Nicotinamide attenuates streptozotocin-induced diabetes complications and increases survival rate in rats: role of autonomic nervous system

This research was carried out in accordance with the ARRIVE procedures for animal experiments [17]. All experimental procedures were approved by the Ethics Committee for Animal Experimentation from the University of São Paulo Medical School (protocol no. 338/12). This investigation was conducted in accordance with the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health. Male Wistar rats (330–390 g) were housed in collective polycarbonate cages (four animals per cage) with a controlled temperature (22 °C) and a 12:12-h light-dark cycle. Rats were randomly divided into 4 four groups (6–8 rats/group): control (Cont), diabetic (Diab), control pre-treated with nicotinamide (Cont+NicA), and diabetic pre-treated with nicotinamide (Diab+NicA).

On the day of diabetes induction, all rats underwent a 6-h fasting. Cont+NicA and Diab+NicA rats received a single intraperitoneal injection of 100 mg/kg of nicotinamide (Sigma-Aldrich) 15 min before diabetes induction [18, 19]. Diab and Diab+NicA rats were made diabetic by a single injection of freshly prepared STZ (50 mg/kg body weight) in citrate buffer (0.1 M, pH 4.5) in the tail vein [20]. Control animals received an intraperitoneal injection of NaCl 0.9% and citrate buffer without STZ. After 15 days of induction, glycemia was measured and only animals with levels higher than 180 mg/dl were included in the study. All groups were followed-up for 5 weeks.

All animals were weighed once per week throughout the protocol period. To determine the body mass index, Lee index was calculated by cube root of body weight (g) × 10/naso-anal length (mm) [21, 22].

At the beginning and at the end of the protocol, all animals underwent 4 h of fasting. A drop of blood from the tail vein was collected to measure plasma glucose using a glucometer, and another drop was collected for triglyceride measurement with Accu-Chek e Accutrend GTC (Hoffman-La Roche Ltd., IN, EUA), respectively.

ITT was performed at the end of the protocol. Baseline glucose levels were determined according to Furuya et al. 2003. Insulin (0.75 units/kg) was injected subcutaneously. Then, blood samples were collected from the tail vein at 0 (just before the insulin injection), 4, 8, 12 and 16 min (after the insulin injection) for the glucose assay. Plasma glucose was measured from blood samples collected from the tail vein with the use of a glucometer (Accu-Check; Hoffman-La Roche Ltd.) [23‐25]. Glycemia values for 4 to 16 min were used to calculate the constant decrease of plasma glucose (kITT) [26].

Echocardiographic evaluations were blindly performed by an expert and in accordance to the guidelines of the American Society of Echocardiography. Rats were anaesthetized (50 mg/kg ketamine and 12 mg/kg xylazine, intraperitoneal injection) and images were obtained with a 10–14 mHz linear transducer in a SEQUOIA 512 (ACUSON Corporation, Mountain View, CA, USA) for measurements of left ventricular function and morphology. Echocardiographic parameters were measured as previously described in detail elsewhere [27, 28].

One day after the echocardiographic evaluation, two polyethylene catheters filled with 0.06 ml of saline were implanted into the femoral artery and femoral vein (PE-10). Rats were anesthetized with isoflurane (1 a 2.5%), for direct measurements of arterial pressure and drug administration, respectively. After this procedure, the animals were allowed to recover in individual cages. The rats were taken to the recording room at least 30 min before the beginning of the experiment, and a quiet environment was maintained to avoid any stress. On the next day, the arterial cannula was connected to a strain-gauge transducer (Blood Pressure XDCR; Kent Scientific, Torrington, CT), and arterial pressure signals and pulse interval (heart rate) were digitally recorded over a 30-min period in conscious animals, moving freely during the experiments using a data acquisition system (WinDaq, 2 kHz, DATAQ, Springfield, OH, USA) as previously described [21, 29]. Beat-by-beat time series of systolic, diastolic, and mean arterial pressures were generated. Heart rate was measured from successive diastolic pulse intervals. This basal acquisition was used to evaluate heart rate variability and systolic arterial pressure variability.

For baroreflex sensitivity evaluation, sequential bolus injections of increasing doses of phenylephrine (0.25 to 32 μg/kg) and sodium nitroprusside (0.05 to 1.6 μg/kg) were given to increases or decreases in mean arterial pressure responses (for each drug), ranging from 5 to 40 mmHg.32. Peak increases or decreases in mean arterial pressure after phenylephrine or sodium nitroprusside injection and the corresponding peak reflex changes in heart rate were recorded for each dose of the drug. A 3–5 min interval between doses was necessary for arterial pressure to return to baseline. Baroreflex sensitivity was expressed as bradycardic response and tachycardic response, expressed as bpm/mmHg, as described elsewhere in beats per minute per millimeter of mercury [27, 29, 30]. After the hemodynamic recordings, rats were euthanized by decapitation. The heart, pancreas and gastrocnemius were rapidly removed, rinsed in ice-cold 0.9% NaCl solution and weighed.

The heart rate variability was performed by linear methods in time and frequency domain by Cardioseries® v2.4. Temporal series of pulse interval and systolic blood pressure from baseline recordings were analyzed and total heart rate variability (HRV) and systolic arterial pressure variability (SAPV) were calculated. The mean square root of differences between consecutives PI (RMSSD), an index of parasympathetic modulation in time domain, was also evaluated. For frequency domain, the interpolated wave of these same basal periods were divided in segments of 512 beats with overlap of 50% and processed by Fast Fourier Transform. One spectrum was obtained by each segment and the oscillatory components were quantified in low frequency (LF; 0.20 to 0.75 Hz), which indicates sympathetic modulation predominance, and high frequency (HF, 0.75 to 3.00 Hz), which indicates parasympathetic modulation predominance [31].

At the end of the protocol Cont and Cont+NicA groups presented body weight gain in relation to initial values. In contrast, Diab group presented body weight loss throughout the protocol, and nicotinamide prevented body weight loss in diabetic rats (Fig. 1A). Moreover, diabetic rats presented decreased Lee index when compared to other groups, and nicotinamide in diabetic rats reestablished this parameter (Fig. 1B). Both diabetic groups presented reduction in pancreas weight in relation to non-diabetic rats (p < 0.05) (Cont:1.48 ± 0.11 g; Cont+NicA:1.86 ± 0.21 g; Diab:1.23 ± 0.06 g; Diab+NicA:1.11 ± 0.07 g). Similarly, gastrocnemius muscle weight was also lower in diabetic groups (Diab and Diab+NicA; 1.16 ± 0.10 g and 1.51 ± 0.11 g, respectively) when compared to controls (Cont and Cont+NicA; 2.29 ± 0.19 and 2.04 ± 0.26 g, respectively). Absolute cardiac mass was not different among groups (Cont:1.50 ± 0.05 g; Cont+NicA:1.72 ± 0.08 g; Diab:1.44 ± 0.09 g; Diab+NicA:1.34 ± 0.07 g).

Before induction, all groups had similar glycemia values (week 0) (Fig. 2A). As expected, STZ induced a sharp increase in glycemia (Diab: 541.28 ± 18.68 mg/dl) when compared to non-diabetic rats, and although nicotinamide reduced glycemia in STZ rats (440.87 ± 20.96 mg/dl), these values were still higher than Cont and Cont+NicA groups (110.00 ± 3.48 and 108.50 ± 1.52 mg/dl, respectively). Concomitantly, Diab rats presented an important deficiency in the insulin tolerance test (0.21 ± 0.05 mg/dl/min), which was totally prevented by nicotinamide, as observed in Diab+NicA group (2.30 ± 0.47 mg/dl/min) (Fig. 2B). STZ also induced an increase in triglycerides levels (180.42 ± 34.70 mg/dl), while nicotinamide in STZ-induced rats attenuated this increase (111.00 ± 5.12 mg/dl) (Fig. 2C).

Results from hemodynamic and cardiovascular autonomic modulation are presented in Table 1. Diab rats presented resting bradycardia and decreased mean blood pressure levels when compared to control groups, while Diab+NicA rats presented heart rate and blood pressure values equivalent to non-diabetic groups. Heart rate and blood pressure variabilities were not different among groups. Despite the absence of significant alterations in the LF component, which represents sympathetic predominance in autonomic modulation, Diab group presented reduced absolute HF component, indicating parasympathetic impairment. Positive effect of nicotinamide in diabetic rats was observed in parameters of parasympathetic modulation, as absolute HF component and RMSSD.

Results of baroreflex sensitivity evaluated by heart rate changes to vasoactive drugs revealed a reflex impairment of the bradycardic response in Diab vs. Cont. nicotinamide treatment in STZ rats increased both tachycardic and bradycardic responses in relation to Diab (Fig. 3).

Echocardiography analysis results are shown in Table 2. Left ventricular diameter during systole was not different among groups; however, left ventricular diameter in diastole was diminished in Diab rats vs. Cont. Diabetes did not interfere in interventricular septum thickness in diastole, in posterior wall in diastole, and in relative wall thickness. Functional left ventricular analysis revealed diastolic dysfunction with preserved ejection fraction in Diab rats. Isovolumetric relaxation time was higher in Diab group in relation to Cont group, and Diab+NicA rats showed lower values than Diab. The ratio between E wave (early diastolic transmitral flow velocity) and isovolumetric relaxation time (E/IVRT), which was reduced in Diab group, was similar to Cont in Diab+NicA group. No differences in systolic function parameters were observed among groups.

Mortality among studied rats was evaluated for 5 weeks. Our results showed that nicotinamide was an effective intervention to increase survival rate in STZ-induced diabetic rats. Figure 4 depicts the survival curve for all studied groups.

Pearson’s correlation analysis using all diabetic rats showed strong associations between the index of reflex bradycardia and fasting glycemia (r = 0.75, p = 0.003), and the rate constant for plasma glucose disappearance (kITT) (r = − 0.75, p = 0.003). In addition, RMSSD was also associated with kITT (r = 0.61, p = 0.04). Moreover, body weight gain was associated with both bradycardia and tachycardia indexes with (r = 0.66, p = 0.013; r = 0.59, p = 0.03, respectively), as well as with kITT (r = 0.78, p = 0.001) and RMSSD (r = 0.58, p = 0.03).