This work has demonstrated a novel role for Myosin VI in proper synaptic development, vesicle localization, and synaptic transmission at the Drosophila NMJ. Severe jar loss of function mutants exhibited significant defects in larval locomotor behavior prompting our investigations of Myosin VI at the NMJ. Imaging of synaptic morphology revealed that Myosin VI is important for NMJ development: a reduction in NMJ length and a decrease in bouton number were observed at jar loss of function mutant nerve terminals.
Although Myosin VI function in the vesicle cycle has been implicated in mammalian cells, this report provides the first evidence that Myosin VI is important for maintaining normal peripheral vesicle localization at the bouton . In Drosophila, there are four types of boutons, which are the sites of neurotransmitter release at the NMJ, and they differ in their morphological and chemical properties . Of interest for this study were the largest synaptic boutons found at type I axon terminals, which are present at all NMJs of mature larvae . Visualization of synaptotagmin staining using confocal imaging revealed a mislocalization of synaptic vesicles in jar mutant boutons. An increasing number of jar mutant boutons, corresponding to the severity of Myosin VI loss of function, were found to exhibit diffuse staining over the entire bouton area as opposed to the doughnut-shaped staining pattern present in control boutons. Bouton centre occupancy has previously been observed at Drosophila NMJs of larvae lacking synapsin, a phosphoprotein that reversibly associates with vesicles, using FM1-43 loading under low frequencies . EM analysis confirmed that in synapsin knockouts there was a spread of vesicles into the bouton centre, accompanied by a reduction in the size of the reserve pool . Thus, synapsin is thought to function in maintaining the peripheral distribution of vesicles in Ib boutons . Likewise, the unexpected diffuse synaptotagmin staining of jar mutant boutons suggests Myosin VI participates in restricting vesicles to the bouton periphery. It is possible that Myosin VI is functioning as a regulator of the actin cytoskeleton at the synapse. Mutant studies have revealed that the presynaptic actin cytoskeleton is required for proper synaptic morphogenesis . Myosin VI has already been shown to function in regulating the actin cytoskeleton during the process of spermatid individualization, by acting either as a structural cross-linker or as an anchor at the front edge of the actin cone, and during nuclear divisions in the syncytial blastoderm [15, 46]. However, live imaging of actin dynamics at the synaptic boutons revealed no major defects in the actin cytoskeleton at jar loss of function mutant nerve terminals.
To assess whether the morphological defects in vesicle localization observed at jar mutant synapses impact synaptic transmission, electrophysiological assays with different stimulation paradigms were used to recruit vesicles from different functional pools. Our data add to the knowledge of this protein's physiological role at synapses. Myosin VI mutant mouse hippocampal neurons exhibit defects in the internalization of the α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid-type glutamate receptor, responsible for fast glutamatergic transmission, suggesting Myosin VI normally plays a role in AMPAR endocytosis . In addition, basal synaptic transmission is reduced in Myosin VI deficient mouse hippocampal slices compared to wild-type controls . Electrophysiological experiments also indicate that Myosin VI mediates glutamate release induced by brain-derived neurotrophic factor, which is known to modulate synaptic transmission and plasticity in the mammalian central and peripheral nervous system .
Our study is the first to show that Myosin VI's role in synaptic transmission involves mobilization of vesicles from different functional pools, indicating that Myosin VI is important for synaptic plasticity. At the Drosophila NMJ, three pools of vesicles with differential release properties have been identified using FM1-43 staining loaded by various stimulation protocols (, reviewed by ). The immediately releasable pool (IRP), representing approximately 1% of all vesicles at the NMJ, consists of vesicles docked and primed at active zones for immediate release and experiences rapid depletion within a few stimuli [48, 49]. The readily releasable pool (RRP), making up 14 to 19% of all vesicles at the NMJ, is mobilized by moderate stimulation of ≤3 Hz and maintains exo/endocytosis at these stimulation frequencies . The reserve pool (RP) represents the vast majority of vesicles, 80 to 90%, and is mobilized upon depletion of the RRP . Recruitment from the RP occurs with high frequency stimulation of ≥10 Hz . Spontaneous release was reduced in the most severe jar loss of function mutants. Evoked response at 1 Hz stimulation was also reduced in the jar maternal null mutant. Although less severe jar mutants exhibited a significant decrease in bouton number, they did not experience an accompanying reduction in evoked potential amplitude at low frequency stimulation, suggesting that other homeostatic mechanisms are important for maintaining synaptic strength . The impaired synaptic response in the jar maternal null mutant may be due to a reduction in the probability of RRP vesicle release or in RRP size. If Myosin VI functions to anchor synaptic vesicles, it may act on the RRP to ensure vesicles are localized in manner that makes them readily available for release. Thus, in jar maternal null mutants the reduction in EJP amplitude may occur because a significant number of vesicles were displaced from areas of higher probability release. Alternately, RRP pool size may be reduced at jar mutant synapses.
Different synaptic vesicle pool properties, such as rate of recruitment of the RP in response to high frequency stimuli, may translate to changes in short-term synaptic plasticity . The increase in EJP amplitude observed at 10 Hz stimulation in 1 mM Ca2+ saline may be attributable to enhanced mobilization of the RP for jar
/Df(3R)crb87-5 NMJs . Filamentous actin has been implicated in RP mobilization as cytochalasin D, an inhibitor of actin polymerization, has been shown to reduce RP dynamics . This suggests translocation from the RP to the RRP may be mediated by an actin-based myosin motor protein. If Myosin VI functions as a synaptic vesicle tether to regulate recruitment from the RP pool, RP vesicles would be more readily mobilized and transitioned into the RRP upon high frequency stimulation in jar loss of function mutants. Consistent with the idea that RP vesicles were more rapidly incorporated into the RRP, a greater initial depression is observed at jar
/Df(3R)crb87-5 mutant synapses during high frequency stimulation in 10 mM Ca2+ saline corresponding to the depletion of vesicles at high calcium concentrations. Taken together, the data suggest that Myosin VI mediates synaptic transmission and short-term plasticity by regulating the mobilization of synaptic vesicles from different functional pools. In mammalian cells, Myosin VI has been implicated as a mediator of vesicle endocyctosis and has been shown to transport uncoated vesicles through the actin-rich periphery to the early endosome . Our experiments, however, indicate that endocytosis is not likely affected at jar mutant synapses. Typically, endocytotic mutants are unable to maintain synaptic transmission in response to high frequency stimulation [52, 53], whereas our Myosin VI loss of function mutants exhibited enhanced EJP amplitude observed at 10 Hz stimulation in 1 mM Ca2+ saline. Additional experiments are required to confirm that Myosin VI is functioning as a vesicle tether. Fluorescence recovery after photobleaching analysis can be used to examine the effect of Myosin VI on synaptic vesicle mobility. If Myosin VI is functioning as a vesicle tether, synaptic vesicle mobility is expected to be increased in jar mutants compared to controls.