טכניון מכון טכנולוגי לישראל
הטכניון מכון טכנולוגי לישראל - בית הספר ללימודי מוסמכים  
Ph.D Thesis
Ph.D StudentPolsky Alon
SubjectSynaptic Integration in Cortical Pyramidal Neurons:
Coordinated Experimental and Modeling Studies
DepartmentDepartment of Medicine
Supervisor Professor Jackie Schiller
Full Thesis textFull thesis text - English Version


Abstract

The main processing devices in the brain are dendrites of principle neurons. They receive process and transform the information into output action potentials (APs). However, little is known on the rules that govern dendritic synaptic integration, intra-dendritic interactions and the local dendritic environment.

In the work presented here we combined theoretical and state-of-the-art elctrophysiological and imaging approaches to examine the inner-workings of dendrites of layer 5 pyramidal neurons. We mainly concentrated on understanding non-linear processing in thin basal dendrites which receive ~80% of all synaptic contacts to the neuron. Our findings indicate that despite the small dimensions of the basal tree, it is electrically remote from the soma. Axo-somatic action potentials significantly attenuate as they propagate to basal dendrite. Furthermore, synaptic potentials significantly attenuate as they propagate from their site of initiation to the soma.

With respect to active non linear processing, we found that basal dendrites are only very weakly excitable with respect to sodium and calcium spikes. The main regenerative event is mediated via the synaptic receptors themselves, the NMDA-R channels.

These NMDA-spikes are generated by coordinated spatial-temporal synaptic activity; on average, about 10-20 synapses can initiate an NMDA-spike as long as they are activated within the distal part of the dendrite and within a time window of about 200msec. We found that every dendrite integrates synaptic inputs to generate NMDA-spikes independently, showing that dendrites can function as semi-autonomous computational subunits. When activated, NMDA-spikes are the key element that controls synaptic plasticity, and are triggers of high frequency somatic firing, which is a special signal that can lead to an effective generation of NMDA-spikes in postsynaptic neurons.

To conclude, these findings greatly expand our understanding of synaptic input integration and dendritic function in one of the most important cortical neurons. The theoretical implications of our findings are profound; previous work had shown that parallel dendritic computation abilities can significantly increase the computation capability of the neuron. The products of this computation can propagate between different cells in the network, conveyed by bursts of APs, creating a new layer of information that can be transmitted between cells. The finding that a small number of precisely activated synapses can have a dramatic effect on the activity of a cell puts a new perspective on the rules of neural connectivity and coordinated activity within the neural network.