Rule-Based Model Specification with Applications to Motoneuron Dendritic Processing
Shapiro, Nicholas Pabon
MetadataShow full item record
With the recent discoveries of phenomena such as plateau potentials, bistability, and synaptic amplification the focus of motoneuron research has been directed to the dendritic processes giving rise to these latent behaviors. The common consensus is that the mechanism behind bistability (an L-type calcium channel generating a persistent inward current, PIC; Schwindt and Crill 1980, Hounsgaard and Kiehn 1985, 1989) is also responsible for the amplification of synaptic input in motoneurons. However, modeling studies utilizing only calcium-based PICs (Powers 1993, Booth et al. 1997, Elbasinouy et al. 2005) have been unable to reproduce the high degree of synaptic amplification observed in experimental preparations (Prather et al. 2001, Lee et al. 2003, Hultborn et al. 2003). The present work examines a theoretical amplification mechanism (electrotonic compression), based on a sodium PIC of dendritic origin, which acts to supplement the synaptic amplification due to the calcium PIC. The current goal is to test the "goodness-of-fit" of electrotonic compression with established mechanisms and behaviors. The findings of this modeling study support the concept of a dendritic sodium PIC which acts to reduce the attenuation of synaptic currents enroute to the motoneuron soma. Furthermore, it is suggested that the ratiometric expression of ion channels giving rise to this mechanism takes the form of a distribution "rule" applied ubiquitously across the dendritic tree, while the plateau-producing L-type calcium channels undergo a more discretized or regional distribution. This study demonstrates the power inherent to the controlled expansion of morphological complexity in an already complex model. While modeling studies are suitable testbeds for the evaluation of theoretical and/or experimentally intractable facets of physiology, great care and consideration should be given to the specification of models with high dimensionality. With the continual progression of our knowledge-base and computational capabilities, we can expect that more and more empirical observations will find their way into models of increasing complexity wherein the layers of embedded hypotheses are frequently implicit. It is therefore imperative that the neural modeling discipline adopt more rigorous methodologies to both accommodate and rein-in this growing complexity.