Xanthine oxidoreductase-mediated injury is amplified by upregulated AMP deaminase in type 2 diabetic rat hearts under the condition of pressure overload

Highlights

• AMPD is upregulated in type 2 diabetic (T2DM) hearts.

• Pressure overload unmasked latent diastolic dysfunction in T2DM hearts.

• Activity and substrates of XOR were increased in pressure-overloaded T2DM hearts.

• Upregulated AMPD leads to ATP depletion and ROS production.

• Inhibition of XOR restored mitochondrial and ventricular functions in T2DM hearts.

Abstract

Background

We previously reported that upregulated AMP deaminase (AMPD) contributes to diastolic ventricular dysfunction via depletion of the adenine nucleotide pool in a rat model of type 2 diabetes (T2DM), Otsuka Long-Evans-Tokushima Fatty rats (OLETF). Meanwhile, AMPD promotes the formation of substrates of xanthine oxidoreductase (XOR), which produces ROS as a byproduct. Here, we tested the hypothesis that a functional link between upregulated AMPD and XOR is involved in ventricular dysfunction in T2DM rats.

Methods and results

Pressure-volume loop analysis revealed that pressure overloading by phenylephrine infusion induced severer left ventricular diastolic dysfunction (tau: 14.7 ± 0.8 vs 12.5 ± 0.7 msec, left ventricular end-diastolic pressure: 18.3 ± 1.5 vs 12.2 ± 1.3 mmHg, p < 0.05) and ventricular-arterial uncoupling in OLETF than in LETO, non-diabetic rats, though the baseline parameters were comparable in the two groups. While the pressure overload did not affect AMPD activity, it increased XOR activity both in OLETF and LETO, with OLETF showing significantly higher XOR activity than that in LETO (347.2 ± 17.9 vs 243.2 ± 6.1 μg/min/mg). Under the condition of pressure overload, myocardial ATP level was lower, and levels of xanthine and uric acid were higher in OLETF than in LETO. Addition of exogenous inosine, a product of AMP deamination, to the heart homogenates augmented XOR activity. OLETF showed 68% higher tissue ROS levels and 47% reduction in mitochondrial state 3 respiration compared with those in LETO. Overexpression of AMPD3 in H9c2 cells elevated levels of hypoxanthine and ROS and reduced the level of ATP. Inhibition of XOR suppressed the production of tissue ROS and mitochondrial dysfunction and improved ventricular function under the condition of pressure overload in OLETF.

Conclusions

The results suggest that increases in the activity of XOR and the formation of XOR substrates by upregulated AMPD contribute to ROS-mediated diastolic ventricular dysfunction at the time of increased cardiac workload in diabetic hearts.

Introduction

Diabetic cardiomyopathy is defined as cardiac dysfunction encompassing structural, functional and metabolic changes in the absence of coronary artery disease and hypertension [1,2]. A large population study revealed that the risk of hospitalization for heart failure in patients with type 2 diabetes was higher than that in non-diabetic subjects despite good control of other cardiovascular risk factors, while the increase in risk of myocardial infarction was negligible in those patients [3]. It has been reported that diastolic dysfunction precedes systolic dysfunction in patients with diabetic cardiomyopathy [1] and diastolic dysfunction appears to predict cardiovascular mortality in type 2 diabetic patients, even in those without clinically manifest heart failure at rest [4]. Thus, treatment for diastolic dysfunction in diabetic patients would be of clinical importance, though no specific therapy is currently available for this disorder.

We recently reported that upregulation of AMP deaminase (AMPD) activity by reduced micro RNA (miR)-301b is one of the mechanisms underlying augmentation of diastolic dysfunction at the time of pressure overload in diabetic hearts [5,6]. AMPD activity was 2.5-fold higher in Otsuka Long-Evans-Tokushima fatty rats (OLETF), an established model of type 2 diabetes (T2DM) [5,6], than in LETO, non-diabetic control rats, and the upregulated AMPD activity was associated with an increase in the level of IMP and decreases in the levels of adenine nucleotide pool and ATP during ventricular pressure overloading [5]. Overexpression of flag-AMPD3, an exclusively dominant isoform in rats, resulted in reduction of ATP level in cardiomyocytes, indicating a causal relationship between increased AMPD activity and ATP depletion [6]. However, roles of increased IMP and inosine nucleosides by AMPD upregulation in diabetic cardiomyopathy have not been examined.

In the present study, we hypothesized that upregulated AMPD augments the pathological role of xanthine oxidoreductase (XOR), via increases in both XOR activity and formation of its substrates, in diabetic cardiomyopathy. The rationale for the hypothesis is three-fold. First, while XOR is synthesized in its constitutively active dehydrogenase form, it can be converted through sulfhydryl group oxidation or limited proteolysis to xanthine oxidase, a form that produces cytotoxic reactive oxygen species (ROS) [7,8]. Experimental studies using animal heart failure models demonstrated that administration of XOR inhibitors improved mechano-energetic coupling and left ventricular (LV) function, attenuated LV remodeling and promoted survival [[9], [10], [11], [12]]. Notably, the magnitude of functional improvement with XOR inhibitors depends on the initial level of XOR activity [10], indicating that XOR inhibitors would be particularly effective in hearts with high XOR activity. Furthermore, some observational studies and meta-analyses showed that elevated serum uric acid level is an independent predictor of poor cardiac function and high mortality in patients with heart failure [13,14]. Second, our previous study showed that upregulated activity of AMPD in type 2 diabetic hearts results in elevation of the levels of IMP, inosine and xanthine, substrates of XOR, in the LV myocardium under the condition of pressure overload [5]. Third, increase in XOR activity under diabetic conditions has been reported in non-cardiac tissues [[15], [16], [17], [18]]. We tested our hypothesis by using OLETF as a model of T2DM as in previous studies [5,6,19,20] since it has been shown that OLETF have many similarities to T2DM patients in terms of metabolic and cardiovascular phenotypes [1,5,21,22,23].