|Home | About | Journals | Submit | Contact Us | Français|
Late endocytic organelles including lysosomes are highly dynamic acidic organelles. Late endosomes and lysosomes directly fuse for content mixing to form hybrid organelles, from which lysosomes are reformed. It is not fully understood how these processes are regulated and maintained. Here we show that the Caenorhabditis elegans ARL-8 GTPase is localized primarily to lysosomes and involved in late endosome-lysosome fusion in the macrophage-like coelomocytes. Loss of arl-8 results in an increase in the number of late endosomal/lysosomal compartments, which are smaller than wild type. In arl-8 mutants, late endosomal compartments containing endocytosed macromolecules fail to fuse with lysosomal compartments enriched in the aspartic protease ASP-1. Furthermore, loss of arl-8 strongly suppresses formation of enlarged late endosome-lysosome hybrid organelles caused by mutations of cup-5, which is the orthologue of human mucolipin-1. These findings suggest that ARL-8 mediates delivery of endocytosed macromolecules to lysosomes by facilitating late endosome-lysosome fusion.
The late endosomal–lysosomal system is composed of dynamic acidic organelles that repeatedly undergo fusion and fission (Storrie and Desjardins, 1996 ; Mullins and Bonifacino, 2001 ; Luzio et al., 2007 ). Late endosomes fuse with lysosomes transiently or completely, allowing their content to mix; thereby endocytosed macromolecules are degraded by various lysosomal hydrolases. The complete fusion between late endosomes and lysosomes creates hybrid organelles, from which lysosomes can be reformed via fission (Bright et al., 1997 , 2005 ; Pryor et al., 2000 ). The fusion/fission events in late endosomal–lysosomal system also mediate the delivery of newly synthesized lysosomal proteins from the trans-Golgi network (TGN) to lysosomes via late endosomes (Luzio et al., 2003 ). Although it is clear that the fusion/fission must be judiciously regulated to maintain the normal late endosomal–lysosomal system, the molecular mechanisms that regulate the processes are poorly understood.
Many steps of membrane traffic such as budding, transport, and fusion are regulated by small GTPases of the Rab and Arf families (Zerial and McBride, 2001 ; Kahn et al., 2005 ; Gillingham and Munro, 2007 ; Stenmark, 2009 ). Among more than 60 and 20 members of Rab and Arf families, respectively, in mammals, Arl8 is the first small GTPase reported to be primarily localized to lysosomes with some colocalization with late endosomal markers (Bagshaw et al., 2006 ; Hofmann and Munro, 2006 ). Arl8 is highly conserved among multicellular animals, but not in the yeast Saccharomyces cerevisiae (Okai et al., 2004 ). Recent reports have shown that overexpression of Arl8 affects lysosome motility in mammalian cells (Bagshaw et al., 2006 ; Hofmann and Munro, 2006 ); however, the physiological function of Arl8 in the late endosomal–lysosomal system remains uncertain.
In the present study, we performed the genetic and cell biological analysis of an arl-8 mutant using C. elegans and show that ARL-8 is involved in late endosome-lysosome fusion. Loss-of-function of ARL-8 gave rise to supernumerary late endosomal and lysosomal compartments, which were smaller than wild type. Formation of the supernumerary compartments in arl-8 mutants was largely reduced by blocking early-to-late endocytic traffic. Endocytosed macromolecules were transported to late endosomal compartments in arl-8 mutants; however, the compartments failed to fuse with lysosomal compartments enriched in the aspartic protease ASP-1. Furthermore, arl-8 functions upstream to cup-5: loss-of-function of ARL-8 strongly suppressed the enlargement of late endosome-lysosome hybrid organelles caused by mutations of cup-5, the C. elegans orthologue of human mucolipin-1. On the basis of these findings, we propose that ARL-8 mediates delivery of endocytosed macromolecules to lysosomes by facilitating late endosome-lysosome fusion.
Worm cultures and genetic crosses were basically performed according to standard protocols (Brenner, 1974 ). C. elegans strains used in this study are listed in Supplemental Figure S8. The strains were maintained at 15 or 20°C.
The allele arl-8(tm2504) was isolated by the TMP/UV method (Gengyo-Ando and Mitani, 2000 ). The PCR primer sets used in the screening are: external sets, Y57G11C#F10[CTCGACTGGATTCGCAGTCT] and Y57G11C#R7[ATTAGGCTGCAATGCTAGAG]; internal sets, Y57G11C#F11[CTTTAGAGCGGCCAATTCAC] and Y57G11C#R8[CATCCCACAGTGAAAGCTAC]. The isolated strain was outcrossed four times to the wild-type N2 strain and maintained over the balancer chromosome nT1[qIs51]. The homozygous arl-8(tm2504) mutant was distinguished from the heterozygote on the basis of no green fluorescent protein (GFP) fluorescence (nT1[qIs51]) of pharynx when scoring phenotypes.
To construct pYB101 (arl-8::GFP, translational C-terminal GFP fusion), a DNA fragment containing the arl-8 promoter region and the entire exon/intron without a stop codon (1.7 kb) was amplified by PCR with primers KK-388[GAAATAAGCTTGTCGCTGAGAGT] and KK-406[CGGGATCCGCGTTGAGCTTTCGAGTGA] and cloned into C. elegans GFP vector pPD95.77 (a generous gift from A. Fire, Stanford University School of Medicine). To analyze the subcellular localization of ARL-8 in coelomocytes, pYB109 (punc-122::arl-8::mCherry) was constructed as follows: A DNA fragment of mCherry was cloned into NotI/BglII site of a expression vector controlled by unc-122 promoter (pFX-unc-122p; Gengyo-Ando et al., 2006 ) and then a DNA fragment containing the entire exon/intron of arl-8 without a stop codon was cloned into NotI site of the same vector. NotI site in front of the first ATG of the resultant vector was removed by PCR with primers IN-9[TTGGCTATGGTGAATAAGGTTCTCGACT] and IN-8[CATATTGTGAGCCCAATGAAGTAAAATTTC]. To construct pYB102 (arl-8/WT), a DNA fragment containing arl-8 promoter region, the entire exon/intron and 3′ untranslated region (UTR; 2.3 kb) was amplified by PCR with primers KK-388 and TF-18[GGATCCCTGTACAATCACAATTCA] and cloned into pGEM-T Easy vector. pYB104 (arl-8/T34N) and pYB105 (arl-8/Q75L) were made by PCR with primers TF-1[GGAAAAAACACATTTGTCAATGTTATTGCT]/TF-2[AGAGTTTTGGAGCCCAACAAG] and TF-15[GATATCGGGGGCCTTCCACG]/TF-16[CCATAATTTGATCGTCACGTTGCC] using pYB102 (arl-8/WT) as a template DNA, respectively. To construct pYB107 (parl-8::human Arl8a) and pYB108 (parl-8::human Arl8b), human cDNA fragments of Arl8a and Arl8b without the first ATG and stop codon were amplified by PCR with primers TF-35[ATCGCTTTGTTCAACAAGCTGCTG]/TF-36[GCTTCTCCGTGACTTCGAGT] and TF-33[CTGGCGCTCATCTCCCGCCT]/TF-34[GCTTCTTCTAGATTTTGAATGCTGAATAA], respectively. The cDNA fragment was ligated to the DNA fragment amplified by PCR with primers TF-38[TAAACTAGGCACAATCCACGACT] and TF-37[CATCTTGAATATTGATGTTGAATGAGGAA] using pYB102 as a template to make a modified version of pYB102 lacking arl-8 coding region. To express ARL-8::mCherry under heat-shock promoter control, a DNA fragment coding ARL-8::mCherry was amplified by PCR with primers KK-642[GCGGTACCATGTTGGCTATGGTGAATAAGG] and KK-643[GCGGTACCTTACTTGTACAGCTCGTCCA] using pYB109 as a template and cloned into KpnI site of pPD49.78 and pPD49.83 (generous gifts from A. Fire). To express ASP-1::DsRed, a DNA fragment was amplified by PCR with primers Y39B6B#F1[TGGTACCAGAAATTCAGCCTTCTTGTATG] and Y39B6B#R1 [TGCGGCCGCACAATCCCTTGTGGACGGCGG], and cloned into KpnI/NotI site of the expression vector pFX_DsRedxT (Gengyo-Ando et al., 2006 ). pYB116 (punc- 122::asp-1::mCherry) was constructed as follows: a DNA fragment containing the entire exon/intron of asp-1 without a stop codon was amplified by PCR with primers IN-94[GCGGCCGCATGCAGACCTTCGTTTTGCTCG]/IN-95 [GCGGCCGCACAATCCTTGTGGACGGCG]. The DNA fragment was then cloned into NotI site of pFX-unc-122p containing mCherry cDNA. pYB117 (sand-1 RNAi construct) was generated by cloning the DNA fragment of sand-1 into the BglII/PstI sites of vector pPD129.36. A sand-1 DNA fragment was amplified by PCR with primers IN-116[GAAGATCTGATTCATCCTCGTGATTTGAAC]/IN-117[ATTGGCTGCAGGGCTGACATACAGACTTGC]. To express ARL-8 under heat-shock promoter control, a DNA fragment containing the entire exon/intron of arl-8 was amplified by PCR with primers KK-642[GCGGTACCATGTTGGCTATGGTGAATAAGG] and IN-83[GGGGTACCTTAGCGTTGAGCTTTCGAGTG] using N2 genomic DNA as a template and cloned into the KpnI site of pPD49.78 and pPD49.83. We performed germline transformation experiments by injecting various constructs with coinjection markers (Mello et al., 1991 ). See also Supplemental Figure S8 for strains and plasmids used in this study.
The traffic of Texas-Red and Alexa488-conjugated bovine serum albumin (TR-BSA and Alexa488-BSA, Invitrogen, Carlsbad, CA) in coelomocytes was monitored as described previously (Zhang et al., 2001 ). TR-BSA or Alexa488-BSA (1 mg/ml in water) was injected into the body cavity of adult worms and incubated at 15 or 20°C. At the indicated times, worms were subjected to microscopic analysis. For each time point, similar results were obtained with more than five coelomocytes of different worms.
To evaluate rescue of lysosomal morphologies or endocytic traffic in arl-8(tm2504) mutant coelomocytes, we maintained the transgenic animals (YB1143 or YB1306) at 15°C, shifted the adult animals from 15 to 33°C, incubated them for 30 min, and allowed them to recover at 20°C for the indicated time before scoring.
Animals were mounted on 3% agarose pads with 50 mM NaN3 in M9. Differential interference contrast (Nomarski) and fluorescent images were obtained with a Zeiss Axio Imager M1 microscope system equipped with the Axiovision software (Thornwood, NY). Confocal images were acquired with Nikon Eclipse TE2000-E (Melville, NY) equipped with Confocal scanner unit CSU10 (Yokogawa, Tokyo, Japan)/iXon DV887 (Andor, South Windsor, CT) and processed with Andor iQ, Nikon NIS-Elements, and Adobe Photoshop (San Jose, CA) software.
Worms were dissected using 30-gauge needles in a culture medium (137 mM NaCl, 5 mM KCl, 1 mM MgCl2, 2 mM CaCl2, 10 mM HEPES, 5 mM glucose, 2 mM l-asparagine, 0.5 mM l-cysteine, 2 mM l-glutamine, 0.5 mM l-methionine, 1.6 mM l-tyrosine, 26 mM sucrose; Espelt et al., 2005 ) or CO2-independent medium with glutamine (Invitrogen) to release coelomocytes from the pseudocoeloms. The isolated coelomocytes were incubated in 150 μl of medium containing Alexa488-conjugated BSA (10 μg/ml) or Cascade Blue dextran (20 μg/ml) in a collagen-type I (Cellmatrix type I-P, Nitta Gelatin, Tokyo, Japan) coated glass-base dish with grid (Iwaki, Holliston, MA) for 5–10 min at room temperature (~25°C). After washing four times with ice-cold medium, time-lapse fluorescence images were acquired at room temperature using a TCS-SP5 (Leica Microsystems, Deerfield, IL) equipped with a HCX PL APO 100×/1.46 NA oil objective lens. Laser lines at 405, 488, and 561 nm were used for excitation of Cascade Blue, Alexa488/GFP, and Alexa594/DsRed, respectively. Laser power was attenuated to 1.5–2% of maximum to minimize phototoxicity and photobleaching. To prepare figures, single frames were processed with Adobe Photoshop software.
Anti-ARL-8 antibodies were prepared in rabbits against a synthetic peptide (CDITLQWLIDHSKAQR, corresponding to the C-terminus of ARL-8) conjugated with keyhole limpet hemocyanin and purified from whole serum with the peptide-immobilized affinity column (Pierce, Rockford, IL) according to the manufacturer's instructions. Immunoblotting was performed as described previously (Saito et al., 2002 ; Kajiho et al., 2003 ).
Feeding RNA interference (RNAi) was performed as described previously (Kamath et al., 2001 ; Poteryaev et al., 2007 ). In brief, adult worms (P0) were placed on either control (pPD129.36) or sand-1 (pYB117) RNAi plates and removed after 12–24 h. After further incubation for 72 h, F1 worms were transferred to another RNAi plates seeded with the same bacteria. The coelomocyte phenotype was observed in the F2 generation. The worms were maintained at 20°C during RNAi treatment.
Mammals have two closely related Arl8 homologues, Arl8a and Arl8b, whereas C. elegans has only one Arl8 (arl-8, Y57G11C.13; Li et al., 2004 ). We made transgenic animals expressing C-terminally GFP-tagged ARL-8 under the control of the arl-8 promoter. Figure 1, A–C, shows that ARL-8::GFP was widely expressed in many tissues including neurons, muscle, intestine, and coelomocytes in adult worms. Expression of the ARL-8::GFP fusion proteins could rescue the maternal embryonic lethality and coelomocyte phenotypes of arl-8 mutants (Supplemental Figure S1), suggesting that the GFP fusion proteins are functional and are likely to reflect localization of endogenous ARL-8.
ARL-8::GFP labeled the subsets of intracellular compartments in C. elegans coelomocytes (Figure 1C). These cells are macrophage-like scavenger cells that contain many membrane-bound compartments of different sizes and possess high endocytic and degradation activities (Grant and Sato, 2006 ; Fares and Greenwald, 2001a ). To define the subcellular localization of ARL-8 in coelomocytes, we performed double labeling of ARL-8::mCherry with endocytic compartment GFP-markers. ARL-8::mCherry did not overlap with 2xFYVE::GFP (early endosomes; Nicot et al., 2006 ) or RME-8::GFP (early/late endosomes; Zhang et al., 2001 ; Figure 1, D and E). In contrast, most ARL-8::mCherry colocalized well with GFP::RAB-7 (late endosomes/lysosomes; Poteryaev et al., 2007 ) and LMP-1::GFP (lysosomes; Treusch et al., 2004 ; Figure 1, F and G). More precisely, ARL-8::mCherry showed virtually complete overlap with LMP-1::GFP, whereas there were some GFP::RAB-7–positive compartments lacking ARL-8::mCherry (arrows in Figure 1F). These results indicate that, as the mammalian counterpart, C. elegans Arl8 localizes primarily to lysosomes.
To investigate the physiological role of ARL-8 in vivo, we isolated a mutant carrying a deletion in the arl-8 gene, arl-8(tm2504). Western blot analysis using an anti-ARL-8 antibody indicates that the deletion is a strong loss-of-function or null mutation (Supplemental Figure S1). Animals heterozygous for arl-8(tm2504) displayed no observable phenotype and were mainly used as control animals in this study. Homozygous arl-8 animals from heterozygous arl-8(tm2504) mothers developed to fertile adults; however, no viable embryos were produced by hermaphrodites homozygous for arl-8(tm2504), suggesting that ARL-8 is required maternally for embryonic development (not shown). The embryonic lethality was efficiently rescued by introduction of wild-type arl-8 (arl-8/WT) or a constitutively active mutant (arl-8/Q75L) but not by a putative nucleotide-free mutant (arl-8/T34N), suggesting that function of ARL-8 as a guanine nucleotide-binding protein is important for normal embryogenesis (Supplemental Figure S1). Furthermore, expression of human Arl8 proteins in arl-8 mutants could also rescue the embryonic lethality, indicating that Arl8 plays a conserved role in multicellular animals. The basis of the embryonic lethality is under investigation and we focus on the coelomocyte phenotypes in this study.
Figure 2A shows the localization of endocytic compartment markers in arl-8(tm2504) coelomocytes. The morphologies of 2xFYVE- and RME-8-positive compartments in arl-8(tm2504) coelomocytes were indistinguishable from those in control ones. In contrast, the arl-8(tm2504) mutation caused an increase in the number of RAB-7– or LMP-1–positive compartments, most of which were smaller in size compared with those of control animals (Figure 2, B and C). These data suggest that the morphologies of late endosomes/lysosomes, but not early endocytic organelles, are specifically affected in arl-8(tm2504) coelomocytes. To test if the increased number of RAB-7– or LMP-1–positive compartments reflected an increase in concentration of RAB-7 or LMP-1, we quantified fluorescence intensities of GFP::RAB-7 in coelomocytes and found that GFP::RAB-7 levels were not increased but rather reduced slightly in arl-8 mutant coelomocytes than those in control (Supplemental Figure S2).
The aberrant morphologies of RAB-7– or LMP-1–positive compartments in arl-8 mutant coelomocytes were not so obvious until adult stage (not shown), raising the possibility that progressive developmental abnormalities in arl-8 mutants irreversibly impaired their morphologies in coelomocytes. Thus, we examined whether heat-shock promoter–driven expression of arl-8 in adult stages of arl-8 mutants can rescue the phenotypes. We found a decrease in the abnormal size and number of those organelles in the mutant after a brief heat shock compared with siblings that did not have heat-shock–induced expression of ARL-8::mCherry (Figure 2D). These results indicate that the morphologies of RAB-7– or LMP-1–positive compartments are not irreversibly impaired in adult arl-8 mutants but can be restored upon arl-8 expression.
To test if endocytic membrane traffic is involved in formation of the supernumerary membrane compartments in arl-8 mutants, we analyzed a genetic interaction of arl-8 with sand-1, which is required for early-to-late endosomal traffic (Poteryaev et al., 2007 ). In sand-1 mutants, no or little endocytosed macromolecules exit from early endosomes, which results in accumulation of endocytic contents in enlarged endosomes (Poteryaev et al., 2007 ). Indeed, RNAi against sand-1 induced enlarged RME-8–labeled endosomes with no apparent changes in 2xFYVE-labeled endosomes (Supplemental Figure S3). We found that formation of supernumerary LMP-1–labeled compartments in arl-8 mutants was largely suppressed by sand-1(RNAi) (Figure 3), suggesting that the majority of the supernumerary membrane compartments in arl-8 mutants are late endosomal compartments generated via early-to-late endosomal traffic.
We next analyzed endocytic traffic in arl-8 mutant coelomocytes. Coelomocytes actively take up endocytosis markers such as fluorescently labeled proteins that are delivered into the pseudocoelom (body cavity) by transgene expression or microinjection (Fares and Greenwald, 2001a ). Previous studies have shown that TR-BSA injected into the pseudocoelom travels through the RME-8–labeled compartments and finally reaches the LMP-1–labeled compartments in wild-type coelomocytes (Zhang et al., 2001 ; Treusch et al., 2004 ; Gengyo-Ando et al., 2007 ; Poteryaev et al., 2007 ). To analyze early endocytic trafficking in arl-8(tm2504) coelomocytes, we injected TR-BSA into animals expressing RME-8::GFP in coelomocytes. Figure 4A shows that the endocytosed TR-BSA accumulated in the RME-8::GFP–positive compartments within 15 min after injection and then appeared in distinct compartments negative for RME-8::GFP in both control and arl-8(tm2504) coelomocytes by 45 min, indicating that early endocytic traffic is almost normal in arl-8 mutants.
We next investigated late endocytic trafficking in arl-8(tm2504) coelomocytes. ASP-1 is a C. elegans aspartic protease, which is homologous to human lysosomal enzyme cathepsin D (Tcherepanova et al., 2000 ). In control coelomocytes, ASP-1::DsRed localized to virtually all of LMP-1–positive compartments (Figure 4B, top panel). After leaving RME-8–positive compartments, endocytosed macromolecules were transported to RAB-7–positive/ASP-1–negative compartments and then reach to ASP-1–enriched compartments in control animals (Figure 4B, bottom panels, and Supplemental Figure S4), which appears to represent transport from late endosomes to lysosomes. In arl-8 mutants, ASP-1 localized to LMP-1–positive compartments as in control coelomocytes; however, there were number of LMP-1–positive compartments with no or little ASP-1 (Figure 4C, top panels). Interestingly, endocytosed Alexa488-BSA failed to reach to ASP-1–enriched compartments in arl-8 mutants (Figure 4C, bottom panels). In control coelomocytes, endocytosed TR-BSA was finally transported to most of RAB-7– or LMP-1–positive compartments, whereas it localized to a subpopulation of those compartments in arl-8 mutants (Figure 5, A and B). Collectively, these results suggest that endocytosed macromolecules are delivered to RAB-7/LMP-1–positive late endosomal compartments in arl-8 mutants; however, transport from those compartments to lysosomes is compromised.
To analyze the dynamics of endocytic traffic in coelomocytes, we developed a method for time-lapse imaging of isolated coelomocytes from adult animals. We applied Alexa488-BSA to the extracellular medium and recorded the endocytic traffic of Alexa488-BSA to ASP-1–positive compartments in wild-type coelomocytes. Endocytosed Alexa488-BSA accumulated in small compartments outside ASP-1–positive compartments at early time point and then underwent complete fusion with ASP-1–positive compartments, leading to the increase in Alexa488-fluorescence in the ASP-1–positive compartments (Figure 6A, Supplemental Figure S5, and Supplemental Movies 1 and 2). In arl-8 mutants, endocytic compartments concentrating Alexa488-BSA were formed and occasionally contacted with ASP-1–positive compartments as in wild type. Only a small fraction of the interaction events led to appearance of Alexa488-BSA in ASP-1–positive compartments (Supplemental Figure S6); however, no apparent complete fusion between these two compartments was observed during the time courses of our live cell imaging (Figure 6B and Supplemental Movie 3). Quantitative analysis of fusion events revealed that the frequency of complete fusion between these compartments was greatly decreased in arl-8 mutants (Figure 6C), suggesting that arl-8 mutants have a defect in fusion between the Alexa488-BSA–containing compartments and ASP-1–positive compartments.
We next investigated whether transient expression of arl-8 can promote fusion between Alexa488-BSA–containing compartments and ASP-1–positive compartments in arl-8 mutants. arl-8 was expressed under heat-shock promoter in arl-8(tm2504) mutant coelomocytes where Alexa488-BSA and ASP-1::mCherry had been localized to distinct compartments before heat shock. We found that Alexa488-BSA and ASP-1:: mCherry double-positive compartments were formed within 1 h after brief heat shock, and the population of those compartments was increased time-dependently in arl-8 mutants harboring the transgene for heat shock–inducible expression of arl-8 (Figure 7, A and B). Such a phenotype was never observed in siblings without the transgene. These results suggest that arl-8 is involved in promotion of fusion between Alexa488-BSA–containing compartments and ASP-1–positive compartments.
Recent studies have shown that delivery of endocytosed macromolecules to lysosomes is mediated by fusion between late endosomes and lysosomes (Luzio et al., 2007 ). Late endosomes fuse with lysosomes to form hybrid organelles, from which lysosomes can be reformed via fission. We investigated a genetic interaction of arl-8 with cup-5, which is involved in reformation of lysosomes from late endosome-lysosome hybrid organelles (Fares and Greenwald, 2001b ; Hersh et al., 2002 ). cup-5 is the C. elegans orthologue of human mucolipin-1 implicated in mucolipidosis type IV. cup-5 mutant coelomocytes display enlarged vacuoles with properties of late endosome-lysosome hybrid organelles due to a defect in the reformation of lysosomes from the hybrid organelles (Piper and Luzio, 2004 ; Treusch et al., 2004 ). If arl-8 would promote fusion between late endosomes and lysosomes, arl-8 would be required for formation of the enlarged vacuoles in cup-5 mutants. Consistent with this prediction, formation of the enlarged vacuoles of cup-5(ar465) mutants was strongly suppressed by the arl-8(tm2504) mutation (Figure 8A). The same results were obtained using the cup-5 null allele n3194, the quantification of which is shown in Figure 8B. ASP-1::mCherry localized primarily to enlarged vacuoles in cup-5 mutants, whereas it localized to a subpopulation of small compartments in cup-5; arl-8 mutant coelomocytes as observed in arl-8 mutants. These results are consistent with the idea that ARL-8 facilitates late endosome-lysosome fusion.
In the present study, we have shown that arl-8(tm2504) mutants display the unique late endosomal/lysosomal phenotypes that are unprecedented in previous studies: Loss of arl-8 resulted in supernumerary late endosomal and lysosomal compartments that were smaller than wild type. Formation of the supernumerary compartments was largely suppressed by blocking of early-to-late endosomal traffic using sand-1(RNAi). Endocytosed macromolecules were transported to late endosomal compartments in arl-8 mutants; however, the compartments fail to fuse with ASP-1–enriched lysosomal compartments. Finally, loss-of-function of ARL-8 suppresses the enlargement of late endosome-lysosome hybrid organelles caused by mutations of cup-5. A proposed model of endocytic pathway in coelomocytes is shown in Supplemental Figure S7.
There is now compelling evidence that the transport of endocytosed macromolecules to lysosomes is mediated by fusion between late endosomes and lysosomes (Futter et al., 1996 ; Bright et al., 1997 ; Mullock et al., 1998 ; Pryor et al., 2000 ; Ward et al., 2000 ; Bright et al., 2005 ). We have shown that endocytosed Alexa488-BSA–containing compartments completely fused with ASP-1–enriched compartments in wild-type coelomocyte, which may correspond to late endosome-lysosome fusion in mammalian cells. In arl-8 mutants, endocytosed macromolecules were transported to RAB-7–positive/ASP-1–negative compartments as in wild-type coelomocytes; however, these compartments failed to undergo complete fusion with ASP-1–enriched compartments, suggesting a defect in fusion between late endosomes and lysosomes in arl-8 mutants. It should be noted that RAB-7–positive/ASP-1–negative compartments, where endocytosed macromolecules were transported, were positive also for LMP-1 in arl-8 mutants, whereas those of wild-type had no or little LMP-1 (not shown). Late endosomal compartments in arl-8 mutants may thus have distinct properties from those in wild type, although we cannot rule out the possibility of transient localization of LMP-1 to RAB-7–positive/ASP-1–negative compartments in wild-type coelomocytes.
How could ARL-8 regulate fusion between late endosomes and lysosomes? ARL-8 may indirectly regulate SNARE function required for late endosome-lysosome fusion through effecter molecules in a similar manner to those observed for Rab GTPases (Grosshans et al., 2006 ). Interaction of the Rab5 effector EEA1 (a coiled-coil tethering factor) with syntaxin-13 (t-SNARE) is required for homotypic early endosome fusion. The class C-VPS/HOPS tethering complex, an effector of the yeast Rab Ypt7p, binds SNAREs required for vacuole fusion. Alternatively, given that mammalian Arl8 has been shown to be involved in a microtubule-dependent lysosomal transport (Bagshaw et al., 2006 ; Hofmann and Munro, 2006 ), ARL-8 may promote late endosome-lysosome fusion by regulating lysosome motility along microtubules. Indeed, Rab GTPases orchestrate membrane fusion not only by recruiting the components of the fusion machinery, but also by controlling motor-based vesicle transport (Jordens et al., 2005 ). Interaction of Rab5 with Vps34 leads to local early endosomal production of phosphatidylinositol-3-phosphate, which not only attracts EEA1 but also recruits KIF16B (a PX-domain containing kinesin), thereby stimulating plus-end transport of the endosomes (Hoepfner et al., 2005 ). Rab7 also regulates dynein/dynactin-motor recruitment to late endosomes via interaction with the two effectors, RILP and ORP1L (Johansson et al., 2007 ). Although ARL-8/Arl8 localizes primarily to lysosomes in C. elegans coelomocytes/mammalian cells, we cannot rule out the possibility that loss of arl-8 affects late endosomal biogenesis and/or functions critical for fusion with lysosomes because late endosome morphologies appear to be affected in arl-8 mutants. Understanding the molecular event in late endosome-lysosome system that involves ARL-8 requires identification of effectors and other interacting proteins for ARL-8.
Late endosome-lysosome fusion mediates transport of newly synthesized lysosomal enzymes to lysosomes in mammalian cells (Luzio et al., 2003 ). Besides complete fusion, transient fusion between late endosomes and lysosomes (“kiss-and-run”) has also been shown to mediate endocytic delivery to lysosomes in mammalian cells (Bright et al., 2005 ). Considering that transient fusion, together with complete fusion, was rarely observed in arl-8 mutants, there appears to be little, if any, contribution of these fusion events to transport of ASP-1 to LMP-1–positive compartments in arl-8 mutants. ASP-1 may be transported directly to LMP-1–positive lysosomes in a similar manner to the yeast ALP pathway, in which hydrolytic enzymes bypass endosomes and reach directly to the vacuoles, the equivalent of the mammalian lysosome (Cowles et al., 1997 ; Piper et al., 1997 ; Stepp et al., 1997 ).
We previously reported that mammalian Arl8 localized to spindle midbody in mitotic PC12 cells, together with perinuclear localization in the interphase cells and that Arl8 may be involved in chromosome segregation in cultured mammalian and fly cells (Okai et al., 2004 ). No apparent mitotic spindle localization of Arl8 was, however, observed in most mammalian cell lines other than PC12 cells we tested, and no apparent defect in chromosome segregation was detected in arl-8(tm2504) mutant embryos (not shown). Arl8 is thus not likely to make significant contribution to chromosome segregation in most cell types, although we cannot rule out the possibility that Arl8 have a role in the mitotic spindle of certain cell types and/or under some circumstances.
Besides the late endosomal–lysosomal system, what cellular processes would arl-8 be involved in? Considering that phagosome-lysosome fusion contributes to the removal of apoptotic cells during programmed cell death in C. elegans (Kinchen et al., 2008 ; Lu et al., 2008 ; Mangahas et al., 2008 ; Xiao et al., 2009 ), arl-8 may function in clearance of apoptotic cells. In C. elegans embryos, ARL-8 may be localized to lysosome-related yolk granules and involved in the degradation of yolk proteins, which are important sources of nutrition and energy during embryogenesis (Fagotto, 1995 ). Lysosome-related organelles such as yolk granules are a group of cell type–specific compartments that share various features with lysosomes, but are distinct in function, morphology, and composition (Dell'Angelica et al., 2000 ). Gut granules are also intestine-specific lysosome-related organelles in C. elegans (Hermann et al., 2005 ). In mammals, lysosome-related organelles include melanosomes, lytic granules, major histocompatibility complex class II compartments, platelet-dense granules, basophil granules, and neutrophil azurophil granules. Given that recent organelle proteomics studies in mammals have revealed that Arl8 is detected in some lysosome-related organelles (Bagshaw et al., 2005 ; Hu et al., 2007 ), this GTPase may also be crucial for the biogenesis and functions of these lysosome-related organelles.
We gratefully acknowledge B. Grant (Department of Molecular Biology and Biochemistry, Rutgers University) for the RME-8::GFP–expressing strain and R. Y. Tsien (HHMI, University of California, San Diego) for mCherry plasmids. J. Rothman and E. Witze for providing insightful and invaluable comments on this manuscript. E. Kuranaga and Tokyo University Leica Imaging Center for expert technical assistance of time-lapse analysis. and present and former members of the Katada laboratory for useful discussions. Some C. elegans stains used in this study were provided by the Caenorhabditis Genetics Center, which is funded by the National Institutes of Health National Center for Research Resources. This work was supported in part by research grants from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT) of Japan (K.K., T.K., K.G.-A., and S.M.), Japan Society for the Promotion of Science (JSPS; K.K., T.K., K.G.-A., and S.M.), and Japan Science and Technology Agency (JST; K.G.-A. and S.M.).
This article was published online ahead of print in MBoC in Press (http://www.molbiolcell.org/cgi/doi/10.1091/mbc.E09-12-1010) on May 19, 2010.