Feline Lindhout

VAP–SCRN1 interaction regulates dynamic endoplasmic reticulum remodeling and presynaptic function 127 4 in generic structural ER proteins such as atlastin-1, reticulon-2, REEP1, and REEP2 are causative for the axonopathy HSP (Hazan et al, 1999; Zhao et al, 2001; Nishimura et al, 2004; Zuchner et al, 2006; Montenegro et al, 2012; Esteves et al, 2014; Yalcin et al, 2017). Together, these findings indicate that the delicate ER network in axons is more sensitive to maintenance defects. Therefore, ubiquitous ER disruptions may result in more profound phenotypes on particularly axonal functions. Here, we demonstrated that the observed ER abrogations with decreased VAP-SCRN1 interactions were accompanied with impaired SV cycling at presynaptic sites. Consistent with this, recent studies in Drosophila neurons showed that loss of ER-shaping protein atlastin or reticulon also resulted in impaired ER structures as well as decreased neurotransmitter release (Summerville et al, 2016; De Gregorio et al, 2017). Together, these findings imply that ER structure and function are engaged in modulating SV cycling. Notably, investigating the direct relation between these phenotypes is challenging, as it is not feasible to specifically isolate the local function of ER at presynaptic sites since all ER membranes and lumen in the cell are continuous. Nevertheless, considering that the neuronal ER network remains fully continuous despite its complex cellular morphology may actually highlight the functional relevance of this continuity for its role at presynaptic sites. Toward this end, it would be interesting to direct future work on investigating whether the continuity of the neuronal ER network is important for the reported ER-mediated functions at presynaptic sites. VAP-SCRN1 interactions modulate presynaptic Ca 2+ dynamics and SV cycling In this study, we reported that the ER defects observed with VAP or SCRN1 depletion are accompanied with reduced presynaptic Ca 2+ influx and SV cycling. Considering that axonal ER structures are exclusively comprised of smooth ER, we hypothesized that defects in maintaining Ca 2+ homeostasis, which is a key function of smooth ER, could provide a mechanistic link between the observed phenotypes on ER integrity and SV cycling. Previous reports already hinted for a correlation between ER-mediated Ca 2+ homeostasis and neurotransmitter release. In a recent study, a feedback loop between ER Ca 2+ concentration, presynaptic Ca 2+ influx, and SV exocytosis was identified in dissociated rodent neurons (de Juan-Sanz et al, 2017). Additionally, in Drosophila HSP models, both impaired ER integrity and affected SV cycling were rescued upon Ca 2+ bath application (Summerville et al, 2016). In this report, we found that loss of VAP-SCRN1 interactions in neurons results in elevated basal Ca 2+ levels at presynaptic sites. This implies that the cytoplasmic-extracellular Ca 2+ concentration gradient is reduced, which could explain the reduction in evoked Ca 2+ response and subsequent decreased SV cycling. Consistently, previous reports in non-neuronal cells showed that VAP-mediated membrane contact sites regulate Ca 2+ homeostasis and that SCRN1 controls Ca 2+ -dependent processes (Way et al, 2002; Lin et al, 2015; Paillusson et al, 2017). It remains poorly understood how prolonged increase in basal Ca 2+ levels leads to reduced SV cycling. We speculate that chronic elevation of basal Ca 2+ levels could result in compensatory responses that may lead to downscaled synaptic strength. This could be accomplished at different levels, e.g.,