Cardiac Autonomic Neuropathy as a Result of Mild Hypercaloric Challenge in Absence of Signs of Diabetes: Modulation by Antidiabetic Drugs

Abstract

Cardiac autonomic neuropathy (CAN) is an early cardiovascular complication of diabetes occurring before metabolic derangement is evident. The cause of CAN remains elusive and cannot be directly linked to hyperglycemia. Recent clinical data report cardioprotective effects of some antidiabetic drugs independent of their hypoglycemic action. Here, we used a rat model receiving limited daily increase in calories from fat (HC diet) to assess whether mild metabolic challenge led to CAN in absence of interfering effects of hyperglycemia, glucose intolerance, or obesity. Rats receiving HC diet for 12 weeks showed reduction in baroreceptor sensitivity and heart rate variability despite lack of change in baseline hemodynamic and cardiovascular structural parameters. Impairment of cardiac autonomic control was accompanied with perivascular adipose inflammation observed as an increased inflammatory cytokine expression, together with increased cardiac oxidative stress, and signaling derangement characteristic of diabetic cardiomyopathy. Two-week treatment with metformin or pioglitazone rectified the autonomic derangement and corrected the molecular changes. Switching rats to normal chow but not to isocaloric amounts of HC for two weeks reversed CAN. As such, we conclude that adipose inflammation due to increased fat intake might underlie development of CAN and, hence, the beneficial effects of metformin and pioglitazone.

Figure 1

Daily calorie intake, body weight variation, and changes in blood glucose levels as a function of time in different treatment groups. (a) HC-fed rats (red) consumed higher calories on a daily basis compared to control rats on normal chow (black). Depicted data are mean ± SEM for eight control and 32 HC-fed rats. Daily calorie intake was significantly higher on most treatment days, P < 0.05 as estimated by two-way ANOVA. (b) No differences in weight gain were detected among the different treatment groups over twelve weeks of treatment. (c) No significant differences in random blood glucose levels were detected among the different treatment groups over twelve weeks of treatment. (d) At the end of the treatment period, different groups responded similarly to an oral glucose load. Data depicted in (b–d) represent mean ± SEM of values obtained from eight rats in each treatment groups.

Figure 2

Noninvasive hemodynamic parameters of different treatment groups. No significant differences were detected among the different treatment groups in SBP (a), DBP (b), MAP (c), or HR (d) at baseline, 4, 10, and 12 weeks of treatment. Depicted data represent mean ± SEM of values obtained from eight rats in each treatment groups.

Figure 3

Echocardiographic parameters representing structural and functional aspects of the left ventricle. No significant differences were detected among the different treatment groups in LVDd (a), LVPWd (b), LV mass (c), EDV (d), or EF (e) at baseline, 10 and 12 weeks of treatment. Depicted data represent mean ± SEM of values obtained from eight rats in each treatment groups. All represented parameters are normalized to body weight (BW).

Figure 4

Representative tracings of the MAP and HR response of control (left) and HC-fed (right) rats to different doses of PE (a) and SNP (b). Vertical scale bars represent HR and MAP as indicated while horizontal scale bars represent time (60 s).

Figure 5

The effect of HC feeding and treatment with metformin or pioglitazone on BRS. (a) The pressor effect of different doses of PE in different treatment groups. ∗ and # denote P < 0.05 versus response at corresponding PE doses in control or HC-fed rats, respectively. Statistical significance was determined by two-way ANOVA followed by Sidak's post hoc test. (b) The reflex bradycardic response to different doses of PE in different treatment groups. (c) Best fit regression lines for the correlation between changes in MAP in response to increasing PE doses and reflex change in HR in different treatment groups. (d) Slope of the best fit regression line of the ΔMAP versus ΔHR relationship representing BRS in response to PE treatment. ∗ and # denote P < 0.05 versus slope in control or HC-fed rats, respectively. Statistical significance was determined by ANOVA followed by Dunnett post hoc test. (e) Best fit regression lines for the correlation between changes in MAP in response to increasing SNP doses and reflex change in HR in different treatment groups. (f) Slope of the best fit regression line of the ΔMAP versus ΔHR relationship representing BRS in response to SNP treatment. Depicted data represent mean ± SEM of values obtained from eight rats in each treatment groups.

Figure 6

Changes in time domain and frequency domain HRV parameters in different treatment groups. (a), (b), (c), and (d) represent changes in SDNN, rMSSD, power spectral density of LF, and power spectral density of HF, respectively, among different treatment groups. ∗ and # denote P < 0.05 versus corresponding values in control or HC-fed rats, respectively. Statistical significance was determined by ANOVA followed by Dunnett post hoc test. Depicted data represent mean ± SEM of values obtained from eight rats in each treatment groups.

Figure 7

Cellular and molecular changes in ventricular tissue in response to HC feeding and the effect of treatment with metformin or pioglitazone. (a) Representative micrographs of histopathological staining, TGF-β immunostaining, and DHE staining of ventricular midsection. Data presented are serial sections obtained from the same tissue and are representative of tissues harvested from four rats in each group. Scale bars are 50 μm. (b) Changes in phosphorylation of Erk1/2 (above) and AMPK (below) in rat ventricles in response to HC feeding and treatment with metformin or pioglitazone. The depicted blots are representative of experiments performed on protein extracts from tissues harvested from four rats in each group. The bar graphs represent comparison of the normalized intensity of the phosphorylated protein bands. ∗ denotes P < 0.05 versus corresponding values in control rats. Statistical significance was determined by ANOVA followed by Dunnett post hoc test.

Figure 8

Cellular and molecular changes in brainstem in response to HC feeding and the effect of treatment with metformin or pioglitazone. (a) Representative micrographs of DHE and TGF-β immunostaining of brainstem sections. Data presented are serial sections obtained from the same tissue and are representative of tissues harvested from three rats in each group. Scale bars are 50 μm. (b) Changes in phosphorylation of Erk1/2 (above) and AMPK (below) in brainstem in response to HC feeding and treatment with metformin or pioglitazone. The depicted blots are representative of experiments performed on protein extracts from tissues harvested from three rats in each group. The bar graphs represent comparison of the normalized intensity of the phosphorylated protein bands.

Figure 9

The effect of HC feeding and treatment with metformin or pioglitazone on inflammatory cytokine expression in perivascular adipose, ventricular, and brainstem tissues. (a) Changes in IL-1β (left) and TNF-α (right) mRNA levels in perivascular adipose tissue. Values were determined in triplicates of mRNA extracts from five different animals. (b) Representative Western blots showing changes in IL-1β expression levels in ventricular tissue (right) and brainstem tissue (left) in response to HC feeding and metformin or pioglitazone treatment. The depicted blots are representative of experiments performed on protein extracts from tissues harvested from three rats in each group. The bar graphs represent comparison of the normalized intensity of the protein bands. ∗ and # denote P < 0.05 versus corresponding values in control or HC-fed rats, respectively. Statistical significance was determined by ANOVA followed by Dunnett post hoc test. Depicted values represent mean ± SEM.

Figure 10

The effects of switching to normal chow (NC) versus isocaloric feeding with HC (HC-L) on BRS and HRV. (a) The pressor effect of different doses of PE. ∗ and # denote P < 0.05 versus response at corresponding PE doses in control or HC-fed rats, respectively. Statistical significance was determined by two-way ANOVA followed by Sidak's post hoc test. (b) The reflex bradycardic response to different doses of PE. (c) Best fit regression lines for the correlation between changes in MAP in response to increasing PE doses and reflex change in HR. (d) Slope of the best fit regression line of the ΔMAP versus ΔHR relationship representing BRS in response to PE treatment. (e) and (f) represent changes in time domain parameters of HRV, SDNN (e), and rMSSD (f). (g) and (h) represent changes in the power spectral density of LF (g) and HF (h). Data depicted represent mean ± SEM of values obtained from eight rats in control and HC groups and five rats in NC and HC-L groups. ∗ and # denote P < 0.05 versus slope in control or HC-fed rats, respectively. Statistical significance was determined by ANOVA followed by Dunnett post hoc test.