CDS Thesis Seminar: Dendritic Spines Dance Before They Die: Spine Motility, Cell Adhesion, and Structural Plasticity

Shreesh P. Mysore, Control and Dynamical Systems
California Institute of Technology

Despite stable memories, neuronal synapses are in a constant state of flux both biochemically, and structurally. Dendritic spines are the tiny, mushroom-shaped, postsynaptic sites of most excitatory synapses, and they show a wide range of structural dynamics (descriptively called spine motility) --- they constantly twitch, change size, and are lost and gained, even in adult brains. Recent studies have associated changes in dendritic spines with bidirectional changes in synaptic plasticity. However, little is known about the progression of spine morphological changes leading up to drastic forms of motility --- spine loss and gain --- that can mediate changes in the underlying connectivity pattern between neurons. Also, though spine dynamics have been studied separately at timescales from seconds to days, the relationship between structural changes at different timescales is unknown.

I have developed a unified approach to characterize spine motility, and used it to reliably uncover novel phenomena in, and relatively subtle forms of, spine dynamics in experiments with submicron-resolution, time-lapse, confocal microscopy in cultured rat hippocampal neurons. Using this approach, I have examined the progression of structural changes that culminate in spine loss  and synapse elimination in the context of the disruption of the synaptic adhesion molecule, N-cadherin. I show that interfering with the structural supports of synapses causes spines to first be more motile and to shrink in length, and then to be lost. For the first time, I show that early structural changes can predict later synapse elimination, suggesting that early dynamics may be readouts for future changes in the neural wiring diagram. I also address some of the related mechanistic questions. I then place my data in the broader context of structural plasticity, and briefly mention the computational modeling work I have done on structural plasticity in barn owls. I end with thoughts on the tight structure-function coupling in the brain, and the potential implications of such a coupling for the design of artificially intelligent machines.