Effects of altered cellular ultrastructure on energy metabolism in diabetic cardiomyopathy: an in silico study

Figures

• 

  Figure 1.. Segmented organelle areas in cardiomyocyte cross-sections and their area fractions. (a) Segmentation of organelle areas in a representative cardiomyocyte cross-section from TEM micrograph of a control animal. Segmentations are overlaid over original transverse micrograph. (b) Segmentation of organelle areas in a representative cross-section from a diabetic animal. (c) Pie charts showing average organelle area fractions in the control and diabetic group. Both groups consist of n = 3 animals, with at least five cross-sections analysed for each animal. The p-values correspond to the null hypothesis that organelle fractions do not change in diabetes.

• 

  Figure 2.. Schematic diagram of the biophysical model of cardiac bioenergetics. The rectangular boxes coloured in cyan represent the state variables of the model. The state variables are molar concentrations of different metabolites present in myofibrils, mitochondrial IMS (suffix - IMS) and mitochondrial matrix (suffix - X), except for ψ which is mitochondrial membrane potential. All other boxes in the figure represent reaction fluxes involving the state variables. Metabolites present in the myofibrils and mitochondrial IMS also partake in diffusive fluxes across the mitochondria–myofibril interface. The electronic supplementary material, table S1 provides expanded names of the state variables and reaction fluxes used in this figure.

• 

  Figure 3.. Model predicted spatial distributions of key metabolic parameters in the myofibrillar regions of representative control and diabetic cross-sections. Each column of the figure presents results from three distinct simulation sets—control mitochondrial metabolism with control ultrastructure (set CC—column 1), diabetic mitochondrial metabolism with control ultrastructure (set CD—column 2) and diabetic mitochondrial metabolism with diabetic ultrastructure (set DD—column 3). The regions belonging to mitochondria, glycogen and t-tubule are shown in black as part of the background. Each row of the figure represents a particular metabolic parameter: (a) distributions of myofibrillar ADP/ATP ratios; (b) distributions of myofibrillar Pi; (c) distributions of myofibrillar VATPase; and (d) distributions of myofibrillar VCK/VATPase ratio.

• 

  Figure 4.. Relationship between cardiac ultrastructure and energy metabolism. The scatterplots show the spatial average and MAD of metabolic parameters in each cross-section as a function of either corresponding average mito/myoglobal ratio or MAD of mito/myolocal ratio. Each scatter plot contains results from simulation set CC (19 control cross-sections), CD (19 control cross-sections) and DD (21 diabetic cross-sections). The legends show the Pearson correlation coefficient (r) separately for each simulation set. The p-value corresponding to the null hypothesis of no correlation is indicated as either *(0.05 ≥ p > 0.01) or **(0.01 ≥ p). The bar plots located next to the scatterplots show the spatially averaged ADP/ATP ratios further averaged over three animals corresponding to each simulation set. The p-values indicated as *(0.05 ≥ p > 0.01) or **(0.01 ≥ p) correspond to the null hypothesis of unchanged value from CC. (a) (i) ADP/ATP ratios in each cross-section and (ii) their simulation averages. (b) (i) Average Pi concentration (µM) in each cross-section and (ii) their simulation averages. (c) (i) Average VATPase (µmol l−1. myofibrils s−1 in each cross-section and (ii) their simulation…