NETWORK OF TUFTED LAYER-5 PYRAMIDAL NEURONS

Tufted layer 5 (TL5) pyramidal neurons are important projection neuron s from the cerebral cortex to subcortical areas. Recent and ongoing ex periments aimed at understanding the computational analysis performed by a network of synaptically connected TL5 neurons are reviewed here. The experiments employed dual and triple whole-cell patch clamp record ings from visually identified and preselected neurons in brain slices of somatosensory cortex of young (14- to 16-day-old) rats. These studi es suggest that a local network of TL5 neurons within a cortical modul e of diameter 300 mu m consists of a few hundred neurons that are exte nsively interconnected with reciprocal feedback from at least first-, second- and third-order target neurons. A statistical analysis of syna ptic innervation suggests that this recurrent network is not randomly arranged and hence each neuron could be functionally unique. Synaptic transmission between these neurons is characterized by use-dependent s ynaptic depression which confers novel properties to this recurrent ne twork of neurons. First, a range of rates of depression for different synaptic connections enable each TL5 neuron to receive a unique mixtur e of information about the average firing rates and the temporally cor related action potential (AP) activity in the population of presynapti c TL5 neurons. Second, each AP generated by any neuron in the network induces a change (defined as an iteration step) in the functional coup ling of the neurons in the network (defined as network configuration). It is proposed that the network configuration is iterated during a st imulus to achieve an optimally orchestrated network response. Hebbian, anti-Hebbian and neuromodulatory-induced modifications of neurotransm itter release probability change the rates of synaptic depression and thereby alter the iteration step size. These data may be important to understand the dynamics of electrical activity within the network.