|
There is much interest in understanding the intrinsic biophysical and network mechanisms by which neurons are tuned to the behaviorally relevant characteristics of sensory input. We investigated the effects of KCNQ (M) and small conductance calcium activated (SK) potassium channels on the excitability of electrosensory pyramidal cells in vitro. Our results show that both channels contribute in a quantitatively similar fashion to the medium component of the afterhyperpolarization (AHP) and thus oppose burst firing in these cells. Both channels had qualitatively similar effects and decrease the time constant of spike frequency adaptation. However, M current blockade increased the subthreshold membrane resistance whereas SK channel blockade did not. This indicates that, as seen in other systems, M currents were activated in the subthreshold regime whereas SK channels were only activated by spiking activity. Previous studies have shown that both channels had differential effects of the f-I curve: SK channels had a divisive effect whereas M currents had a subtractive effect. But perhaps the most surprising result is the opposite effects that these channels have on the tuning of pyramidal cells to time varying input. Whereas SK channel blockade increased the tuning to low (<20 Hz) temporal frequencies, M channel blockade instead decreased the tuning to low temporal frequencies. This is contrary to what was predicted from previous modeling studies. In order to understand these results better, we built a mathematical model incorporating both M and SK channels. A detailed analysis of the model's response to time varying inputs revealed the mechanism by which both currents could differentially affect temporal frequency tuning. Blocking M currents led to a change in the membrane resistance and shifted the neuron from the subthreshold to the suprathreshold regime through the shift in rheobase current and thereby shifted the frequency tuning from low to higher frequencies. Instead, blocking SK channels did not alter the rheobase current and instead caused a low frequency (~10 Hz) oscillation in the membrane potential which enhanced the low frequency tuning of the neuron. We then verified these modeling predictions experimentally and found them to be accurate. We conclude that membrane conductances that have quantitatively similar effects on neural excitability and spike frequency adaptation can have opposite consequences on frequency tuning. |
|