Sarcolemmal and mitochondrial membrane potentials measured ex vivo and in vivo in the heart by pharmacokinetic modelling of [ 99m Tc]sestamibi

Figures

• 

  Figure 1. Molecular structures of the radiotracers A, [99mTc]sestamibi. B, [99mTc]tetrofosmin.

• 

  Figure 2. Model framework Three compartment model describing radiotracer passage through the heart: plasma (cp(t), black), cytosol (cc(t), magenta) and mitochondria (cm(t), green). The bolus input function U(t) enters the plasma compartment via flow (rate constant k f). Exchange across the sarcolemmal membrane is described by k 1 and k‐ 1, and across the mitochondrial membrane by k 2 and k‐ 2. Sarcolemmal and mitochondrial membrane potentials are represented by Em and ΔΨm.

• 

  Figure 3. Example fitting for n = 1 control dataset A, representative time‐activity curve following bolus injection of [99mTc]sestamibi into a control rat heart perfused with Krebs–Henseleit buffer (KHB). Initial least squares fit to experimental dataset. Compartment concentrations are shown in black, magenta and green; the cyan line indicates the best fit. The best fit parameters are used as the initial input vector for the MCMC procedure. B, correlation matrix of the fitting parameters showing parameters that are positively or negatively correlated. C, representative MCMC random walk of the experimental data in (A) showing three independent trajectories (10,000 iterations) for the sarcolemmal (Em) and mitochondrial (ΔΨm) membrane potentials with a different starting vector to assess parameter convergence. D, histograms of the MCMC trajectories in (C). If the chains converge to a stable equilibrium, then the histograms are centred at the same mean value. The burn in period of 1000 datapoints is discarded and the mean ± SD of the MCMC fitting calculated for the remaining 27,000 iterations. This corresponds to n = 1 with associated fitting error.

• 

  Figure 4. Effect of extracellular potassium concentration on sarcolemmal and mitochondrial membrane potentials A, representative time‐activity curve following bolus injection of [99mTc]sestamibi into rat hearts perfused with Krebs–Henseleit buffer (KHB) containing 25 mm K+. The cyan line shows the best model fit; individual compartments are plasma (cp(t), black), cytosol (cc(t), magenta) and mitochondria (cm(t), green). B, MCMC random walk of the experimental data in (A) showing three independent trajectories (10,000 iterations) for the sarcolemmal (Em) and mitochondrial (ΔΨm) membrane potentials. C, histograms of the MCMC trajectories in (B). D, left ventricular developed pressure (LVDP) in hearts perfused with control KHB and increasing [K+]. Data are presented as the mean ± SD. E, mean values of sarcolemmal (Em) and mitochondrial (ΔΨm) membrane potentials derived from MCMC fitting in control (n = 4) and those perfused with hyperkaliaemic buffers (n = 3). Data are presented as the mean ± SD (see text for error estimate). F, values of Em from (E) replotted against extracellular [K+], compared with theoretical predictions from the Goldman–Hodgkin–Katz equation (continuous line) and the…

• 

  Figure 5. Effect of mitochondrial depolarization with CCCP on sarcolemmal and mitochondrial membrane potentials A, representative time activity curve following bolus injection of [99mTc]sestamibi into rat hearts perfused with Krebs–Henseleit buffer (KHB) containing 600 nm CCCP. The cyan line shows the best model fit; individual compartments are plasma (cp(t), black), cytosol (cc(t), magenta) and mitochondria (cm(t), green). B, MCMC random walk of the experimental data in (A) showing three independent trajectories (10,000 iterations) for the sarcolemmal (Em) and mitochondrial (ΔΨm) membrane potentials. C, histograms of the MCMC trajectories in (B). D, left ventricular developed pressure (LVDP) in hearts perfused with control KHB, 300 nm CCCP and 600 nm CCCP. Data are presented as the mean ± SD. E, mean sarcolemmal (Em) and mitochondrial (ΔΨm) potentials derived from MCMC fitting of [99mTc]sestamibi in control (n = 4), 300 nm CCCP (n = 3), 600 nm CCCP (n = 3) and combined 300 nm CCCP + 25 mm K+ (n = 3) conditions. F, mean Em and ΔΨm values obtained with [99mTc]tetrofosmin in control (n = 3), 300 nm CCCP (n = 3) and 600 nm CCCP (n = 3) hearts. Data are presented as the mean ± SD (see text for…

• 

  Figure 6. In vivo estimation of sarcolemmal and mitochondrial membrane potentials A, dynamic time‐series of planar SPECT scintigraphy images following bolus injection of [99mTc]sestamibi in a healthy anaesthetized rat, with a region of interest (ROI) placed over the heart. B, static planar SPECT image acquired over 10 min at 2.5 h post injection. C, representative time‐activity curve for the heart ROI. The cyan line shows the best model fit; plasma (cp(t), black), cytosol (cc(t), magenta) and mitochondria (cm(t), green) are shown separately. D, MCMC random walk of the experimental data in (C) showing three independent trajectories (10,000 iterations) for the sarcolemmal (Em) and mitochondrial (ΔΨm) membrane potentials. E, histograms of the MCMC trajectories in (D). F, mean sarcolemmal and mitochondrial membrane potentials measured in vivo were Em=−61±8mV and ΔΨm=−151±13mV, respectively (n = 4). Data are presented as the mean ± SD (see text for detail of error estimate).