This article is mentioned in:

Introduction

Diabetes mellitus (DM) is one of the most important risk factors for heart failure and increased morbidity and mortality (1). There is growing evidence that the increased risk of heart failure may occur independently of accelerated coronary artery disease and arterial hypertension, suggesting that other mechanisms associated with diabetes underlie the development of cardiomyopathy (2,3).

In addition, a number of studies have demonstrated that hyperglycemia directly causes cardiac damage, contributing to the development of diabetic cardiomyopathy (4,5). However, numerous pathophysiological stimuli are involved in its development and mediate cardiac injury, leading to systolic and diastolic left ventricular (LV) dysfunction (6). The pathophysiology of diabetic cardiomyopathy includes, for example, microangiopathy, endothelial dysfunction, cardiac fibrosis and the disruption of intracellular Ca2+ transport, all triggered by the diabetic milleu (7–9). Moreover, structural changes in extracellular matrix (ECM) regulation and the accumulation of cardiac fibrosis, an intensified production of oxidative stress and an overwhelming cardiac inflammatory immune response play a crucial role in the pathogenesis of diabetic cardiomyopathy (6,7).

Previous studies have demonstrated the impaired LV performance under basal conditions during the chronic stage of streptozotocin (STZ)-induced diabetic cardiomyopathy. Furthermore, in previous studies, we identified possible pathophysiological mechanisms, resulting in the cardiac phenotype of diabetic cardiomyopathy. Among others, we found that cardiac fibrosis, endothelial dysfunction, cardiac inflammation, as well as neurohumoral activation are greatly involved in the development of cardiac dysfunction under those conditions (7,9–11).

The rat/mouse model of diabetic cardiomyopathy induced by an injection of STZ is a well-established model for investigating this condition. The cardiac phenotype under basal conditions in the chronic stage has been sufficiently characterized. Thus, in the current study, by time course analysis, we investigated LV performance by in vivo pressure-volume loops and adverse cardiac remodeling using a rat model of STZ-induced diabetic cardiomyopathy. We measured LV function, cardiac immune cell invasion, oxidative stress, as well as cardiac remodeling under diabetic conditions.

Materials and methods

Animal characteristics and study design

Six-week-old Sprague-Dawley rats (n=22) were maintained on a 12:12 h light-dark cycle and fed with standard chow ad libidum. In 12 rats, diabetes was induced by an intraperitoneal (i.p.) single injection of STZ (70 mg/kg) diluted in 0.1 M sodium citrate buffer, pH 4.5 (Sigma-Aldrich Chemie Gmbh, Munich, Germany); the rats were then randomly divided into 2 experimental groups (STZ 2 weeks and STZ 6 weeks). Hyperglycemia was measured 48 h later using a reflectance meter (Acutrend; Roche Diagnostics GmbH, Mannheim, Germany). All diabetic animals displayed a blood glucose level >550 mg/dl associated with severe polyuria and polydipsia. For the controls, 6 vehicle-treated animals (treated with citrate buffer only) were used. All animal experiments were carried out in accordance with the Guide for the Care and Use of Laboratory Animals published by the US NIH (NIH Publication no. 85-23, revised 1996).

Surgical procedures and hemodynamic measurements

Two (STZ 2 weeks) and 6 (STZ 6 weeks) weeks after the STZ injection, the animals were anesthetized (pentobarbital 60 mg/kg, i.p. injection; and buprenorphine 0.1 mg/g, i.p. injection), intubated and artificially ventilated. A 2 French micro-conductance catheter (Aria SPR 858; Millar Instruments, Inc., Houston, TX, USA) was positioned in the LV for the continuous registration of LV pressure-volume (PV) loops in a closed-chest model. Calibration of the volume signal was obtained using the hypertonic saline (10%) wash-in technique. Indices of cardiac function were derived from PV data obtained both at the basal steady state and during transient preload reduction by temporary occlusion of the abdominal vena cava. All hemodynamic measurements were performed during apnea. After the euthanization of the mice, LV tissues were excised, immediately snap-frozen in liquid nitrogen and stored at −80°C for biological and immunohistological analyses.

Immunohistological measurements

The total collagen content of the Sirius red (Polysciences, Inc., Warrington, PA, USA)-stained sections was measured under circularly polarised light as previously described (12). The data were then quantified by digital image analysis as the percentage of area fraction.

As previously described (13,14), LV tissue of the left ventricle was embedded in Tissue-Tek (Dako) and immunohistochemistry was performed with specific antibodies directed against CD3 (Bioss Inc., Woburn, MA, USA), CD8a (Bioss Inc.), nitrotyrosin (Sigma-Aldrich Chemie Gmbh), α-smooth muscle actin (SMA) (Abcam, Cambridge, UK) and matrix metalloproteinase (MMP)-2 (Chemicon, Temecula, CA, USA). Quantification was performed by digital image analyses (13,14). In brief, the ratio between the heart tissue area and the specific chromogen-positive area was calculated (area fraction, %). The numbers of infiltrating cells were calculated by measuring the number of cells/area of heart tissue (cells/mm2).

Statistical analysis

Statistical analysis was performed using SPSS version 12.0 software. Data are expressed as the means ± SEM. Statistical differences were assessed using the Kruskal-Wallis test in conjunction with the Mann-Whitney U post hoc test. Bonferroni correction was applied to the post hoc Mann-Whitney U test to adjust for multiple comparisons. Pearson’s correlation co-efficient was used for linear regression analysis and Spearman’s correlation co-efficient was used for non-linear correlations. Regression analyses and curve fitting were performed to determine exact correlations. Differences were considered statistically significantly at a value of p<0.05.

Results

Animal characteristics and hemodynamic measurements after the induction of diabetes

Two and 6 weeks after STZ-induced type I diabetes, animal characteristics, systolic and diastolic LV function were determined in all experimental groups using the conductance catheter technique (Table I).

Table I. Animal characteristics and hemodynamic results 2 and 6 weeks after the induction of diabetes.

Table I.

Animal characteristics and hemodynamic results 2 and 6 weeks after the induction of diabetes.

Items	Control	STZ 2 weeks	STZ 6 weeks
Characteristics
Body weight (g)	491±10	374±9a	253±7a
Blood glucose (mmol/l)	4.1±0.3	>30.5	>30.5
Mean blood pressure	113±7	94±8a	73±9a,b
Heart weight (mg)	1223±16	1009±19a	839±21a
LV weight (mg)	883±20	697±16a	586±18a
Heart weight/body weight ×10−3	2.49±0.06	2.67±0.09a	3.31±0.07a
LV weight/heart weight	0.726±0.02	0.68±0.03a	0.69±0.02a
Global LV function
Heart rate (bpm)	378±13	257±10a	229±9a
LV end systolic pressure (mmHg)	116±4	97.4±6a	76±7a,b
Cardiac output (μl/min)	95872±10723	50893±8983a	53211±5213a
Ejection fraction (%)	52.7±4.8	45.2±5.5	49±3.9
Systolic LV function
LV end systolic volume (μl)	256±42	272±52	268±32
LV contractility (mmHg/sec)	7234±365	5671±331a	4178±245a,b
LV end systolic pressure (mmHg)	115.3±3.9	94.8±4.2a	76±6.1a,b
Diastolic LV function
LV end diastolic volume (μl)	489±46	446±61	477±51
LV diastolic relaxation (mmHg/sec)	−6598±195	−3723±201a	−2874±202a,b
LV diastolic relaxation time (msec)	13.6±0.8	21.4±0.7a	23.5±0.9a
LV end diastolic pressure (mmHg)	4.1±1	5.3±0.8	6.1±1
LV pressure half time (msec)	8.2±0.3	12.8±0.7a	14.2±1.1a
Stiffness constant β	0.0054±0.001	0.0086±0.003	0.0142±0.003a

a p<0.05 vs. control;

b p<0.05 vs. control and STZ 2 weeks group. STZ, streptozotocin; LV, left ventricular.

At the early time point of 2 weeks, a significant decrease (−33%) in heart rate was observed in the diabetic animals. At the later time point of 6 weeks, no change in heart rate was observed (Fig. 1A). This was associated with a reduction in diastolic LV function indicated by a significant reduction of LV relaxation at the early time point (−42%). Six weeks after STZ-induced diabetes, an additional decrease in LV relaxation (−38%) was observed (Fig. 1D). In addition, STZ-induced diabetes resulted in a significant increase (+38%) in cardiac stiffness at the later time point (Fig. 1E). Systolic LV function, indexed by LV contractility, displayed a significant reduction of −22% at the early time point of the disease, leading to an additional decrease of another −26% after 6 weeks (Fig. 1B); the end systolic pressure exhibited the same pattern (Fig. 1C). The initial decrease (−18%) was followed by a second significant (−26%) decrease, indicating global LV dysfunction. All these results were associated with a significant reduction (−47%) in cardiac output after 2 and 6 weeks (Fig. 1F).

Figure 1. Heart rate, left ventricular (LV) contractility, end systolic pressure, LV relaxation, cardiac stiffness and cardiac output as representative hemodynamic parameters of control rats and rats with streptozotocin (STZ)-induced diabetes in a time course analysis. Data are from 6–8 rats/group. Data are expressed as the means ± SEM. *p<0.05.

ECM alterations and remodeling after the induction of diabetes

To examine the turn over of the ECM assembly, cardiac tissue was stained by Sirius red measuring the total content of collagen in the cardiac tissue. Moreover, α-SMA and MMP-2 protein expression levels were measured by specific staining. The diabetic animals showed an initial increase (1.9-fold) in total collagen, but there was no additional increase at the later time point (Fig. 2). By contrast, the investigation of α-SMA and MMP-2 protein expression levels revealed an initial 16.6-fold (p<0.05) and 11.6-fold (p<0.05) increase after 2 weeks, respectively. Six weeks after the induction of diabetes by STZ, we observed an additional 1.6-fold (p<0.05) and 5-fold (p>0.05) increase in α-SMA and MMP-2 protein expression levels in the cardiac tissue, respectively (Fig. 3C and D).

Figure 2. Representative images of Sirius red staining in the cardiac tissue of the control rats and rats with streptozotocin (STZ)-induced diabetes. Data are from 6–8 rats/group. Data are expressed as the means ± SEM. *p<0.05.

Figure 3. Cardiac immune cell infiltration of CD3+ and CD8a+ T lymphocytes and protein expression levels of nitrotyrosin, α-SMA and MMP-2 in the cardiac tissue of the control rats and rats with streptozotocin (STZ)-induced diabetes. Data are from 6–8 rats/group. Data are expressed as the means ± SEM. *p<0.05.

Cardiac immune cell infiltration and oxidative stress response after the induction of diabetes

To investigate the immune cell infiltration by T cells, we measured the number of CD3+ and CD8a+ immune cells in the cardiac tissue by specific immunological staining, as well as the content of nitrotyrosin, to evaluate oxidative stress response in the cardiac tissue. The number of CD3+ immune cells significantly increased at 2 and 6 weeks after the induction of diabetes (Fig. 3A; p<0.05). Of note, the number of CD8+ cells in the diabetic mice displayed an initial 3.6-fold (p<0.05) increase followed by a significant decrease after 6 weeks (Fig. 3B). By contrast, the level of nitrotyrosin showed a constant 7.8-fold (p<0.05) increase followed by another 1.4-fold (p<0.05) increase in the animals with STZ-induced diabetes compared with the controls (Fig. 3E).

Correlations between hemodynamic and morphological parameters after the induction of diabetes

We examined the possible correlations between hemodynamic and morphological parameters in this experimental setting. We investigated the correlation between the total content of collagen and systolic LV function, indexed by LV contractility and end systolic pressure. We found that an increase in total collagen content correlated with a significant reduction in systolic LV function (r=0.55, r=0.6; Fig. 4A and B). Concerning the diastolic LV function, we investigated the pathophysiological correlation between the protein content of α-SMA and LV stiffness. We determined a correlation between the increased number of α-SMA-positive myofibroblasts and the increase in cardiac stiffness (r=0.7; Fig. 4C).

Figure 4. Correlations between protein expression levels of collagen content/left ventricular (LV) contractility, collagen content/end systolic pressure, as well as protein expression levels of α-SMA/cardiac stiffness in the control rats and rats with streptozotocin (STZ)-induced diabetes. Data are from 6–8 rats/group.

Moreover, we performed correlation analyses to determine the correlation between cardiac immune cell infiltration and the parameters of systolic and diastolic LV function. We found that an increased cardiac CD3+ immune cell infiltration correlated with a reduction in LV contractility (r=0.83), end systolic pressure (r=0.77) and an increase in cardiac stiffness (r=0.54) (Fig. 5A–C). In addition, we also investigated possible correlations between the number of CD8a+ immune cells and these hemodynamic parameters. It was also found that an increased number of CD8a+ immune cells correlated with a reduction in LV contractility (r=0.48), end systolic pressure (r=0.35) and an increase in cardiac stiffness (r=0.31) (Fig. 5D–F). However, the correlation between CD8a+ immune cell infiltration and hemodynamic parameters was weaker when compared with the number of infiltrating CD3+ immune cells and the hemodynamic parameters (Fig. 5). Furthermore, we observed that a significant increase in CD3+ and CD8a+ immune cells correlated with an increase in collagen and α-SMA protein content in the cardiac tissue after the induction of diabetes (Fig. 6).

Figure 5. Correlations between CD3+ and CD8a+ immune cells and left ventricular (LV) contractility, end systolic pressure, as well as cardiac stiffness in control rats and rats with streptozotocin (STZ)-induced diabetes. Data are from 6–8 rats/group.

Figure 6. Correlations between CD3+ and CD8a+ immune cells and the protein content of collagen and α-SMA in control rats and rats with streptozotocin (STZ)-induced diabetes. Data are from 6–8 rats/group.

Discussion

The salient finding of this study is that the model of STZ-induced diabetic cardiomyopathy is a robust model for the investigation of cardiac inflammation and remodeling processes. Our data demonstrate that experimental diabetic cardiomyopathy is characterized by an increase in cardiac inflammation and changes in the regulation of the ECM over a time period of 6 weeks following STZ-induced-diabetes.

Diabetic cardiomyopathy is associated with LV dysfunction (12,15). The results of the present study revealed an impairment in diastolic and systolic LV function at 2 and 6 weeks after the induction of diabetic cardiomyopathy. These hemodynamic results are in line with those from previous studies using the same experimental rat model (7,9,11,16). However, we also described the concrete time course of the impairment in systolic and diastolic LV performance over a period of 2–6 weeks, indicated by a decrease in LV contractility, end systolic pressure, LV relaxation and an increase cardiac stiffness, all resulting in a significant reduction in cardiac output.

Our hemodynamic findings identified an early diastolic LV dysfunction, which showed a clear progression over time in this animal model. Active and passive diastolic LV relaxation was affected by the STZ injection. LV relaxation as a marker for active LV relaxation was significantly impaired 2 weeks after the induction of diabetes, whereas cardiac stiffness as a marker for passive LV relaxation displayed a tendency of deterioration after 2 weeks, but was only impaired 6 weeks after the induction of diabetes. The end diastolic pressure was unaffected in this experimental setting; there was no significant increase in end systolic pressure following STZ-induced diabetes (Table I). This may be explained by chronic dehydration in the animals with STZ-induced diabetes, leading to a reduction in afterload. In previous studies, we demonstrated that an induction of cardiac inflammation under diabetic conditions was associated with an impairment in systolic LV function (17,18). In the current study, we confirmed the hemodynamic profile of systolic and diastolic LV dysfunction under diabetic conditions.

Cardiac inflammation is one of the hallmarks of heart failure (6,9). Intensified pro-inflammatory cytokine expression levels, as well as increased immune cell infiltration, such as cytotoxic T lymphocytes and macrophages, has been observed in the inflamed heart in diabetic cardiomyopathy (19–21). As regards this finding, we examined the invasion of CD3+ and CD8a+ immune cells into the heart in this disease model. The induction of diabetes led to an uninterrupted increase in CD3+ immune cell invasion over the 6-week observation period. In addition, the induction of diabetes by STZ led to a significant increase in the number of CD8a+ T lymphocytes 2 weeks after the STZ injection. Of note, 6 weeks after the induction of diabetes, the cardiac amount of this cell population was reduced when compared to the time point of 2 weeks after the STZ injection, indicating an important role of CD8a+ immune cells predominantly during the early stages of diabetic cardiomyopathy.

A large body of evidence indicates that LV remodeling accompanied with changes in ECM regulation is an important factor for LV function in diabetic cardiomyopathy (11,16). In this study, the induction of diabetes led to a significant increase in cardiac fibrosis, indexed by an increase in total collagen content at the time points of 2 and 6 weeks after the induction of diabetes. Moreover, we matched the values of total collagen and LV contractility. We showed that the total collagen content correlated with a reduction in LV contractility and end systolic pressure, suggesting that the corrrelation between systolic LV function and myocardial fibrosis plays a pathophysiological role in this experimental setting.

In the clinical course of heart failure, MMPs are upregulated by intense cardiac inflammation and may contribute to cardiac remodeling under diabetic conditions (22,23). We therefore analysed the protein expression levels of MMP-2 and found increased protein expression levels of MMP-2 in this experimental setting.

However, the persistence of an abnormally high number of myofibroblasts is a hallmark of fibrotic disease in other organs, as well as the heart (24,25). In the current study, we identified that STZ-induced diabetes led to an increase in the number of α-SMA-positive myofibroblasts in the heart 2 weeks after the STZ injection. However, an additional marked increase in the number of α-SMA-positive myofibroblasts was documented at the time point of 6 weeks after the STZ injection. Concerning these findings, we also observed a correlation between the number of α-SMA-positive myofibroblasts and cardiac stiff-myofibroblasts stiffness, as a strong marker for diastolic LV dysfunction. Of note, an increase in the number of α-SMA-positive myofibroblasts correlated with an increase in the cardiac stiffness index, suggesting a pathophysiological impact of myofibroblasts in adverse myocardial remodeling during diabetic conditions in rats. Moreover, we verified that an increased cardiac immune cell invasion of CD3+ and CD8+ cells correlated with an increased protein content of collagen and α-SMA in the cardiac tissue at 2 and 6 weeks after the induction of diabetes. In addition, we also showed that an increased infiltration of these immune cell populations was significantly associated with an impairment in systolic and diastolic LV performance. These results indicate an important role of CD3+ and CD8a+ immune cells in the development and progression of LV dysfunction and adverse cardiac remodeling under diabetic conditions.

DM is associated with an exponential increase in oxidative damage (26). Previous studies have demonstrated that nitrotyrosin, as a maker of oxidative stress, can participate in adverse remodeling, contributing to the development of heart failure (27,28). In line with this finding, we observed that the STZ injection led to an increase in nitrotyrosin protein expression levels, suggesting an important role of oxidative stress in the pathogenesis of diabetic cardiomyopathy.

Previous studies have reported the effects of exogenous insulin therapy. It has also been shown that insulin treatment alone cannot normalize heart function under diabetic conditions (29–31). However, future studies should include a third experimental group with insulin treatment.

In conclusion, the current study displayed the cardiac phenotype of rats with STZ-induced diabetes rats in a time course analysis. The induction of diabetes by STZ led to an impairment in systolic and diastolic LV function, associated with an increase in immune cell invasion and adverse cardiac remodeling. This study reveals an important role of the maintenance of cardiac structure, by regulating the ECM assembly, in diabetic cardiomyopathy. We hope that these new findings of cardiac performance and remodeling will increase our understanding of the pathophysiology and development of diabetic cardiomyopathy.

Acknowledgements

We thank Kerstin Puhl, Georg Zingler and Nadine Orrin for their excellent technical assistance. This study was funded by FP7-Health-2010, MEDIA (261409).

References