IPN Seminar Speaker: Ruben Tikidji-Hamburyan

Title: Cellular-synaptic mechanism of network desynchronization in unrefined visual thalamus

Dr. Ruben Tikidji-Hamburyan is a senior research scientist in the Department of Physiology & Pharmacology at George Washington University. 

Visual perception requires establishing precise connections between neurons in different brain areas during development. Many of the main computational elements of the visual cortex, such as visual topography and orientation-selectivity, are present at the eye-opening. Formation of these elements requires positional (spatial) information (SI) provided by retinal spontaneous wave-like activity that correlates spiking among neighboring ganglion cells (rGC) and encodes spatial information in inter-neuron spike intervals. The initial formation and refinement of thalamocortical and corticocortical synapses happen in parallel with the thalamus, and therefore, the SI must pass the thalamic network even though thalamic circuits are not fully refined. Foundational work by Butts et al. showed that the SI is encoded by the correlations of rGC at specific slow (>100ms) timescales. However, at this age, thalamocortical relay neurons (TCs) receive 10 to 20 inputs from rGC, which should cause correlation in input currents and, as a result, strong correlations between TCs spikes at the millisecond timescale. To understand how SI is transferred in the developing thalamus, we established a detailed biophysical model calibrated to reproduce neuronal and synaptic dynamics at postnatal day 7 of mouse development. We showed that fast correlations are suppressed by the slow “immature” dynamics of developing neurons in combination with the synaptic currents dominated by slow NMDA receptors. Such fast correlations would be detrimental to cortical development because they reduce the SI in TCs firing. The decorrelation mechanism is a peculiar phenomenon because it suppresses correlations in a specific timescale, leaving informative correlations intact. Moreover, adult neuronal dynamics or synaptic composition removes this suppression, and therefore, accurate replication of synaptic and neuronal dynamics is essential for modeling information processing in the developing visual system.