Do retinal waves correlate thalamo-cortical activity?
The guiding theory of early visual development posits that retinal waves affect brain development by correlating activity both spatially and temporally. When examined on the relatively slow time scale of calcium dynamics that imaging allows (100s of milliseconds), this appears to be true in visual cortex. What about on the faster time scale of electrical signaling? This paper measured single cell firing in the developing visual cortex, and found that firing is "correlated" in the sense that neuronal firing within a cortical column is constrained, mostly, to the retinal wave. However, within these ~10 second long firing bursts during retinal waves, cells are far less correlated than expected by chance, suggesting activity in early development is actively decorrelated on a short time scale (10s of milliseconds). Firing becomes more correlated as the animal ages. The trajectory is reversed in the LGN, becoming less correlated with age. This is pretty surprising, and not consistent with a simple refinement model of cortical circuit development. This finding triggers some specific questions about our preconceptions of visual development. What's more relevant for functional development - the slow dynamics of calcium influx, or the fast responses of synaptic signaling? What is the significance of the opposite pattern of development between LGN and cortex? Does this mean retinal topography in each must follow different rules? Read the paper here.