Modeling Drosophila motoneurons to examine the functional effect of Na channel splice variants

Abstract

Neurons have diverse electrophysiological characteristics controlled by voltage-gated ion channels. It is not known how much of the diversity of neuronal activity is caused by differential channel gene expression as opposed to alternate splicing of these genes. The contribution of alternate splicing to neural activity and therefore neuronal function can be addressed more easily in invertebrates because of their smaller genome. Specifically, the fruitfly Drosophila melanogaster represents a very powerful molecular genetic model system that has been instrumental for our understanding of early development of the nervous system. Recently in Drosophila, several voltage-gated sodium channel (DmNav) splice variants have been identified [1,2]. Splice variants can be expressed in the oocytes of the South African clawed frog Xenopus Laevis. The expression of these channels in Xenopus oocytes allows the electrophysiological characterization and the construction of computational ion channel models. If these models are built with sufficient detail, the functional effect of the splice variants on neuronal activity can be analyzed. To achieve this, we build a novel computational model of the Drosophila motoneuron. As a first step, we show that repetitive firing in response to current injections can be achieved in a minimal model neuron that includes a well-characterized fast inactivating sodium (DmNav10) channel, and slow (non-inactivating delayed rectifier) and fast (A-type, inactivating) potassium channels. To build this model, we combined potassium channel data recorded from 3rd instar Drosophila larvae motoneurons and sodium channel data recorded from Xenopus oocytes. In the oocytes, recordings contain artifacts of an endogenous calcium-dependent chloride, Ca(Cl), channel [3] and a space clamp problem. The space clamp is caused by the oocytes' large size, which is required for sufficient expression of DmNav splice variants. We address this problem with a spatial model of leak and Ca(Cl) currents in the oocyte (with a similar approach to [4]). Drosophila motoneurons also have a calcium channels and calcium-dependent potassium (BK and SK) channels. Morphological localization of various channels makes modeling a challenge. In summary, we present solutions to several obstacles in modeling fast kinetics of DmNav channels and also putting them together in a full model of a Drosophila motoneuron.