Healthy male Wistar rats, 21 days old, acquired at the animal distribution center of UNICAMP, were used in this study. The experimental protocol was approved by the Ethics Committee for Animal Experimentation of the Biology Institute, IB/UNICAMP (Protocol n˚ 1075-1). During the entire experimental period, the animals were housed in the Animal Facility of the University of Campinas (UNICAMP) under standard conditions of temperature (21˚C ± 2˚C) and humidity (55% ± 10%) with a 12-hour light/dark cycle and food and water provided ad libitum. Animal care were in accordance with basic principles of animal experimentation (Ordinance IB 05/2017 and 18/2016). Body weight and food consumption were measured at the beginning and the end of the experimental period. The control group (C) received a commercial diet (Labina-Purina-Paulínia, SP, Brazil), and the other groups received the commercial diet plus 1.0%, 2.5% and 5.0% MSG (1 g MSG + 99 g food, 2.5 g MSG + 97.5 g food and 5.0 g MSG + 95 g food, respectively) for a period of 70 days. At the end of this period, diabetes was induced in all rats. Then, the animals (weighing 326.0 g ± 16.0 g) were separated into eight groups (n = 6) and maintained on the same diets in metabolic cages for an additional 21 days. The eight groups were as follows: C (nondiabetic/commercial diet), C-MSG 1.0% (nondiabetic/1.0% MSG diet), C-MSG 2.5% (nondiabetic/2.5% MSG diet), C-MSG 5.0% (nondiabetic/5.0% MSG diet); D (diabetic/commercial diet), D-MSG 1.0% (diabetic/1.0% MSG diet); D-MSG 2.5% (diabetic/2.5% MSG diet) and D-MSG 5.0% (diabetic/5.0% MSG diet). Diabetes was induced by streptozotocin (STZ). For this purpose, a freshly prepared solution of STZ (50 mg/kg bw) in 0.1 M citrate buffer, pH 4.5, was injected intraperitoneally into rats that had faster overnight. After 24 hours, each streptozotocin-treated rat was evaluated for hyperglycemia by measuring its glucose concentration using the Accutrend GCT method (Roche Diagnostics®-Berna, Switzerland), and a blood sample was collected from the tail vein. All animals showed marked hyperglycemia (glucose > 200 mg/dl). Insulin (Novolin L 100 IU/mL; Novo Nordisk A/S; Bagsvaerd, Dinamarca) was injected intraperitoneally every two days. The experimental protocol used in this study is shown in Figure 1.

Anesthetized rats (sodium pentobarbital, 40 mg/kg bw) were kept in the supine position with spontaneous breathing for the ECG recording. The electrodes were connected to computer channels (Heart Ware System, Heart Ware International

Inc, Framingham, Massachusetts, USA), and six standard waves (I, II, III, aVR, aVL and aVF) were recorded with 2N sensitivity at a speed of 50 mm/second. The P wave amplitude, PR interval and QT interval were measured for three consecutive beats from the beginning of the QRS complex to the point of the return of the T wave to the isoelectric line, defined as the TP segment. The QT interval was corrected for heart rate using Bazett’s formula [27] .

The hearts of the rats were collected for morphological evaluation at the end of the study. Fragments of the left ventricle were removed, stored in a 37% formaldehyde solution, and subjected to routine histological examination using the hematoxylin-eosin method [28] .

The GraphPad Prism program, version 6.0 (2012), was used for statistical analyses with a random block design according to the animals’ weights. Two-way variance analysis (ANOVA) was used to determine statistical significance, and the Tukey test was used to compare the means of significant values, with a probability of 5%.

The treatment of young adult rats with STZ produces a diabetic state that is characterized by weight loss, polydipsia, polyuria, glucosuria, polyphagia, hypoinsulinemia and hyperglycemia [29] [30] . The pathophysiology of STZ-induced diabetes includes a cardiomyopathy that is frequently associated with contractile dysfunction and heart rhythm disturbances [31] . In this study, diabetes was induced with STZ and confirmed by elevated blood glucose level (491.4 ± 48.8 mg/dL).

The presence of MSG in the diet, at concentrations up to 5.0%, had no effect on the body weight in the nondiabetic or diabetic rats (Figure 2), as previously described [30] . Contradictory results with respect to the relationship between body weight gain and the consumption of an MSG-containing diet have been reported in rats. Hermanussen (2006) [10] reported that the ingestion of MSG increased body weight gain and food consumption. On the other hand, in a study using Sprague-Dawley rats fed with diets of varying caloric density, fat content and carbohydrate content, Kondoh and Torii (2008) [32] reported a significant difference in body weight gain between rats that drank a 1.0% MSG solution and those that drank only water; they concluded that the voluntary ingestion of an MSG solution reduced weight gain and had no effect on food intake.

The sinoatrial node (SA) normally serves as the cardiac pacemaker, initiating electrical impulses and setting the heart rate. There were no significant differences in heart rate (HR) between the control and MSG-fed animals in the nondiabetic and diabetic groups. Nevertheless, the HR of diabetic rats was lower (272.6 ± 14.6 beats/min) than that of nondiabetic rats (329 ± 6.8 beats/min) (Figure 3). These results corroborate those of Howarth et al. (2005) [31] , who reported that the HR declined rapidly after the administration of STZ, reaching a new steady state after 1 week, at which it remained for up to 20 days. A decrease in physical activity may partly underlie this reduction in HR.

The reduction in HR in the STZ-induced diabetic rats may be caused by a prolongation of the action potentials in the sinoatrial node, which in turn may be induced by the altered expression and/or function of the ion channels. In addition, STZ itself may directly contribute to heart rhythm disturbances [33] . Despite these observations, the addition of MSG to the diet at the concentrations studied did not change the action potential generation rates in the cardiac pacemakers in nondiabetic and diabetic rats.

The amplitude of the P wave and the PR interval (Figure 4) were not significantly different between the control and MSG-fed rats in the nondiabetic and diabetic groups. Furthermore, there was no significant difference between nondiabetic and diabetic animals fed with the diets containing MSG. Impulses from the SA node spread out across the atrial muscle and are conducted from one muscle cell to the next. The P wave represents atrial depolarization, and the PR interval includes the P wave and the period of electrical standstill created by the impulse crossing the AV node [22] . Thus, diets containing up to 5.0% MSG did not alter atrial depolarization or the conduction of electrical stimuli from the atria to the ventricle.

The main clinical manifestation of diabetic cardiomyopathy is a higher incidence of cardiac arrhythmias, including ventricular fibrillation, and a higher occurrence of sudden death because of alterations in ventricular repolarization. These alterations may cause changes in cardiovascular physiology and structure, which can be recorded by ECG. One of the most common problems detected in diabetic patients is the prolongation of the QRS complex, QT interval and QTc [34] . The QRS complex represents ventricular depolarization, and it is the largest component of the sinus rhythm complexes because a large voltage is required for ventricular depolarization. A specialized conduction pathway composed of the His bundle, bundle branches, hemi-branches and Purkinje fibers ensures that impulses travel quickly from the AV node to the ventricular muscle. The QT interval, representing the total ventricular depolarization and repolarization time, is measured from the beginning of the Q wave to the end of the T wave and should be less than half the time of the preceding R-R interval. Prolonged QT intervals represent delayed repolarization, which may cause tachydysrhythmias and sudden cardiac death [35] .

The results obtained for the QRS complex, QT interval and QTc are illustrated in Figure 5. There were no differences in any of these parameters between the

control and MSG-fed rats in both the nondiabetic and diabetic groups. However, the rats in the diabetic group showed significant increases in the QRS complex, QT interval and QTc when compared with the rats in the nondiabetic group (P <  0.05), although this result was irrespective of the addition of MSG to the diet. These findings are consistent with the results of previous studies with STZ-induced diabetic rats [31] . A prolonged QRS complex indicates a prolongation of ventricular depolarization, and a prolonged QT interval indicates prolongation of the events between depolarization and repolarization [31] . In fact, the direct actions of STZ may also contribute to these heart rhythm disturbances.

The prevalence of QT prolongation has been reported to be as high as 16% and 26% in patients with type 1 and type II diabetes, respectively. Diabetic patients with more pronounced QT abnormalities tend to be older and have higher blood pressure levels and more cardiovascular complications [36] .

Although prolongation of the QRS complex and a prolonged QT interval could be the result of cardiac hypertrophy [27] , the histological evaluations of the rat hearts did not reveal any significant differences between the nondiabetic and diabetic groups. All groups presented irregular nuclei located in the center of the cells and spaces between the fibers, both of which are typical characteristics of a normal heart (Figure 6 and Figure 7). In addition, the cardiac cells had similar sizes in all the studied groups. Therefore, there was no evidence of cardiac hypertrophy observed in the rats from either the nondiabetic group or the diabetic group.

Given these data, we believe that the changes in the QRS complexes and, consequently, in the QT and QTc intervals in the diabetic animals fed the control diet or the diets containing up to 5.0% MSG are the consequences of hydroelectrolytic alterations that are typical of diabetes. In fact, such electrocardiographic alterations may be explained by the polyuria generally observed in diabetics, which is caused by osmotic diuresis as a result of the increase in glycemia [37] . This excessive urinary loss causes a decrease in potassium and other electrolytes, which could be responsible for the increase in the membrane repolarization period represented by a prolonged QT interval. The increased QT interval may also be the result of changes in voltage-dependent potassium ion channels [27] [31] . In fact, diabetes can alter the magnitude of the potassium channels involved in the repolarization process of the cardiomyocyte membrane [38] , thereby affecting cardiac tissue.