use your browser's BACK button to return to
last screen once you are finished reading abstract


Poirazi, Panayiota
Coauthors(s): Panayiota Poirazi Bartlett Mel Department of Biomedical Engineering University of Southern California
USC
Biomedical Engineering
Mail Code 1451 USC Los Angeles, CA 90089
lnc.usc.edu


Sublinear vs. Superlinear Synaptic Integration? Tales of a Duplicitous Active Current

Hippocampal pyramidal cells contain a high concentration of A-type K+ channels in their dendrites. This fast inactivating outward current can be viewed as the mirror image of the voltage-dependent Na+ channel, in that it activates transiently to oppose rapid depolarizations, blunting the peaks of EPSP's and spikes which occur in its presence. A-channels are known, for example, to be responsible for the decremental back-propagation of action potentials in CA1 cell dendrites. The A-current has also been implicated with regard to the sublinear summation of EPSP's in both CA1 and CA3 pyramidal cells in vitro (Cash & Yuste, 1998, Urban & Barrionuevo, 1998). The summation of two simple EPSP's in a cell otherwise at rest is likely to be a poor model for the process of synaptic integration in vivo, where synaptic inputs are commonly activated at high frequencies on dendritic branches whose repertoire of active responses (e.g. calcium/sodium/NMDA spikes) are not likely to come into full play in response to punctate test inputs. For this reason, the contribution of any given active current to synaptic integration in vivo cannot be easily extrapolated from the highly stylized stimulus conditions used in typical in vitro experiments. At worst, the functional role ascribed to a given channel could be diametrically opposed in the two cases. Using a biophysically detailed compartmental model of a CA1 pyramidal cell, we contrasted the influence of the A-current and the NMDA current on the responses of a thin dendritic branch to either: (1) an increasing number of synapses on the branch stimulated once synchronously, and (2) an increasing number of synapses on the branch stimulated asynchronously at 100 Hz. We found that whereas the A-current imposed a sublinear increase in branch response (i.e. peak depolarization) for single test pulses, it led also to a powerful --expansive-- nonlinearity in the branch response to high-frequency synaptic inputs. In contrast, we found that the NMDA current counteracted the compressive nonlinearity provided by the A-current for single test pulses, and in the case of high-frequency synaptic input, led to a much steeper slope in the branch input-output nonlinearity. We conclude that, whereas the nonlinear boosting effects of the NMDA channel are consistent for simple vs. complex synaptic stimuli, the A-current can contribute to either a compressive or an expansive synaptic integration nonlinearity depending on stimulus conditions.