The lysosomal membrane protein SCAV-3 maintains lysosome integrity and adult longevity

Lysosomes degrade extra- and intracellular material delivered by endocytosis, phagocytosis, or autophagy and recycle catabolites for reutilization in cellular metabolism. There is increasing recognition that lysosomes are advanced organelles that perform much broader functions than cargo digestion and recycling and are involved in fundamental cellular processes such as plasma membrane repair, immune responses, nutrient sensing and signaling, and cell death. As a result, impairment of lysosome function contributes to the pathogenesis of many diseases including lysosomal storage diseases, neurodegenerative disorders, and cancer (Appelqvist et al., 2013; Settembre et al., 2013).

The lysosomal lumen contains ∽60 different soluble acid hydrolases that act in coordination to degrade all types of macromolecules (Settembre et al., 2013). With such a high enrichment of catabolic enzymes, lysosomes have long been considered “suicide bags.” Lysosomal rupture causes leakage of the luminal contents into the cytosol, leading to membrane trafficking defects, abnormal energy metabolism, and cell death by means of apoptosis or necrosis (de Duve, 1983; Repnik et al., 2014). Thus, maintenance of lysosome integrity is clearly important for cell viability. On the other hand, induction or amplification of cell death events by lysosomal membrane permeabilization and the consequent release of lysosomal contents are considered alternative strategies for treating apoptosis- or multidrug-resistant cancers (Kreuzaler and Watson, 2012; Groth-Pedersen and Jäättelä, 2013; Kallunki et al., 2013).

As a safeguard, the lysosome is limited by a 7- to 10-nm single phospholipid bilayer, which serves as the mechanical barrier that encloses the potent luminal hydrolases and protects the other cellular constituents from unwanted degradation (Saftig et al., 2010). The lysosomal membrane contains heavily glycosylated membrane proteins, which may form a continuous carbohydrate layer at the luminal leaflet to prevent the membrane from being degraded by the lysosomal hydrolases (Fukuda, 1991). These major lysosomal glycoproteins include lysosomal-associated membrane protein (LAMP)-1, LAMP-2, lysosomal integral membrane protein (LIMP)-1/CD63, and LIMP-2. LAMP-1 and LAMP-2 have partially overlapping functions in vivo, but LAMP-2 seems to carry out additional unique functions such as acting as a receptor of chaperone-mediated autophagy (Eskelinen, 2006; Orenstein and Cuervo, 2010). LAMPs are thought to play an important protective role against lysosomal hydrolases owing to their very high abundance and the extensive glycosylation of their large intraluminal domain. However, in LAMP-1/LAMP-2 double-deficient MEF cells, lysosomal integrity is not affected (Eskelinen et al., 2004). In oncogene-transformed embryonic fibroblasts, decreased LAMP levels contribute to the sensitization of the cells to lysosomal cell death induced by various anticancer drugs. The cell death process is characterized by lysosomal membrane permeabilization and the subsequent release of cathepsins into the cytosol, suggesting a role of LAMPs in promoting lysosomal membrane stability in transformed cells (Fehrenbacher et al., 2008). Nevertheless, depletion of lamp1 and lamp2 has no effect on the sensitivity of nontransformed fibroblasts to agents that trigger lysosomal leakage (Fehrenbacher et al., 2008). Moreover, deglycosylation of LAMP-1 and LAMP-2 by endoglycosidase H, which removes Asn-linked glycans, causes rapid degradation of LAMPs but has no measurable effect on lysosomal pH, osmotic stability, density, or cell survival (Kundra and Kornfeld, 1999). Thus, the function of LAMPs in maintaining lysosomal membrane integrity under physiological conditions remains elusive.

It is worth noting that endoglycosidase H treatment, which was performed in living cells with special genetic and pharmacologic manipulations to prevent formation of complex oligosaccharide chains that are insensitive to this enzyme, does not affect the stability of LIMP-2, another major glycoprotein on the lysosomal membrane, implying that LIMP-2 may be involved in preserving the integrity of lysosomes (Kundra and Kornfeld, 1999). LIMP-2 is a ubiquitously expressed type III transmembrane protein that spans the membrane twice, with both N and C termini exposed to the cytosol and a large luminal domain containing complex carbohydrate structures (Vega et al., 1991). LIMP-2 has been identified as a receptor that mediates the mannose 6-phosphate–independent lysosomal targeting of β-glucocerebrosidase (β-GCase), the enzyme defective in Gaucher disease (Reczek et al., 2007). In this process, the luminal domain of LIMP-2 binds to β-GCase in the ER at neutral pH; the two proteins disassociate upon arrival at the lysosome because of the low pH (Blanz et al., 2010; Zachos et al., 2012 ,Neculai et al., 2013). In addition to mediating β-GCase transport, LIMP-2 was found to serve as a receptor for several enteroviruses that associate with hand, foot, and mouth disease and neurological diseases (Gonzalez et al., 2014). Moreover, overexpression of LIMP-2 in mammalian cells causes enlarged vacuoles that share both endosomal and lysosomal features, suggesting that LIMP-2 may be involved in the biogenesis of lysosomes/endosomes (Kuronita et al., 2002). Consistent with this, LIMP-2–deficient mice display phenotypes that may be caused by impaired membrane trafficking (Gamp et al., 2003; Saftig et al., 2010). However, it is unclear whether LIMP-2 plays a role in maintaining the integrity of lysosomes.

Lysosomes are the final degradative compartment downstream of various proteolytic pathways. Lysosome functions appear to decline during aging, which may account at least in part for the reduced protein degradation in older organisms and may contribute to various age-related pathologies (Hochschild, 1971; Cuervo and Dice, 2000). Consistent with this, age-related changes in lysosomes, such as increased volume, accumulation of indigestible materials, and fragility, are observed in senescent organisms (Hochschild, 1971; Terman and Brunk, 1998; Cuervo and Dice, 2000). It is unknown, however, whether lysosome activities are modulated during aging by longevity-regulating pathways. One such pathway is the insulin/insulin-like growth factor 1 (IGF-1) signaling cascade, which controls aging in Caenorhabditis elegans, insects, and mammals and extends the lifespan of these organisms when attenuated (Anisimov and Bartke, 2013). In worms, reduced insulin/IGF-1 signaling, such as that caused by mutation in the daf-2 gene, which encodes the sole C. elegans insulin/IGF-1 receptor, leads to significantly increased adult longevity. This process involves a phosphorylation cascade that ultimately results in nuclear translocation of the DAF-16/FOXO transcription factor and subsequent transcriptional regulation of its target genes (Murphy and Hu, 2013).

In this study, we identify SCAV-3, the C. elegans homologue of human LIMP-2, as a key regulator of lysosome integrity and dynamics. Loss of scav-3 causes lysosomal damage and shortened lifespan, which can be efficiently suppressed by inhibiting insulin/IGF-1 signaling in a DAF-16–dependent manner. Thus, lysosome integrity is modulated by the insulin/IGF-1 signaling pathway and is crucial for longevity.

We coexpressed the lysosomal DNase II NUC-1 and the lysosomal membrane protein LAAT-1 to visualize lysosomes in live C. elegans (Guo et al., 2010; Liu et al., 2012). In wild type, NUC-1 and LAAT-1 labeled both vesicular and tubular structures at embryonic, larval, and adult stages in multiple tissues (Fig. 1, A–A″, C–C″, and E–E″; and Fig. S1, A–C). The vesicular and tubular lysosomes are mobile and undergo a variety of dynamic changes (Figs. 1 J and S1 L). In qx193, a recessive mutation generated by γ-irradiation, however, we observed significantly enlarged globular lysosomes (Fig. 1, B–B″, D–D″, and F–F″; and Fig. S1 D). In wild type, 82% and 73% of lysosomes in fourfold embryos and L1 larvae are small, with volumes of 0.02–0.4 µm³ (Fig. 1, G and H). The mean lysosome volume in embryos and larvae is 0.39 and 0.31 µm³, respectively (Fig. 1 I). In qx193 mutants, however, 56% and 44% of lysosomes in fourfold embryos and L1 larvae are big, ranging in volume from 0.4 µm³ to more than 10.2 µm³ (Fig. 1, G and H). The mean volumes reached 1.2 and 1.05 µm³ in embryos and larvae, respectively, which are 2.1 and 2.4 times bigger than in wild type (Fig. 1 I). qx193 mutants lacked tubular lysosomes at both embryonic and larval stages, whereas tubule-like structures that were much thicker than wild-type tubules were present in qx193 adults (Fig. 1, B–B″, D–D″, and F–F″; and Fig. S1 D). Thus, both globular and tubular lysosomal structures are affected in qx193 mutants. In contrast, reporters of ER (GFP::TRAM), Golgi membrane (MANS::GFP), early endosomes (YFP::2xFYVE, HGRS-1::GFP), and recycling endosomes (GFP::RAB-10, GFP::RAB-11, and GFP::RME-1) in qx193 mutants showed similar patterns to those in wild type, indicating that these intracellular organelles are not affected (Fig. 2). In addition to morphological changes, lysosomes were obviously less dynamic in qx193 than in wild type (Fig. S1 M). We measured the Pearson’s correlation coefficient to compare the colocalization of lysosomes in two time-lapse image frames taken 20 s apart (French et al., 2008). We found that LAAT-1::GFP–labeled lysosomes were more colocalized in qx193, resulting in higher Pearson’s correlation in qx193 than in wild type (Fig. 1 J) and indicating that lysosomal dynamics are affected in qx193.

We found that qx193 causes a 648-bp deletion in scav-3, which encodes a conserved protein homologous to the human lysosomal integral membrane protein LIMP-2 (Fig. S2, A and L). SCAV-3 and LIMP-2 contain two predicated transmembrane domains and share 76% sequence similarity and 28% sequence identity (Fig. S2 L). Two scav-3 deletion mutants, ok1286 and tm2659, and seven other alleles of scav-3 isolated from a forward genetic screen all showed lysosome phenotypes similar to qx193 (Fig. S2, A–G). We generated a SCAV-3::GFP reporter driven by the scav-3 promoter and found that it fully rescued the lysosome phenotypes of qx193 (Fig. S2, H–I″). SCAV-3::GFP is widely expressed during embryonic, larval, and adult stages in various tissues including pharynx, hypodermis, sheath cells, vulva, and tail region (Fig. 3, A–F′). SCAV-3::GFP was not evident in the intestine and was only very weakly expressed in body wall muscle cells, consistent with less severely affected lysosomes in these two cell types in scav-3(lf) mutants (Fig. S1, E and F). SCAV-3::GFP labeled both puncta and tubular structures that were positive for NUC-1 and LAAT-1, suggesting that SCAV-3 localizes to lysosomes (Fig. 3, G–H″). SCAV-3(AAA)::GFP, in which the C-terminal leucine-based lysosomal targeting motif (Leu⁵¹⁶-Val⁵¹⁷-Leu⁵¹⁸) was replaced by three alanines, mostly stained plasma membranes, further suggesting that SCAV-3 is a lysosomal membrane protein (Fig. 3, I and J; and Fig. S2 L).

Human LIMP-2 is an abundant lysosomal membrane protein that performs multiple functions including transporting β-GCase to lysosomes (Reczek et al., 2007). We found that the C. elegans β-GCase GBA-3 localized to both vesicular and tubular lysosomes in wild type and appeared in the enlarged lysosomes in multiple scav-3 mutant strains, including those with deletion mutations that cause large protein truncations and those with missense mutations that abolish lysosomal targeting of SCAV-3 (Fig. 4, A and B; Fig. S1, O–R″; and Fig. S2, J–L; Materials and methods). This suggests that loss-of-function of scav-3 does not affect lysosomal localization of GBA-3. In addition, GBA-1 (glucosidase), CPL-1 (cathepsin L), CPR-6 (cathepsin B), Y105E8B.9 (β-glucuronidase), and ASM-1 (acid sphingomyelinase) were all transported to lysosomes in scav-3(qx193) mutants as in wild type, suggesting that lysosomal targeting of these hydrolases is not affected by loss of scav-3 function (Fig. 4, C–L).

SCAV-3 contains a large intraluminal loop that is probably highly glycosylated, as in LIMP-2 (Vega et al., 1991). We next investigated whether SCAV-3 may play a structural role to protect the integrity of lysosome membranes. To do this, we expressed GFP fusions of galectin 3 (Gal3) and Gal9 in worms. Gal3 and Gal9 are β-galactoside–binding proteins that bind luminal glycoproteins of endosomes or lysosomes with ruptured membranes and thus detect endosomal or lysosomal damage (Paz et al., 2010; Thurston et al., 2012; Maejima et al., 2013). We found that GFP::Gal3 and GFP::Gal9 were diffuse throughout the cytosol in wild-type worms but accumulated as vesicular and irregular membrane-like structures in scav-3(qx193) mutants (Fig. 5, A–G; and Fig. S3, A–B″). The GFP::Gal3 and GFP::Gal9 speckles contained diffuse and faint NUC-1::mCherry, whereas GFP rings were seen surrounding NUC-1 puncta with brighter mCherry fluorescence (Fig. 5, B–B″; and Fig. S3, B–B″, arrows). Similar GFP::Gal3 patterns were observed in scav-3(qx193) when lysosomes were labeled by CPL-1::mCherry or CPR-6::mCherry (Fig. S3, C–D″). We performed time-lapse analyses in scav-3(qx193) mutants to follow dynamic changes of Gal3 and lysosomes labeled by multiple luminal hydrolases. We found that GFP::Gal3 appeared initially around lysosomes containing bright NUC-1::mCherry, GBA-3::mCherry, CPL-1::mCherry, or CPR-6::mCherry, which was followed by gradual enrichment of GFP with the concomitant loss of mCherry fluorescence (Fig. 6, A–D; and Videos 1, 2, 3, and 4). Collectively, these data suggest that loss of scav-3 causes damage to lysosome membranes, leading to leakage of the lysosomal hydrolases NUC-1, GBA-3, CPL-1, and CPR-6 into the cytosol and association of cytosolic Gal3 and Gal9 with glycoproteins at the luminal leaflet of lysosomes. Consistent with leakage of luminal contents after membrane damage, we found that 44–47% of lysosomes in scav-3(qx193) contained very weak NUC-1–, CPL-1–, or CPR-6–mCherry, whereas 100% of wild-type lysosomes had bright mCherry fluorescence (Figs. 5 H and S3 E). Moreover, GFP::Gal3-positive structures in scav-3(qx193) contained faint or undetectable NUC-1::mCherry or CPL-1::mCherry, whereas GFP::Gal3 was absent from lysosomes with bright mCherry fluorescence (Figs. 5 I and S3 F). Furthermore, GFP::Gal3-positive structures in scav-3(qx193) were not stained by LysoTracker red, indicating loss of interior acidity after lysosomal membrane damage (Fig. 5, J–L).

To further examine lysosome damage in scav-3(lf), we performed transmission electron microscopy (TEM) analyses. In wild-type worms, lysosomes appeared as dense, spherical, membrane-enclosed vesicles (Fig. 7, A and B). In scav-3(qx193), however, we observed three types of abnormal lysosome that were not seen in wild type. First, we found significantly enlarged membrane-bound vesicles that were filled with membranous and granular contents (Fig. 7, C and D; 45.1%, n = 51), suggesting accumulation of undigested cargo in lysosomes. Second, we observed large vesicular structures that contained similar cargo but had clear membrane damage (Fig. 7, E–H, red arrows; 19.6%, n = 51). Finally, we saw aggregation of dense granules and membranous structures reminiscent of undigested cargo in the enlarged lysosomes, but very little or no surrounding membranes were detected (Fig. 7, G and H; 35.3%, n = 51). These ultrastructural data indicate that loss of scav-3 function causes damage to lysosome membranes. To further corroborate this, we purified lysosomes from wild type and scav-3(qx193) mutants and examined acid phosphatase activity in the lysosome suspension (Fig. S4 P). We observed higher available acid phosphatase activity in the suspension of scav-3(qx193) lysosomes than wild type, suggesting that membrane integrity or permeability is affected in scav-3(qx193) lysosomes (Fig. 5 M). Collectively, these data indicate that loss of scav-3 function affects lysosome integrity and function, causing cargo accumulation and membrane damage.

LAMPs are the most abundant glycoproteins on the lysosome membrane and are thought to play a protective role against lysosomal hydrolases (Eskelinen, 2006). C. elegans has two LAMP-like proteins, LMP-1 and LMP-2 (Kostich et al., 2000; de Voer et al., 2008). LMP-1 was widely distributed but enriched in the intestine and pharynx, whereas LMP-2 expression appeared to be restricted to the intestine (Fig. 3, K–T). In contrast to the predominantly lysosomal localization of SCAV-3, both cell surface and intracellular distribution of LMP-1 and LMP-2 were observed (Fig. 3, K–S). Loss of lmp-1 and lmp-2 did not affect lysosome morphology, distribution, or dynamics in hypodermis, sheath cells, or body wall muscles but caused enlarged LAAT-1–positive vesicles in the intestine, consistent with previous findings that loss of lmp-1 affects lysosome-related organelles in the intestine (Fig. S1, G–K and N; Kostich et al., 2000). We found that worms lacking lmp-1 or lmp-2 or both contained very few GFP::Gal3-positive structures in hypodermal cells, whereas diffuse GFP::Gal3 was seen in other tissues including pharynx, intestine, and body wall muscle (Fig. 8, A–E and I; and Fig. S3, L–N). Moreover, loss of LMP-1 and LMP-2 either individually or simultaneously did not further increase the number of GFP::Gal3-positive damaged lysosomes in scav-3(lf) (Fig. 8, F–I). These data suggest that lysosomal integrity is not severely affected by loss of LMP-1 and LMP-2. In addition, loss of the lysosomal Ca²⁺ channel CUP-5 or the lysosomal amino acid transporters CTNS-1 and LAAT-1 impairs lysosome function, causing accumulation of enlarged lysosomes, but lysosome integrity was not affected in these mutants (Fig. 6, E–F″; and Fig. S4, A–C; Treusch et al., 2004; Liu et al., 2012).

We found that overexpression of LMP-1 or LMP-2, but not CTNS-1 or LAAT-1, greatly reduced the number of GFP::Gal3-positive structures in scav-3(qx193) and increased the ratio of lysosomes with intact membranes revealed by TEM (Fig. 7, K and L; Fig. 8, J–L and P; and Fig. S4, D–G). These data suggest that the increased level of LMP-1 and LMP-2 compensates for the loss of SCAV-3 to maintain lysosome membrane integrity. Overexpression of LMP-1, however, did not reverse the abnormal lysosomal morphology and dynamics in scav-3(qx193) mutants, suggesting that SCAV-3 has specific effects on membrane stability (Fig. S5, A, B, E, and F). We next investigated whether glycosylation may contribute to the protective role of SCAV-3 and LMP-1. SCAV-3 and LMP-1 contain 10 and five putative glycosylation sites, respectively, among which six and three are conserved in human LIMP-2 and LAMP1 (Fig. S2, L and M). Peptide N-glycosidase F (PNGase F) treatment, which removes both mature and immature N-linked oligosaccharides, greatly reduced the apparent molecular weight of SCAV-3, suggesting that SCAV-3 is highly glycosylated (Fig. S4 M). PNGase F treatment also removed several high molecular mass bands of LMP-1 (Fig. S4 N). Mutation of each individual glycosylation site of SCAV-3 did not affect protein targeting or rescue activity (not depicted), whereas in SCAV-3 mutants, the lack of all six of the conserved glycosylation sites (6N) or the three conserved sites at Asn 233, 278, and 456 (3N-2) severely affected protein stability and targeting, causing greatly reduced protein level, diffuse localization, and significantly decreased rescue activity (Fig. S4, H–M and O). This indicates that glycosylation is critical for the protein stability and targeting of SCAV-3. We found that SCAV-3(3N-1), in which the other three conserved glycosylation sites at Asn 53, 80, and 118 were replaced by Ala, was expressed at a level similar to the wild-type protein. In the absence of PGNase F treatment, the apparent molecular weight of SCAV-3(3N-1) was slightly lower than that of the wild-type protein, which suggests that glycosylation is partially affected (Fig. S4 M). SCAV-3(3N-1) localized to lysosomes but did not completely rescue the lysosome integrity defect in scav-3(qx193) mutants, suggesting that glycosylation may contribute to the protection of lysosome membranes by SCAV-3 (Fig. 8, N, O, Q, and S–S″). Mutations in Asn 27, 50, and 60 of LMP-1 [LMP-1(3N)] did not obviously affect protein expression or targeting but may impair glycosylation (Fig. 8, T–T″; and Fig. S4 N). Compared with LMP-1, overexpression of LMP-1(3N) was significantly less effective at suppressing the lysosome damage caused by scav-3(lf), suggesting that glycosylation is important for the protection of lysosome integrity by LMP-1 (Fig. 8, L, M, and R).

In cultured mammalian cells, damaged lysosomes caused by lysosomotropic agents or light-activated photosensitizers are cleared through autophagy (Hung et al., 2013; Maejima et al., 2013). We examined lysosome damage in wild-type and scav-3(lf) worms that carry mutations in autophagy genes including lgg-1, atg-2, atg-9, and atg-18 (Tian et al., 2010). We found that GFP::Gal3 was diffuse in autophagy-defective mutants, as in wild-type, indicating that loss of autophagy genes does not cause lysosome damage (Fig. 9, A–C, G–I, and P). Moreover, the number of GFP::Gal3-positive structures in scav-3(lf) mutants was not altered by mutations in autophagy genes (Fig. 9, D–F, J–L, and P). We observed that the GFP::Gal3-positive damaged lysosomes appeared in adults but not embryos or larvae of scav-3(lf), and the size and number gradually reduced as the worms grew, resulting in far fewer GFP::Gal3-positive structures at day 6 of adulthood than day 1 (Fig. 9, D, J, M, P, and Q). Mutations in autophagy genes did not affect this reduction in GFP::Gal3-positive damaged lysosomes in scav-3(lf) (Fig. 9, M–O and Q). Thus, the appearance and persistence of damaged lysosomes caused by loss of scav-3 are not altered in autophagy-defective mutants.

We found that loss of scav-3 function caused significantly shortened adult lifespan. scav-3(qx193) mutants, which developed normally through embryonic and larval stages, had a mean lifespan of 12.1 d, which is 42% shorter than wild type (20.8 d; Fig. 10 A). The maximum lifespan of scav-3(qx193) was also shorter than wild-type worms (Fig. 10 A). The shortened lifespan phenotype of scav-3(qx193) was efficiently rescued by expression of SCAV-3::GFP, leading to a restored mean lifespan of 22.5 d (Fig. 10 A). These data suggest that lysosome integrity and function maintained by SCAV-3 is important for longevity.

Inhibiting insulin/IGF-1 signaling extends the C. elegans lifespan. Decreased activity of the insulin/IGF-1 receptor DAF-2 leads to significantly extended lifespan, which requires the function of DAF-16, a FOXO transcription factor (Murphy and Hu, 2013). We found that daf-2(e1370ts), a temperature-sensitive (ts) mutation of daf-2 that lives twice as long as wild-type worms, significantly extended the lifespan of scav-3(qx193) (Fig. 10 B). The lifespan extension of daf-2 scav-3 was completely suppressed in daf-16;daf-2 scav-3 triple mutants, which exhibited a shortened lifespan similar to scav-3(lf) and daf-16(lf) (Fig. 10 B). This indicates that DAF-16 mediates the lifespan extension of scav-3(lf) by daf-2(lf). daf-2(e1370ts) scav-3(qx193) worms contained greatly reduced numbers of GFP::Gal3-positive structures at both early and late stages of adulthood (Fig. 10, F–O). Moreover, many more lysosomes with intact membranes were observed by TEM in daf-2(e1370ts) scav-3(qx193) worms (75%, n = 60; Fig. 7, M and N). These data suggest that inhibiting insulin/IGF-1 signaling suppresses lysosome damage in scav-3(lf). The Gal3::GFP-positive structures, which decreased significantly in daf-2(e1370ts) scav-3(qx193), reappeared in daf-16;daf-2scav-3, indicating that daf-16 is required for suppression of lysosome damage by daf-2(lf) (Fig. 10, H, K, N, and O). In contrast, inactivation of HSF-1 or SKN-1, the two transcription factors that are also regulated by insulin/IGF-1 signaling, had no effect on the suppression of lysosome damage by daf-2(e1370ts) (Fig. S5, L–Q). Collectively, these data suggest that inhibiting insulin/IGF-1 signaling suppresses lysosome damage through the DAF-16 transcription factor, leading to lifespan extension of scav-3(lf). Overexpression of LMP-1 and LMP-2, which led to a significant reduction in damaged lysosomes in scav-3(lf), also restored the lifespan of scav-3(lf), further emphasizing that maintenance of lysosome integrity is important for longevity (Fig. 10, C–E). Inhibition of lysosomal damage by overexpression of LMP-1 and LMP-2 was not affected by loss of daf-16, suggesting that the effect of LMP-1 and LMP-2 is independent of or bypasses the activity of DAF-16 (Fig. S5, R–U).

Reducing insulin/IGF-1 signaling leads to increased resistance to a wide variety of stresses. We found that tunicamycin treatment induced a strong ER unfolded protein response, indicated by P_hsp-4GFP expression in wild type and scav-3(qx193) but not in daf-2(e1370ts) or daf-2(e1370ts) scav-3(qx193) (Fig. 10, P–W; Calfon et al., 2002). Moreover, significantly more daf-2(e1370ts) and daf-2(e1370ts) scav-3(qx193) worms survived after heat-shock treatment than wild type or scav-3(qx193), indicating increased thermotolerance (Fig. 10 X). These data suggest that inhibiting insulin/IGF-1 signaling increases stress resistance in scav-3(lf). Reduced DAF-2 activity or overexpression of LMP-1 led to decreased lysosome volume in scav-3(qx193) adults but did not alter the abnormal lysosomal morphology and dynamics in scav-3(qx193) larvae, suggesting that lysosome function is not fully restored in these animals (Fig. S5, A–K).

The lysosome membrane is the physical barrier that separates the luminal acidic milieu from the cytoplasmic environment, thus protecting cellular constituents from unwanted degradation. The mechanisms that preserve lysosomal membrane remain unclear. Here we found that SCAV-3, the C. elegans homologue of human LIMP-2, plays an essential role in maintaining lysosome integrity. Overexpression of LMP-1 or LMP-2 efficiently suppressed lysosome damage caused by loss of scav-3, indicating that C. elegans LAMPs can act as structural proteins and substitute for the function of SCAV-3 in maintaining the stability of lysosome membranes. In contrast, loss of other lysosome membrane proteins, such as the calcium channel CUP-5 and the amino acid transporters LAAT-1 and CTNS-1, impairs lysosome function and causes enlargement of lysosomes but has no effect on membrane integrity, which further suggests that SCAV-3 plays a specific role in preserving lysosomal membrane stability.

Extensive glycosylation in the luminal portion of abundant lysosomal membrane proteins has been suggested to perform a protective role by forming a continuous carbohydrate layer at the luminal leaflet (glycocalyx) to prevent membrane damage by the lysosomal hydrolases (Lewis et al., 1985; Fukuda, 1991). Expression of SCAV-3(3N-1) and LMP-1(3N), which contained mutations in the conserved glycosylation sites and caused partially impaired glycosylation, did not affect protein stability or targeting but led to reduced preservation of lysosomal membrane stability. This suggests that glycosylation of SCAV-3 and LMP-1 may contribute to the formation of the glycocalyx that protects lysosomal membranes from the damage by luminal hydrolases. However, mutations that abolished all six of the conserved glycosylation sites (6N), or the three conserved sites at Asn 233, 278, and 456 (3N-2) of SCAV-3, severely affected protein stability and targeting, thus precluding us from further analyzing their effects on membrane protection. More in-depth analyses will therefore be needed to demonstrate the function of SCAV-3 in glycocalyx formation. The accumulation of damaged lysosomes caused by loss of scav-3 is more evident in the hypodermis, where SCAV-3 is highly expressed, but LMPs are not. Future studies are also needed to identify cellular mechanisms responsible for the cell type–specific requirement for SCAV-3 and to reveal the function of LMPs in protecting lysosome membrane integrity.

We found that C. elegans lysosomes are mobile and undergo a variety of dynamic changes, a process that requires SCAV-3 function. Overexpression of LMP-1 efficiently suppressed membrane damage but not the abnormal morphology or defective dynamics of scav-3(lf) lysosomes, suggesting that SCAV-3 may regulate lysosome integrity and dynamics by distinct mechanisms. Future investigations should address the mechanisms by which SCAV-3 functions in these two processes.

Mammalian LIMP-2 was identified as a receptor for transporting β-GCase, the enzyme defective in Gaucher disease, to lysosomes in an mannose 6-phosphate–independent manner (Reczek et al., 2007). In worms, however, lysosomal localization of β-GCase is not affected by loss of scav-3, suggesting the involvement of different targeting mechanisms. LIMP-2 deficiency has been implicated in the pathogenesis of Gaucher disease and Parkinson disease because of its role in transporting β-GCase to lysosomes (Gonzalez et al., 2014). Moreover, mutations in SCARB2, the gene encoding LIMP-2, have been identified as the cause of action myoclonus–renal failure, a fatal inherited form of progressive myoclonic epilepsy associated with renal failure or neurological symptoms. The underlying pathogenesis, however, remains unclear (Berkovic et al., 2008; Gonzalez et al., 2014). Our findings reveal the essential role of SCAV-3 in the maintenance of lysosomal membrane integrity and dynamics, which may contribute to the pathogenesis of diseases caused by or associated with LIMP-2 deficiency.

In scav-3(lf) mutants, abnormal lysosomal morphology and dynamics were observed from embryonic to adult stages, whereas damaged lysosomes did not appear until adulthood. This suggests that lysosomes may enter a damage-prone stage before becoming ruptured. We found that inhibition of insulin/IGF-1 signaling efficiently suppressed lysosome damage in scav-3(lf) in a DAF-16–dependent manner, indicating that the DAF-2/IGFR-DAF-16/FOXO pathway plays an important role in modulating lysosomal membrane stability. We found that down-regulation of daf-2 signaling increases stress resistance in scav-3(lf). Down-regulated daf-2 signaling may alleviate the stress on scav-3(lf) lysosomes by reducing lysosomal substrate loading, thus maintaining membrane stability. In addition, the insulin/IGF-1 pathway may control expression of lysosomal genes that increase the physical stability of membranes or promote cargo digestion, thus stabilizing the damage-prone lysosomes.

Lysosomes are the final destinations of damaged cellular structures and proteins that accumulate during aging. We found that SCAV-3 is essential for maintaining lysosome integrity and normal adult lifespan, and that both lysosomal damage and shortened lifespan in scav-3(lf) are rescued or extended by reduced daf-2 signaling or LMP-1 and LMP-2 overexpression. This suggests that lysosome integrity is an important component in lifespan regulation. The suppression of lysosome damage by daf-2 mutation is dependent on DAF-16 but not HSF-1 or SKN-1, indicating specific roles of DAF-16 in this process. Given the evolutionary conservation of the insulin/IGF-1 signaling pathway in metazoans, similar mechanisms may be used in mammals to modulate lysosome integrity in longevity.

Lysosomes damaged by lysosomotropic agents or light-activated photosensitizers in cultured mammalian cells are cleared by autophagy, leading to recovery of low pH and degradation capacity (Hung et al., 2013; Maejima et al., 2013). In scav-3(lf) mutants, however, appearance and disappearance of GFP::Gal3-positive damaged lysosomes were unchanged in autophagy-defective mutants. It is conceivable that loss of scav-3 may affect autophagic removal of damaged lysosomes because of the lack of healthy lysosomes. Alternatively, autophagy-independent mechanisms may be involved. Future investigations should address whether and how autophagy-dependent and/or -independent mechanisms regulate clearance of damaged lysosomes in vivo.

Strains of C. elegans were cultured and maintained using standard protocols (Brenner, 1974). The N2 Bristol strain was used as the wild-type strain except in polymorphism mapping, in which Hawaiian strain CB4856 was used. The following strains were used in this work: linkage group (LG) I: hsf-1(sy441), and daf-16(mu86); LG II: laat-1(qx42), lgg-1(bp500), and ctns-1(ok813); LG III: daf-2(e1370ts), cup-5(bp510), and scav-3(qx193, tm3659, ok1286, qx232, qx237, qx260, qx266, qx267, qx271, qx281); LG V: atg-9(bp564), atg-18(gk378), and zcIs4; and LG X: atg-2(bp576), lmp-2(tm6600), and lmp-1(nr2045). daf-2(e1370ts) was maintained at 20°C. L4 larvae were moved to 25°C for 24–144 h before the damaged lysosome phenotype was examined.

Transgenic animals carrying extrachromosomal arrays (qxEx) were generated using standard microinjection methods, and genome-integrated arrays (qxIs) were obtained by γ-ray irradiation to achieve stable expression from arrays with low copy numbers (Evans, 2005). The reporter strains used in this study are listed as follows: qxIs354 (P_ced-1LAAT-1::GFP), qxIs257 (P_ced-1NUC-1::mmCherry), qxIs352 (P_ced-1LAAT-1::nmCherry), qxIs430 (P_scav-3SCAV-3::GFP), qxIs439 (P_hyp-7GFP::TRAM-1), qxIs468 (P_myo-3LAAT-1::GFP), qxIs582 (P_hyp-7sfGFP::Gal3), qxIs594 (P_myo-3sfGFP::Gal3), qxIs599 (P_vha-6sfGFP::Gal3), qxIs612 (P_hspNUC-1::sfGFP::mmCherry), qxEx3928 (P_hyp-7MANS::GFP), qxEx4342 (P_hyp-7GFP::RAB-10), qxEx4345 (P_hyp-7GFP::RAB-11), qxEx7016 [P_scav-3SCAV-3(AAA)::sfGFP], qxEx4274 (P_vha-6LAAT-1::GFP), qxEx7170 (P_hyp-7sfGFP::Gal9), qxEx6312 (P_vha-6sfGFP::Gal3), qxEx6953, 6966, 6967 (P_hyp-7LMP-1::ceBFP), qxEx6751, 6752, 6758 (P_hyp-7LMP-2::ceBFP), qxEx6975, 6976, 6977 (P_hyp-7LAAT-1::CeBFP), qxEx6978, 6979, 6980 (P_hyp-7CTNS-1::CeBFP), qxEx7189 (P_lmp-1LMP-1::sfGFP), qxEx7190 (P_lmp-2LMP-2::sfGFP), qxEx6981 (P_hspsfGFP::Gal3), qxEx7396, 7397, 7398 [LMP-1(3N): P_hyp-7LMP-1(N27A, N50A, N60A)::CeBFP], qxEx7399, 7400, 7401 [SCAV-3(3N-1): P_scav-3SCAV-3(N53A, N80A, N118A)::CeBFP], qxEx7402, 7403, 7404 [SCAV-3(3N-2): P_scav-3SCAV-3(N233A, N278A, N456A)::CeBFP], qxEx7405 [SCAV-3(4N): P_scav-3SCAV-3(N53A, N233A, N278A, N456A)::CeBFP], qxEx7406 [SCAV-3(6N): P_scav-3SCAV-3(N53A, N80A, N118A, N233A, N278A, N456A)::CeBFP], qxEx7407 [SCAV-3(3N-1): P_scav-3SCAV-3(N53A, N80A, N118A)::sfGFP], qxEx7408 [SCAV-3(3N-2): P_scav-3SCAV-3(N233A, N278A, N456A)::sfGFP], qxEx7409 [SCAV-3(4N): P_scav-3SCAV-3(N53A, N233A, N278A, N456A)::sfGFP], qxEx7410 [SCAV-3(6N): P_scav-3SCAV-3 (N53A, N80A, N118A, N233A, N278A, N456A)::sfGFP], qxEx7411 [P_scav-3SCAV-3(G279R)::sfGFP], qxEx7412 [P_scav-3SCAV-3(S382F)::sfGFP], and yqEx632 (P_cpl-1CPL-1::mCherry).

We obtained bIs46 (P_vit-2GFP::RME-1) from B. Grant (Rutgers University, New Brunswick, NJ); opIs334 (P_ced-1YFP::2xFYVE) from K.S. Ravichandran (University of Virginia, Charlottesville, VA) and M.O. Hengartner (University of Zurich, Zurich, Switzerland); P_hgrs-1HGRS-1::GFP from R. Legouis (Institute for Integrative Biology of the Cell, Gif-sur-Yvette, France); kxEx147 (P_gba-3GBA-3::mCherry), kxEx152 (P_asm-1ASM-1::mCherry), and kxEx141 (P_cpr-6CPR-6::mCherry) from G. Hermann (Lewis and Clark College, Portland, OR); and scav-3(tm3659) from S. Mitani (Tokyo Women’s Medical University, Tokyo, Japan).

The qx193 mutation was isolated through backcrossing the genome-integrated array of LAAT-1::GFP generated by γ-ray irradiation. Other alleles of scav-3 including qx232, qx237, qx260, qx266, qx267, qx271, and qx281 were isolated from a forward genetic screen using the abnormally enlarged lysosome phenotype. qx193 was mapped to the right arm of linkage group III (LG III) between genetic map positions 17.26 [Snp-Y49E10(2)] and 21 (Snp-UCE3-1426) by single nucleotide polymorphism mapping. Transformation rescue experiments were performed, and a DNA fragment containing the scav-3 gene fully rescued the qx193 defect. The sequence of the scav-3 gene was determined in all scav-3 alleles. qx193 contains a 648-bp deletion that removes most of exon 5, all of exon 6, and part of exon 7, causing an early stop codon that results in a truncated protein containing the first 346 amino acids. The ok1286 allele of scav-3 contains a 1,146-bp deletion and a 2-bp insertion that removes most of the second and third exons and generates an early stop codon, resulting in a truncated protein containing only the first 118 amino acids. The qx237 and qx266 alleles contain a G-to-A transition that results in a premature stop codon after Ser 321. qx271 and qx281 contain C-to-T mutations that cause a premature stop codon after Arg 372 and substitution of Ser 382 with Phe, respectively. qx260 contains a G-to-A mutation that results in substitution of Gly 279 with Arg. qx232 and qx267 contain G-to-A transitions at the splicing sites located at the boundaries of exon 2 and intron 2 and exon 3 and intron 3, respectively, which cause premature stop codons after Ile 133 and Ile 189. The tm3659 allele of scav-3 contains a 694-bp deletion that removes the entire fourth and fifth exons and generates an early stop codon, resulting in a truncated protein containing only the first 238 amino acids. All mutants were backcrossed with N2 animals at least four times before further analyses.

The bacteria-feeding protocol was used in RNAi experiments as described before (Kamath and Ahringer, 2003). In skn-1 RNAi experiments, 30 L1 or L2 larvae were cultured on the RNAi plates, and lysosome damage was examined in adult animals at the same generation because most F1 progeny die because of inactivation of skn-1. The DNA sequence that is targeted by the dsRNA in the RNAi experiments is skn-1 (T19E7; 590–2,157 nt).

Aged adult worms (>50) were soaked in 100 µl LysoTracker red in M9 (1:200; Invitrogen) for 1.5 h at 20°C in the dark. Worms were then transferred to a fresh OP50-seeded NGM plate and allowed to recover at 20°C for 4 h in the dark before examination by fluorescent microscopy.

Differential interference contrast and fluorescent images were captured with an Axioimager A1 (ZEISS) equipped with epifluorescence (Filter Set 13 for GFP [excitation BP 470/20, beam splitter FT 495, emission BP 503–530] and Filter Set 20 for Cherry [excitation BP 546/12, beam splitter FT 560, emission BP 575–640]) and an AxioCam monochrome digital camera (ZEISS). Images were processed and viewed using Axiovision Rel. 4.7 software (ZEISS). A 100× Plan-Neofluar objective (NA1.30) was used with Immersol 518F oil (ZEISS). Confocal images were captured by an LSM 5 Pascal (ZEISS) inverted confocal microscope with 488-nm (emission filter BP 503–530) and 543-nm (emission filter BP 560–615) lasers, and images were processed and viewed using LSM Image Browser software (ZEISS). All images were taken at 20°C.

To quantify the fluorescence intensity of LMP-1::sfGFP, 15 images were taken of each cell type with equal exposure time and quantified using ImageJ 1.42q (National Institutes of Health).

L1 larvae or adult worms 24 h after the L4/adult molt were mounted on agar pads in M9 buffer with 5 mM levamisole, which prevents animals from moving but does not affect the hypodermis. Fluorescent images were captured using a 100× objective (CFI Plan Apochromat Lambda; NA 1.45; Nikon) with immersion oil (type NF) on an inverted fluorescence microscope (Eclipse Ti-E; Nikon) with a spinning-disk confocal scanner unit (UltraView; PerkinElmer) with 488-nm (emission filter 525 [W50]) and 561-nm (dual-band emission filter 445 [W60] and 615 [W70]) lasers. To follow lysosome dynamics, images were captured every 2 s (wild type) or 5 s (scav-3(qx193), daf-2 scav-3, and scav-3+LMP-1) for 5 min. To examine dynamic changes in worms coexpressing sfGFP::Gal3 and mCherry fusions of NUC-1, GBA-3, CPL-1, or CPR-6, images were captured every 1 min for 1–3 h. The collected images were viewed and analyzed using Velocity software (PerkinElmer).

Time-lapse images of L1 larvae expressing LAAT-1::GFP or NUC-1::mCherry were captured by spinning-disk microscopy. The colocalization of two frames taken 20 s apart was analyzed by Velocity software. At least 10 independent videos were followed and quantified in each strain. Selected images from the representative videos are shown in Figs. 1 J, S1 N, and S5 F.

Fluorescent images of embryos, larvae, or adults expressing LAAT-1::GFP or NUC-1::mCherry in 10–15 z-series (0.5 µm/section) were captured by spinning-disk microscopy. Serial optical sections were analyzed, and the volume of LAAT-1::GFP– or NUC-1::mCherry–positive lysosomes was quantified by Velocity software. At least 10 animals were quantified in each strain at each stage.

To examine damaged lysosomes, fluorescent images were captured using an Axioimager A1 microscope equipped with epifluorescence and an AxioCam monochrome digital camera or an LSM 5 Pascal inverted confocal microscope with equal exposure time. Damaged lysosomes were quantified by scoring the number of sfGFP::Gal3-positive membrane-like structures in hypodermal cell 7 that covers the whole midbody of the worm. Leakage of lysosomal hydrolases was quantified by scoring the percentage of lysosomes that contained strong mCherry fluorescence and were negative for sfGFP::Gal3 (intact) and those that contained faint mCherry fluorescence and were positive for sfGFP::Gal3 (damaged) in wild type and scav-3(qx193) expressing mCherry fusions of the lysosomal hydrolases NUC-1, CPL-1, or CPR-6. At least 15 animals were scored in each strain at each stage.

The fluorescence intensity of lysosomes in scav-3(qx193) coexpressing sfGFP::Gal3 and NUC-1::mCherry or CPL-1::mCherry or stained by LysoTracker red was quantified by linescan analysis. Fluorescent images of lysosomes were captured by spinning-disk microscopy. A line of 6 to 12 µm in length was manually drawn through a lysosome containing bright GFP or mCherry fluorescence (dashed lines in Fig. 5 [I and L] and Fig. S3 F), and the fluorescence pixel intensity along the line was calculated by Velocity software.

Lifespan assays were performed at 20°C as described previously (Hansen et al., 2005). In brief, worms at the L4 stage (day 0, >100) were placed on NGM plates seeded with OP50, with 10 worms per plate. The animals were transferred to new plates when their progeny grew up to the L3 stage, and dead animals were counted every 2 d. Animals that crawled off the plate, exploded, bagged, or became contaminated were discarded.

Worms at day 1 after L4 were treated with tunicamycin at 5 µg/ml. The expression of P_hsp-4GFP was examined 10 h after treatment to assess the ER unfolded protein response. To examine thermotolerance, worms (>100) at day 1 of adulthood were incubated at 37°C for 2.5 h and recovered at 20°C for 12 h before survival rate was quantified.

Lysosomes were purified from C. elegans at day 1 of adulthood using a Lysosome Isolation kit (LYSISO1; Sigma-Aldrich) as described previously with minor modifications (Liu et al., 2012). In brief, the crude lysosomal fraction was purified on a density gradient built up with Optiprep Density Gradient Medium Solution and Sucrose (bottom to top: 27, 19, 16, and 8%). After centrifugation, the enrichment of lysosomes in the different fractions (bottom to top: B1–4) was determined by examining the processing of cathepsin using anti–CPL-1 antibodies after normalization to the same total protein concentration. The purified lysosome fraction (B3; Fig. S4 P) from wild type and scav-3(qx193) was normalized to the same total protein concentration and suspended in the extraction buffer provided in the Lysosome isolation kit. The available acid phosphatase activity in the lysosome suspension was examined using an acid phosphatase assay kit (CS0740; Sigma-Aldrich). Samples were incubated at 25°C for 30 min, and the reactions were stopped by the addition of 0.5 N NaOH before measurement of the absorbance at 405 nm. Activities were compared with the total activities obtained in the presence of 0.1% Trion X-100 and are presented as a percentage of the total activity in Fig. 5 M. At least five independent experiments were performed in each strain, and the mean percentage of total activity is shown.

Adult worms expressing wild-type or mutant versions of SCAV-3::GFP or LMP-1::BFP that lack multiple glycosylation sites were lysed by several cycles of freeze-thawing. The resulting lysates were denatured before incubating with PNGase F for 1 h at 37°C following the manufacturer’s instructions (New England BioLabs, Inc.). The proteins were resolved by SDS-PAGE and revealed by anti–SCAV-3 or anti–LMP-1 antibodies.

Adult C. elegans (day 1 of adulthood) were rapidly frozen using a high-pressure freezer (EM PACT2; Leica Biosystems). Freeze-substitution was performed in anhydrous acetone containing 1% osmium tetroxide. The samples were kept sequentially at −90°C for 72 h, −60°C for 8 h, and −30°C for 8 h and were finally brought to 20°C for 10 h in a freeze-substitution unit (EM AFS2; Leica Biosystems). The samples were washed three times (1 h each time) in fresh anhydrous acetone and were gradually infiltrated with Embed-812 resin in the following steps: resin/acetone 1:3 for 3 h, 1:1 for 5 h, 3:1 overnight, and 100% resin for 4 h. Samples were then kept overnight and embedded at 60°C for 48 h. The fixed samples were cut into 70-nm sections with a microtome EM UC6 (Leica Biosystems) and electron stained with uranyl acetate and lead citrate. Sections were observed with a JEM-1400 (JEOL) operating at 80 kV.

The SD was used as y-axis error bars for bar charts plotted from the mean value of the data. Data derived from different genetic backgrounds were compared by Student’s two-tailed unpaired t test, one-way analysis of variance (ANOVA) followed by Tukey’s post-test, two-way ANOVA followed by Bonferroni post-test, or Kaplan–Meier method followed by log-rank test as indicated in the figure legends. Data were considered statistically different at P < 0.05. P < 0.05 is indicated with single asterisks and P < 0.0001 with double asterisks (0.001 in a two-way ANOVA analysis).

P_scav-3SCAV-3::GFP reporter was created by a PCR fusion-based method (Hobert, 2002). In the first round of PCR, scav-3 genomic DNA including the 2.7-kb promoter region was amplified using primers PWZ853 and PWZ855 and N2 genomic DNA as the template. GFP and the 3′UTR of unc-54 were amplified from pPD49.26-GFP using PHWD478/PWZ512. The two PCR products were fused in the second round of PCR using PWZ854/PPFG199. The worm sfGFP (super-fold GFP) reporter was constructed by amplifying the sfGFP fragment from cc-HS4C3-sfGFP (H.E. Buelow, Albert Einstein College of Medicine, New York, NY) with primers PDFZ623/PDFZ624 and inserting it into pPD49.26 through the KpnI–EcoRV sites. P_scav-3SCAV-3::sfGFP was generated by first ligating the scav-3 genomic DNA amplified with primers PWZ862 and PDFZ721 into pPD49.26-sfGFP through the NheI–MluI sites, followed by insertion of the scav-3 promoter amplified by primers PWZ895 and PWZ896 at the HindIII–PstI sites. To generate P_scav-3SCAV-3(AAA)::sfGFP, the AAA mutation of SCAV-3 was introduced into P_scav-3SCAV-3::sfGFP by site-directed mutagenesis using primers PYL379 and PYL380. To express RAB-10 and RAB-11 in hypodermis, RAB-10 and RAB-11.1 fragments were digested from P_vha-6GFP::RAB-10 and P_vha-6GFP::RAB-11.1 and cloned into pPD49.26-P_hyp-7GFP1 through the KpnI site. P_hyp-7MANS::GFP was constructed by amplifying the MANS cDNA from a C. elegans cDNA library (Invitrogen) using primers PYTX95 and PYTX96 and ligating the product into pPD49.26-P_hyp-7GFP3 through the NheI–KpnI sites. To construct P_hyp-7GFP::TRAM, TRAM cDNA was amplified from the C. elegans cDNA library (Invitrogen) with primers PYJ45 and PBL727 and cloned into pPD49.26-P_hyp-7GFP1 through the KpnI site. To generate P_hyp-7sfGFP::Galectin3 (Gal3), Gal3 was amplified from ptf-Galectin3 (T. Yoshimori, Osaka University, Osaka, Japan) with primers PDFZ627/PDFZ628 and ligated into pPD49.26-P_hyp-7sfGFP1 through the EcoRV–SacI sites. The hyp-7 promoter (P_hyp-7) was then replaced by P_vha-6, P_myo-3, or P_hsp through the BamHI site to generate sfGFP::Gal3 reporters specifically expressed in the intestine cells (P_vha-6), body wall muscle (P_myo-3), or pharynx (P_hsp). To construct P_hyp-7sfGFP::Galectin9 (Gal9), Gal9 was amplified from pENTR221-Galectin9 (Invitrogen) with primers PDFZ898/PDFZ899 and cloned into pPD49.26-P_hyp-7sfGFP1 through the EcoRV site. To generate P_hyp-7LMP-1::CeBFP and P_hyp-7LMP-2::CeBFP, LMP-1 and LMP-2 were amplified with primers PSZ46/PSZ47 and PXCW192/PXCW193, respectively, and ligated into pPD49.26-P_hyp-7CeBFP3 through the KpnI site. P_hyp-7CTNS-1::CeBFP and P_hyp-7LAAT-1::CeBFP were constructed by amplifying CTNS-1 and LAAT-1 fragments using primers PPFG227/PPFG228 and PBL516/PBL515, respectively, and cloning them into pPD49.26-P_hyp-7-CeBFP3 through the KpnI site. To construct P_Y105E8B.9Y105E8B.9::mCherry, the promoter region of y105e8b.9 was amplified by primers PJCL257/PJCL258 and cloned to pPD49.26-cherry via the SphI–XmaI sites. The full-length genomic DNA of y105e8b.9 amplified by PJCL249/PJCL439 was then inserted at the KpnI site. To generate P_hspNUC-1::sfGFP::mmCherry, mmCherry was cloned into pPD49.26-sfGFP via the EcoRV–SacI sites, followed by insertion of NUC-1 at the XmaI–KpnI sites and P_hsp at the BamHI site. To generate P_lmp-1LMP-1::sfGFP and P_lmp-2LMP-2::sfGFP, LMP-1 or LMP-2 fragments were cloned into pPD49.26-sfGFP3 through the KpnI site, followed by insertion of the promoter of lmp-1 or lmp-2 amplified using primers PBL520/PBL521 or PDFZ989/PDFZ990 at the PstI–XmaI sites. To generate P_scav-3SCAV-3(G279R)::sfGFP and P_scav-3SCAV-3(S382F)::sfGFP, site-directed mutagenesis was performed to introduce mutations into P_scav-3SCAV-3::sfGFP by primers PYL365/PYL366 and PYL367/PYL368, respectively. P_scav-3SCAV-3::CeBFP was constructed by first ligating the scav-3 genomic DNA amplified with primers PWZ862 and PDFZ721 into pPD49.26 through the NheI–MluI sites, followed by insertion of CeBFP amplified by primers PYL404 and PWZ985 at the MluI–SacI sites. The scav-3 promoter was then amplified by primers PYL405 and PDFZ999 and inserted at the PstI–XmaI sites. Mutations in the glycosylation sites of SCAV-3 were introduced into P_scav-3SCAV-3::sfGFP and P_scav-3SCAV-3::CeBFP by site-directed mutagenesis using primers PDFZ1079/PDFZ1080 (N53A), PDFZ1081/PDFZ1082(N80A), PDFZ1083/PDFZ1084(N118A), PDFZ1087/PDFZ1088(N233A), PDFZ1091/PDFZ1092(N278A), and PDFZ1097/PDFZ1098 (N456A). Mutations in the glycosylation sites of LMP-1 were introduced into P_hyp-7LMP-1::CeBFP by site-directed mutagenesis using primers PDFZ1069/PDFZ1070 (N27A), PDFZ1071/PDFZ1072(N50A), and PDFZ1073/PDFZ1074 (N60A). See Table 1 for a list of primers used for plasmid construction.

Fig. S1 shows that loss of scav-3 affects lysosome morphology and dynamics. Fig. S2 shows the molecular cloning of scav-3. Fig. S3 shows that scav-3(qx193) mutants accumulate damaged lysosomes. Fig. S4 shows that glycosylation is important for the stability and lysosomal targeting of SCAV-3. Fig. S5 shows that overexpression of LMP-1 or reduced activity of daf-2 does not reverse the abnormal lysosome morphology and dynamics in scav-3(qx193). Videos 1–4 show appearance of damaged lysosomes in scav-3 mutants expressing GFP::Gal3 and NUC-1::mCherry (Video 1), GBA-3::mCherry (Video 2), CPL-1::mCherry (Video 3), and CPR-6::mCherry (Video 4).

We thank Drs. K. Ravichandran, M.O. Hengartner, B. Grant, S. Mitani, G. Hermann, T. Yoshimori, and H. E. Buelow for strains and plasmids and Dr. Isabel Hanson for editing services.

This work was supported by the Chinese Ministry of Science and Technology (2016YFA0500203), the National Basic Research Program of China (2013CB910100 and 2014CB849702), the National Natural Science Foundation of China (31325015), and an International Early Career Scientist grant from the Howard Hughes Medical Institute to X. Wang. Some strains were provided by the Caenorhabditis Genetics Center, which is funded by the National Institutes of Health Office of Research Infrastructure Programs (P40OD010440).

The authors declare no competing financial interests.