestingly, in SO diabetic hearts (Figure 5B), patches ofyocytes showed a decrease in the cytoplasmic intensity of he-atoxylin and eosin staining, with loss of myocyte nuclei and

ig. 3. Bax and Bcl-2 expressions in the heart subsequent to HF feedingnd diabetes. (A) Inset: Representative Southern blot of Bax RT-PCRroducts. Bars represent the mean � standard error of the mean ofax/�-actin ratio. (B) Inset: Representative Southern blot of Bcl-2 RT-CR products. Bars represent the mean � standard error of the mean ofcl-2/�-actin ratio. (C) Ratio of Bax/Bcl-2 expression. @Significantlyifferent from normal chow fed control. *Significantly different fromespective control. †Significantly different from sunflower oil fed control.Significantly different from sunflower oil fed diabetic (P � 0.05). HF,igh fat; NC, normal control; O.D.U., optical density unit; PC, palmontrol; PD, palm diabetic; SC, sunflower control; SD, sunflower diabetic.

nflammatory cell infiltration, characteristic of myocardial necrotic w

amage.27 PO-fed diabetic hearts (Figure 5A) and all other groupsmicrographs not shown) demonstrated minimal inflammatory cellnfiltration and uniform staining of the cytoplasm. Substantiationf cardiac necrosis by measurement of serum LDH showed thatO diabetic animals had high circulating levels of this enzymearker (Figure 5, bottom).

Figure 6 demonstrates cardiac immunohistochemistry of cTnT,component of the cardiac troponin complex, that is released fromecrotic cardiomyocytes into the serum. The serum levels of cTnTave been used extensively to identify myocardial necrosis inifferent conditions such as myocardial infarction, acute myocar-itis, and doxorubicin-induced cardiomyopathy. Thus, loss of car-iomyocyte cTnT has been directly correlated to myocardial ne-rosis and focal damage.22,23 Intense and uniform cTnT staining

ig. 4. Oxidative stress-related parameters in animals fed a normal chow orHF diet for 4 wk in the presence and absence of hyperglycemia. (Top)ardiac lipid peroxidation was estimated by a thiobarbituric acid-reactive

ubstances assay that measures MDA. @Significantly different from nor-al chow-fed control. *Significantly different from respective control.

Significantly different from sunflower oil-fed diabetic (P � 0.05). (Bot-om) Nitrotyrosine (NT) localization (arrows, red fluorescence) in cardiac

yocytes (blue fluorescence). All nuclei were counterstained green. (A)O-fed controls demonstrated no staining for NT. SO-fed diabetics (B) andO-fed controls (C) demonstrated sparse localization of NT. (D) PO-fediabetics demonstrated maximum NT staining. Magnification, 400�. HF,igh fat; MDA, malondialdehyde; PO, palm oil; So, sunflower oil.

as observed in the I bands of myocytes in normal control and

daiclib

D

Icpac

epoctu3rmamIcat

Ftcodemonstrated minimal inflammatory cell infiltration and uniform staining of the cytoplasm. HF, high fat; ISO, isoproterenol; LDH, lactate dehydrogenase.

Nutrition Volume 20, Number 10, 2004 921Dietary Fats and Cell Death in the STZ-Diabetic Heart

iabetic HF controls (data not shown) and PO diabetic hearts (solidrrows, Figure 6A, confocal image; and Figure 6B, brightfieldmage). Interestingly, only SO diabetic groups demonstrated de-reases in immunostaining, with non-discernible I bands at someocations (open arrows, Figure 6C, confocal image). These changesn fluorescence correlated with focal disruption of myofibrils in therightfield image (open arrows, Figure 6D).

ISCUSSION

n humans or in experimental STZ-induced diabetes in rats or mice,ardiomyocyte apoptosis has been directly linked to myocardialathology.2,5,6 Although glucotoxicity may be a key regulator ofpoptosis in the diabetic heart, the importance of FA, more spe-

ifically palmitic acid, to this apoptotic process recently has been v

mphasized in vitro.12 Although this setting partly resemblesalmitic acid derived from plasma, it does not account for FAsbtained from other sources such as hydrolysis of lipoproteins byardiac lipoprotein lipase (LPL).28 This is especially relevant givenhat 1) LPL-derived FA is the principal source of FA for cardiactilization; 2) activity of cardiac LPL increases after diabetes,8 and) incubation of circulating lipoproteins with LPL predominantlyeleases palmitic acid (47.5% of total FAs released).29 Thus, a trueeasure of the effects of palmitic acid on apoptosis should include

n examination of its in vivo effects. Feeding PO successfullyagnified palmitic acid content within lipoproteins and the heart.

n these animals, a fraction (0.02% to 0.03%) of the cardiomyo-ytes underwent apoptosis, in marked contrast to results obtainedfter in vitro incubations of cardiomyocytes with palmitic acid (upo 40% of the total cell population perished).9 It is possible that in

24

ig. 5. (Bottom) Quantitative estimation of serum LDH. Serum was collected at death, and LDH activity was measured with an appropriate kit. ISO injectedo a normal chow-fed rat was used to generate a positive control. #Significantly different from sunflower oil-fed diabetic. †Significantly different from normalhow-fed control (P � 0.05). (Top) Micrographs of myocytes showing a decrease in the cytoplasmic intensity of hematoxylin and eosin staining in sunfloweril diabetic hearts (B), with loss of myocyte nuclei and inflammatory cell infiltration, characteristic of myocardial necrotic damage. (A) Palm oil diabetic hearts

ivo phagocytes rapidly clear apoptotic cells (within 24 h); thus,

duaa

efirdimoipssot

hSRin

oftcigoMh

tcaePofadZhIra

FTflo

922 Ghosh et al. Nutrition Volume 20, Number 10, 2004

etermination of apoptosis in the intact heart could easily benderestimated. Alternatively, in the intact heart, protective mech-nisms could be triggered that dampen the detrimental effects ofugmented palmitic acid.

In this regard, cardiac stearoyl coenzyme A desaturase, annzyme that synthesizes monounsaturated FAs such as oleic acidrom saturated FAs, is increased after diabetes.30 In the presence ofncreased saturated FAs, an increased stearoyl coenzyme desatu-ase activity could explain the high oleic acid content in POiabetic hearts. Interestingly, oleic acid has been shown to beneffective10 or decreases12 cardiac apoptosis when incubated si-ultaneously with palmitic acid. Moreover, in Chinese hamster

vary cells, oleic acid increases the incorporation of palmitic acidnto TG, thereby preventing an accumulation of excess cytotoxicalmitic acid in the free form.31 More recently, it was demon-trated that long-term feeding of Wistar rats with canola oil, a richource of oleic acid, prevents cardiomyocyte cell death.32 Whetherleic acid performs a similar function in PO diabetic hearts has yeto be determined.

Albeit low, PO-fed diabetic hearts demonstrated significantlyigher levels of apoptosis compared with isocaloric controls fedO. Palmitic acid, by its conversion to ceramide, can generateOS such as peroxynitrite (ONOO�).11,33 Peroxynitrite in turn can

nduce nitration of tyrosine residues, leading to the formation of19

ig. 6. Cardiac necrosis illustrated by cardiac troponin T immunostaininghe troponin T band is regularly distributed in the I bands of palm oil duorescence, focal disruption, and loss of I bands (open arrow, C and D),il diabetic hearts.

itrotyrosine, MDA, DNA strand breaks, and apoptosis. Levels p

f MDA and nitrotyrosine-positive cells were the highest in PO-ed diabetic rat hearts and may explain the augmented apoptosis inhis group. Interestingly, SO diabetic animals did not demonstrateomparable lipid peroxidation as PO diabetic rats, despite anncrease in the highly oxidizable linoleic acid. Bcl-2 has been sug-ested to play a significant protective role in attenuating cardiacxidative stress by blocking lipid peroxidation and appearance ofDA.34 Because SO feeding increased Bcl-2 expression, it could

ave conferred protection against exaggerated lipid peroxidation.Another role for the Bcl-2 family of proteins is maintenance of

he balance between cell death and survival by regulating cyto-hrome C release from the mitochondria, caspase activation, andpoptosis.24,25 For instance, increased expression of Bax throughxcess FFAs or oxidative stress, conditions that are prevalent inO diabetic hearts, promotes apoptosis by forming death-inducingligomers upon apoptotic stimuli.35 Other members of the sameamily such as Bcl-2 can heterodimerize with Bax and preventpoptosis. Thus, estimation of Bax/Bcl-2 ratio has been used toetect pro- or anti-apoptotic environments. In insulin-resistantucker diabetic fatty rats, an augmented Bax/Bcl-2 mRNA ratioas been linked to an increased lipoapoptosis in pancreatic islets.17

n this study, PO diabetic hearts showed an increased Bax/Bcl-2atio that could contribute to the exaggerated apoptosis in thesenimals. At least in SO diabetic animals, expression of BCl-2 may

focal images (A and C) matched with brightfield microscopy (B and D).c heart myocytes (solid arrows, A and B). Reduction in the intensity oftive of severe myocardial damage and necrosis, is observed in sunflower

in coniabetiindica

rotect against apoptosis but not necessarily necrotic cell

dSts

adpebmaironmp

adeoihsoi

A

TtRT

R

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

3

3

3

3

3

3

3

3

3

3

4

4

4

Nutrition Volume 20, Number 10, 2004 923Dietary Fats and Cell Death in the STZ-Diabetic Heart

eath.36,37 It should be noted that, although the Bax/Bcl-2 ratio inO groups were unchanged compared with normal chow-fed con-

rols, the increase in apoptosis in SO groups could be ascribedimply to cytotoxic effects of augmented intracellular FAs.38

Necrosis, an alternate mode of cell death, is energy independentnd is characterized by membrane leakage, cell swelling, andissolution of cellular structures.24 Unlike apoptosis, the partici-ation of different FAs toward cardiac necrosis has not beenntirely elucidated. In the heart, linoleic acid can decrease mem-rane stability and/or produce a proinflammatory response, thatakes cells more prone to necrotic damage.39 In humans, serum

nd adipose tissue linoleic acid have been strongly correlated to anncreased risk of myocardial infarction and necrosis.40 A similarelation between cardiac linoleate and necrosis may be proposed inur SO diabetic rats. However, given the recent assertion thatecrosis may not simply be an “accidental” form of cell death,41

olecular mechanisms other than linoleic acid could also induce aronecrotic environment.

In conclusion, a PO diet in conjunction with hyperglycemiaugmented apoptosis compared with SO feeding through an oxi-ative stress-dependent pathway and overexpression of Bax. How-ver, the rate of this phenomenon was much lower than thatbserved in vitro. Interestingly, use of a PUFA-rich diet led tonduction of rapid myocardial necrotic changes in the diabeticeart. Given the current dietary recommendation of substitutingaturated fats with “heart-friendly” vegetable oils such as sunfloweril,42 our results question the beneficial role of an -6 PUFA-rich dietn the cardiovascular system, especially after diabetes.

CKNOWLEDGMENTS

he technical help of Mr. Roger Dyer and the financial support ofhe Health Research Foundation/Canadian Institutes of Healthesearch for Graduate Research Scholarships to Sanjoy Ghosh andhomas Pulinilkunnil are gratefully acknowledged.

EFERENCES

1. Rodrigues B, Cam MC, McNeill JH. Myocardial substrate metabolism: implica-tions for diabetic cardiomyopathy. J Mol Cell Cardiol 1995;27:169

2. Frustaci A, Kajstura J, Chimenti C, et al. Myocardial cell death in humandiabetes. CircRes 2000;87:1123

3. Fiordaliso F, Li B, Latini R, et al. Myocyte death in streptozotocin-induceddiabetes in rats in angiotensin II–dependent. Lab Invest 2000;80:513

4. Shizukuda Y, Reyland ME, Buttrick PM. Protein kinase C-delta modulatesapoptosis induced by hyperglycemia in adult ventricular myocytes. Am J PhysiolHeart Circ Physiol 2002;282:H1625

5. Kajstura J, Fiordaliso F, Andreoli AM, et al. IGF-1 overexpression inhibits thedevelopment of diabetic cardiomyopathy and angiotensin II–mediated oxidativestress. Diabetes 2001;50:1414

6. Cai L, Li W, Wang G, et al. Hyperglycemia-induced apoptosis in mouse myo-cardium: mitochondrial cytochrome C–mediated caspase-3 activation pathway.Diabetes 2002;51:1938

7. Fiordaliso F, Leri A, Cesselli D, et al. Hyperglycemia activates p53 and p53-regulated genes leading to myocyte cell death. Diabetes 2001;50:2363

8. Rodrigues B, Cam MC, Jian K, et al. Differential effects of streptozotocin-induced diabetes on cardiac lipoprotein lipase activity. Diabetes 1997;46:1346

9. Rabkin SW, Huber M, Krystal G. Modulation of palmitate-induced cardiomyo-cyte cell death by interventions that alter intracellular calcium. ProstaglandinsLeukot Essent Fatty Acids 1999;61:195

0. Sparagna GC, Hickson-Bick DL, Buja LM, et al. A metabolic role for mitochon-dria in palmitate-induced cardiac myocyte apoptosis. Am J Physiol Heart CircPhysiol 2000;279:H2124

1. Listenberger LL, Schaffer JE. Mechanisms of lipoapoptosis implications forhuman heart disease. Trends Cardiovasc Med 2002;12:134

2. Dyntar D, Eppenberger-Eberhardt M, Maedler K, et al. Glucose and palmitic acidinduce degeneration of myofibrils and modulate apoptosis in rat adult cardiomy-ocytes. Diabetes 2001;50:2105

3. Portillo MP, Tueros AI, Perona JS, et al. Modifications induced by dietary lipidsource in adipose tissue phospholipid fatty acids and their consequences in lipid

mobilization. Br J Nutr 1999;82:319

4. Chapman MJ, Goldstein S, Lagrange D, et al. A density gradient ultracentrifugalprocedure for the isolation of the major lipoprotein classes from human serum. JLipidRes 1981;22:339

5. Innis SM, Dyer RA. Brain astrocyte synthesis of docosahexaenoic acid from n-3fatty acids is limited at the elongation of docosapentaenoic acid. J Lipid Res2002;43:1529

6. Scarabelli TM, Knight RA, Rayment NB, et al. Quantitative assessment ofcardiac myocyte apoptosis in tissue sections using the fluorescence-based tuneltechnique enhanced with counterstains. J Immunol Methods 1999;228:23

7. Shimabukuro M, Wang MY, Zhou YT, et al. Protection against lipoapoptosis ofbeta cells through leptin-dependent maintenance of Bcl-2 expression. Proc NatlAcad Sci USA 1998;95:9558.

8. Mihara M, Uchiyama M. Determination of malonaldehyde precursor in tissues bythiobarbituric acid test. Anal Biochem 1978;86:271

9. Virag L, Szabo E, Gergely P, et al. Peroxynitrite-induced cytotoxicity: mecha-nism and opportunities for intervention. Toxicol Lett 2003;113:140–141

0. Roe CR. Diagnosis of myocardial infarction by serum isoenzyme analysis. Ann.Clin Lab Sci 1977;7:201

1. Sumitra M, Manikandan P, Kumar DA, et al. Experimental myocardial necrosisin rats: role of arjunolic acid on platelet aggregation, coagulation and antioxidantstatus. Mol Cell Biochem 2001;224:135

2. Herman EH, Lipshultz SE, Rifai N, et al. Use of cardiac troponin T levels as anindicator of doxorubicin-induced cardiotoxicity. Cancer Res 1998;58:195

3. Herman EH, Zhang J, Lipshultz SE, et al. Correlation between serum levels ofcardiactroponin-T and the severity of the chronic cardiomyopathy induced bydoxorubicin. J Clin Oncol 1999;17:2237

4. Haunstetter A, Izumo S. Apoptosis: basic mechanisms and implications forcardiovascular disease. Circ Res 1998;82:1111

5. Clerk A, Cole SM, Cullingford TE, et al. Regulation of cardiac myocyte celldeath. Pharmacol Ther 2003;97:223

6. Cai L, Kang YJ. Oxidative stress and diabetic cardiomyopathy: a brief review.Cardiovasc Toxicol 2001;1:181

7. Chow LH, Boughner DR, Dunmore-Buyze J, et al. Enhanced detection ofcardiacmyocyte damage by polarized light microscopy. Use in a model ofcoxsackievirus B3-induced myocarditis. Cardiovasc Pathol 2001;10:83

8. Augustus AS, Kako Y, Yagyu H, et al. Routes of FA delivery to cardiac muscle:modulation of lipoprotein lipolysis alters uptake of TG-derived FA. Am JPhysiolEndocrinol Metab 2003;284:E331

9. Sato K, Takahashi T, Takahashi Y, et al. Preparation of chylomicrons and VLDLwith monoacid-rich triacylglycerol and characterization of kinetic parameters inlipoproteinlipase-mediated hydrolysis in chickens. J Nutr 1999;129:126

0. Miyazaki M, Ntambi JM. Role of stearoyl-coenzyme A desaturase in lipidmetabolism. Prostaglandins Leukot Essent Fatty Acids 2003;68:113

1. Listenberger LL, Han X, Lewis SE, et al. Triglyceride accumulation protectsagainst fatty acid-induced lipotoxicity. Proc Natl Acad Sci USA 2003;100:3077

2. Aguila MB, Mandarim-de-Lacerda CA. Heart and blood pressure adaptations inWistar rats fed with different high-fat diets for 18 months. Nutrition 2003;19:347

3. Zhang DX, Zou AP, Li PL. Ceramide-induced activation of NADPH oxidase andendothelial dysfunction in small coronary arteries. Am J Physiol Heart CircPhysiol 2003;284:H605

4. Hattori R, Hernandez TE, Zhu L, et al. An essential role of the antioxidant geneBcl-2 in myocardial adaptation to ischemia: an insight with antisense Bcl-2therapy. Antioxid Redox Signal 2001;3:403

5. Piro S, Anello M, Di Pietro C, et al. Chronic exposure to free fatty acids or highglucose induces apoptosis in rat pancreatic islets: possible role of oxidative stress.Metabolism 2002;51:1340

6. Virag L, Szabo C. BCL-2 protects peroxynitrite-treated thymocytes frompoly(ADP- ribose) synthase (PARS)-independent apoptotic but not from PARS-mediated necrotic cell death. Free Radic Biol Med 2000;29:704

7. Shinoura N, Yoshida Y, Asai A, et al. Relative level of expression of Bax andBcl-XL determines the cellular fate of apoptosis/necrosis induced by the over-expression of Bax. Oncogene 1999;18:5703

8. Suliman A, Lam A, Datta R, et al. Intracellular mechanisms of TRAIL: apoptosisthrough mitochondrial-dependent and -independent pathways. Oncogene 2001;20:2122.

9. Toborek M, Lee YW, Garrido R, et al. Unsaturated fatty acids selectively inducean inflammatory environment in human endothelial cells. Am J Clin Nutr 2002;75:119

0. Valek J, Hammer J, Kohout M, et al. Serum linoleic acid and cardiovasculardeath in postinfarction middle-aged men. Atherosclerosis 1985;54:111

1. Proskuryakov SY, Konoplyannikov AG, Gabai VL. Necrosis: a specific form ofprogrammed cell death? Exp Cell Res 2003;283:1

2. Simopoulos AP. Essential fatty acids in health and chronic disease. Am J Clin

Nutr 1999;70:560S