Proceedings of The Physiological Society
University of Oxford (2011) Proc Physiol Soc 23, PC249
Effects of Physiological Parameter Variation in a Computational Rabbit Ventricular Cell Model
P. Gemmell1, K. Burrage1,2, B. Rodriguez1, T. Quinn1,3
1. Computing Laboratory, University of Oxford, Oxford, United Kingdom. 2. Department of Mathematics, Queensland University of Technology, Brisbane, Queensland, Australia. 3. National Heart And Lung Institute, Imperial College London, London, London, United Kingdom.
INTRODUCTION: Computational cardiomyocyte models are provided with a single, deterministic parameter set. This parameter set is composed of variables that are determined either directly by experimentation, or based on phenomenological fitting of several parameters at once, comparing the output of the model to experimental data. Ion current conductances are especially subject to this, as experimental methods are unable to measure them directly, forcing indirect measurement using drug block, which in itself can produce artefacts in the data. Consequently, the parameter set of a model may not reflect physiological reality. Furthermore, a deterministic parameter set does little to address physiological variation between individuals. Thus, we investigated the effect of variation in ion channel peak conductance in a state-of-the-art rabbit ventricular cell model by conducting a simultaneous sweep of peak conductance values ranging from -30 to +30% for six different ion channels (Ito, IKr, IKs, IK1, ICa,L, INaK). METHODS: Using the Nimrod/G grid computing system, each combination of peak conductances was used for a simulation run to steady state (15,625 total simulations). The resulting membrane potential, [Ca2+]i transient, and individual ion currents were analysed to compare action potential and [Ca2+]i transient shape, measure the resulting ion currents, and extract commonly used biomarkers. RESULTS: The results indicate that: (1) there is a non-linear relationship between peak conductance and the corresponding ion channel current, (2) the interaction between the currents makes prediction of the relation between ion channel conductance and the resulting action potential difficult, (3) similar action potentials and [Ca2+]i transients can be achieved using vastly different - but still physiological - combinations of ion channel conductance, and (4) the biomarkers currently used in the literature are not necessarily accurate in predicting the goodness-of-fit of model output to experimental data. CONCLUSION: This investigation implies that conductance parameter variation may be a source of physiological deviation between individuals. Furthermore, the response of ion channel conductance to drug block may be counter-intuitive, as a result of non-linear interactions between various intracellular ion concentrations and ion channel populations.
Where applicable, experiments conform with Society ethical requirements