Preserved endothelium-dependent dilatation of the coronary microvasculature at the early phase of diabetes mellitus despite the increased oxidative stress and depressed cardiac mechanical function ex vivo

The prevalence of Type 2 Diabetes (T2D) is increasing at an alarming rate assuming epidemic dimensions in industrialized societies [1]. Individuals with T2D have increased risk for developing cardiovascular diseases (CVDs), which is the main cause of early mortality and morbidity in the Western world [2]. Insulin resistance and T2D usually result from excess intake of deleterious nutrients such as saturated and trans fatty acids [3]. Consequent metabolic changes such as hyperinsulinemia and hyperglycemia [4] can provoke vascular lesions and endothelial dysfunctions at both micro- and macro-circulations [5‐8]. The vulnerability of the coronary circulation to the diabetic milieu can lead to endothelial dysfunction at this bed, which consists a significant biomarker of early coronary artery disease independently of atherosclerosis [9].

An accurate experimental model in order to clarify the mechanisms responsible for the pathophysiology of diabetes evolution and its complications is the inbred Zucker diabetic fatty (ZDF) rat. The homozygous fa/fa Zucker rat exhibits hyperphagia caused by a non-functioning leptin receptor. This leads to the development of obesity, hyperglycemia, hyperinsulinemia and finally diabetes at a young age [10, 11]. Previous studies on these rats in the later stages of diabetes have demonstrated that chronic hyperglycemia and hyperlipidemia can result in inflammation [12, 13], increased oxidative stress and vascular dysfunction [14, 15]. Coppey et al.[16] have shown that the endothelium-mediated responses to acetylcholine (Ach) are attenuated in epineurial arterioles of the sciatic nerve in diabetic ZDF rats. A key feature is the reduced production of nitric oxide (NO), a compound which mediates endothelium-dependent vasorelaxation and inhibits inflammation. In T2D, its bioavailability can be diminished either by the impaired insulin signaling either by the action of reactive oxygen species (ROS) [17].

Although the consequences of the later stages of T2D on the cardiac and endothelial function are well characterized, less is known concerning the early period of the disease where an interventional treatment may be more effective. Furthermore, most studies are focused on the endothelial function and perfusion of large arteries [18‐20] and few on the coronary function of resistance vessels. The primary function of the coronary microcirculation is to optimize nutrient and oxygen supply to the heart in response to any metabolic demand by coordinating the resistances within different microvascular domains, each governed by distinct regulatory mechanisms [21]. Coronary resistance arteries are capable of adapting to acute or chronic increases in blood flow leading to an increased NO-mediated relaxation and a consequent enlargement of their diameter. Furthermore, endothelial dysfunction in resistance arteries seems to precede that of large arteries [22]. Although there is strong evidence indicating that T2D is associated with impaired vasodilator responses of both peripheral and coronary vessels, Oltman et al.[22] have demonstrated that in diabetic young (8- to 12 wk old) ZDF rats the coronary arteriolar dilation to Ach of isolated microvessels is preserved. However, these in vitro studies isolate the coronary system from the cardiac environment and its influences.

Thus, this work aimed at studying the ex vivo cardiac and coronary vascular functions of young ZDF rats and at characterizing the levels of oxidative stress in their organs. The endothelial function of the intact coronary microvasculature was assessed in terms of endothelium-dependent and -independent vasodilatations in an ex vivo heart perfusion model at this phase of T2D. The NO production in aortas and hearts was evaluated indirectly by estimating the degree of phosphorylation of the endothelial NO-synthase (eNOS) at serine 1177 (Ser1177) and the iNOS levels in the heart. Finally, the fatty acid profile of cardiac membrane phospholipids was evaluated since any modification at this level leads to functional changes in lipid-protein interactions and related signaling pathways.

Even though several studies have examined the effects of diabetes on the vascular function, most of them used techniques of isolated vessels and most of them examined the late stages of diabetes [22, 43]. This is the first study that focuses on the effects of T2D on the intact coronary microvasculature at the early phase of the disease. This study addressed cardiac mechanical function in an isolated heart model that provided also the opportunity to study the vascular functionality in the intact coronary circulation. This allowed the analysis of the coupling of cardiac and coronary function, which is not feasible in isolated vessels.

The ZDF rat has been well characterized as experimental model of T2D. The ZL rats in our study ate a normal amount of diet (approximately 20 g/day) and exhibited a low blood glucose concentration (1 g/l) and proportion of HbA1c (approximately 4%) between the 7th and 11th wk of life. Their insulinemia was also low (1 μg/l) at the moment of sacrifice. In contrast, the ZDF animals consumed greater amounts of food (more than 30 g/day) that resulted to a higher body weight during the whole course of the experiment. Their blood analysis revealed a glycemia reaching 5 g/l at the 9th wk of age and a proportion of HbA1c close to 9% representing an already established diabetic state. The insulinemia at the 11th week was almost 4 times higher than that of the ZL control animals, indicating functional β cells in the Langerhans islets despite the high blood glucose concentration. Furthermore, their plasma triglycerides and cholesterol levels were approximately 2 times higher than those of the lean animals. All these characteristics associated with the fact that the visceral fat mass was abnormally high, clearly demonstrate that the ZDF animals displayed a severe insulin resistance responsible for the development of type-2 diabetes, which corresponds to a stage of early human type 2 diabetes.

In our study, the presence of diabetes provoked an enhanced cytosolic and mitochondrial oxidative stress in the hearts of the ZDF rats as observed by the lactate-to-pyruvate and aconitase-to-fumarase ratios respectively. It seems though that the respiratory chain complexes (RCC) were not implicated in the development of this mitochondrial oxidative stress, especially since the complex IV (CIV) activity was increased in the ZDF animals. It has been previously proposed that an up-regulation of CIV activity without any other changes in the other RCC or citrate synthase activities may serve to reduce any production of oxidative stress by the electron transport chain and improve the electron flux [44], but no measurements of mitochondrial respiration were performed in this study in order to directly evaluate the mitochondrial function that could be related to the observed depressed cardiac mechanical work. However, it is probable that a lack of mitochondrial antioxidant defenses could result to the increased mitochondrial oxidative stress observed n this study. The increased oxidative stress was also demonstrated in cardiac level by the TBARS results as already shown in the literature [22]. This could have been the result of the observed increase in the n-3 PUFA content of the cardiac membrane phospholipids of the ZDF hearts, since n-3 PUFAs are highly susceptible to peroxidation. A hyperglycemia-induced ROS production could have resulted to the observed increased lipid peroxidation due to the elevated n-3 PUFA content of membrane phospholipids.

The evaluation of the global plasma antioxidant capacity of ZDF rats was significantly decreased despite the increase of the GPx enzyme activity even though specific activities of other antioxidant enzymes such as thioredoxin reductase or other peroxidases were not determined. This indicates an increased presence of ROS in their plasma as evidenced also by the disappearance of the plasma thiol groups, even though it did not reach absolute significance (p = 0.079, ANOVA). The observed hyperglycemia and hyperinsulinemia could have also induced the NADPH oxidase activity and the consequent production of H₂O₂[45] that could explain the increased activity of GPx. The elevated activity of GPx may thus reflect a protective response against increased oxidative stress, since oxidative stress-induced antioxidant adaptive response could be particularly important in high ROS environments. Thus, an increased oxidative stress was already present in the cardiac tissue and plasma at the early phase of T2D despite the effort of the organism (CIV and GPx activities) to eliminate it.

The development of T2D also induced changes in the fatty acid profile of cardiac membrane phospholipids that may influence lipid-protein interactions, inflammation and related metabolic processes. In particular, an increase in the SFAs in cardiac membranes was observed to the detriment of MUFAs. This increased degree of saturation could negatively affect the membrane fluidity and increase its rigidity. However, an increase in the PUFAs content, in particular the C20:4 n-6 (arachidonic acid, AA), C22:5 n-3 (docosapentaenoic acid, DPA) and C22:6 n-3 (docosahexaenoic acid, DHA) contents, was observed probably as an effort to maintain a proper membrane fluidity degree. Moreover, these increases seem to result from stimulation of the desaturation and elongation enzymes in the organism of the ZDF rats rather from an increase in the concentration of the initial phospholipids of the PUFAs metabolisms (e.g. C18:2 n-6 and C20:5 n-3 for n-6 and n-3 respectively). It has been observed that tissues such as heart, kidney and liver from diabetic rats are characterized by a decrease in arachidonylated phospholipids and an increase in phospholipids containing linoleic acid (LA, C18:2 n-6). However, these modifications are mostly related to the later stages of diabetes. The n-3 and n-6 PUFAs of membrane phospholipids are also responsible for the production of anti- and pro-inflammatory molecules respectively. The low ratio EPA (eicosapentanoic acid)/AA found in our ZDF rats predisposes to a balance of eicosanoids favouring platelet aggregation and inflammatory mediator signaling [46]. The development of a compensatory mechanism might thus be in question as the levels of n-3 PUFAs were increased in the ZDF group. Low levels of EPA + DHA have been related to increased risk for sudden cardiac death [46] and hearts with high DHA content present very low in vivo and in vitro vulnerability to arrhythmia [47]. The ZDF hearts have high levels of EPA + DHA in order to reduce pro-inflammatory eicosanoids and cytokines. These modifications in the PUFAs levels of the cardiac membrane phospholipids probably help the heart to resist to any sudden cardiac damage at this early phase of diabetes [47].

In our study, we reported a strong decrease in the ex vivo cardiac function as already shown in pre-diabetic [48] and diabetic [49, 50] states, even though Daniels et al.[51] did not found an impaired in vivo cardiac function in db/db mice until 18 weeks of age. This could reflect the fact that in our study we have used an ex vivo heart perfusion model where the perfusion buffer is similar for both control and diabetic groups while the diabetic hearts in vivo are submitted to different plasma substrate and hormone concentrations than the control ones. In our ex vivo model, the RPP was significantly reduced, mainly because of the decreased heart rate. The T2D-induced reduction of the heart rate has already been commonly shown in the diabetic state [52] and it has been explained by an abnormal functioning of the cells involved in the generation and transfer of the electric influx triggering the cardiac contraction [53]. We also observed a noticeable diabetes-induced decrease in the coronary flow. This decrease could also be responsible for the reduction of the heart rate and cardiac mechanical work through insufficient oxygen supply. This could not be explained by an increase in the vascular tone triggering vasoconstriction and limitation of the oxygen and substrate supply since no abnormalities of the vascular function were observed according to the results of the vascular reactivity. Finally, the two phenomena could be synergistic and lead to the decreased RPP. In contrast, the left ventricle developed pressure was increased, which could compensate for the decreased heart rate. This phenomenon could be related to the action of sarcoplasmic reticulum Ca²⁺ ATPase (SERCA) whose levels have been shown to be increased in early diabetes and especially stimulated by the presence of insulin [54] but were not evaluated in this work. Thus, our data confirm a decrease in the ex vivo cardiac function, and particularly in the heart rate, at this model of heart perfusion, which is a common characteristic of all types of diabetes.

The underlying mechanism, which could explain the observed decreased cardiac mechanical work, has been already characterized. Several metabolic modifications in our study suggest that the reduced ex vivo cardiac function was due to this mechanism. In this study, we found an increased plasma triglyceride concentration that could allow the excess free fatty acid uptake and stimulation of the peroxisome proliferator-activated receptor alpha (PPARα) [50]. This would lead to increased β-oxidation and mitochondrial oxygen consumption [55]. The resulting excessive mitochondria-related ROS production, as evidenced by the aconitase-to-fumarase ratio in our study, would favour the expression of protein 53 (p53). The observed increased activity of the cytochrome c oxidase suggests an increased expression of the cytochrome c oxidase 2 (SCO2). Consequently, ectopic lipid accumulation may occur in the cardiomyocytes through increased expression of the fatty acid translocase protein FAT/CD36. Lipotoxicity then contributes to cardiac cell damages and myocardial dysfunction. A severe intramyocardial lipid accumulation, even at 8 wk of age [56] and an increased fatty acid oxidation [49] have been observed in ZDF rats. Thus, the altered myocardial substrate utilization affecting the mitochondrial function and stimulating the above described mechanism could be one of the factors responsible for the development of the T2D-induced ex vivo cardiac mechanical dysfunction leading to reduction of oxygen demand and subsequent decrease in the coronary perfusion. In our study though we did not observe any decrease in the left ventricle developed pressure that could have resulted from this mechanism. Instead, the heart rate was the parameter mostly affected by the diabetes in our study. However, the conditions of the perfusion model did not allow us to evaluate correctly the left ventricle developed pressure. A future study of heart perfusion at stable heart rate (pacing) in order to evaluate the cardiac contractility and relaxation by measuring the maximal rate of the ventricular pressure rise (dP/dt_max) and fall (dP/dt_min) respectively could enlighten our knowledge concerning this mechanism. Taken together, these observations suggest that the lipid accumulation and alterations of substrate utilization in ZDF rats may affect firstly the cells responsible for the cardiac contraction that are involved in the generation and transfer of the electric influx. Furthermore, hyperglycemia and insulin resistance, two states that characterized the ZDF rats in this study, have been related to damages in cardiac nodal cells and to cardiac electrophysiological properties.

The effects of the T2D on the function of the coronary resistance arteries were also evaluated in this ex vivo model of isolated perfused heart through measurement of the coronary reactivity. This parameter was estimated through changes in the global coronary tone which mainly reflects the function of the arteriole network, since atherosclerosis does not occur in the rat [57]. The effects of T2D on the heart can thus be evaluated independently of the development of coronary artery disease. In this study, the EDD of diabetic hearts was fully maintained while the EID was even enhanced. We also evaluated indirectly the NO production through measurements of the expression and phosphorylation of the cardiac and aortic eNOS, since the phosphorylation of the enzyme at this site has been shown to increase NO production [58], and that of the iNOS levels, since it has been shown that under diabetic conditions stress-induced iNOS is able to produce an abnormal amount of NO [59]. In our study though, diabetes did not modify neither cardiac iNOS levels nor the expression and phosphorylation of eNOS at Ser1177, even though studies concerning other phosphorylation sites or the cGMP signaling pathway were not performed. It has been recently shown that the formation of superoxide from uncoupled bovine eNOS in endothelial cell can be stimulated by this kind of phosphorylation [60]. The presence of increased oxidative stress in the diabetic hearts of the ZDF rats could have led to the depletion of tetrahydrobiopterin (BH₄) and uncoupling of the eNOS. This could result to increased production of ROS from the enzyme even though there was no change in the phosphorylation of eNOS at Ser1177. The consequent superoxide production could lead to H₂O₂ production, which could have participated to the maintained EDD of our study since it has been shown that this molecule can act as endothelium-hyperpolarizing factor (EDHF) [61]. The maintained EDD was a surprising finding given the huge amount of studies associating T2D and dysfunctions of the coronary microcirculation [62‐64]. Factors contributing to these discrepancies are the severity of the obesity and diabetic state studied as well as the experimental method used. Oltman et al.[22] have reported a preservation of the coronary arteriolar dilatation to Ach in isolated vessels of pre-diabetic young (8- to 12-wk old) ZDF rats. However, in the present study, the ZDF rats were not in a prediabetic state, but the T2D was already developed as indicated by the blood glucose concentration, which was already high from the 8th wk of age. Thus, it seems that the endothelial function of the intact coronary microvasculature is not affected from the diabetes at this phase. The unaltered eNOS and iNOS activities and the high levels of AA and DHA despite the presence of oxidative stress found in the diabetic hearts could have contributed to this phenomenon. AA [65] and DHA [66] are known for their vasorelaxant effects via the production of prostacyclin (PGI2) and the reduction of calcium influx in vascular smooth muscle cells. A future study of vascular reactivity in the presence of a cyclo-oxygenase (COX) inhibitor such as indomethacin could enlighten this hypothesis. Furthermore, the early diabetes could have provoked an increased expression of SERCA [54] favouring the calcium uptake in vascular smooth muscle cells that could be involved in the increased EID observed in the ZDF diabetic hearts of this study.

As shown by the calculated activity of endothelial cells to induce dilatation and the evaluation of eNOS expression the phenomenon of the maintained Ach-mediated vasodilatation was partly mediated by the activity of endothelial cells. However, the SNP responses were enhanced in the ZDF rats representing an enhanced function of the smooth muscle cells of the coronary system contributing to the maintained endothelium-dependent dilatation. This enhanced function may be due to a modified NO response, which could increase guanylate cyclase activity as already shown in cases of obesity and hypertension [67, 68].

These vascular alterations may reflect a compensatory adaptation of the cardiovascular system to support increased cardiac work since cardiac output and stroke volume are increased in obese and diabetic states [69, 70]. Taken also under consideration the decreased ex vivo cardiac mechanical function observed in this study, this adaptation seems to be essential to adjust organ perfusion during physiological processes such as exercise and pathological processes such as ischemic diseases [11]. Otherwise, the heart would not be able to respond to the increased metabolic demands. These results come in agreement with data of Oltman et. al[22] that showed that there is no change in the Ach response of small coronary arteries in ZDF rats at the age of 8–12 weeks but there is a progressive impairment until the age of 40 weeks.

Cardiovascular function was evaluated in young diabetic ZDF rats using an ex vivo heart perfusion model. Our data suggest that at the early phase of diabetes, increased oxidative stress in tissue and plasma is already present and probably responsible for the observed ex vivo cardiac mechanical dysfunction. However, the heart tries to resist by preserving the EDD of the coronary microvasculature. A number of other adaptations seem to take place at this phase of the disease such as the increased GPx and CIV activities and the increase in the n-3 PUFAs content of the myocardial membrane. This would help the heart to keep an adequate perfusion and respond to any acute cardiac incident at this phase. Thus, therapeutic interventions at this early phase of the disease aiming at increasing the heart rate and maintaining the observed adaptations could be an option for delaying or decreasing the late-stage complications of the diabetes.