Dynein light chain 1 is required for autophagy, protein clearance, and cell death in Drosophila

Abstract

Autophagy is a catabolic pathway that is important for turnover of long-lived proteins and organelles, and has been implicated in cell survival, tumor progression, protection from infection, neurodegeneration, and cell death. Autophagy and caspases are required for type II autophagic cell death of Drosophila larval salivary glands during development, but the mechanisms that regulate these degradation pathways are not understood. We conducted a forward genetic screen for genes that are required for salivary gland cell death, and here we describe the identification of Drosophila dynein light chain 1 (ddlc1) as a gene that is required for type II cell death. Autophagy is attenuated in ddlc1 mutants, but caspases are active in these cells. ddlc1 mutant salivary glands develop large fibrillar protein inclusions that stain positive for amyloid-specific dyes and ubiquitin. Ectopic expression of Atg1 is sufficient to induce autophagy, clear protein inclusions, and rescue degradation of ddlc1 mutant salivary glands. Furthermore, ddlc1 mutant larvae have decreased motility, and mutations in ddlc1 enhance the impairment of motility that is observed in a Drosophila model of neurodegenerative disease. Significantly, this decrease in larval motility is associated with decreased clearance of protein with polyglutamine expansion, the accumulation of p62 in neurons and muscles, and fewer synaptic boutons. These results indicate that DDLC1 is required for protein clearance by autophagy that is associated with autophagic cell death and neurodegeneration.

Fig. 1.

ddlc1 is required for salivary gland degradation. (A) Control animals lack salivary glands 12 h AHE (red circles; n = 16). (B and C) ddlc1 mutant animals fail to degrade their salivary glands 12 h AHE, with some being intact (B) and others being fragmented (C) (n = 28). (D and E) Expression of Ddlc1 in salivary glands of ddlc1 mutants rescues the degradation defect phenotype, with a small number having fragmented salivary glands (D) and most having none (E) (n = 22). (F) Percentages of animals with intact, fragmented, and no salivary glands 12 h AHE.

Fig. 2.

ddlc1 mutants have altered caspase activity and autophagy. (A and B) TUNEL assay to detect DNA fragmentation (arrows) in control (A) and ddlc1 mutant (B) salivary glands 2 h AHE (n = 5 animals/genotype). (C) Average caspase activity (±SEM) of pupal lysates collected 4 h after puparium formation and measured by cleavage of fluorogenic caspase-3 substrate Z-DEVD-AMC (n = 3). WT, wild-type Canton S; WT + C.I., Canton S + caspase inhibitor DEVD-CHO; p35, daGAL4 × uas-p35: Ex, precise excision of the P element; ddlc1, ddlc1/Y. (D) Percent animals with salivary gland cell fragments and intact glands in paraffin sections of pupae 12 h AHE. Genotypes: Control (fkhGAL4/+) (n = 16), p35 (fkhGAL4/uas-p35) (n = 13), ddlc1 mutant (ddlc1/Y) (n = 28), and p35 expression in ddlc1 mutant (ddlc1/Y; fkhGAL4/uas-p35) (n = 24). (E and F) Representative fluorescence microscopy images of salivary glands expressing GFP-LC3 (green) and DNA (blue). (E) Control (P-excision; ^+/+; fkhGAL4/uas-GFP-LC3) salivary glands have large numbers of puncta 2 h AHE (n = 20 glands). (F) ddlc1 mutant (ddlc1/Y; ^+/+; fkhGAL4/uas-GFP-LC3) salivary glands have few puncta 2 h AHE (n = 17 glands). (G) Average number of GFP-LC3 puncta (±SEM) in the genotypes shown in E and F and in ddlc1 mutant salivary glands 12 h AHE (n = 23 glands). (H) Average number of GFP-LAMP1 puncta (±SEM) in salivary glands of control (P-excision; ^+/+; +/tub-GFP-LAMP1) 6 h after puparium formation, 2 h AHE, and in ddlc1 mutant (ddlc1/Y; ^+/+; +/tub-GFP-LAMP1) 2 h AHE, and 12 h AHE (n = 21 glands).

Fig. 3.

ddlc1 mutant salivary glands contain protein inclusions that are reduced when autophagy is induced. (A) WT salivary glands 2 h AHE lack inclusions by TEM, whereas (B) ddlc1 mutant salivary glands possess large cytoplasmic inclusions 2 h AHE. (C) WT salivary glands have no thioflavin-S foci 2 h AHE (n = 10 animals), whereas (D) ddlc1 mutant salivary glands possess numerous green thioflavin-S–stained foci 2 h AHE (n = 10 animals). (E) WT salivary glands exhibit no colocalization of Ref(2)P (green) and ubiquitin (red) 2 h AHE (n = 4 glands), whereas (F) Ref(2)P and Ubiquitin are colocalized in ddlc1 mutant salivary glands 2 h AHE (n = 5 glands) (DNA, blue). (G) Expression of Atg1 in ddlc1 mutant salivary glands eliminates thioflavin-S–stained foci (compare with D) (n = 10), and (H) attenuates fibrillar inclusions by TEM. (Scale bars, 1 μm in A, B, and H.)

Fig. 4.

ddlc1 enhances motility and protein clearance defects in a model of SBMA. (A) Average number (±SEM) of grid lines passed by the posterior end of WT larvae (^+/+) or larvae expressing human AR in motor neurons containing 20 glutamines (AR Q20 = w; uas-hAR Q20/d42GAL4), or 52 glutamines (AR Q52 = w; uas-hAR Q52/d42GAL4) and raised on food containing dihydroxytestosterone (DHT; n = 3 × 15 larvae). *P < 0.01 and **P < 0.001, significant difference from control. (B) Average number (±SEM) of grid lines passed by the posterior end of control female sibling (ddlc1/+), ddlc1 mutant (ddlc1/Y), Atg12 RNAi (y w, uas-Atg12-IR/+; d42GAL4/+), AR 52Q-expressing (w; uas-hAR Q52/d42GAL4), or ddlc1 mutant expressing AR 52Q (ddlc1/Y; uas-hAR Q52/d42GAL4) larvae in the absence or presence of DHT (n = 3 × 15 larvae). *P < 0.01 and **P < 0.001, significant difference from ddcl1/+ control; and ***P < 0.001, significant difference from AR Q52 expression + DHT alone. (C–E) Ref(2)P staining (green) in third instar larval motor neurons of (C) WT (n = 8 animals), (D) ddlc1 mutant (n = 8 animals), and (E) ddlc1 mutant expressing hAR Q52 in the presence of DHT (n = 8 animals). (F) Reduced ddlc1 function decreases the turnover of polyQ-expanded AR. Western blots showing the temporal profile of AR Q52 protein level after 1-h pulse of expression. A logarithmic plot of AR Q52/Tubulin ratios was used to determine the line of best fit, and the half-life of AR Q52 was increased 1.73-fold in ddlc1 mutant larvae.

Fig. 5.

ddlc1 mutants accumulate Ref(2)p in muscles and have decreased numbers of synaptic boutons. (A–E) Ref(2)P staining (green) of third instar larval muscles of (A) WT (n = 8 animals), (B) ddlc1 mutant (n = 8 animals), (C) ddlc1 mutant larvae expressing AR Q52 in the presence of dihydroxytestosterone (DHT; n = 8 animals), (D) ddlc1 RNAi (Ddlc1-IR) expression by the muscle c57 GAL4 driver (n = 5 animals), and (E) ddlc1 RNAi (Ddlc1-IR) expression by the motor neuron d42 GAL4 driver (n = 5 animals). (F–H) Quantification of bouton numbers in neuromuscular junctions. (F) WT (n = 9 animals) and (G) ddlc1 mutant (n = 11 animals) stained with HRP antibody. (H) Average number of boutons (±SEM) in the genotypes shown in F and G. *P < 0.00001, significant difference from control.