Purchase this article with an account.
N.L. Kamiji, A. Ishihara, S. Usui; A Model–Based Ionic Conductance Estimation Method for Retinal Neurons . Invest. Ophthalmol. Vis. Sci. 2005;46(13):3938.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: Conductance changes induced by neuromodulators, neurotransmitters, second messengers and other mechanisms play an important role in information processing within the retina. To clarify such processes, we propose a new parameter optimization method that estimates the ionic conductances of individual neurons without pharmacologically isolating the current. Methods: Each ionic conductance of the recorded neuron was simultaneously estimated to fit voltage– and/or current–clamp responses of a neuron model based on the ionic current mechanism. We first evaluated this new technique with simulated voltage–clamp responses under background white noise. In a second experiment using dissociated bipolar cells of the goldfish retina, we analyzed the ionic conductance changes by comparing the results before and after TEA (tetraethylammonium chloride) and Cs (cesium chloride) application. Results: In the model–based test, conductances estimated from simulated responses were highly reproducible (error 0.04 ±0.04 nS, N=20). In experiments where TEA (35 mM) blocked the delayed rectifier potassium current IKv, was applied, our method demonstrated that the maximum conductance (0.40 ±0.13 mS/cm2, N=5) was significantly decreased by 90.7 ±5.34% (N = 5), as expected. In addition, inhibition of the hyperpolarized activated current Ih by perfusion with Cs (10 mM) also elicited a decrease of 99.8 ±0.30% (N = 4) in the maximum conductance (0.057 ±0.010 mS/cm2, N=4). These results are consistent with previous studies (Kaneko and Tachibana, 1985). Furthermore, simulated responses using the estimated values reproduced the current–voltage characteristics of the recorded responses very well. Conclusions: The proposed parameter tuning method provided a consistent numerical measure of the response to the drug applied. Thus this novel technique, which utilizes a simple experimental protocol in conjunction with computational models, could give additional insights into the mechanism and role of conductance changes induced by neuromodulators and other mechanisms. Furthermore, since the time for estimation is relatively short (< 1 s), this method can be accomplished on–line and in real time for conductance change measurement. Additionally, this method would be useful to analyze detailed neural mechanisms using the tuned model for closed–loop experiments that provide information not accessible by other techniques.
This PDF is available to Subscribers Only