Rtnl1 loss causes localized depletion of presynaptic ER lumenal marker. (A) A series of 12 serial ATUMtome 50-nm sections through a ROTO-stained larval bouton, visualized by scanning EM and inverted grayscale images. An example is shown of a cisterna (large arrows in sections 5–10) that appears to be continuous with tubules in the same and neighboring sections (smaller arrows), in a region with few synaptic vesicles. Tubules are identified as darkly staining puncta in successive sections; their lumen is too small to detect using this method (Terasaki, 2018; Yalçın et al., 2017). Scale bar in panel 12, 500 nm. The cisterna highlighted by the large arrow and its attached tubules (arrows in “Side View”) were used to generate a reconstruction, shown on the right and viewed from three different angles. EM section 1 is on the left of the reconstruction, section 12 on the right. Scale bars, 70 nm. Larval genotype is Rtnl1¹⁸/WT. (B) Representative confocal projections showing GFP::HDEL in Type Ib muscle 1 NMJs of WT and Rtnl1¹⁸ larvae. Scale bars, 10 μm. (C) Magnified views of the areas inside broken lines in B. Scale bars, 5 μm. Arrowheads indicate regions with low levels of GFP::HDEL, which we interpret as tubules, and arrows indicate examples of puncta which we interpret as cisternae. (D) Insets show the lines drawn along GFP::HDEL signal to plot intensity profiles. (E) GFP::HDEL intensity across the whole NMJ, within an ROI containing all puncta, and in the NMJ excluding the puncta ROI. Plots show individual larval datapoints and mean ± SEM; intensity levels indicate arbitrary units (au) after normalization to control (WT); intensity values are relative to CD4::tdTom (Han et al., 2011) signal. Larval datapoints were calculated and are shown and compared using Student’s t tests as in Fig. 2. Genotypes are Ib-GAL4, UAS-CD4::tdTom/UAS-BiP::sfGFP::HDEL, in either a Rtnl1⁺ (WT) or Rtnl1¹⁸ background. Transheterozygous Rtnl1¹⁸/Rtnl1¹ mutant larvae also showed a similar phenotype to Rtnl1¹⁸ (Fig. S5 B).