KIF1A/UNC-104 Transports ATG-9 to Regulate Neurodevelopment and Autophagy at Synapses

Abstract

Autophagy is a cellular degradation process important for neuronal development and survival. Neurons are highly polarized cells in which autophagosome biogenesis is spatially compartmentalized. The mechanisms and physiological importance of this spatial compartmentalization of autophagy in the neuronal development of living animals are not well understood. Here we determine that, in Caenorhabditis elegans neurons, autophagosomes form near synapses and are required for neurodevelopment. We first determine, through unbiased genetic screens and systematic genetic analyses, that autophagy is required cell autonomously for presynaptic assembly and for axon outgrowth dynamics in specific neurons. We observe autophagosome biogenesis in the axon near synapses, and this localization depends on the synaptic vesicle kinesin, KIF1A/UNC-104. KIF1A/UNC-104 coordinates localized autophagosome formation by regulating the transport of the integral membrane autophagy protein, ATG-9. Our findings indicate that autophagy is spatially regulated in neurons through the transport of ATG-9 by KIF1A/UNC-104 to regulate neurodevelopment.

Figure 1. ATG-9 is required for AIY synaptic vesicle clustering

(A) Schematic of AIY interneurons (red) in the nerve ring (indicated by bracket) in the head of the worm. Reprinted with permission from wormatlas.org (Z. Altun). (B) Distribution of synaptic vesicles (visualized with GFP::RAB-3) in a representative wild type animal and represented in a schematic diagram (C). Throughout the manuscript, Zone 1 corresponds to an asynaptic region of the AIY neurite, Zone 2 (enclosed by dashed box) corresponds to a synaptic rich region in the dorsal turn of the neurite and Zone 3 corresponds to a synaptic region with intermittent presynaptic clusters. We focus the characterization of our phenotypes to Zone 2, as described previously (Colón-Ramos et al., 2007; Stavoe et al., 2012; Stavoe and Colón-Ramos, 2012). (D) Schematic of ATG-9 transmembrane domains with bp564, wy56 and gk421128 lesions and corresponding protein effects indicated. (E) Distribution of synaptic vesicles in an atg-9(wy56) mutant animal and represented in a cartoon diagram (F). (G) Quantification of the AIY presynaptic phenotype in wild type, three atg-9 mutant backgrounds, heterozygous atg-9 animals, atg-9 (wy56/bp564) transheterozygotes, and atg-9(bp464) mutant animals that contain a pan-neuronal (Punc-14) atg-9 rescuing construct. Error bars represent 95% confidence interval. ****, P< 0.0001 between indicated groups by Fisher’s exact test. (H–I) Distribution of synaptic vesicles in atg-9(bp564) (H) and atg-9(gk421128) (I) mutant animals. Each image is a maximal projection of a confocal z-stack; the asterisk denotes the location of the cell body and the dashed box encloses AIY Zone 2. Scale bar (in B for B, E, H, and I), 5 μm.

Figure 2. The autophagy pathway is required for AIY synaptic vesicle clustering

(A) Schematic of the autophagosome biogenesis pathway, including the general stages: induction (red), nucleation (blue), elongation (green), retrieval (yellow), closure, fusion and maturation (cyan). (B) Quantification of the penetrance of the AIY presynaptic defect in wild type and autophagy pathway mutant animals. For all genotypes quantified, n>100 animals. Error bars represent 95% confidence interval. ****, P< 0.0001 between autophagy mutants and wild type by Fisher’s exact test. (C) Quantification of the relative distribution of GFP::RAB-3 in AIY Zone 2 in wild type and representative autophagy pathway mutants. For all genotypes quantified, n>28 neurons. Error bars represent SEM. **, P< 0.01; ****, P< 0.0001 between autophagy mutants and wild type by one-way ANOVA with Tukey’s post-hoc analysis. (D) Distribution of synaptic vesicles (visualized with GFP::RAB-3) in a representative wild type animal. (D′) Inset in (D) is enlarged Zone 2 region from (D). (E–T) Distribution of synaptic vesicles (visualized with GFP::RAB-3) in AIY Zone 2 of atg-9(gk421128) (E), unc-51(e369) (F), epg-9(bp320) (G), atg-13(bp414) (H), epg-8(bp251) (I), lgg-1(bp500) (J), lgg-2(tm5755) (K), lgg-3(tm1642) (L), atg-3(bp412) (M), atg-5(bp484) (N), atg-4.1(gk127286);atg-4.2(gk430078) double mutants (O), atg-16.1(gk668615);atg-16.2(gk145022) double mutants (P), atg-2(bp576) (Q), epg-6(bp242) (R), mtm-3(tm4475) (S), and epg-5(tm3425) (T) mutant animals. Each image is a maximal projection of a confocal z-stack; the asterisk denotes the location of the cell body and the dashed box encloses AIY Zone 2 in (D). Only the AIY Zone 2 region is depicted in (E–T), similar to the region depicted in (D′). Scale bar in D, 5 μm; scale bar in D′ for D′-T, 1 μm. See also Figures S1, S2 and Table S1.

Figure 3. The autophagy pathway acts cell-autonomously to instruct AIY presynaptic assembly

(A) Distribution of synaptic vesicles (visualized with GFP::RAB-3) in a representative wild type animal. (B–C) Distribution of synaptic vesicles (visualized with mCh::RAB-3, pseudocolored green) in an atg-9(bp564) mutant animal (B) and an atg-9(bp564) mutant expressing a panneuronal ATG-9::GFP rescuing construct (C). (D–L) Distribution of synaptic vesicles in AIY Zone 2 in atg-9(bp564) (D), lgg-1(bp500) (G), atg-2(bp576) (J) mutant animals and mutant animals expressing a corresponding rescuing array (E, H, and K). (F, I, L) Quantification of rescue of autophagy mutant animals expressing unstable rescuing transgenes. Error bars represent 95% confidence interval. **, P< 0.01; *, P< 0.05 between indicated groups by Fisher’s exact test. Each image is a maximal projection of a confocal z-stack; in A–C, the asterisk denotes the location of the cell body and the dashed box encloses AIY Zone 2. Scale bar (in A for A–C), 5 μm, (in D for D–E, G–H, J–K), 1 μm. See also Figure S3.

Figure 4. The autophagy pathway is required for active zone assembly and F-actin organization

(A–I) Distribution of synaptic vesicles in AIY Zone 2 (visualized with mCh::RAB-3), and localization of active zones in AIY (visualized with GFP::SYD-1) in wild type (A–C), atg-9(bp564) (D–F) and epg-9(bp320) (G–I) mutant animals. (J–N) F-actin organization in AIY Zone 2 (visualized with UtrCH::GFP) in wild type (J), atg-9(bp564) (K), atg-13(bp414) (L), epg-8(bp251) (M), and atg-2(bp576) (N) mutant animals. (O) Quantification of penetrance of F-actin Zone 2 enrichment defect in AIY. For all genotypes quantified, n>50 animals. Error bars represent 95% confidence interval. ****, P< 0.0001 between mutants and wild type by Fisher’s exact test. Each image is a maximal projection of a confocal z-stack; only AIY Zone 2 is depicted. Scale bar (in A for A–N), 1 μm. See also Figure S4.

Figure 5. Autophagosome biogenesis occurs at AIY presynaptic sites during development

(A–C) Transmission electron micrographs of autophagic vacuole (AV)-like organelles in embryonic neuronal processes of C. elegans during the developmental time of axon outgrowth and synaptogenesis. Images were originally acquired by Richard Durbin (Durbin, 1987) and represent 500-minute-old embryo PR5_(PVPR) (A and B) and 550-minute-old embryo RDE (C). In all images, dashed lines outline cross-section of neurite; arrowheads indicate autophagosomes in neurite. Neurite and AV-like organelle in (A) is pseudocolored blue (neurite) and pink (AV) in (B). (D) Transmitted light image of a C. elegans embryo for reference. (E–F) Embryonic expression of GFP::LGG-1 (E) and GFP::LGG-1(G116A) (F) in AIY in a wild type animal. LGG-1(G116A) is a point mutant incapable of associating with autophagosomes (Mizushima et al., 2010; Zhang et al., 2015). (G–L) Distribution of autophagosomes visualized with GFP::LGG-1 (G, I–L) or GFP::LGG-1(G116A) (H) in AIY in wild type (G, H), atg-9(bp564) (I), atg-2(bp576) (J), epg-5(tm3425) (K) and unc-104(e1265) (L) mutant animals. (M) Quantification of the penetrance of animals with LGG-1 puncta in the neurite of AIY in wild type and mutant backgrounds. For all categories quantified, n>100 animals. Error bars represent 95% confidence interval. *, P< 0.05; ***< 0.001 between mutants and wild type by Fisher’s exact test. (N) Quantification of autophagosome (AV) biogenesis in the AIY neurite (as described in Experimental Procedures) in wild type and mutant backgrounds (n>10 videos per genotype). Error bars represent SEM. ***, P< 0.001 between mutants and wild type by one-way ANOVA with Tukey’s post-hoc analysis. (O–P) Time series depicting autophagosome biogenesis (O) (see also Supplemental Movie S1) and autophagosome retrograde movement (P) (see also Supplemental Movie S4). In E–L and O–P, arrowheads denote the location of LGG-1 puncta in the neurite and arrows denote the location of LGG-1 puncta in the cell body. Each image is a maximal projection of a confocal z-stack. Scale bars (in A for A–B, in C), 200 nm; (in D, in E for E–F, in G for G–L), 5 μm; (in O, in P), 1 μm. See also Figure S5 and Supplemental Movies S1–S4.

Figure 6. ATG-9 transport to AIY presynaptic zones requires KIF1A/UNC-104

(A) Schematic of the nematode nerve ring (green, encircled by a dotted line) in C. elegans. (B) Schematic of the atg-9 genomic region in a wild type genome (top) and in the eGFP CRISPR knock-in (bottom). (C–D) Location of the nerve ring, referenced with a transmitted light image (C) and visualized with GFP::RAB-3 (pseudocolored red) expressed panneuronally with Prab-3 in a wild type animal (D). (E) Visualization of the subcellular localization of panneuronal ATG-9 (ATG-9::GFP rescuing array expressed with Punc-14). (F–K) Distribution of endogenous ATG-9 as visualized by CRISPR insertion of eGFP at the C-terminus of the genomic atg-9 (schematized in B) in adult (F–H) and embryo (I–K) in wild type (G and J) and unc-104(e1265) mutant (H and K) animals. Anterior end of the embryo (head) is indicated with an asterisk in I–K. (L–Q) Distribution of synaptic vesicles (visualized with mCh::RAB-3) (L and O) and ATG-9::GFP (M and P) in AIY (merge in N and Q) in wild type (L–N) and unc-104(e1265) mutant animals (O–Q). In L–Q, the asterisk denotes the location of the AIY cell body; the dashed box encloses AIY Zone 2. Each image is a maximal projection of a confocal z-stack. Scale bar (in D for C–H, in J for I–K and in L for L–Q), 5 μm.

Figure 7. The autophagy pathway regulates the rate of PVD axon outgrowth

(A) Diagram of AIY (dark green) and PVD (orange) neurons in C. elegans. Schematics of PVD axon morphology (red) and presynaptic sites (light green) in wild type (left box) and autophagy mutant animals (right box). (B) Quantification of PVD axon length in wild type and autophagy mutant L4 animals. Error bars represent SEM. ****, P< 0.0001 between mutants and wild type by one-way ANOVA with Tukey’s post-hoc analysis. (C–H) PVD morphology (visualized with cytoplasmic mCh) and presynapses (visualized with SAD-1::GFP) in the axon of PVD in wild type (C), atg-9(bp564) (D), atg-13(bp414) (E), epg-8(bp251) (F), lgg-1(bp500) (G), and unc-104(e1265) (H) mutant L4 animals. (I–L) Distribution of ATG-9::GFP (I and J) and synaptic vesicles (visualized with mCh::RAB-3) (K and L) in PVD in wild type (I and K) and unc-104(e1265) mutant animals (J and L). Arrowheads denote the tip of the axon in wild type animals (I and K). Each image is a maximal projection of a confocal z-stack; the asterisk denotes the location of the cell body; the bracket denotes length of the PVD axon; and the dotted bracket (J and L) denotes the PVD axon (not visible due to absence of presynaptic protein localization). Scale bar (in C for C–L), 5 μm. See also Figure S6.