Autophagosomal Syntaxin17-dependent lysosomal degradation maintains neuronal function in Drosophila

Abstract

During autophagy, phagophores capture portions of cytoplasm and form double-membrane autophagosomes to deliver cargo for lysosomal degradation. How autophagosomes gain competence to fuse with late endosomes and lysosomes is not known. In this paper, we show that Syntaxin17 is recruited to the outer membrane of autophagosomes to mediate fusion through its interactions with ubisnap (SNAP-29) and VAMP7 in Drosophila melanogaster. Loss of these genes results in accumulation of autophagosomes and a block of autolysosomal degradation during basal, starvation-induced, and developmental autophagy. Viable Syntaxin17 mutant adults show large-scale accumulation of autophagosomes in neurons, severe locomotion defects, and premature death. These mutant phenotypes cannot be rescued by neuron-specific inhibition of caspases, suggesting that caspase activation and cell death do not play a major role in brain dysfunction. Our findings reveal the molecular mechanism underlying autophagosomal fusion events and show that lysosomal degradation and recycling of sequestered autophagosome content is crucial to maintain proper functioning of the nervous system.

Figure 1.

Syx17, usnp, and VAMP7 are required for autophagy in Drosophila. (A–C) Syx17 (A), usnp (B), or VAMP7 (C) depletion in GFP-marked fat cell clones leads to formation of numerous small, mostly perinuclear mCherry-Atg8a dots, unlike the larger, brighter, evenly distributed punctae in surrounding control cells of starved larvae. (D–F) Knockdown of Syx17 (D), usnp (E), or VAMP7 (F) in LAMP1-GFP–marked cells blocks starvation-induced punctate LysoTracker red (LTR) staining. (G–J) Starvation leads to formation of LTR dots in control larvae (G). No LTR punctae form in starved Syx17 mutants (H), whereas LTR staining is restored in mutants expressing a Syx17 transgene (I). No LTR dots appear in VAMP7 mutants (J). (K) Expression of a second, independent usnp RNAi transgene also blocks LTR puncta formation. (L–O) Tandem-tagged mCherry-GFP-Atg8a is transported to autolysosomes that appear as mCherry-positive puncta (magenta in L) in control cells of starved larvae. Silencing of Syx17 (M), usnp (N), or VAMP7 (O) results in the formation of numerous dots positive for both mCherry and GFP (white). Dot plots in L′–O′ show intensity and colocalization profiles of mCherry and GFP dots. Pearson correlation coefficients shown at the top indicate strong colocalization of GFP and mCherry in M′–O′. (P–R) Quantification of data presented in A–C (P), D–F and K (Q), and G–J (R); n = 10 for all genotypes. mCh, mCherry. Error bars mark SDs. *, P < 0.05; ***, P < 0.001. Bar, 20 µm.

Figure 2.

Autophagosomes accumulate upon loss of Syx17, usnp, or VAMP7. (A–C) Increased numbers of Atg8a-positive autophagosomes are seen in Syx17 mutant (A), usnp (B), and VAMP7 (C) RNAi cells in starved larvae. (D–F) p62 aggregates accumulate in Syx17 mutant (D), usnp (E), and VAMP7 (F) RNAi cells in starved larvae. Note that Syx17 mutant cells are marked by lack of GFP expression in A and D, whereas RNAi cells express LAMP1-GFP in B, C, E, and F. (G) Western blots show increased autophagosome-associated, lipidated Atg8a-II levels in starved Syx17 and VAMP7 mutant larvae compared with controls. Accumulation of p62 in Syx17 and VAMP7 mutants is comparable to Atg7 mutants that are unable to lipidate Atg8a. Both the 34- and 40-kD isoforms of Syx17 disappear in Syx17[LL] mutants based on rat anti-Syx17 immunoblots, whereas faint bands are visible in Syx17[WH] mutant larvae. (H) Atg8a-II and p62 accumulate in well-fed Syx17 mutant adults compared with controls or Syx16 mutants (an additional control). Both Syx17-specific bands are missing from Syx17 mutant adults. (I–L) Numerous large autolysosomes (AL) and few double-membrane autophagosomes (arrowheads) are visible in ultrastructural images of control fat body cells from starved animals (I). Loss of Syx17 (J), usnp (K), or VAMP7 (L) function leads to accumulation of autophagosomes and lack of autolysosomes. (M–O) Quantification of data presented in A–C (M), D–F (N), and I–L (O). AP, autophagosome. n = 10 for A–F, and n = 4 for I–L. Error bars mark SDs. ***, P < 0.001. Bars: (A–F) 20 µm; (I–L) 1 µm.

Figure 3.

Syx17 binds to usnp and VAMP7. (A) Coimmunoprecipitation shows that Syx17-FLAG binds to both HA-usnp and HA-VAMP7 in cultured Drosophila cells. Note that usnp facilitates the interaction of overexpressed VAMP7 with Syx17 (both lacking transmembrane domains). (B) Endogenous usnp coimmunoprecipitates with endogenous Syx17. (C) Endogenous Syx17 coimmunoprecipitates with endogenous usnp. gp, guinea pig; IP, immunoprecipitation; IB, immunoblotting.

Figure 4.

Syx17 is recruited to completed autophagosomes. (A and B) Endogenous Syx17 colocalizes with both endogenous (A) and GFP-tagged (B) Atg8a in starved fat body cells. (C and D) No colocalization is observed between Syx17 and the phagophore marker Atg5 (C) or late endosomes and lysosomes, labeled by GFP-dLAMP (D). (E) Syx17 and Atg8a do not colocalize in Atg2 mutants that accumulate Atg8a-positive stalled phagophores. Insets show merged images (top), Syx17 channels (middle), and relevant green channels (bottom) enlarged from boxed areas in A–E. (F) Immunogold labeling reveals that Syx17 is associated with the outer membrane of autophagosomes (AP). Bars: (A and C–E) 20 µm; (B) 20 µm; (F) 100 nm.

Figure 5.

Impaired autophagosome maturation leads to locomotion defects in 2-d-old adult flies. (A and B) Atg8a-positive autophagosomes and p62 aggregates accumulate in Syx17 mutant brains (B) compared with similarly aged controls (A). (C–E) No autophagosomes are found by EM in neurons of control adult flies (C). Large-scale accumulation of autophagosomes (arrowheads) is obvious in Syx17 mutant neurons (D). Autophagosome accumulation is rescued in Syx17 mutant neurons by transgenic expression of Syx17 (E). (F and G) Quantification of data presented in A and B (F) and C–E (G); n = 9 for A and B, and n = 4 for C–E. Error bars mark SDs. ***, P < 0.001. (H) Syx17 mutant adults perform poor compared with controls in climbing tests. Expression of Syx17 in mutants rescues locomotion defects. Neuron-specific expression of caspase inhibitors p35 or DIAP1 have no influence on climbing performance of adult flies, and these do not rescue the defects of Syx17 mutant adults. n = 90 for all genotypes. Error bars mark SDs. ***, P < 0.001. (I–K) TUNEL assays reveal apoptotic DNA fragmentation in Syx17 mutant brains (J) compared with controls (I), which is rescued by expression of Syx17 (K). Bars: (A, B, and I–K) 20 µm; (C–E) 1 µm.