Max Planck Institute for Dynamics and Self-Organization -- Department for Nonlinear Dynamics and Network Dynamics Group
Personal tools
Log in

Neuronal sodium channels: surface density and kinetics

Models of Na+-channels form the basis of our biophysical understanding of action potential (AP) initiation and hence of information encoding. Due to this important role, Na+-channels have been studied for decades1. However, even widely accepted and used Na+-channel models do not pass a fundamental test of consistency (Fig. 1): the current, which a model generates in response to a recorded somatic AP waveform, should be compatible with the notion that the AP upstroke is mainly shaped by the somatic Na+-current: Any large discrepancy between V˙mem (Vmem)  and INa(Vmem )  during the upstroke suggests a faulty model (Fig.1).

This promted us to develop a new model, derived from our own high-resolution measurements, the first measurements to combine minimally invasive cell-attached recordings at single channel resolution with appropriate time resolution necessary to resolve the kinetics of Na+-channel activation (Fig. 2A). We obtained a Markov model for Na+-channel gating that reliably describes the currents evoked by voltage steps (Fig. 2B). Importantly, the model also passes the consistency test (Fig. 2C), although it has not been optimized for this task at all.

The acurate description of somatic Na+-channels (Figure2C) even allows for a quantitative interpretation of the dynamics of action potential waveform. Having directly measured most of the parameters in this equation: , a strict lower bound for the Na+-channel density Nchan/A of 10 channels per can be derived; about twice as much as previously thought. A more elaborate estimate yields even

30 Na+-channels per . We currently use the acurate model and these somatic channel density to obtain a mechanistic understanding of information encodig capabilities of neurons.


Figure 1: The Na+-channel model [2] used in the BlueBrainProject is driven with a recorded AP (phaseplot in black). The predicted somatic Na+-current (blue) is inconsistent with the AP waveform, indicating a faulty model.


Figure 2: A Single channel resolution yields single channel current amplitudes iNa(V )  and open probability Popen B Modelled waveforms (gray) closely match measured (black) activation kinetics (bottom) and inactivation properties (top) C The new model passes the consistency test (see Figure 1).


1. Hodgkin and Huxley [1] pioneered the combination of quantitative analysis of unclamped action potential plus voltage-clamp characterization of selective conductances. Intricate morphology and diverse ion channel populations vastly complicate equivalent studies in mammalian cortical neurons.


[1]    A. L. Hodgkin and A. F. Huxley, J Physiol 117, 4 500 (1952)

[2]    C. M. Colbert and E. Pan, Nat Neurosci 5, 6 533 (2002)

Members working within this Project:

 Andreas Neef 
 Fred Wolf