Synaptic takeover: Coordinated proliferation and elimination of competing synapses imaged in living mice

During mammalian development, synaptic circuitry in the nervous system is refined by the permanent elimination of some synaptic inputs converging on individual postsynaptic cells. This form of plasticity is due to axonal branch trimming rather than cell death. Physiological studies during the developmental period of synaptic refinement have suggested that maintained inputs increase their synaptic strength as losing inputs weaken. But, it is not understood how synaptic potentiation and weakening occur concurrently on the same postsynaptic cell.; This thesis focuses on naturally occurring synaptic refinements at the developing neuromuscular junction, which was chosen because of its unparalleled accessibility and because of the development of transgenic mice that express one or more fluorescent proteins in subsets of motor neurons. By developing a technique to image individual neonatal neuromuscular junctions repeatedly over the course of several days, I was able to monitor synaptic dynamics of two differently colored presynaptic axons converging at the same neuromuscular junction, with the postsynaptic site labeled a third color. Multiply innervated junctions in mice at one week after birth underwent a protracted process that led to the withdrawal of all but one of the inputs within the next several days. The eliminated input generally atrophied but often contained varicosities. In addition, it appeared to resorb into the parent axon but on occasion, it fragmented and remnants of this losing input remained at the junction for many hours. Concurrently, in most junctions, the remaining input grew into regions of the neuromuscular junction that were being vacated, thus conserving the established postsynaptic specialization of the muscle fiber. Hence, in this form of synaptic plasticity, new synapse formation by one axon accompanies synapse loss by the other indicating that these opposing processes are linked. However, synaptic takeover does not appear to be the driving force that causes axon withdrawal because in rare instances synapse loss occurred even though the vacated postsynaptic sites were not taken over by the remaining input. Synaptic takeover seems to be a mechanism permitting large scale and rapid increases in axon terminal area, and hence synaptic strength, without requiring new postsynaptic site formation. This process provides a potential mechanism for simultaneous proliferation and elimination of synapses in the developing brain. Moreover because this process links potentiation and synaptic weakening it is an attractive model for the particular kinds of synaptic changes thought to underlie learning and memory.