Feline Lindhout

128 4 by decreasing bouton size, lower number of synapses and SVs, and downregulation of proteins involved in the SV cycle machinery. Consistent with this idea, we observed less and smaller boutons in VAP knockdown neurons. In addition to the Ca 2+ -mediated effects, the smooth ER in axons could also modulate SV cycling by controlling lipid homeostasis. The presynaptic membrane is comprised of a unique presynaptic lipid composition that is required for proper overall presynaptic function (Lauwers et al, 2016). Possibly, this presynaptic lipid composition could be facilitated and maintained by enabling lipid delivery at VAP-mediated membrane contact sites. Taken together, it would be interesting to direct future research in exploring the possible roles of VAP-SCRN1 interactions in controlling Ca 2+ homeostasis and presynaptic lipid composition, and on how this could modulate the tightly spatiotemporal controlled SV cycle. Molecular function of VAP-interacting protein SCRN1 It remains unclear how the SCRN1 interaction with VAP at the ER membrane may control the observed phenotypes on ER integrity. The C69 protease domain of SCRN1 did not show proteolytic activity, whereas this was observed for family members SCRN2 and SCRN3. Thus, it is unlikely that the observed VAP-mediated functions involve enzymatic activity of SCRN1. However, we did observe oligomerization of SCRN1, which may hint for a scaffolding function of SCRN1. Possibly, as scaffolding protein, SCRN1 could promote stabilization of VAP interactions at membrane contact sites. Consistent with this, we observed increased stabilization of ER structures uponwild-type SCRN1 expression. This highlights the importance of controlling the endogenous levels of SCRN1 at the ER membrane in order to balance between ER dynamics and stability. This is supported by the observation that endogenous SCRN1 proteins were not robustly localized at ER structures, whereas SCRN1 proteins were fully recruited to VAP at the ER membrane upon elevating VAP and SCRN1 levels. Similarly, it was previously shown that many other proteins containing a FFAT(-like) motif also did not fully coincide with ER structures (Murphy & Levine, 2016). Together, these findings imply that SCRN1, as well as many other FFAT(-like) proteins, undergoes continuous cycles of competitive binding and unbinding to the limited available VAP-binding pockets. As such, we propose that controlling endogenous intracellular SCRN1 levels, and thus competition with other VAP interactors, could be a novel mechanism to tune ER dynamics and subsequently presynaptic function. Moreover, unlike the ubiquitously expressed VAPs, SCRN1 expression is highly enriched in brain tissue. Therefore, SCRN1 may be engaged in controlling VAP- mediated functions in brain tissue specifically. In summary, we propose that VAP-SCRN1 interactions act as a novel control mechanism for dynamic ER remodeling, and consequently Ca 2+ homeostasis and SV cycling. Future work is required to better understand the molecular function of VAP-SCRN1 interactions in mediating ER integrity and Ca 2+ -driven SV cycling. Finally, investigating additional ER-mediated control mechanisms that are engaged in modulating presynaptic function is required to obtain more insights into the precise function of the dynamic and continuous neuronal ER network in controlling SV cycling.