1 Volume 26, Issue 7, Pages 862-871 (April 2016)The Conserved VPS-50 Protein Functions in Dense-Core Vesicle Maturation and Acidification and Controls Animal Behavior  Nicolas Paquin, Yasunobu Murata, Allan Froehlich, Daniel T. Omura, Michael Ailion, Corinne L. Pender, Martha Constantine-Paton, H. Robert Horvitz  Current Biology  Volume 26, Issue 7, Pages (April 2016) DOI: /j.cub Copyright © 2016 Elsevier Ltd Terms and Conditions

2 Figure 1 VPS-50 Regulates the Behavioral State of C. elegans(A) The locomotory behavior of C. elegans is modulated by the presence of food and past feeding experience. Well-fed wild-type animals move more slowly on a bacterial lawn (red bars) than in the absence of bacteria (blue bars), and food-deprived animals (green bars) slow even more than well-fed animals [1]. vps-50 mutants (n3925 and n4022) moved as though they were food deprived even when well fed, and this behavioral defect was rescued by a transgene expressing a GFP-tagged wild-type copy of vps-50 from its endogenous promoter. n = 6 plates for the wild-type; n = 3 plates for all other genotypes. Means ± SDs. ns, not significant. (B) vps-50 mutants show normal pumping rates, indicating that they do not have a feeding defect. (C) vps-50 mutants develop at a rate similar to that of wild-type animals. We followed synchronized animals after recovery from L1 arrest. All wild-type and vps-50 mutant worms observed had a developed vulva after 44 hr; 12/12 wild-type animals were gravid after 49 hr, whereas 11/12 vps-50 mutants were gravid after 49 hr. (D) Analyses of the in utero retention time of eggs, assayed by the distribution of the developmental stages of newly laid eggs. vps-50 mutants retained eggs in utero for an abnormally long period of time, as seen by a shift to later stages of their newly laid eggs. The egg-laying defect of vps-50 mutants was rescued by transgenes expressing vps-50 under its endogenous promoter or a pan-neuronal promoter, but not under a body-wall muscle promoter. See also Figure S1. Current Biology  , DOI: ( /j.cub ) Copyright © 2016 Elsevier Ltd Terms and Conditions

3 Figure 2 VPS-50 Functions in Neurons(A) Fluorescence and Nomarski micrographs of regions of C. elegans transgenic animals expressing VPS-50::GFP (nIs388) using the vps-50 promoter. vps-50 was expressed in most, if not all, neurons. Occasional expression was observed in some pharyngeal muscles. VNC, ventral nerve cord. #To enhance visualization, we used unc-104(e1265) mutant animals to increase the fluorescence level in cell bodies. Scale bars, 10 μm. (B) vps-50 functions in neurons to regulate locomotion. vps-50 was expressed pan-neuronally (Prab-3) and in body-wall muscles (Pmyo-3) to identify its site of action. Pan-neuronal expression of vps-50, or of its murine homolog mVps50, rescued the locomotion defect of vps-50(n4022) animals. n = 3 plates for all genotypes. (C) vps-50 functions in cholinergic neurons to regulate locomotion on food. vps-50 was expressed in cholinergic neurons (Punc-17), in GABAergic neurons (Punc-47), and in a subset of sensory neurons (Ptax-2) to identify its site of action. n = 5 plates for the wild-type and vps-50(n4022); n = 3 plates for all other genotypes. (B and C) Significance is defined by comparison to the equivalent state for vps-50(n4022) mutants. Means ± SDs. (D) The localization of VPS-50::GFP to synapse-rich areas of the nerve ring depends on UNC-104/KIF1A, a molecular motor that transports synaptic vesicles and their associated proteins to synapses. Immunohistochemistry against GFP and synaptobrevin (SNB-1) and DAPI staining are shown. The dotted lines indicate the synapse-rich nerve ring, and clusters of neuronal cell bodies are marked. Scale bar, 10 μm. Current Biology  , DOI: ( /j.cub ) Copyright © 2016 Elsevier Ltd Terms and Conditions

4 Figure 3 mVPS50 Is Widely Expressed in Mammalian Brain NeuronsAnti-VPS50 antibodies specifically recognize mVPS50 (see Figure S2 for validation of the anti-mVPS50 antibody). (A) mVPS50 is enriched in brain tissue. Protein extracts from dissected adult mouse tissues were analyzed for mVPS50 expression by immunoblotting. (B) mVPS50 is expressed in most brain regions, with strong expression in cortex and hippocampus. (C) mVPS50 is expressed in mouse cortex throughout postnatal (P) development. Protein extracts from dissected mouse cortex were analyzed for mVPS50 expression at different developmental stages. (D) mVPS50 is expressed broadly in mouse hippocampal neurons. Transgenic mice that express VGAT-Venus in inhibitory neurons were used for immunohistochemical studies of sagittal hippocampal slices [9]. The dentate gyrus and CA3 are shown. The mVPS50 immunostaining overlaps with the neuron-specific marker NeuN, including both Venus-positive inhibitory neurons and Venus-negative excitatory neurons. Scale bar, 100 μm. See also Figure S2. Current Biology  , DOI: ( /j.cub ) Copyright © 2016 Elsevier Ltd Terms and Conditions

5 Figure 4 mVPS50 Associates with Synaptic and Dense-Core Vesicles(A) mVPS50 does not colocalize with the cis-Golgi apparatus (GM130) in mouse primary cultured cortical neurons. (B) mVPS50 partially colocalizes with the trans-Golgi apparatus (Golgin-97) in mouse primary cultured cortical neurons. (C) mVPS50 partially colocalizes with Chromogranin C-containing dense-core vesicles (ChrC) in mouse primary cultured cortical neurons. (D) mVPS50 partially colocalizes with neuropeptide Y-containing dense-core vesicles (NPY) in mouse primary cultured cortical neurons. (A–D) Endogenous mVPS50, GM130, Golgin-97, ChrC, and NPY were detected by immunofluorescence. Scale bars, 10 μm (A and B) and 5 μm (C and D). (E) mVPS50 significantly cofractionates with the synaptic vesicle protein synaptophysin and the neuropeptide Chromogranin C. Extracts from adult mouse cortex were fractionated, and fractions were probed by immunoblotting for mVPS50, the NMDA glutamate receptor subunit GluN1, the postsynaptic density protein PSD-95, the neuropeptide Chromogranin C, the synaptic vesicle membrane protein synaptophysin, the cytoplasmic V-ATPase A and H subunits (the mammalian homolog of C. elegans VHA-15), and the vesicle coat protein clathrin heavy chain. (F) mVPS50 is a soluble protein. The synaptic vesicle and cytosol (LS1) fraction from adult mouse cortex was further fractionated using sucrose gradient centrifugation, and fractions were probed by immunoblotting. See also Figure S3. Current Biology  , DOI: ( /j.cub ) Copyright © 2016 Elsevier Ltd Terms and Conditions

6 Figure 5 vps-50 Disruption Reduces Neuropeptide Levels and Impairs Neuropeptide Processing (A) Modulation of locomotion in response to the presence of food and past feeding experience. Like vps-50 mutants, unc-31 (CADPS2) and rab-2 (Rab2) mutants, which are defective in dense-core vesicle release and maturation, respectively, behave as though they had been food deprived even when well fed; egl-3 (PC2) mutants are more similar to wild-type animals. n = 4 plates for the wild-type; n = 3 plates for all other genotypes. (B) Mutant animals defective in the synthesis of biogenic amines (cat-2: dopamine; tph-1: serotonin; tdc-1: tyramine and octopamine) behave differently when well fed or food deprived, unlike vps-50 mutants. n = 10 plates for the wild-type; n = 3 plates for all other genotypes. (C) vps-50 mutants, like rab-2 mutants, show reduced FLP-3::Venus neuropeptide levels in the C. elegans dorsal nerve cord. unc-31 and egl-3 mutants show elevated neuropeptide levels at synapses. (D) vps-50 mutants do not significantly accumulate FLP-3::Venus neuropeptides in neuronal cell bodies. (E) vps-50 mutants have reduced levels of FLP-3::Venus neuropeptides in coelomocytes. (F) vps-50 mutants do not have a dense-core vesicle maturation defect identical to that of rab-2 mutants based on the dense-core vesicle marker IDA-1::GFP in the C. elegans dorsal nerve cord. (G) vps-50 mutants, like rab-2 mutants, abnormally accumulate IDA-1::GFP in ventral nerve cord cell bodies. (H) vps-50 mutants have reduced levels of the neuropeptide FLP-3 as well as reduced FLP-3 processing. Immunoblots of C. elegans protein extracts showing the levels of processed and unprocessed neuropeptide FLP-3::Venus and the constant levels of the synaptobrevin reporter mCherry::SNB-1. The transgene ceIs61 expressed both FLP-3::Venus and mCherry:SNB-1 from the unc-129 promoter. Mutants for the proprotein convertase 2 homolog EGL-3 had no detectable fully processed FLP-3 neuropeptides. (I) vps-50 mutants show reduced NLP-21::Venus neuropeptide levels in the C. elegans dorsal nerve cord. Scale bars, 5 μm. Bar graphs (A and B), means ± SDs. Bar graphs (C–G and I), n values (number of animals) are indicated on the bars; representative fluorescence micrographs and quantification are shown; means ± SEMs. See also Figure S4. Current Biology  , DOI: ( /j.cub ) Copyright © 2016 Elsevier Ltd Terms and Conditions

7 Figure 6 Disruption of vps-50 in C. elegans or of Its Murine Homolog mVps50 in Mouse Cultured Neurons Similarly Impairs Synaptic Vesicle Acidification (A) VPS-50 can associate with the V-ATPase subunit VHA-15 in the yeast two-hybrid assay. Growth occurred on media lacking histidine (dark triangle). AD, activation domain; DBD, DNA-binding domain. (B) vps-50 mutants have a synaptic vesicle acidification defect. Representative micrographs and quantification of synapto-pHluorin (SpH) fluorescence levels. SpH fluorescence is quenched by acidic pH; thus, increased fluorescence levels correspond to increased pH. Mutants defective in the V-ATPase complex subunit gene unc-32 were used as an acidification-defective control [22]. (C) vps-50 mutants did not have elevated levels of synaptobrevin (SNB-1) at synapses, indicating that the higher SpH (an SNB-1::pHluorin fusion) fluorescence levels observed in vps-50 and unc-32 mutants were not caused by elevated levels of SpH at synapses. Representative fluorescence micrographs and quantification of mCherry::SNB-1 levels at synapses. (D) Knockdown of mVps50 in mouse primary cultured cortical neurons led to a synaptic vesicle acidification defect. Quantification of SypHy (synaptophysin-pHluorin fusion) fluorescence levels with or without knockdown of mVps50. Higher fluorescence levels correspond to higher pH. SypHy expression levels in wild-type and mVps50 knocked-down neurons are similar, as addition of NH4Cl to increase the intravesicular pH to 7.4 led to similar SypHy fluorescence levels in both. (E) vps-50 mutants have a dense-core vesicle acidification defect. pHluorin and mCherry fluorescence intensities were quantified from the FLP-31–5::mCherry::pHluorin reporter. An increased pHluorin/mCherry fluorescence ratio indicates increased pH. Mutants defective in the V-ATPase complex subunit gene unc-32 were used as an acidification-defective control. (F) Knockdown of mVps50 in mouse primary cultured cortical neurons reduces the amount of V-ATPase (V1) subunits A and B in the synaptosomal fraction. Levels of V-ATPase soluble subunits A and B were quantified in total protein extracts and in synaptosomal fractions in wild-type and mVps50 knocked-down primary cultured neurons. n = 6. Scale bars, 5 μm. Bar graphs, n values (number of animals or neurons) are indicated on the bars; means ± SEMs. See also Figure S5. Current Biology  , DOI: ( /j.cub ) Copyright © 2016 Elsevier Ltd Terms and Conditions