All experiments followed the European Union recommendations concerning the care and use of laboratory animals for experimental and scientific purposes. All animal work was approved by the local board of ethics for animal experimentation (Cometh) and notified to the research animal facility of our laboratory (authorization n° 38 07 23). The performed research was in compliance with the ARRIVE guidelines on animal research 23 .

Ten ZDF and eleven Zucker lean (ZL) male rats were obtained from Charles Rivers (L’Arbresle, France) at 7 weeks (wk) of age. Rats were housed in a temperature- and humidity-controlled facility on a 12-h light:dark cycle. The two groups were fed ad libitum with a standard carbohydrate diet (A04, Safe, Augy, France), they had free access to water and their body weight and food intake were recorded twice weekly. The composition of the chosen diet by weight is 60% assimilable glucides (52% mainly starch and cellulose), 16% proteins and 3% fat. After analysis of the fatty acid composition of our diet we found a formula with approximately 24% of saturated fatty acids (SFAs), 23% of monounsaturated fatty acids (MUFAs), 48% of n-6 polyunsaturated fatty acids (PUFAs) and 4.5% of n-3 PUFAs. Plasma glucose and glycated hemoglobin (HbA1c) concentrations were also measured weekly via the tail vein.

On the day of the experiment, the rats were weighed and heparinized (1500 I.U./kg) intraperitoneally 30 minutes (min) before their decapitation. Blood samples were collected for further biochemical analysis and their retroperitoneal and mesenteric adipose tissues were weighed for determination of the abdominal fat mass 24 . Perfusion experiments were performed twice a day by alternating the rat phenotypes. The first experiment was performed between 8:00 a.m. and 8:30 a.m. and the second one between 12:30 p.m. and 1:00 p.m.

All rats underwent ex vivo Langendorff assessment of their cardiac function. For this reason, a rapid thoracotomy was performed and the heart was immediately collected in Krebs-Henseleit solution maintained at 4°C. It was then rapidly (less than one minute to avoid problems of cellular damages or preconditioning) perfused at constant pressure according to the Langendorff mode with a Krebs–Henseleit buffer containing (in mM) NaCl 119, MgSO₄ 1.2, KCl 4.8, NaHCO₃ 25, KH₂PO₄ 1.2, CaCl₂ 1.2 and glucose 11 mM as sole energy substrate. The perfusion buffer used was similar for both groups in order to minimize the variables studied. The buffer was maintained at 37°C and continuously oxygenated with carbogen (95% O₂/5% CO₂). A latex balloon connected to a pressure probe was inserted into the left ventricle and filled until the diastolic pressure reached a value of 7–8 mmHg. This allowed the monitoring of heart rate, systolic, diastolic and left ventricle developed pressures throughout the perfusion protocol. A pressure gauge inserted into the perfusion circuit just upstream the aortic cannula allowed the evaluation of the coronary pressure. The heart was perfused at constant pressure of 59 mmHg 25 for 30 minutes and the coronary flow for each heart was evaluated by weight determination of 1-min collected samples at the 25th min of perfusion. After this period, the heart was perfused at constant flow conditions, for which the flow rate was adjusted in order to obtain the same coronary flow as in the preparation at constant pressure. The systolic, diastolic and left ventricle developed pressures as well as the heart rate was determined after 10 min of perfusion at forced flow in order to allow a satisfying stabilization of the heart. The left ventricle developed pressure was calculated by subtracting the diastolic pressure to the systolic pressure. The rate-pressure product (RPP) was defined as the product of left ventricle developed pressure and heart rate and was used as indicator of the cardiac mechanical work 26 . All the parameters were recorded and analyzed with a computer using the HSE IsoHeart software (Hugo Sachs Elektronik, March-Hugstetten, Germany).

After the evaluation of the cardiac function at constant flow, we assessed the effects of diabetes on the coronary reactivity. After the 10-min equilibration period at constant flow, the coronary tone was raised by using the thromboxane analog U46619 (30nM), which was constantly infused into the perfusion system near the aortic cannula at a rate never exceeding 1.5% of the coronary flow. This allowed the obtainment of a coronary pressure between 120 and 130 mmHg. In our model of perfusion at forced flow, the aortic pressure equaled the coronary pressure and changes in the coronary tone triggered modifications of the aortic pressure. Changes in aortic perfusion pressure were thus used to monitor changes in coronary tone. Furthermore, this experimental model allows the evaluation of the coronary microvasculature reactivity since the coronary resistance vessels determine the overall coronary pressure. Relaxation responses to Ach (4, 10, 20, 40, 60, 80 and 100 pmoles) and sodium nitroprusside (SNP, 100, 200, 400, 600, 800 and 1000 pmoles) injections were determined reflecting the endothelial-dependent vasodilatation (EDD) and endothelium-independent vasodilatation (EID) respectively.

The dilatation amplitude was calculated as the ratio between the maximal decrease in the coronary pressure and the coronary pressure just before the injection of the dilatation agents. Since the heart weight and coronary volume were subjected to intra- and inter-group variations, a correction was performed to normalise the input-function of the vasodilatation agents according to the coronary flow. The dose–response curve between the amount of vasodilatation agent injected and the maximal vasodilatation was then fitted to a logarithm function for each heart which allowed the fulfillment of statistical analyses. Moreover, the vasodilatation activity of the endothelial cells was also estimated from the corrected EDD and EID curves. For each heart and each injected Ach dose, the amount of SNP (reflecting the amount of vasodilator agents) necessary to obtain the same percentage of Ach-induced vasodilatation was extracted from the EID curve according to the formula: endothelial cell vasodilatation activity (ECVA) = e ^{[(% Ach-induced dilatation - b)/a]}, where a and b are the coefficients of the theoretical EID curve. The results were expressed in pmole equivalents of nitroprusside. At the end of the perfusion protocol, the hearts were freeze-clamped and stored at -80°C until the biochemical analyses were performed.

Lactate and pyruvate released in the coronary efflu-ents were spectrophotometrically assayed according to Bergmeyer 29 . The lactate to pyruvate ratio was calculated to estimate the cytosolic redox potential (NADH/NAD⁺) 30 31 32 . This is a highly specific assay using the enzyme lactate dehydrogenase (LDH) to catalyze the reversible reaction of pyruvate and NADH to lactate and NAD⁺. The catalytic action of LDH permits spectrophotometric measurement at 340 nm (spectrophotometer ULTROSPEC^TM 2100 pro, Amersham Biosciences, Uppsala, Sweden) of lactate production in terms of the generation of NADH in the reaction shown above. To measure lactate, the reaction is carried out from right to left with excess NAD⁺. To force the reaction to completion in this direction, it is necessary to trap formed pyruvate with hydrazine. The increased absorbance at 340 nm due to NADH formation becomes a mole-to-mole measure of the lactate originally present in the sample.

Lipid peroxidation was assessed by measuring the concentration of thiobarbituric acid reactive substances (TBARS) in cardiac homogenates 33 . TBARS were determined using the fluorimetric determination of malondialdehyde – thiobarbituric acid complex after acid hydrolysis at 95°C and extraction with n-butanol. Briefly, tissue homogenate aliquots were placed in polyethylene tubes mixed with TBA (thiobarbituric acid)/perchloric acid 7% (2:1, v/v) and incubated at 95°C. After cooling, n-butanol was added for the extraction and then the aliquots were centrifuged for 10 min at 35000 g. The supernatant was used to read the fluorescence at excitation and emission wavelengths of 532 and 553 nm respectively. The TBARS calculated were normalized to the content of polyunsaturated fatty acids of cardiac membrane phospholipids since it differed between groups. The protein thiol groups and FRAP assays were also evaluated in cardiac homogenates as described previously for the plasma.

The ratio between the activities of aconitase and fumarase of the myocardium was calculated as an indicator of mitochondrial ROS production. Mitochondrial aconitase is sensitive to inactivation by superoxide due to the susceptibility of its iron-sulfur core to oxidation; however, fumarase is unaffected. Thus, the activity ratio of aconitase to fumarase was calculated as an indicator of the presence of mitochondrial ROS 34 . Aconitase and fumarase activities were determined according to Gardner et al. 34 , but were measured after extraction with a medium supplemented with citrate sodium (1 M) in order to stabilize the aconitase activity ex vivo. Values of aconitase and fumarase activities were determined on the same extract for each biological sample.

Activities of the NADH-ubiquinone oxido-reductase (complex I), succinate-ubiquinone oxido-reductase (complex II), ubiquinol cytochrome c reductase (complex III), cytochrome c oxidase (complex IV), NADH cytochrome c reductase (activity of complex I + III) and succinate cytochrome c reductase (activity of complex II + III) were determined as previously described 35 . Heart samples (100 mg) were homogenized at 4°C with 0.9 ml of a potassium phosphate buffer 100 mM, pH 7.4. The homogenates were centrifuged (1,500 × g, 5 min, 4°C), and the resulting supernatants were stored at -80°C until the determination of the various enzymatic activities. Activity of the citrate synthase was determined according to Faloona and Srere 36 . The activities of the respiratory chain complexes and citrate synthase were expressed in units per mg of proteins.

The expressions of total eNOS, phosphorylated eNOS at Ser1177 and iNOS were evaluated by Western blot. Frozen samples were homogenized in ice-cold lysis buffer containing 20 mM Tris (pH 7.8), 137 mM NaCl, 2.7 mM KCl, 1 mM MgCl2, 1% Triton X-100, 10% (w/v) glycerol, 10 mM NaF, 1 mM ethylenediaminetetraacetic acid, 5 mM Na pyrophosphate, 0.5 mM Na₃VO₄, 1 μg/ml leupeptin, 0.2 mM phenylmethylsulfonyl fluoride and 1 mM benzamidine. The homogenates were centrifuged at 5,000 g for 20 min at 4°C, and the protein concentration in the supernatant was determined in each aliquot. Protein extracts (50 μg/lane) were loaded onto a 10% SDS gel and separated by electrophoresis. Extracts from the control group were loaded on both gels, and the amount of protein was accordingly compared pairwise. Proteins were transferred to nitrocellulose membranes. The membranes were incubated overnight at 4°C with rabbit antibodies against total eNOS (1:150, Thermoscientific, Illkirch, France), iNOS (1:1,000, BD Biosciences Pharmingen, Le Pont de Claix, France) and phosphospecific mouse antibodies against eNOS Ser1177 (1:1,000, BD Biosciences Pharmingen, Le Pont de Claix, France). After being washed in tris buffered saline (TBS)-Tween, the membranes were incubated with horseradish peroxidase-conjugated anti-mouse IgG for eNOS Ser1177 (1:3000, Jackson Immunoresearch, Montluçon, France) and anti-rabbit IgG for total eNOS and iNOS (1:20000, Jackson Immunoresearch, Montluçon, France) for 1 h at room temperature, followed by additional washing. Proteins were visualized by enhanced chemiluminescence with ECL advanced Western blotting detection kit (Amersham Biosciences, Brumath, France) and quantified using densitometry and Image J software. PAN-Actin (1:1000, Cell Signaling Technology, St-Quentin-en-Yvelines, France) was used as a loading control.

Total RNA from rat hearts was isolated by using Tri reagent according to instruction recommended by the manufacturer. Briefly, 50 mg of tissue was homogenized in 1 ml of Tri reagent (Sigma Chemical Co, Via Coger, Paris, France) and the aqueous phase was collected after chloroform addition. RNA was precipitated with isopropanol and washed with 75% ethanol. The RNA pellet was dried and redissolved in diethylpyrocarbonate water. The concentration and purity of RNA were determined by measurement of absorbance at 260:280 nm. RNA samples (2.5 μg of total RNA) were analyzed by 1% agarose electrophoresis for control of RNA degradation.

1 μg of total RNA of each sample was primed with oligo(dT) and reverse transcribed with SuperscriptIII reverse transcriptase (RT) (Life Technologies SAS, Saint Aubin, France) to produce cDNA. PCR amplification of the cDNA from the reverse transcription reaction was carried out using specific primer pairs for Nos2 (nitric oxide synthase 2, inducible; GenBank accession number: NM_012611) and the house-keeping gene Arbp (acidic ribosomal phosphoprotein; GenBank accession number: NM_022402.2). Sequences of the primers for analysis of mRNA were for Nos2 (forward): CAG GTT GAG GAT TAC TTC TTC CA; Nos2 (reverse): TGT CAG AGT CTT GTG CCT TTG with PCR product length of 132 bp and for Arbp (forward): CCT GCA CAC TCG CTT CCT A; Arbp (reverse): TGA TGG AGT GAG GCA CTG AG with PCR product length of 95 bp (Eurogentec France SASU, Angers, France). These primers were intron-spanning in order to avoid genomic DNA contamination. Quantitative real-time PCR (qRT-PCR) was then performed with a LightCycler FastStart DNA Master SYBR Green I kit (Roche, Diagnostics, Meylan, France) on a LightCycler 1.5 Instrument (Roche, Diagnostics, Meylan, France) in capillaries of 20 μl volume/capillary by adding 4 μl of Master Mix solution, 0.5 μM of Nos2 or 0.4 μM Arbp primers, 5 μl of the sample and completing up to 20 μl with RNAse free water. The assays were performed in duplicates. The capillaries were then appropriately sealed, centrifuged for few seconds at high speed and then placed into the LightCycler. The thermal cycle conditions were 95°C for 10 min (pre-incubation) followed by 45 cycles of amplification that were run at 95°C for 10 s, at 55°C for 10 s and at 72°C for 10 s for Nos2 and at 95°C for 10 s, at 55°C for 5 s and at 72°C for 6 s for Arbp. Cycle threshold values (Ct) were analyzed and the level of expression of Nos2 gene was standardized against that of Arbp gene detected in the same sample by using the 2^-ΔΔCt method 37 .

The phospholipid fatty acid composition was determined in cardiac homogenates as previously described 38 . The lipids were extracted according to Folch et al. 39 . The phospholipids were separated from non-phosphorus lipids using a Sep-pack cartridge 40 . After transmethylation, the fatty acid methyl esters were separated and analyzed by gas chromatography.

Blood glucose concentrations were determined with a glucose analyzer (ACCU-CHECK Active, Softclix). Plasma insulin concentrations were determined using a radioimmunoassay kit (ICN Pharmaceuticals, Orangeburg, SC). Plasma triglyceride and cholesterol levels were measured using commercially available kits from Biomérieux (Craponne, France) and Roche (Boulogne-Billancourt, France), respectively. HbA1c levels were evaluated with the kit Bayer Healthcare’s analyser A1cNow® determined in blood samples (5 μl) drawn from the rat fingers. Proteins were measured using the bicinchoninic acid method with a commercially available kit (Thermo Scientific, Rockford, IL).

Results are presented as mean ± S.E.M. Animal weight, heart dry weight, glycemia, activity of respiratory chain complexes, aconitase-to-fumarase and lactate-to-pyruvate ratios and data describing the oxidative stress and the cardiac mechanical and vascular function (developed pressure, heart rate, rate pressure product, coronary pressure, and coronary flow) were contrasted across the two groups by one-way analysis of variance (ANOVA). Measures related to the action of the vasodilatation agents were treated with repeated-measures ANOVA to test the effect of the diabetes of ZDF rats (external factor), that of the amount of dilatation agent (internal factor) and their interaction. When required, group means were contrasted with a Fisher’s LSD test. A probability (p) less than 0.05 was considered significant. Statistical analysis was performed using the NCSS 2004 software.

As shown in Figure 1, the food intake was always higher in the ZDF compared to the ZL group (+85% at the 9th wk of age), which provoked their increased body weight (+21% at the day of the sacrifice). Consequently, based on the analysis of the diet ingredients, the ZDF group consumed greater amounts of fat than the lean group. The abdominal fat mass was partly responsible for the increased body weight since the mesenteric, retroperitoneal and visceral adipose tissues were significantly heavier as shown in Table 1. However, the heart weight of the ZDF rats did not differ of that of the ZL animals (Table 1).

The number of experiments was 11 and 10 for the ZL and ZDF groups respectively. The weight of mesenteric, retroperitoneal, visceral and abdominal adipose tissues is expressed in g of wet weight. The abdominal adipose tissue weight normalized to the body weight is expressed in g of wet weight per g of body weight. The heart weight is expressed in mg of dry weight. The heart-to-body weight ratio is expressed in mg of dry weight per g of body weight. ZL: Zucker lean rats; ZDF: Zucker diabetic fatty rats; AT: adipose tissue; BW: body weight; *: significantly different.

Figure 2 shows that the blood glucose concentration of the ZDF rats was increased as soon as the 7th wk of age and reached a value close to 5 g/l at the 9th wk (+276% compared to the ZL rats). The hyperglycemia triggered an increase in HbA1c already significant at the 9th wk (+56%) and a huge augmentation of the insulinemia (+244% at the moment of the sacrifice). The plasma triglyceride and cholesterol concentrations were also significantly higher for the ZDF rats at the moment of the sacrifice.

The mitochondrial-derived oxidative stress was estimated in cardiac homogenates by the aconitase-to-fumarase ratio. As shown in Figure 3A, the ratio was significantly reduced in the ZDF group (-41%), indicating an increase in the cardiac mitochondrial oxidative stress. The lactate-to-pyruvate ratio in the coronary effluents (Figure 3B), reflecting the cytosolic redox potential, was also decreased in the ZDF group (-34%). This decrease in the cytosolic redox potential may indicate a diminished capacity of the system to buffer ROS and thus an increased presence of oxidizing species as its primary outcome 30 31 32 . In the heart, the global antioxidant power or the protein thiol groups were not modified by the diabetic state of the ZDF rats (Figure 4A). However, there was a significant increase of 105% in the lipid peroxidation as shown by the TBARS concentration normalized to the PUFAs content of the cardiac membrane phospholipids (Figure 4B).

In the plasma, even though the antioxidant enzyme GPx was significantly increased in the ZDF group (+21.5%, Figure 3C), the global antioxidant power as estimated by the FRAP assay was significantly decreased (-15,6%, Figure 3D). This was accompanied by a decrease in the plasma thiol groups (Figure 3E), which however did not reach significance.

Citrate synthase was significantly increased (+4.5%) in the ZDF group as shown by the values for the ZL versus the ZDF rats in Table 2. When normalized to the amount of myocardial proteins, the activity of the cytochrome oxidase was increased in the ZDF group (+20%). No modifications concerning the activities of the other respiratory chain complexes were observed.

The number of experiments was 11 and 10 for the ZL and ZDF groups respectively. ZL: Zucker lean rats; ZDF: Zucker Diabetic fatty rats; CI: NADH:ubiquinone oxidoreductase; CII: succinate-ubiquinone oxido-reductase; CIII: ubiquinol-cytochrome-c reductase; CIV: cytochrome c oxidase; CI + III: NADH cytochrome c reductase; CII + III: succinate cytochrome c reductase; CS: citrate synthase. The results are expressed in mU/mg of proteins. *: significantly different.

The fatty acid composition of cardiac membrane phospholipids was modulated by the development of diabetes (Table 3). The SFAs were significantly increased in the ZDF group (+24%), especially the 18:0 (+33%). This increase partly occurred at the detriment of the MUFAs. Indeed, all the MUFAs were reduced (-36, -27 and -42% for the 16:1n-7, 18:1n-9 and 18:1n-7, respectively). The n-6 PUFAs were also reduced, not only in their totality (-18%) but also regarding the 18:2n-6 (-52%). However, the 20:3 n-6 and 20:4 n-6 were significantly increased (+152 and +47%, respectively). The important reduction of the n-6 PUFAs was accompanied by an increase in n-3 PUFAs (+98%). This was particularly true for the 22:5 n-3 and 22:6 n-3 levels (+130 and + 97%, respectively). Finally, the n-6 to n-3 PUFA ratio of cardiac phospholipids was reduced by the occurrence of diabetes (-60%).

Values are expressed as relative amounts of the total fatty acid content. The analysis was performed on 5 samples randomly selected in each group. ZL: Zucker lean rats; ZDF: Zucker diabetic fatty rats; DMA: dimethylacetal; SFA: saturated fatty acids; MUFA: monounsaturated fatty acids; PUFA: polyunsaturated fatty acids; EPA: eicosapentanoic acid; AA: arachidonic acid; DHA: docosahexaenoic acid; *: significantly different.

The results of the ex vivo cardiac function are shown in Table 4. The measured parameters were recorded when the heart was perfused at constant flow before the infusion of U46619. In the ZDF group, the RPP was reduced (-35%) compared to the ZL group. This was due to a reduction of the heart rate (-42%), since the left ventricle developed pressure was slightly increased (+18.2%). The changes in the RPP were consequently related to the observed decrease in coronary flow (-25%), but the coronary pressure was unaffected. The infusion of U46619 raised the coronary pressure from 80 mmHg to a value close to 125 mmHg in both groups.

The number of experiments was 11 and 10 for the ZL and ZDF groups respectively. ZL: Zucker lean rats; ZDF: Zucker Diabetic fatty rats; HR: Heart Rate, LVDP: Left Ventricle Developed Pressure, RPP: Rate x Pressure Product, CF: Coronary Flow, CP: Coronary Pressure. * significantly different.

Figure 5A depicts an EDD, which was similar in the two groups, reaching 15% of dilatation as soon as 40 pmoles of Ach were injected. The EID was significantly increased in the ZDF group (Figure 5B) as soon as the SNP dose of 600 pmoles was injected (+14, +16 and +18% at the doses of 600, 800 and 1000 pmoles of SNP, respectively). Finally, the ECVA was not modified by the occurrence of diabetes (Figure 5C).

The eNOS expression and its phosphorylation at Ser1177 were evaluated in aortic and cardiac homogenates of both groups. No difference was observed between groups in the expression and phosphorylation of the enzyme neither in the aorta (Figures 6A and B) nor in the heart (Figures 6C and D).

The Nos2 mRNA and protein expression were evaluated in cardiac homogenates of both groups. No difference was observed between groups at mRNA (Figure 7A) or protein level (Figure 7B).

Surprisingly, after using anti-iNOS antibody, a band of 95-kDa protein instead of 130 kDa was revealed. This could represent a breakdown product of iNOS but the harvest storage and analysis conditions were designed to minimize proteolysis. These experiments were repeated several times but the 130-kDa band never appeared which is consistent with findings by other authors that tried to detect iNOS in cardiac or skeletal muscle tissue 41 42 . It could thus be possible that this 95-kDa band could represent a novel isoform or an alternatively spliced form of iNOS as previously proposed 41 . Furthermore, the evaluation of its mRNA expression indicated the presence of Nos2 mRNA in the samples and gave results that were in accordance with the Western blot findings.