of work: We have continued our work on the cell biological mechanisms involved in activity-dependent, Hebbian synapse elimination in an in vitro neuromuscular synaptic system. Previous experiments showed that conjoint but opposing actions of protein kinase C (PKC) and protein kinase A (PKA) were essential for the stimulus specific loss of unstimulated inputs to target cells activated by other inputs. This years findings include: 1) Collaborative experiments in the rat or mouse in vivo confirm the importance of PKC to the process of synapse elimination that occurs post-natally in the intact animal. Some dissociation between postsynaptic receptor loss and neurite retraction was evident in vivo, although a general correlation between these two processes was evident. Mice in which the theta isoform of PKC was knocked out show a delay in the synapse elimination that occurs at the neuromuscular junction, but eventually loss of multiple innervation does occur. 2) When synapses form in vitro between nerve and muscle from PKC theta k.O. animals, stimulation of PKC no longer produces synapse loss. 3) When a PKC activating phorbol ester such as PMA is placed in the center, synaptic chamber of our 3 compartment system, we see loss of synapses as expected if PKC acted in the muscle. Similarly, a PKC blocker is effective if applied only in the center, synaptic compartment. Also consistent with a muscle locus of PKC action are the results of experiments in which PKC deficient muscle were combined with normal nerve in the side neuronal chambers. These preparation showed a marked decrement in the synapse elimination produced by PKC activation. More surprising were results with normal muscle and PKC theta knockout nerve. These preparations also showed a marked deficit in PKC induced synapsed loss, entirely comparable to that shown with the muscle K.O., normal nerve combination. This suggests that presynaptic PKC function must be combined with postsynaptic PKC to produce synapse loss. 4) PKA also probably has both pre- and post-synaptic action. PKA mediates the stabilization and strengthening of stimulated inputs and we have shown that post-synaptic injection of PKI, an inhibitor of PKA, in conjunction with electrical activation of the synapse results in a major loss of synaptic connectivity. When a PKA blocker, H-89, is applied to the side chamber only, we also see a major activity-dependent loss of synapses. This loss is due to a decrease in the probability of release of neurotransmitter. This sensitivity to stimulation takes some 20-30 minutes to develop, which we interpret as being the time taken for some PKA dependent material to be transported from the cell body to the synapse where it is needed for maintaining transmitter output. 5) We have examined the possibility that the Glia Derived Neurotrophic Factor (GDNF) may have an effect on synapse stabilization. It has been shown by others that GDNF released from muscle can affect presynaptic function. We have tested whether there may be some effect of GDNF on muscle function, specifically on the acetylcholine receptor (AChR). We find that GDNF treatment of muscle, even in the absence of nerve but also in innervated fibers, increases the rate at which AChR are inserted into receptor clusters. The rate of loss of receptors from the clusters is not affected by GDNF treatment. Some of the cell biological mechanisms by which GDNF is coupled to receptor disposition have been examined. We feel that our results identify some of the critical post-synaptic events mediating Hebbian plasticity, and will be putting some increased focus on possible presynaptic mechanisms in the future.
