Effects of hypoglycemia on myocardial susceptibility to ischemia–reperfusion injury and preconditioning in hearts from rats with and without type 2 diabetes

Type 2 diabetes mellitus (T2DM) increases morbidity and mortality after myocardial infarction (MI) [1, 2]. Albeit controversial, randomized controlled trials have suggested that an intensified glycemic control in patients with T2DM, with an increased incidence of hypoglycemia, is associated with increased mortality [3‐5]. Hypoglycemia is a common adverse effect of treatment of T2DM with insulin and sulphonylureas [6]. Studies using continuous glucose monitoring systems have revealed a higher incidence of hypoglycemia than previously appreciated [7, 8]. The mechanistic link between hypoglycemia and increased cardiovascular mortality remains unclear. Reduced myocardial tolerance to ischemia and reperfusion (IR) or reduced capacity for activation of cardioprotection may be underlying mechanisms of increased mortality after MI in patients with T2DM. Cardioprotection can be activated by ischemic preconditioning (IPC), by which repetitive sublethal episodes of ischemia induce resistance towards myocardial IR injury [9]. Previous studies of IPC and myocardial IR susceptibility during hyperglycemic conditions have demonstrated attenuated efficacy of IPC compared to normoglycemia [10‐12] while hyperglycemia per se, does not seem to influence on myocardial infarct size [13]. However, glucose fluctuations aggravate cardiac susceptibility to IR injury [14] suggesting that hypoglycemia may have impact on infarct size. An altered myocardial susceptibility to IR or capacity for cardioprotection during hypoglycemia may represent a mechanistic link between hypoglycemia and impaired outcome in patients with diabetes after MI.

A major mechanism involved in the influence of varying circulating glucose concentration on cardioprotection seems to involve the activity of hexosamine biosynthetic pathway (HBP), which is sensitive to changes in circulating glucose and glutamine concentrations [15], accordingly referred to as a “nutrient-sensing” pathway. Moreover, elevating O-linked β-N-acetylglucosamine (O-GlcNAc) glycocylation, the resultant of the HBP, promotes cell survival during cellular stress, whereas decreasing the levels of O-GlcNAc reduces cell survival [16]. O-GlcNAcylation seems associated with cardioprotection by IPC [17, 18]. We hypothesized that hypoglycemia modifies IR injury and capacity for cardioprotection by IPC differently in diabetic and non-diabetic hearts through simultaneous changes in myocardial glucose uptake and HBP activity during reperfusion.

The aims of the present study were to compare myocardial IR susceptibility, the cardioprotective efficacy of IPC and their associated changes in myocardial glucose uptake and O-GlcNAc levels in diabetic and non-diabetic during normo- and hypoglycemia.

The present study demonstrates that myocardial susceptibility to IR is augmented during hypoglycemia in both rats with and rats without diabetes. However, the cardioprotective effect of IPC is preserved during hypoglycemia in rats with diabetes hearts contrary to rats without diabetes. The underlying mechanisms of IPC induced cardioprotection are associated with increased myocardial glucose uptake and O-GlcNAc levels in animals with diabetes as well as without diabetes. Increased myocardial susceptibility to IR during hypoglycemia may be a mechanistic link between hypoglycemia and impaired outcome after MI in patients with diabetes.

Myocardial susceptibility to IR in diabetic hearts remains controversial because studies in animal models of type 1 and 2 diabetes have yielded conflicting results [21, 26‐32]. The discrepancy may be explained by different species and models and also by the age of animals and duration of diabetes [25]. In the present study, we confirm our previous findings of reduced susceptibility to IR during normoglycemia in a T2DM animal model with recent onset of diabetes [21, 25]. Our findings, that infarct size was increased in both hearts from animals with and without diabetes during hypoglycemia and that no difference in infarct size was observed between animals with and without diabetes, are supported by previous findings of increased IR susceptibility in animals without diabetes during hypoglycemia in the brain [33] and heart [34]. We now extended these findings to a clinically more relevant model of hypoglycemia in animals with diabetes. The absent difference in infarct size between animals with and without diabetes during hypoglycemia indicates that the endogenous cardioprotection observed in hearts from animals with diabetes during normoglycemia at onset of diabetes [25, 35] seems to be lost during hypoglycemia. Together, our findings emphasize the importance of glucose concentration during IR, when evaluating myocardial susceptibility to IR in hearts from animals with or without diabetes. Myocardial susceptibility to IR during hypoglycemia is of particular importance in diabetic hearts because patients with diabetes frequently suffer episodes of hypoglycemia due to treatment with glucose lowering drugs.

In contrast to our previous findings [21], IPC reduced myocardial infarct size and improved post ischemic left ventricular function in both animals with and without diabetes during normoglycemia. The diabetic heart may still be amenable to protection by IPC with an intensified stimulus that overcomes the increased threshold for the necessary activation of pro-survival kinases [30]. Consequently, we used a more aggressive stimulus by two cycles of 5-min ischemia and 5-min reperfusion in contrast to four cycles of 2-min ischemia followed by 3-min reperfusion as used previously [21]. IPC retained cardioprotection during hypoglycemia in diabetic hearts by reducing infarct size, albeit to a lesser extent than during normoglycemia. In contrast, the cardioprotective effect of IPC in non-diabetic animals during hypoglycemia was abolished.

Because osmolarity changes with variations in circulating glucose concentrations, changes in osmolarity may influence infarct size. A change in plasma glucose from 11 to 3 mmol/l would lead to a 2.5% reduction in osmolarity of the perfusion buffer. The influence of osmolarity on infarct size has varied in previous studies. Kersten et al. [36] demonstrated that increases in serum osmolarity obtained by administration of raffinose did not influence infarct size or interfere with the ability of IPC to protect against infarction. In contrast, Zalesak et al. showed that a hyperosmotic environment blunted the efficiency of IPC against IR injury and improved ischemic tolerance in non-preconditioned isolated rat hearts suggesting that increased osmolarity, similar to that in the hyperglycemic conditions, may play a pivotal role in a failure of IPC to induce cardioprotection in the diabetic myocardium [37]. Since glycemic levels and osmolarity were altered similarly by hypoglycemia and infarct size was equally increased in animals with and without diabetes, we cannot establish whether the increment was caused by hypoglycemia or a reduction in osmolarity. However, an increment in infarct size due to reduced osmolarity would be discrepant to previous findings. The different responses to IPC in animals with and without diabetes during hypoglycemia cannot be related to the minor reductions in osmolarity as the reductions were identical in the two groups.

Consistent with previous findings during non-ischemic conditions [25, 38], we report reduced myocardial glucose uptake during reperfusion and stabilization in hearts from animals with diabetes. Stimulation of glycolysis and glucose oxidation during reperfusion improves post-ischemic left ventricular functional recovery [39‐41]. Cardioprotection by IPC is associated with increased myocardial glucose uptake during reperfusion in non-diabetic hearts [41, 42] as confirmed in hearts from animals without diabetes during normoglycemia in the present study. However, the absent infarct sparring effect of IPC in hearts from animals without diabetes during hypoglycemia was associated with an absent modulation of myocardial glucose uptake during reperfusion by IPC in animals without diabetes. In contrast, IPC increased myocardial glucose uptake during reperfusion in hearts from animals with diabetes at normo- as well as hypoglycemia. These findings parallel cardioprotection afforded by IPC in hearts from animals with diabetes during normo- and hypoglycemia. Together, our findings support the notion that an underlying mechanism of the effects behind IPC involves myocardial glucose uptake as only IPC generated increments in myocardial glucose uptake translated into cardioprotection. This notion is further supported by findings of increased myocardial glucose uptake associated with improved functional recovery after ischemia in animals with diabetes treated with rosiglitazone [43].

O-linked β-N-acetylglucosamine glycocylation is a recently detected posttranslational modification of nuclear, cytoplasmic and mitochondrial proteins [44]. O-GlcNAc acts as an intracellular stress sensor, linking glucose metabolism to cellular function at a molecular level [15, 45]. O-GlcNAcylation is dependent upon substrate synthesis in the HBP [46]. Flux through the HBP parallels glucose availability [15, 46] and increased glucose uptake is linked to HBP activation and downstream formation of O-GlcNAc [47]. On the molecular level, O-GlcNAcylation is implicated as a major mechanism of glucose toxicity and insulin resistance in diabetes [48, 49]. In the present study investigating myocardial O-GlcNAc levels during reperfusion, we did not observe differences between animals with and without diabetes, which may be related to the short duration of diabetes in the relatively young animals used in the present study. Augmentation of cardiac O-GlcNAc levels by glucosamine or salidroside administration affords cardioprotection [50, 51]. In addition, O-GlcNAc has been suggested to be involved in the underlying mechanism of IPC [17, 18]. The mechanisms by which acute elevation of O-GlcNAc levels induce cardioprotection seem to remain intact in the hearts from animals with diabetes even during hypoglycemia as IPC induced reduction in infarct size was associated with increased myocardial O-GlcNAc levels in animals with diabetes during both normo- and hypoglycemia. However, in hearts from animals without diabetes, IPC did not reduce infarct size or increase myocardial glucose uptake and did not influence O-GlcNAc levels during hypoglycemia. Accordingly, increased myocardial O-GlcNAc levels and glucose uptake may represent at mechanistic link to cardioprotection afforded by IPC.

The clinical implication of our findings is that myocardial infarct size in patients with T2DM may be larger during hypoglycemia than during normoglycemia similarly to the animals investigated in the present study. This finding may explain the observed increased mortality in clinical trials investigating an intensified glycemic control in patients with T2DM [3‐5], as myocardial infarct size is a prognostic factor after MI [52]. However, infarct size alone does not seem to be the only determining mechanism of the overall higher mortality in patients with T2DM suffering an acute MI compared with patients without diabetes as infarct size at identical glucose concentrations was smaller or equally sized in animals with and without T2DM in the present study.

Limitations of the present study primarily relate to the use of an isolated perfused heart model but we used this to investigate the impact of hypoglycemia on myocardial susceptibility to IR and not the combined effects of hypoglycemia and a blood glucose lowering compound. However, in vivo induction of hypoglycemia by insulin, sulphonylureas or other agents may influence findings, as these agents per se impacts the effects of IR and IPC [53, 54]. Moreover, an uncontrolled systemic response to hypoglycemia, including activation of neurogenic and humoral components, may constitute important limitations. Importantly, such effects can be excluded in our model. We choose specifically to investigate the effects of hypoglycemia on infarct size in a glucose dependent model in the absence of free fatty acids to specifically investigate the impact on MGU without interference from other substrates. However, the presence of free fatty acids as well as other substrates may potentially have influenced our findings. We used a glucose concentration of 11 mmol/l as this is most commonly used concentration in the Langendorff perfused heart model [19]. Because normal postprandial glucose concentration in the rat is up to 10.4 mmol/l [20] a glucose concentration of 11 mmol/l is not considered hyperglycemic in the absence of free fatty acids and other substrates in the experimental model. We used relatively young animals with a short duration of diabetes. This may potentially limit the generalizability to patients with long lasting T2DM.

In conclusion, hypoglycemia increases myocardial infarct size in hearts from animals with and without diabetes. In contrast to hearts from animals without diabetes, hearts from animals with diabetes are amenable to cardioprotection during hypoglycemia. Increased myocardial glucose uptake and O-GlcNAc levels seem to be involved in the cardioprotective mechanisms of IPC irrespective of circulating glucose concentrations.