Abstract:
After adjusting the parameters of the conductance or properties of late sodium current, transient outward current, slow component of delayed rectifier potassium current, sodium calcium exchange current, as well as the intracellular calcium release and uptake current, the heterogeneous action potential, calcium transient and tension development were re- constructed in Endo, M and Epi cell in human left ventricle. The morphology, amplitude and time courses showed good agreement with the experimental recordings. This modified model suggested that the different APs of the three types of myocytes are based on the different ion current density or ion channel properties. The simulation of the calcium tran- sient predicted that the heterogeneous calcium activities are based both on the different shape of APs and calcium related gene expressions. The mechanical function of Epi, M and Endo cells were described by a tension development model and the time to peak in the tension development curves could explain how the inhomogeneous electrophysiology leads to a more homogenized contraction in left ventricle.Limitations in the Modified ModelTheir are some limitations in the modified model, but the principle behavior of the model is in good agreement with the physiological measurement. Some of the limitations are inherited from the Iyer et al. model, and some are from the modification for the hetero- geneity.Limitation in Experimental DataTo formulate the ion currents in cardiac myocytes, the concrete experimental data must be known, like the current density at different clamped voltages, its frequency dependence and other properties. In cell models for human left ventricular myocytes, though, all the research tried to model the electrophysiology based on more recently experimental data in human ventricle, there were still some currents based on animal experiments, because no human data was available at the moment. For example, in the Iyer et al. model, the current-voltage relations of current INaK was adopted from a formulation given for guinea pig. For the same reason, in this simulation for electromechanical heterogeneity, some currents were adapted according to the canine ventricular experiments, like the adaptation of the INaCa scaling factor in midmyocardial cell, the current density through RyR channels in both Epi and M cells and so on. More data on human left ventricle are needed to accomplish the cell model.Many of the ion currents in the Iyer et al. model were based on voltage clamp experiments in which cloned human cardiac ion channels are heterologously expressed. Therefore, till now there is still no obvious evidence to prove that the cloned ion channels had exactly the same properties as the nature ones [79]. Some modification in this work was also based on the gene expression experiment of the ion channels. More proofs of the relation between existing channel proteins and those who really function in cellular membrane were required to support such modification.Weak Influence of IKs on APDFig. 5.10 showed that IKs had a little influence on action potential duration, that the APD changed only about 24 ms when the total current was blocked. Bosch et al. reported a prolongation of about 140 ms in human ventricular cells when total IKs was blocked [77], which is much longer than the simulation. In the studies of the heterogeneity in ventricle, many experiments have proved that IKs was a very important component of ion currents, which prolonged APD in M cells.The low IKs current density or the influences of other ion currents occurred in repolari- zation might cause smaller effects of IKs on APD. Since the current density of IKr was set to match the experiment of Magyar et al., the conductance of IKs was doubled in the simulation to fit the ratio of IKs : IKr in the same experiment [43]. Even though, the difference of APD in different position of the ventricle was still not as noticeable as in experiments.After the modification of late INa, IK1 and Iup in calcium handling, the APD changed more significantly. So APD in the Iyer et al. model might depend more on such currents and intracellular calcium concentration than IKs.Controversial Influence of INaCa on APDThe heterogeneous distribution of INaCa was firstly proposed by Zygmunt et al. in 2000, who have measured the whole cell current of INaCa in Endo, M and Epi cells in canine left ventricle [46]. INaCa can be triggered both by release of calcium from SR or by rapid removal of external sodium. In their experiments, INaCa was largest in M cell and smallest in Endo cell under both conditions. As conclusion, they have drawn that larger INaCa might contribute to the prolongation of APD in M cell, though weaker IKs and larger late INa may play a more important role in this prolongation. The modification of INaCa in the Priebe-Beuckelmann model confirmed this determination and presented that larger INaCa prolonged the APD [80][81]. In our simulation, largest INaCa in M cell and smallest INaCa in Endo cell were obtained by adjusting the conductance of INaCa, but the influence of the current to action potential duration was controversial compared to the report of Zygmunt et al.. Fig. 5.13 shows the INaCa current and the action potential of three types of cells after modifying the conductance of INaCa. M cell presents a larger INaCa but shorter APD. In further, more experimental data is needed to determine the influences of INaCa on APD accurately.Insignificant Difference of Time to Peak in Calcium TransientCompared with the experiments of Cordeiro et al., there are still some disagreements in the calcium activities and tension development in the simulation. The time to peak of calcium transients is shortest in Endo cell after the modification, but the experimental data shows that it is shortest in subepicardium. The relation of time to peak, as well as the duration of tension development in Endo and M cells is reversed as in the experiment.Lower Computational EfficiencyMost ion currents in the Iyer et al. model were described with Markov chain models, which sustain more variables and smaller calculation steps, compared with other models. In the Iyer et al. model there are 67 variables, whereas the Priebe-Beuckelmann model and the ten Tusscher et al. model consist of only 15 and 16 variables, respectively. This leads to a higher model complexity, namely lower computational efficiency, especially when tissue simulations were accomplished.PerspectivesAll the simulations till now were based on the single cell model. The simulation of the whole ventricular wall can be applied to reconstruct the cellular physiology in a tissue, the excitation propagation and contraction in the ventricular wall. With the simulation in the intact ventricular wall, a more realistic electrophysiology of the myocytes in different layers can be also obtained, because for the single cell model, the cells are always stimu- lated by a preset depolarizing current, while in nature heart they are mostly stimulated by the neighboring cells.The modified Iyer et al. model has directed the focus on the electromechanical hetero- geneity of human left ventricle and succeeded to reconstructed the APs, calcium transients and tension development of Endo, M and Epi cells. It can help to reconstruct the virtual heart with precise properties in different cell types of left ventricle. More research in hetero- geneity should be carried out on the basis of the results. Further more, it helps to study the basis and mechanisms of cardiac diseases like ventricular tachycardia.