Báo cáo khoa học: A Caenorhabditis elegans model of orotic aciduria reveals enlarged lysosome-related organelles in embryos lacking umps-1 function potx

A Caenorhabditis elegans model of orotic aciduria reveals enlarged lysosome-related organelles in embryos lacking umps-1 function Steven Levitte 1 , Rebecca Salesky 1 , Brian King 2 , Sage Coe Smith 2 , Micah Depper 1 , Madeline Cole 2 and Greg J. Hermann 1,2 1 Department of Biology, Lewis & Clark College, Portland, OR, USA 2 Program in Biochemistry and Molecular Biology, Lewis & Clark College, Portland, OR, USA Introduction Lysosome-related organelles (LROs) represent a diverse collection of specialized compartments that share features in common with conventional lysosomes [1–3]. LROs perform a variety of important physiologi- cal functions. In mammals, for example, lamellar bodies function in the storage and release of lung Keywords gut granule; lysosome-related organelle; orotic aciduria; UMPS Correspondence G. Hermann, Department of Biology, Lewis & Clark College, 0615 S.W. Palatine Hill Rd, Portland, OR 97219, USA Fax: +1 503 768 7658 Tel: +1 503 768 7568 E-mail: hermann@lclark.edu (Received 16 October 2009, revised 26 December 2009, accepted 5 January 2010) doi:10.1111/j.1742-4658.2010.07573.x Gut granules are cell type-specific lysosome-related organelles found within the intestinal cells of Caenorhabditis elegans. To investigate the regulation of lysosome-related organelle size, we screened for C. elegans mutants with substantially enlarged gut granules, identifying alleles of the vacuolar-type H + -ATPase and uridine-5¢-monophosphate synthase (UMPS)-1. UMPS-1 catalyzes the conversion of orotic acid to UMP; this comprises the two ter- minal steps in de novo pyrimidine biosynthesis. Mutations in the ortholo- gous human gene UMPS result in the rare genetic disease orotic aciduria. The umps-1()) mutation promoted the enlargement of gut granules to 250 times their normal size, whereas other endolysosomal organelles were not similarly affected. UMPS-1::green fluorescent protein was expressed in embryonic and adult intestinal cells, where it was cytoplasmically localized and not obviously associated with gut granules. Whereas the umps-1()) mutant is viable, combination of umps-1()) with mutations disrupting gut granule biogenesis resulted in synthetic lethality. The effects of mutations in pyr-1, which encodes the enzyme catalyzing the first three steps of de novo pyrimidine biosynthesis, did not phenotypically resemble those of umps-1()); instead, the synthetic lethality and enlargement of gut granules exhibited by the umps-1()) mutant was suppressed by pyr-1()).Ina search for factors that mediate the enlargement of gut granules in the umps-1()) mutant, we identified WHT-2, an ABCG transporter previously implicated in gut granule function. Our data suggest that umps-1()) leads to enlargement of gut granules through a build-up of orotic acid. WHT-2 possibly facilitates the increase in gut granule size of the umps-1()) mutant by transporting orotic acid into the gut granule and promoting osmotically induced swelling of the compartment. Abbreviations DAPI, 4¢,6-diamidino-2-phenylindole; DIC, differential interference contrast; GFP, green fluorescent protein; HPS, Hermansky–Pudlak syndrome; LRO, lysosome-related organelle; ODC, orotidine-5¢-monophosphate decarboxylase; OMP, orotidine 5¢-monophosphate; OPRT, orotate phosphoribosyltransferase; RNAi, RNA interference; UMPS, uridine-5¢-monophosphate synthase; V-ATPase, vacuolar-type H + -ATPase. 1420 FEBS Journal 277 (2010) 1420–1439 ª 2010 The Authors Journal compilation ª 2010 FEBS surfactant, and melanosomes act to synthesize and store body pigments [2,4]. Investigation of the genetic basis of Hermansky–Pudlak syndrome (HPS), which is characterized by defects in the formation and function of LROs, has led to the identification of 15 genes act- ing in the trafficking pathways to LROs [5]. Although mutations in HPS genes typically result in a reduced number of LROs, there is a subset of HPS mutations that additionally promote the formation of dramatically enlarged LROs, including melanosomes [6,7] and lamellar bodies [8,9]. Lamellar bodies are similarly enlarged in Tangier disease, which results from defects in the function of the ABC transporter ABCA1 [10]. Most dramatically, nearly every class of LRO is enlarged in patients with Chediak–Higashi syn- drome [11]. In none of these diseases do we clearly understand the mechanistic basis for LRO enlarge- ment, reflecting our lack of insight into the processes that control the stereotypic size and morphology of LROs. Gut granules are intestinal cell-specific LROs found in the nematode Caenorhabditis elegans. In addition to typical lysosomal characteristics, gut granules stain with Nile Red, a marker for hydrophobic material, and contain birefringent and autofluorescent materials, which are uniquely localized to the gut granule [12–16]. Gut granule formation is initiated during early embryogenesis, soon after endoderm specification, and intestinal cells typically contain hundreds of gut granules [12–14]. Gut granule biogenesis requires the activity of conserved genes that function generally in LRO formation, including those encoding the HOPS complex, the AP-3 complex, the ABC transporter PGP-2, and the Rab GTPase GLO-1 [13,15]. Here we describe the results of a genetic screen to identify factors involved in regulating gut granule size, and present a phenotypic, cellular and molecular characterization of one of these genes, umps-1. Results A screen for mutants with enlarged gut granules in embryonic intestinal cells Gut granules are abundant, cell type-specific, LROs that are present within the intestinal cells of C. elegans embryos, larvae, and adults [12,13]. The formation of gut granules is initiated during early embryogenesis, and is directly controlled by the regulatory program governing intestinal cell fate and differentiation in the early C. elegans embryo [14,17,18]. We have been investigating the mechanisms controlling the assembly and morphology of gut granules during embryogenesis in order to identify the primary regulators of these processes. In adult C. elegans intestinal cells, the enlargement of endolysosomal organelles typically results in a Vac (vacuolated appearance) phenotype, characterized by the presence of cytoplasmic vacuoles when visualized with differential interference contrast (DIC) micros- copy. The vacuolization of the adult intestine is associ- ated with enlargement of early endosomes [19], recycling endosomes [20], and late endosomes ⁄ lyso- somes [21,22]. We reasoned that enlargement of gut granules would similarly result in a Vac phenotype. We first analyzed strains known to exhibit enlarged endolysosomal compartments in adult or embryonic intestinal cells for vacuolization of the embryonic intestine. Only one of the mutants, ppk-3()), displayed an embryonic Vac phenotype (Fig. 1D,E; Table 1). We therefore performed a screen for additional mutants that contained vacuoles within embryonic intestinal cells. We identified seven mutants exhibiting a Vac pheno- type. Complementation tests and molecular cloning showed that these mutants were defective in three genes: ppk-3 (one allele), unc-32 (five alleles), and umps-1 (one allele). The ppk-3()) mutant displayed prominent vacuoles in the intestine (Fig. 1D,E; Table 1), as well as in other embryonic cells, as has been reported previously [21]. The ppk-3 gene encodes a phosphatidylinositol-3-kinase that catalyzes the for- mation of phosphatidylinositol 3,5-bisphosphate and is orthologous to PIKfyve in mammals and Fab1p in yeast [21]. Cells lacking the function of these kinases display dramatically enlarged late endolysosomal com- partments [23]. The unc-32()) and umps-1()) mutants contained vacuoles exclusively within embryonic intes- tinal cells (Fig. 1G,H,J,K; Table 1). The unc-32 gene encodes an intestinally expressed V 0 subunit of the vacuolar-type H + -ATPase (V-ATPase) [24,25]. The V-ATPase associates with embryonic gut granules [13], where it functions in acidification [26]. The umps-1 gene encodes UMP synthase (UMPS), which is pre- dicted to function in de novo pyrimidine biosynthesis [27]. We analyzed whether gut granules were enlarged in the vac mutants. Embryonic gut granules contain bire- fringent material [13,14] and the integral membrane ABC transporter PGP-2 [15]. The vacuoles within ppk-3()) intestinal cells did not contain birefringent material (Fig. 1D,E; Table 1). Although PGP-2-marked gut granules were slightly enlarged in ppk-3()) embryos (Fig. 1F; Fig. S1), they did not match the size of vacu- oles present within ppk-3()) embryos (Fig. 1D,F). These observations indicate that the vacuoles visible in S. Levitte et al. UMPS-1 and gut granule size FEBS Journal 277 (2010) 1420–1439 ª 2010 The Authors Journal compilation ª 2010 FEBS 1421 ppk-3()) embryos by DIC microscopy are not gut granules. The ppk-3()) adults displayed a slight enlargement of autofluorescent gut granules (Fig. S1). Thus, ppk-3 plays only a minor role in regulating gut granule size. In contrast, the vacuoles within unc-32()) and umps-1()) embryos contained birefringent material (Table 1; Fig. 1G,H,J,K), and dramatically enlarged PGP-2-containing compartments were present in these mutants (Fig. 1I,L), consistent with gut granule enlargement. Here, we present our analysis of the role that UMPS-1 plays in gut granule formation and morphology; detailed studies of the role that the V-AT- Pase plays in regulating these processes will be described elsewhere. Disrupting the activity of umps-1, a gene that functions in pyrimidine biosynthesis, leads to a Vac phenotype We identified umps-1 as the gene disrupted in the vac mutant zu456 (see Experimental procedures). Promi- nent vacuoles were present within the intestinal cells of umps-1(zu456) embryos from the ‘lima bean’ stage through to hatching (Fig. 1J; Fig. S2). Vacuoles dimin- ished in size and number during the L1 stage, and L2-stage to adult-stage animals exhibited normal intes- tinal morphology (Fig. S2). umps-1(RNAi) led to an embryonic Vac phenotype that was indistinguishable from that caused by umps-1(zu456) (Table 1). Despite the dramatic vacuolization of the embryonic intestine, umps-1(zu456) animals can be maintained as a homo- zygous line. UMPS-1 is orthologous to mammalian UMPS [27], a bifunctional enzyme that catalyzes the two terminal reactions in de novo pyrimidine biosynthesis [28] (Fig. 2A,C). The orotate phosphoribosyltransferase (OPRT) activity of UMPS promotes the conversion of orotic acid to orotidine 5¢-monophosphate (OMP). The OMP decarboxylase (ODC) activity of UMPS catalyzes the formation of UMP from OMP. The C. elegans UMPS-1 protein exhibits both OPRT and ODC activity in vitro [27]. The sequence of umps-1 from zu456 showed a mutation that destroys the pre- dicted translation initiation site (Fig. 2B). Use of the next downstream ATG would result in the formation of a short, out-of-frame peptide. We therefore con- clude that zu456 is probably a null allele of umps-1. The C. elegans gene R12E2.11 codes for a protein homologous to the OPRT domain of human and C. elegans UMPS. In vitro, R12E2.11 has OPRT activ- ity but lacks ODC activity [27], suggesting that it might functionally overlap with UMPS-1. R12E2.11(RNAi) did not result in the formation of embryonic vacuoles (Table 1). In addition, R12E2.11(RNAi) did not obviously alter the forma- tion and size of embryonic vacuoles in umps-1(zu456) A B C D E F G H I JK L DIC Polarization PGP-2 Wild typeppk-3 (n2668)unc-32 (f123)umps-1 (zu456) Fig. 1. Analysis of embryonic vac mutants for gut granule enlargement. Prominent vacuoles visible with DIC microscopy that were lacking in wild type (A) were present within the intestinal cells of ppk-3()) (D), unc-32()) (G) and umps-1()) (J) embryos. The vacuoles (white arrowheads) (D) in ppk-3()) embryos did not contain birefrin- gent material (white arrows) (E), as they did in unc-32()) and umps-1()) embryos (G, H, J, K). (C, F, I, L) PGP-2 staining (marked by white arrows) in pretzel-stage embryos. PGP-2-labeled compartments in ppk-3()) embryos (F) were slightly enlarged in comparison with the wild type (C). In contrast, PGP-2-containing compartments were dramatically enlarged in unc-32()) (I) and umps-1()) (L) embryos. The intestine is flanked by black arrowheads. Embryos are approximately 50 lm in length. UMPS-1 and gut granule size S. Levitte et al. 1422 FEBS Journal 277 (2010) 1420–1439 ª 2010 The Authors Journal compilation ª 2010 FEBS embryos (Table 1) or the persistence of intestinal vacu- oles in umps-1(zu456) larvae (data not shown), sug- gesting that R12E2.11 does not play a major role in regulating the size of intestinal organelles. Moreover, R12E2.11(RNAi) did not result in phenotypes charac- teristic of defects in pyrimidine biosynthesis, suggesting that it is not essential for this process (Table S1 and data not shown). The umps-1(zu456) line, while being viable, exhib- ited partially penetrant embryonic and larval lethality. Fifty-six per cent of umps-1(zu456) embryos failed to hatch (Table S2). In addition, 30% of umps-1(zu456) Table 1. Vacuole formation in embryonic intestinal cells. All strains were grown at 22 °C. Pretzel-stage embryos were scored using DIC microscopy for the presence of vacuoles in embryonic intestinal cells. Large vacuoles were typically ‡ 1.5 lm, and small vacuoles were between 0.8 and 1.4 lm in diameter. Polarization microscopy was used to assess the presence of birefringent material within vacuoles. n, number of embryos scored. Genotype Percentage of embryos with large vacuoles containing birefringent material Percentage of embryos with large vacuoles lacking birefringent material Percentage of embryos with small vacuoles containing birefringent material n Wild type 0 0 0 418 Wild type + 5 mgÆmL )1 uracil 0 0 0 55 Enlarged endolysosomal compartments alx-1(gk275) 00 0 70 cup-5(zu223) a 00 0 31 ppk-3(n2668) 0 100 0 59 ppk-3(ok1150) b 032 0 92 ppk-3(zu443) 0 100 0 90 rab-10(dx2) 00 0 53 rme-1(b1045) 00 0 36 tat-1(kr15) 00 0 54 V-ATPase unc-32(f121) c 26 0 0 125 unc-32(f123) c 28 0 0 68 De novo pyrimidine biosynthesis umps-1(zu456) 100 0 0 > 2000 umps-1(zu456) +5mgÆmL )1 uracil 96 0 4 73 umps-1(zu456) ⁄ umps-1(+) d 00 0 50 umps-1(zu456) · umps-1(+) e 100 0 0 30 umps-1(RNAi) f 75 0 16 227 pyr-1(cu8) 00 0 58 pyr-1(RNAi) f 00 0 32 R12E12.11(RNAi) f 00 0 52 Transgenic rescue g umps-1(zu456)+ WRM0627dD02 53 0 0 43 umps-1(zu456)+ UMPS-1::GFP 0 0 0 31 Double mutants h umps-1(zu456); apt-6(ok429) i 0 0 27 59 umps-1(zu456); glo-1(zu437) j 00 0 22 umps-1(zu456); mrp-4(ok1095) 57 43 0 54 umps-1(zu456); pgp-2(kx55) k 81 0 19 90 umps-1(zu456); wht-2(ok2775) l 63 0 17 104 umps-1(RNAi); wht-2(ok2775) f 80 8 51 umps-1(zu456); wht-2(RNAi) f 36 0 0 42 umps-1(zu456); pyr-1(RNAi) f 10 0 73 umps-1(RNAi); pyr-1(cu8) f 00 0 31 umps-1(zu456); pyr-1(cu8) m 00 0 40 umps-1(zu456); R12E2.11(RNAi) f 100 0 0 31 Mosaic RNAi rrf-1(pk1471) n 00 0 30 rrf-1(pk1471); umps-1(RNAi) 69 0 20 55 S. Levitte et al. UMPS-1 and gut granule size FEBS Journal 277 (2010) 1420–1439 ª 2010 The Authors Journal compilation ª 2010 FEBS 1423 L1-stage larvae did not reach adulthood (Table S2). We found that the overall rate of embryogenesis was delayed in umps-1(zu456) embryos; however, all of the major tissues appeared to be properly specified and to differentiate normally, and there were no obvious developmental defects in umps-1(zu456) embryos prior to the bean stage (data not shown). To determine when embryogenesis was affected in umps-1()) embryos, we monitored the development of individual bean-stage umps-1(zu456) and wild-type embryos. Thirty-five per cent (n = 49) of umps-1(zu456) embryos elongated four-fold, whereas 100% of Table 1. (Continued) Genotype Percentage of embryos with large vacuoles containing birefringent material Percentage of embryos with large vacuoles lacking birefringent material Percentage of embryos with small vacuoles containing birefringent material n rde-1(ne219); [elt-2p::rde-1(+)] n 00 030 rde-1(ne219); [elt-2p::rde-1(+)]; umps-1(RNAi) 00 040 a Embryos scored were the progeny of cup-5(zu223) unc-36(e251) adults derived from a cup-5(zu223) unc-36(e251) ⁄ qC1 line. b Embryos scored were the progeny of + ⁄ szT1[lon-2(e678)]; ppk-3(ok1150) ⁄ szT1 adults. Twenty-five per cent of the embryos were predicted to be ppk- 3()) ⁄ ppk-3()). c The unc-32 alleles analyzed result in zygotic lethality. Therefore, the embryos scored were the progeny of dpy-17(e164) unc- 32()) ncl-1(e1865) ⁄ qC1 dpy-19(e1259) glp-1(q339) adults. Twenty-five per cent of the embryos were predicted to be unc-32()) ⁄ unc-32()). The linked dpy17(e164) ncl-1(e1865) markers did not result in a vacuole phenotype. d The embryos scored were the progeny of umps- 1(+) ⁄ umps-1()) adults. Twenty-five per cent of the embryos were expected to be umps-1()) ⁄ umps-1()). e umps-1(+); mIs11[GFP] males were mated with umps-1(zu456) hermaphrodites, and outcross umps-1()) ⁄ umps-1(+) embryos were recognized by their GFP expression and scored. f The wild type or the indicated strain was grown on plates containing E. coli expressing dsRNA against the listed gene. g Embryos from parents containing extrachromosomal arrays were scored. Owing to lack of segregation of the arrays, not all of the progeny will inherit the transgene [78], so some embryos from parents containing WRM0627dD02 still exhibit the umps-1()) phenotype. Only embryos expres- sing GFP, and therefore having inherited the UMPS-1::GFP array, were scored for intestinal vacuoles. h Of the single mutants ⁄ RNAi exam- ined in the double mutant analysis, only umps-1()) single mutants result in the formation of vacuoles within intestinal cells. i Embryos scored were the progeny of umps-1()); apt-6()) parents, which exhibit 100% maternal effect lethality. j Embryos scored were the progeny of umps-1()) ⁄ umps-1()); glo-1()) ⁄ glo-1(+) parents. The umps-1()); glo-1()) embryos were identified by the loss of the birefringent material phenotype exhibited by glo-1()) embryos [13]. k Embryos scored were the progeny of umps-1()) ⁄ umps-1()); pgp-2()) ⁄ pgp-2(+) parents. Twenty-five per cent of the embryos were expected to be umps-1()); pgp-2()). The double mutants were identified by the loss or reduction in the amount of birefringent material exhibited by pgp-2()) homozygotes. l Embryos scored were the progeny of umps-1()) ⁄ umps-1()); wht- 2()) ⁄ wht-2(+) parents. Twenty-five per cent of the embryos were expected to be umps-1()); wht-2()). m pyr-1(cu8) embryos exhibited reces- sive maternal effect suppression of umps-1(zu456). n The strain was scored when grown on plates expressing F33E2.4-derived dsRNA. F33E2.4 is not required for proper gut granule formation or morphology. W A B C Fig. 2. zu456 disrupts the activity of the bifunctional enzyme UMPS-1, which func- tions in de novo pyrimidine biosynthesis. (A) The C. elegans UMPS-1 protein contains distinct domains that mediate its OPRT and ODC activities. (B) zu456 alters the pre- dicted translation initiation site of umps-1 (underlined in bold); use of the next poten- tial downstream start codon results in the formation of a short out-of-frame peptide. (C) The pathway of de novo pyrimidine bio- synthesis in C. elegans. The proteins that catalyze each reaction are listed beside the arrows. UMPS-1 and gut granule size S. Levitte et al. 1424 FEBS Journal 277 (2010) 1420–1439 ª 2010 The Authors Journal compilation ª 2010 FEBS wild-type embryos (n = 19) did so. Thirty-five per cent (n = 49) of umps-1(zu456) embryos arrested at vari- ous stages between the bean stage and four-fold stage of elongation: 10% arrested at the bean stage, 8% arrested between the 1.5-fold stage and two-fold stage, 12% arrested between the two-fold stage and three- fold stage, and 5% arrested between the three-fold stage and four-fold stage. Interestingly, we found that 30% (n = 49) of umps-1(zu456) embryos lysed, typi- cally prior to elongation. Lysis probably results from umps-1()) embryos being sensitive to the mechanical pressure associated with placing embryos between a 3% agarose pad and a coverslip. These observations indicate that umps-1()) activity is important for embryonic and larval development, and the arrest and lysis phenotypes suggest that umps-1(zu456) compro- mises morphogenesis and the mechanical stability of the embryo. The first three enzymatic activities responsible for de novo pyrimidine biosynthesis in C. elegans are encoded by pyr-1 [29]. The pyr-1()) mutants, like umps-1()) mutants, exhibit partially penetrant embry- onic lethality [31] (Table S2). The lethality of pyr- 1(cu8) is partially suppressed by the addition of uracil [29], which can be converted into UMP via a salvage pathway [30]. Similarly, umps-1(zu456) viability was substantially improved by the addition of uracil to the growth medium (Table S3). Some of the lethality seen in pyr-1()) mutants results from a pharyngeal mor- phogenesis defect that leads to a pharynx-unattached (Pun) phenotype. The Pun phenotype is probably due to loss of de novo formation of UMP that is utilized in proteoglycan synthesis, which is known to be essential for pharyngeal organogenesis [29]. Like pyr-1()) embryos, umps-1()) embryos exhibited a partially penetrant Pun phenotype (Table S1). The phenotypic similarities between umps-1()) and pyr-1()) mutants, together with the recent observation that umps-1(+) activity is necessary for 5-fluorouracil-mediated toxicity in C. elegans [27], a process known to require a func- tional pyrimidine biosynthesis pathway [31], and the in vitro biochemical characterization of UMPS-1 [31], indicate that C. elegans UMPS-1 functions in de novo pyrimidine biosynthesis. Embryonic gut granules are enlarged and not properly formed in umps-1( ) ) embryos We investigated whether the vacuoles present in umps-1()) embryos were enlarged gut granules. The umps-1()) vacuoles contained birefringent material, and PGP-2 was localized to enlarged compartments in umps-1()) embryos, suggesting that they were gut granules (Fig. 1J–L). The integral membrane gut granule-associated proteins PGP-2::green fluorescent protein (GFP) (data not shown) [15] and CDF-2::GFP [32] localized to the limiting membrane of the vacu- oles in umps-1()) embryos (Fig. 3O,P). Comparison of PGP-2-stained compartments in wild-type and umps-1()) pretzel-stage embryos showed average diameters of 0.41 ± 0.02 lm(n = 60) and 2.6 ± 0.05 lm(n = 50), respectively (± standard error of the mean). This represents a more than 250-fold increase in organelle volume in umps-1()) embryos. If the vacuoles in umps-1()) embryos are gut gran- ules, then their formation should depend on genes involved in the formation of gut granules. Mutations disrupting the functions of the Rab GTPase GLO-1 [13], the AP-3 complex subunit APT-6 [13] and the ABC transporter PGP-2 [15] result in a Glo (gut gran- ule loss) phenotype. We constructed umps- 1()); glo()) double mutant embryos, and examined their intestinal cells for vacuoles. The umps-1()); glo-1()) embryos completely lacked vacuoles, and umps-1()); apt-6()) embryos typically lacked vacu- oles (Table 1; Fig. 4D,E). The umps-1( )); pgp-2()) embryos exhibited small vacuoles containing birefrin- gent material (Table 1; Fig. 4F), consistent with the partial defect in gut granule biogenesis seen in pgp-2()) embryos [15]. We conclude that gut granules are enlarged in umps-1(zu456) embryos. The umps-1()) mutation affects the characteristics as well as the size of gut granules. Many gut granules in umps-1()) embryos did not stain with Lysosensor Green DND-189 (Fig. 3H), and none of them were stained by acridine orange (Fig. 3D). Both of these markers of acidification accumulate in wild-type gut granules (Fig. 3B,F). VHA-17, a subunit of the V-ATPase V 0 domain [34], is present on gut granules and the apical surfaces of wild-type intestinal cells (Fig. 3R). Although the apical localization was not altered, VHA-17-labeled compartments similar to those seen in wild type were lacking in umps-1()) embryos (Fig. 5T). Detectable levels of VHA-17 were not asso- ciated with structures resembling enlarged gut granules (Fig. 5T), consistent with the observed defects in gut granule acidification in umps-1()) embryos. Unlike those in wild-type embryos (Fig. 3J), gut granules in umps-1()) embryos did not stain with Nile Red (Fig. 3L). These data demonstrate that the properties of gut granules are dramatically altered in umps-1()) embryos. At present, it is not clear whether this results from a defect in trafficking of material to the gut gran- ule or from a dilution of gut granule constituents due to the dramatic enlargement of gut granule volume and surface area in umps-1()) embryos. S. Levitte et al. UMPS-1 and gut granule size FEBS Journal 277 (2010) 1420–1439 ª 2010 The Authors Journal compilation ª 2010 FEBS 1425 We tested whether the sizes of other endolysosomal compartments were as dramatically altered as those of gut granules in umps-1(zu456) embryos. The morphol- ogy of early endosomal-associated RAB-5::GFP [13] and late endosomal-associated RAB-7::GFP [19] was similar in umps-1(zu456) and wild-type embryos (Fig. 5B,D,F,H). RAB-5::GFP, RAB-7::GFP, and the late endosome ⁄ lysosome-associated LMP-1::GFP proteins, which do not normally associate with gut granules [15,33], were not obviously enriched on the limiting membrane of umps-1()) vacuoles (Fig. 5C,D,G,H,L). Compartments containing LMP- 1::GFP [19] were slightly enlarged in umps-1()) embryos (Fig. 5J,L). Additionally, LMP-1::GFP com- partments in umps-1()) 1.5-fold stage embryos were dispersed throughout the cytoplasm, and did not clus- ter near the apical surfaces of polarized intestinal cells, as seen in wild-type embryos (Fig. S3). It is possible that the altered cytoplasmic distribution of LMP- 1::GFP-containing organelles is a consequence of extremely enlarged gut granules in umps-1()) embryos. LMP-1::GFP is localized to lysosomal com- partments in C. elegans phagocytic cells and coelomo- cytes [35,36]. In C. elegans embryonic intestinal cells, A B C D E F G H I J K L M N O Q P R S T Wild type umps-1 (zu456) DIC Fluorescence DIC Fluorescence DAPI Fluorescence DAPI Fluorescence Acridine orangeLysosensorNile RedCDF-2::GFPVHA-17 Fig. 3. Gut granules are enlarged and their properties are altered in umps-1()) pretzel-stage embryos. In wild-type embryos, gut granules were acidified, being stained by acridine orange (B) and Lysosensor Green (F), contained lipid stained by Nile Red (J), and contained the inte- gral membrane proteins CDF-2::GFP (N) and VHA-17 (R) (gut granules are marked by white arrows in each panel). The vacuoles within umps-1()) embryos did not accumulate acridine orange (D) or Nile Red (L); however, some vacuoles accumulated Lysosensor Green [white arrows in (H)]. The umps-1(zu456) embryos contained greatly enlarged gut granules marked with CDF-2::GFP [white arrows in (P)] and lacked VHA-17-stained compartments within intestinal cells (T). The apical localization of VHA-17 was present in both wild-type and umps-1(zu456) embryos [black arrows in (R) and (T)]. The intestine lies between the black arrowheads in all panels. DAPI, 4¢,6-diamidino-2-phenylindole. UMPS-1 and gut granule size S. Levitte et al. 1426 FEBS Journal 277 (2010) 1420–1439 ª 2010 The Authors Journal compilation ª 2010 FEBS we found that mCherry-tagged CPR-6 and F11E6.1 hydrolases were associated with LMP-1::GFP-contain- ing organelles (Fig. S3). The cpr-6 gene encodes a cathepsin B protease, and F11E6.1 encodes a glucosyl- ceramidase, orthologs of which are found in mamma- lian conventional lysosomes [37]. In umps-1(zu456) A B C D E F Fig. 4. Suppression of umps-1()) vacuole formation. DIC microscopy was used to ana- lyze embryos for intestinal vacuoles, which are prominent in umps-1()) embryos [white arrows in (A)]. The umps-1()); pyr-1()) (B) and umps-1()); wht-2()) (C) embryos lacked vacuoles and elongated normally. The umps-1()); apt-6()) (D) and umps-1()); glo-1(zu437) (E) embryos lacked vacuoles and did not elongate beyond the 1.25-fold stage. The umps-1()); pgp-2()) embryos contained small vacuoles [white arrow in (F)] and arrested elongation prior to the 1.5-fold stage. The umps-1()) embryos display vacuoles from the bean stage through embryogenesis (Fig. S2). White arrowheads (A, B) flank the pharynx of an embryo exhibiting the Pun phenotype. Black arrow- heads flank the intestine in all panels. Wild type umps-1 (zu456) A B C D E F I J K L G H Fig. 5. Analysis of endosomal compartments in umps-1()) embryos. The size and morphology of RAB-5::GFP-labeled endosomes [white arrows in (B) and (D)] and RAB-7::GFP-labeled endosomes [white arrows in (F) and (H)] were similar in wild-type and umps-1()) pretzel-stage embryos. LMP-1::GFP-containing compartments were slightly enlarged in umps-1()) embryos [compare white arrows in (J) and (L)]. Black arrowheads flank the intestine in all panels. S. Levitte et al. UMPS-1 and gut granule size FEBS Journal 277 (2010) 1420–1439 ª 2010 The Authors Journal compilation ª 2010 FEBS 1427 embryos, both proteins were localized to LMP- 1::GFP-labeled compartments, suggesting that these organelles are properly formed in umps-1(zu456) embryos (Fig. S3). Thus, umps-1()) appears to most dramatically affect the formation and morphology of gut granules. A role for the ABC transporter WHT-2 in umps-1( ) ) gut granule enlargement Lysosomal compartments are highly sensitive to osmo- tic stress, showing rapid vacuolization on the accumu- lation of osmotically active material within the lysosomal lumen [38,39]. Therefore, material within the gut granule could have a significant impact on its size. Gut granules contain birefringent material [13,14], cur- rently of unknown composition [33]. As the birefrin- gent material is probably present at a high concentration within the gut granule, we examined its role in the vacuolization of umps-1()) gut granules. Disrupting the function of the ABC transporters MRP-4 and WHT-2 delays the appearance of birefrin- gent material within gut granules, but does not other- wise obviously disrupt gut granule biogenesis [33] (data not shown). The mrp-4()); umps-1()) double mutant embryos displayed normal-sized vacuoles, many of which lacked birefringent material, indicating that the formation of birefringent granules per se is not required for gut granule enlargement (Table 1). We used both wht-2(RNAi) and a wht-2 deletion allele, wht-2(ok2775), to disrupt wht-2(+) activity. In all of the wht-2()); umps-1()) double mutant combi- nations, we examined whether there was a loss of, or a significant reduction in, the number of vacuoles within intestinal cells (Fig. 4C; Table 1). The wht-2(ok2775) allele also partially suppressed the embryonic lethality of umps-1(zu456) (Table S2). Anti-PGP-2 staining showed that gut granule size was reduced from an average diameter of 2.6 ± 0.05 lm(n = 50) in umps-1(zu456) embryos to 0.66 ± 0.04 lm(n = 51) in umps-1(zu456); wht-2(ok2775) double mutants (Fig. 6E). In addition, the gut granules in umps- 1(zu456); wht-2(ok2775) embryos were stained by acridine orange (data not shown). Forty other ABC transporter mutants were unable to suppress the for- mation of vacuoles in umps-1(RNAi) embryos (Table S4). These results indicate that wht-2(+) activ- ity is necessary for the enlargement of gut granules in umps-1()) embryos. The lack of similar suppression by mrp-4()) suggests that wht-2()) mediates this effect through processes independent of the accumula- tion of birefringent material within gut granules. We noticed that many pretzel-stage umps-1()); wht-2()) double mutant embryos exhibited a Pun phenotype. Nearly 50% of umps-1(zu456); wht-2 (ok2775) and umps-1(zu456); wht-2(RNAi) embryos exhibited a Pun phenotype, whereas wht-2( ) ) embryos did not, and only 7% of umps-1(zu456) embryos did (Table S1). We investigated whether the genetic inter- action leading to the Pun phenotype was between wht-2()) and the de novo pyrimidine biosynthetic pathway or was specific to umps-1()). pyr-1(cu8); wht-2(RNAi) embryos did not exhibit a Pun pheno- type (Table S1), indicating that the Pun phenotype of umps-1()); wht-2()) represents a specific genetic interaction between these two genes. The genetic inter- actions between umps-1()) and wht-2()) implicate the WHT-2 ABC transporter in the trafficking of metabo- lites that accumulate in umps-1()) embryos, which ultimately impinge upon gut granule size and pharyn- geal morphogenesis. Analysis of UMPS-1 expression, localization, and function To investigate where UMPS-1 functions and how it might directly regulate gut granule morphology, we expressed a umps-1::gfp gene under control of the 2.7 kb umps-1 promoter. The UMPS-1::GFP fusion rescued the Vac phenotype of umps-1(zu456) embryos (Table 1). UMPS-1::GFP expression was first detected in early pretzel-stage embryos, where it was expressed in the intestine and in a few cells in the head and tail of the animal (Fig. 7A,B). In larval (not shown) and adult stages, UMPS-1::GFP was expressed in the intes- tine and neuronal cells located near the nerve ring and rectum (Fig. 7E,F), which is similar to what has been documented for an umps-1 promoter-driven reporter [27]. The umps-1(zu456) embryos displayed a strict, maternal effect Vac phenotype (Table 1). This could result from metabolic processes involving UMPS-1 at work in the adult intestine that impact on embryonic gut granules. For example, yolk proteins derived from the adult intestine are transferred into oocytes, where they accumulate in the embryonic intestine [40,41]. We performed RNAi on rde-1()); elt-2p::rde-1(+) animals, which are only susceptible to feeding based RNAi in larval and adult intestinal cells [42]. We found that none of the embryos exhibited a Vac phenotype (Table 1), suggesting that inhibiting umps-1(+) in the adult intestine does not impact on embryonic gut granule size. We next considered whether the loss of umps-1 expression in the germline leads to enlarged gut UMPS-1 and gut granule size S. Levitte et al. 1428 FEBS Journal 277 (2010) 1420–1439 ª 2010 The Authors Journal compilation ª 2010 FEBS granules. We analyzed the effects of umps-1(RNAi) on rrf-1(pk1417) animals, which are defective for somatic RNAi but are competent for germline RNAi [43]. We found that rrf-1(pk1417) animals were as sensitive to umps-1(RNAi) as wild-type animals (Table 1). Thus, umps-1 expression in the germline is necessary to prevent the enlargement of embryonic gut granules, suggesting that the maternal effect Vac phenotype of umps-1()) probably results from the maternal contribution of umps-1(+) to embryonic progeny. This could take the form of UMPS-1 protein, umps-1 mRNA, and ⁄ or UMPS-1 metabolic activity in the germline. Mammalian UMPS is localized to the cytoplasm [44,45], and C. elegans UMPS-1 does not contain any obvious organelle targeting or retention motifs, sug- gesting a similar localization. In embryonic intestinal cells, UMPS-1::GFP was distributed throughout the cytoplasm, without any obvious organelle association (Fig. 7A,B). However, we often observed UMPS- 1::GFP near the apical surface of the embryonic intestine (Fig. 7A,B). In adult intestinal cells, UMPS- 1::GFP appeared to be uniformly localized throughout the cytoplasm (Fig. 7C,D). These data suggest that UMPS-1 is a cytoplasmic protein that is not associated with the gut granule. Accumulation of orotic acid probably leads to enlarged gut granules in umps-1( ) ) embryos Mutations that disrupt the function of human UMPS result in orotic aciduria, a disease characterized by megaloblastic anemia, failure to thrive, and urinary excretion of large amounts of orotic acid [6,30,46]. Disrupting the function of the Drosophila UMPS- encoding gene rudimentary-like results in sterility, reduced viability, wing and leg morphological defects, and accumulation of orotic acid [47–49]. Many of the phenotypes resulting from loss of UMPS activity are due to pyrimidine auxotrophy [30,47]. We therefore considered the possibility that A B C D E Fig. 6. Suppression of enlarged gut granules in umps-1()) embryos. Embryos lacking umps-1(+) activity had enlarged gut granules marked with antibodies against PGP-2 [white arrows in (B)]. The pyr-1()) embryos had gut granules that were slightly enlarged [white arrows in (C)] in comparison with wild-type embryos [white arrow in (A)]. The gut granules of umps-1()); pyr-1()) and umps-1()); wht-2()) embryos were dramatically reduced in size [white arrows in (D) and (E)], and were similar in size to gut granules in pyr-1()) embryos [white arrows in (C)]. The intestine of pretzel-stage embryos is flanked by black arrowheads in all panels. S. Levitte et al. UMPS-1 and gut granule size FEBS Journal 277 (2010) 1420–1439 ª 2010 The Authors Journal compilation ª 2010 FEBS 1429 [...]... composition, and function in a murine model of Hermansky–Pudlak syndrome Am J Physiol Lung Cell Mol Physiol 285, L643–L653 Nakatani Y, Nakamura N, Sano J, Inayama Y, Kawano N, Yamanaka S, Miyagi Y, Nagashima Y, Ohbayashi C, Mizushima M et al (2000) Interstitial pneumonia in Hermansky–Pudlak syndrome: significance of florid foamy swelling ⁄ degeneration (giant lamellar body degeneration) of type-2 pneumocytes... of intestinal vacuoles present within ppk-3(n2668) L1-stage larvae were not reduced by hypertonic conditions (data not shown), indicating that exposure to hypertonic medium does not generally alter vacuole morphology ⁄ appearance within larval intestinal cells These data show that increased osmolarity can rapidly and substantially reduce the number of vacuoles in umps-1( zu456) larvae Accumulation of. .. The umps-1( )); pyr-1()) embryos exhibited gut granules that stained with Nile Red and acridine orange (Fig S4), characteristics that are lacking in 1430 Fig 7 Expression and localization of UMPS-1: :GFP Pretzel-stage (A, B) and adult-stage (C–F) umps-1( zu456) animals expressing UMPS-1: :GFP are shown UMPS-1 was expressed in intestinal cells at both stages (the intestine is located between the black arrowheads)... four-fold and were viable In fact, we were able to generate and maintain a viable glo-1(zu437); umps-1( zu456) strain, as long as it was grown on pyr-1(RNAi) feeding plates These results are consistent with gut granules providing a protective function, probably acting to suppress the lethality associated with increased levels of orotic acid in umps-1( )) embryos Discussion Role of the V-ATPase and PPK-3 in. .. Benedetto A, Garnier JM, Schwab Y & Labouesse M (2006) The V0-ATPase mediates apical secretion of exosomes containing Hedgehog-related proteins in Caenorhabditis elegans J Cell Biol 173, 949–961 56 Yan Y, Denef N & Schupbach T (2009) The vacuolar proton pump, V-ATPase, is required for Notch signaling and endosomal trafficking in Drosophila Dev Cell 17, 387–402 57 Wang P, Chintagari NR, Narayanaperumal J, Ayalew... sorting of proteins to the vacuole in Saccharomyces cerevisiae J Biol Chem 267, 3416–3422 62 Ban N, Matsumura Y, Sakai H, Takanezawa Y, Sasaki M, Arai H & Inagaki N (2007) ABCA3 as a lipid transporter in pulmonary surfactant biogenesis J Biol Chem 282, 9628–9634 63 Cheong N, Zhang H, Madesh M, Zhao M, Yu K, Dodia C, Fisher AB, Savani RC & Shuman H (2007) ABCA3 is critical for lamellar body biogenesis in. .. newly hatched L1-stage umps-1( zu456) larvae in media with increasing concentrations of NaCl Incubating umps-1( zu456) larvae in water or 100 mm NaCl did not alter the number or morphology of vacuoles over 45 min (Fig 8A, B; Fig S5) Strikingly, incubation of umps-1( zu456) larvae in 300 mm NaCl led to a five-fold to 10-fold reduction in the number of vacuoles within 15 min (Fig 8C–E) A similar effect was seen... arrowheads) (A, B and E, F), where it is cytoplasmically localized (B, D) In embryonic intestinal cells, UMPS1::GFP was often enriched near the apical surface [white arrows in (A) and (B)] UMPS-1: :GFP was additionally expressed in neuronal cells [white arrowheads in (B) and (F)] and the adult ventral nerve cord [white arrow in (F)] The intestinal lumen is marked with a white arrow in (A) –(D) umps-1( )) embryos. .. umps-1( zu456) embryos Table S1 Analysis of Pun phenotypes Table S2 Analysis of embryonic and larval lethality Table S3 Suppression of umps-1( zu456) lethality by the addition of uracil Table S4 Screening ABC transporters for suppression of vacuole formation in umps-1( )) embryonic intestinal cells This supplementary material can be found in the online online version article Please note: As a service to our authors... formation of intestinal vacuoles at any stage of development (data not shown) However, umps-1( )) larvae and adults do not contain enlarged gut granules (Fig S2), so increased levels of orotic acid might not be expected to alter gut granule size at these stages It is unclear whether orotic acid fed to adults can be transported into oocytes, the route by which it would get into embryos Orotic acid is an anionic . A Caenorhabditis elegans model of orotic aciduria reveals enlarged lysosome-related organelles in embryos lacking umps-1 function Steven. containing birefringent material Percentage of embryos with large vacuoles lacking birefringent material Percentage of embryos with small vacuoles containing birefringent