If synchronization is to play a role as signature of assemblies, it must be possible to synchronize discharges rapidly because of the constraints set by processing speed.
Early simulation studies that used harmonic oscillators rather than single spiking neurons showed that it may take a few cycles before syn-chronicity is established through phase locking (Konig and Schillen 1991). However, later simulations with spiking neurons revealed that networks of appropriately coupled units can undergo sudden state changes whereby the synchronization of discharges and their oscillatory patterning occur promptly and virtually simultaneously (for review, see Singer et al. 1997).
Very rapid synchronization has been observed in the visual cortex of cats. When neurons were activated by the onset of an appropriately oriented grating, their initial responses were better synchronized than expected from mere stimulus locking (Fries et al. 1997b). Comparison of actual response latencies and immediately preceding fluctuations of the local field potential revealed that the response latency shifted as a function of the polarity of the preceding field potential fluctuation. Because these fluctuations were not independent between the different recording sites, response latencies became synchronized. Thus, coordinated fluctuations of excitability act like a dynamic filter and cause a virtually instantaneous synchronization of the very first discharges of responses (Fries et al. 1997b). Since the spatiotemporal patterns of these fluctuations reflect the architecture of intracortical association connections, grouping by synchronization can be extremely fast and still occur as a function of the prewired associational dispositions of the cortical network.
Evidence suggests that an oscillatory patterning of responses may be instrumental for the internal synchronization of neurons, in particular when interactions comprise substantial conduction delays or occur across polysynaptic pathways (König et al. 1995). In vitro experiments in slices of the visual cortex support this conjecture, showing that subthreshold oscillatory modulation of the membrane potential is ideally suited to establish synchronization (Volgushev et al. 1998). In cells with oscillating membrane potential, responses can become delayed considerably, whereby the maximally possible delay interval depends on oscillation frequency and can amount to nearly the duration of one cycle. With such a mechanism, responses to temporally dispersed EPSPs can become synchronized within less than an oscillation cycle in cells exhibiting coherent fluctuations of their membrane potential.
Was this article helpful?