The interaction between membrane kinetics and membrane geometry in the transmission of action potentials in non-uniform excitable fibres: a finite element approach
Introduction
Geometric inhomogeneities in axons, such as abrupt or gradual changes in diameter, have been shown to delay (Manor et al., 1991b) or block the propagation of a single action potential (Berkinblit et al., 1970, Goldstein and Rall, 1974, Zhou and Bell, 1994), as well as selectively blocking action potentials in periodic spike trains (Khodorov et al., 1971). Thus, as suggested by a number of authors (see, Khodorov and Timin, 1975, Manor et al., 1991b, Swadlow et al., 1980, and references therein), axonal inhomogeneities may contribute to the signal processing characteristics of axons. Many naturally occurring spike trains, however, are not periodic and so these initial studies can be usefully extended by considering the effect of geometric inhomogeneities on stochastic spike trains. Furthermore, it is well known that membrane kinetics also interact with geometric inhomogeneities to modulate their effect. For example, a particular axonal morphology may allow action potentials to propagate for one kinetics but not another as implied by the work of Parnas and Segev (1979). These results share a common feature—all are inferences drawn from a mathematical model of the axon. This model, based on partial differential equations, is traditionally solved using either finite difference or compartmental methods (e.g. Berkinblit et al., 1970, Cooley and Dodge, 1966, Manor et al., 1991a, Parnas and Segev, 1979). By contrast, finite element methods have been used with great success to solve problems, which are naturally formulated as partial differential equations over a complex geometric region (e.g. Lindsay et al., 2001). The finite element approach offers significant advantages over the traditional methods through the flexible way in which it resolves local inhomogeneities and boundary conditions for models based on partial differential equations. We use a finite element method to solve the partial differential equations describing propagated action potentials along inhomogeneous axons. The accuracy of this method is demonstrated for Hodgkin and Huxley, 1952, Morris and Lecar, 1981 kinetics by comparing the conduction velocity of action potentials in a uniform axon, estimated by two independent methods. For Hodgkin–Huxley kinetics, the ability of the finite element method to predict this speed and reconstruct faultlessly all the known features of the action potential is a severe test of the method.
Earlier work on the influence that local geometric inhomogeneities in axons exert on the propagation of action potentials is enhanced and extended by including a comparative analysis of the interaction between membrane geometry and ‘matched’ Hodgkin–Huxley and Morris–Lecar kinetics. The finite element procedure modified for axonal segments of infinite length, is also used to investigate the impact of local geometric in homogeneities on trains of propagating action potentials, free from the influence of boundary conditions or reflected waves. The statistical properties of stochastic spike trains are shown to be significantly modified by local geometric inhomogeneities, although a Hodgkin–Huxley membrane model is more resistant to these changes than a Morris–Lecar membrane model.
Section snippets
The finite element method
The mathematical model of an axonal segment consists of a cable equation for an excitable axon and two boundary conditions, one for each end. The traditional cable model assumes that the evolution of the transmembrane potential V(x, t) satisfies the second order partial differential equationwhere x measures axial distance along the segment in centimetres, t measures time in milliseconds and A(x), P(x) are, respectively, the segment
Results
The finite element method is applied to three topics concerned with the propagation of action potentials along excitable fibres, namely, velocity profiles of travelling waves approaching changes of axon geometry; the effect of changes of axon geometry1 on the propagation of
Discussion
In this article, two models of membrane kinetics are used to examine how membrane properties and axonal geometry interact to shape the properties of propagated action potentials and, in turn, those of stochastic spike trains. The membrane potential in the first model, due to Hodgkin and Huxley (1952), is activated by a sodium current and restored by a potassium current, aided by sodium inactivation. In the second model, due to Morris and Lecar (1981), calcium plays the role of sodium, but the
References (25)
- et al.
Digital computer solutions for excitation and propagation of the nerve impulse
Biophys. J.
(1966) - et al.
Changes in action potential velocity for changing core conductor geometry
Biophys. J.
(1974) - et al.
Impulses reflected from dorsal root ganglia and from focal nerve injuries
Brain Res.
(1976) - et al.
Propagation of action potentials along complex axonal trees—model and implementation
Biophys. J.
(1991) - et al.
Effect of geometrical irregularities on propagation delay in axonal trees
Biophys. J.
(1991) The leading edge approximation to the nerve axon problem
Bull. Math. Biol.
(1977)- et al.
Study of propagation along nonuniform excitable fibres
Math. Biosci.
(1994) Synaptic facilitation by reflected action potentials: Enhancement of transmission when nerve impulses reverse direction at axon branch points
Proc. Natl. Acad. Sci. USA
(1998)- et al.
Computer investigation of the features of conduction of a nerve impulse along fibres with different degrees of widening
Biofizika
(1970) - et al.
Reflected waves in an inhomogeneous excitable medium
Siam. J. Appl. Math.
(1996)