The most abundant
isoform in the nervous system, Synaptotagmin 1 is associated with synaptic vesicles and has been proposed to function see more as a Ca2+ sensor for neurotransmitter release (Chapman, 2008). Among Synaptotagmins, Syt4 (Littleton et al., 1999; Vician et al., 1995) occupies an interesting yet poorly understood position. Its expression is regulated by electrical activity (Babity et al., 1997; Vician et al., 1995), it is present in vesicles containing regulators of synaptic plasticity and growth, such as BDNF (Dean et al., 2009), it regulates learning and memory (Ferguson et al., 2001), and in humans the syt4 gene is localized to a locus linked to schizophrenia and bipolar disorder ( Ferguson et al., 2001). At the fly NMJ, spaced stimulation results not only in potentiation of spontaneous neurotransmitter release (Ataman et al., 2008; Yoshihara et al.,
2005) but also in structural changes at presynaptic arbors, the rapid formation of ghost boutons, nascent boutons that have still not developed postsynaptic specializations or recruited postsynaptic proteins (Ataman et al., 2008). However, whether this activity-dependent bouton formation also required Syt4-dependent retrograde signaling AZD5363 price was unknown. Here we demonstrate that retrograde Syt4 function in postsynaptic muscles is required for activity-dependent synaptic growth and that this function depends on exosomal release of Syt4 by presynaptic terminals. To determine whether,
similar to the potentiation of spontaneous release (Barber et al., 2009; Yoshihara et al., 2005), the rapid much formation of ghost boutons in response to spaced stimulation required retrograde signaling, we used an optogenetic approach to inhibit responses in the postsynaptic muscle cell. While body wall muscle preparations bathed in normal saline were stimulated after a spaced stimulation paradigm (Ataman et al., 2008), they were simultaneously hyperpolarized by activating the light-gated Cl− channel Halorhodopsin (NpHR) (Zhang et al., 2007), which was expressed in muscles using the C57-Gal4 driver. Illuminating resting preparations expressing NpHR in muscle resulted in rapid hyperpolarization of the muscle membrane (Figure 1A). Using two electrode voltage clamp, we found that the NpHR current peaked at +46 ± 3.5 nA and decayed to +10.8 ± 1.18 nA within 2 min (n = 10). This was sufficient to induce an ∼50% decrease in the amplitude of evoked excitatory junctional potentials (EJPs; Figures 1B and 1C; see Figure S1A available online; recorded in 0.5 mM Ca2+ saline), probably by shunting the depolarizing current induced by neurotransmitter release. This decrease in EJP amplitude was not due to a leaky UAS-NpHR transgene, because in the absence of Gal4 driver there was no significant change in EJP amplitude (Figure 1C; Figure S1A). A similar result has been previously reported when expressing the EKO K+ channel in muscles (White et al., 2001).