| Summary: | <p>The heart plays a crucial role in the circulatory system, pumping blood to and from the body. The pumping action is due to the rhythmic contraction of the muscle cells of the heart (cardiomyocytes). This ‘beating’ is facilitated by ionic transport across the membrane of each cardiomyocyte due to electrophysiological activity. Mathematical models that describe this electrophysiology at the organ, tissue, and cellular level are critical for the prediction of irregular beating of the heart (arrhythmia). Arrhythmias can be initiated by unintended modulations of cardiac electrophysiology by pharmaceutical drugs. Therefore, in the past decade, drug regulatory bodies have implemented the use of cardiac models for cardiac safety assessment of novel drugs. Investigating the effects of the drug on individual ion channels of the heart is an important aspect of such cardiotoxicity assessment initiatives. The <em>L-type calcium current</em> (ICaL) is one such <em>ionic current</em>, key to cardiomyocyte contraction.</p>
<p>This thesis investigates if it is possible to select a model of ICaL that accurately reflects its biology. This was done by first gathering 73 ICaL models and identifying several choices available for individual components of the model. Next, these models were examined using identical experimental conditions thus revealing wide variability in their predictions. To understand this variability, novel experimental data was generated and a pipeline was set up to explore how data collected from a single cell using <em>automated patch-clamp</em> experiments can be used to build an ICaL model. Experimental recordings of ICaL were identified to be contaminated by <em>rundown</em>, which was shown, to be partly due to calcium-dependent inactivation (CDI) of ICaL. A mechanistic model was then used to establish experimental conditions at which CDI-induced rundown can be minimised. Rundown-corrected data was used to develop a preliminary model of ICaL characterising its voltage-dependence. Cell-specific calibrations of this model closely replicated the training data in addition to making predictions about other ICaL properties. Thus, this thesis sets up a method to develop an ICaL model representative of its underlying biology for a particular <em>context of use</em>.</p>
|
|---|