palmitoyl-protein thioesterase 1 deﬁciency in drosophila ... · are in vivosubstrates of ppt1 are...

Copyright � 2006 by the Genetics Society of AmericaDOI: 10.1534/genetics.105.053306

Palmitoyl-Protein Thioesterase 1 Deficiency in Drosophila melanogaster CausesAccumulation of Abnormal Storage Material and Reduced Life Span

Anthony J. Hickey,*,†,1 Heather L. Chotkowski,* Navjot Singh,* Jeffrey G. Ault,* Christopher A.Korey,‡,2 Marcy E. MacDonald‡ and Robert L. Glaser*,†,3

*Wadsworth Center, New York State Department of Health, Albany, New York 12201-2002, †Department of Biomedical Sciences, State Universityof New York, Albany, New York 12201-0509 and ‡Molecular Neurogenetics Unit, Center for Human Genetic Research,

Massachusetts General Hospital, Boston, Massachusetts 02114

Manuscript received November 9, 2005Accepted for publication January 26, 2006

ABSTRACT

Human neuronal ceroid lipofuscinoses (NCLs) are a group of genetic neurodegenerative diseasescharacterized by progressive death of neurons in the central nervous system (CNS) and accumulation ofabnormal lysosomal storage material. Infantile NCL (INCL), the most severe form of NCL, is caused bymutations in the Ppt1 gene, which encodes the lysosomal enzyme palmitoyl-protein thioesterase 1 (Ppt1).We generated mutations in the Ppt1 ortholog of Drosophila melanogaster to characterize phenotypes causedby Ppt1 deficiency in flies. Ppt1-deficient flies accumulate abnormal autofluorescent storage materialpredominantly in the adult CNS and have a life span 30% shorter than wild type, phenotypes thatgenerally recapitulate disease-associated phenotypes common to all forms of NCL. In contrast, somephenotypes of Ppt1-deficient flies differed from those observed in human INCL. Storage material in fliesappeared as highly laminar spherical deposits in cells of the brain and as curvilinear profiles in cells of thethoracic ganglion. This contrasts with the granular deposits characteristic of human INCL. In addition,the reduced life span of Ppt1-deficient flies is not caused by progressive death of CNS neurons. Nochanges in brain morphology or increases in apoptotic cell death of CNS neurons were detected in Ppt1-deficient flies, even at advanced ages. Thus, Ppt1-deficient flies accumulate abnormal storage material andhave a shortened life span without evidence of concomitant neurodegeneration.

NEURONAL ceroid lipofuscinoses (NCLs) are agroup of fatal genetic neurodegenerative diseases

characterized by the progressive death of central ner-vous system (CNS) neurons and the accumulation ofautofluorescent proteinacious storage material in cells(see Cooper 2003; Haltia 2003; Goebel and Wisniewski

2004; Wisniewski 2005 for reviews). NCL subtypes dif-fer mainly in their age of onset, their rate of progression,and the morphology of the storage material that accu-mulates in the cells of patients. The collective incidenceof the various forms of NCL ranges from 0.1 to 7/100,000(discussed in Wisniewski 2005). There are nine NCLs,each caused by mutation in a specific gene (Mole 2004).Six of these genes are known. Infantile neuronal ceroidlipofuscinosis (INCL) is caused by mutations in theCLN1 gene (now named PPT1), which encodes palmitoyl-protein thioesterase 1 (Vesa et al. 1995). Late-infantileNCL (LINCL) results from mutations in the CLN2

gene, which encodes tripeptidyl peptidase 1 (TPP1)(Sleat et al. 1997). Juvenile NCL and two variant lateinfantile forms of NCL (vLINCL) are caused by muta-tions in the CLN3, CLN5, and CLN6 genes, respectively,all of which encode transmembrane proteins withunknown functions (International Batten Disease

Consortium 1995; Savukoski et al. 1998; Gao et al.2002; Wheeler et al. 2002). Progressive epilepsy withmental retardation (EPMR), the least severe form ofNCL, results from mutations in the CLN8 gene. Patientssuffering from this disease can live out a normal lifespan, but suffer from seizures and progressive mentalretardation (Ranta et al. 1999). The CLN4 and CLN7genes associated with adult NCL and Turkish vLINCL,respectively, have not been identified. It has been dem-onstrated recently, however, that a subset of TurkishvLINCL cases are actually allelic variants of EPMR, whilethe genetic causes of the remaining cases of TurkishvLINCL remain unknown (Ranta et al. 2004). Recently,a novel form of NCL was characterized that has beenclassified as CLN9 (Schulz et al. 2004). The CLN9 genehas yet to be identified. All NCLs can be classified aslysosomal storage disorders (LSDs), because of thecharacteristic inclusions that accumulate in the cells ofpatients with disease (see images in Wisniewski et al.

1Present address: Albany Medical College, Albany, NY 12208.2Present address:Department of Biology, College of Charleston, Charles-

ton, SC 29424.3Corresponding author: Wadsworth Center, New York State Department

of Health, P.O. Box 22002, Albany, NY 12201-2002.E-mail: [email protected]

Genetics 172: 2379–2390 (April 2006)

2001). In addition, a number of the known CLN geneproducts are proteins associated with the lysosome.PPT1 and TPP1 are both proteins with known lyso-somal functions. Battenin, the gene product of CLN3,although its function is unknown, has been demon-strated to localize to the lysosomal membrane (Ezakiet al. 2003).

INCL is the most severe form of NCL and accounts for25% of all NCL cases in the United States (discussed inHofmann et al. 2001). Children born with INCL initiallyappear normal; however, by age 1 they begin to show signsof psychomotor degeneration and progressive vision loss,resulting in complete blindness by age 2. Most childrenwith INCL lose all higher brain functions by age 3 butsurvive until their early to mid teens (Santavuori et al.1973; see also discussion in Wisniewski 2005). CNS neu-rons, mainly those in the cerebral cortex and cerebellum,are selectively killed in INCL, while cells from other tissuesare spared. In addition to neurodegeneration, the cellularaccumulation of autofluorescent storage material in theform of granular osmiophilic deposits (GRODs) is ahallmark of INCL (Haltia et al. 1973). GROD accumu-lation occurs in most, if not all, tissues, including neurons(Mitchison et al. 1998). The primary protein compo-nents of INCL storage material are saposins A and D,proteins involved in the degradation of sphingolipids inthe lysosome (Tyynelaet al. 1993). INCL is unique in thisregard, since the primary protein component that accu-mulates in the storage material of all other NCLs issubunit C of mitochondrial ATPase.

INCL is caused by mutations in the PPT1 gene, whichencodes the soluble lysosomal enzyme palmitoyl-proteinthioesterase 1 (PPT1) (Vesa et al. 1995). PPT1 cleavesthe thioester bond that attaches long-chain fatty acids,predominately palmitate, to the cysteine residues ofS-acylated proteins (Camp and Hofmann 1993; Campet al. 1994). It is clear that the loss of PPT1 activity is whatultimately causes the death of CNS neurons in INCL.What is not clear, however, is how loss of PPT1 activitycauses cell death and why neurons are selectivelysensitive. Many of the disease phenotypes characteristicof INCL clearly result from loss of PPT1’s catabolicfunction in the lysosome, such as the accumulation ofGRODs as well as lipid thioesters and the secondarydysregulation of other lysosomal enzymes (Lu et al.1996; Prasad and Pullarkat 1996). An attractivemodel for the underlying etiology of INCL is thatdisruption of normal lysosomal homeostasis causes theprogressive accumulation of a toxic product or processthat can ultimately kill cells and to which neurons areparticularly sensitive (see discussions in Gupta et al.2001; Lu et al. 2002). The fact that the products of atleast four of the CLN genes are known or suspected tohave lysosomal functions supports this hypothesis. Re-cent evidence suggests that the toxic process may beprogressive dysfunction of the endoplasmic reticulum,leading to activation of the unfolded protein response

and neuronal apoptosis (Zhang et al. 2006). In additionto a catabolic role in the lysosome, however, there isevidence that PPT1 may have extralysosomal functions,specifically in neurons; loss of these functions couldcontribute to disease. PPT1 may localize to presynaptictermini of mammalian synapses, suggestive of a possiblefunction in turnover or regulation of palmitoylatedproteins needed for synaptic transmission (Heinonen

et al. 2000; Lehtovirta et al. 2001; Ahtiainen et al.2003; see, however, Virmani et al. 2005). Consistent witha specific role in the nervous system, PPT1 expression isboth temporally and spatially regulated during devel-opment of the mammalian CNS (Isosomppi et al. 1999;Suopanki et al. 1999). There is also evidence that somefraction of PPT1 may localize to lipid rafts in the plasmamembrane and is involved in regulating apoptosis viaceramide-mediated signaling pathways (Dawson et al.2002; Goswami et al. 2005). While the extralysosomallocalization of PPT1 is intriguing and could be relatedto the underlying etiology of INCL, more in vivoevidence is needed to demonstrate that PPT1 actuallyfunctions at these extralysosomal locations. Finally, it isclear that palmitoylated proteins are a substrate of PPT1in vitro, and while they are also likely to be substratesin vivo, the identities of specific S-acylated proteins thatare in vivo substrates of PPT1 are not known. This limitsour understanding of both the normal in vivo functionsof Ppt1 and the way(s) in which the loss of those func-tions leads to neurodegeneration in INCL.

Drosophila has been used successfully to investi-gate numerous human neurological diseases, includinglysosomal storage disorders (Bonini and Fortini 2003;Driscoll and Gerstbrein 2003; Bilen and Bonini 2005;Dermaut et al. 2005; Huang et al. 2005; Myllykangas

et al. 2005). These fly models can produce neurodegen-eration phenotypes that recapitulate aspects of theirrespective human diseases and provide tractable modelsystems with which to investigate disease etiology andexplore potential therapeutic modalities (Marsh andThompson 2004; Wolfgang et al. 2005). We previouslycharacterized the Ppt1 ortholog in Drosophila as well aspalmitoyl-protein thioesterase 1 enzyme activity as aninitial step in assessing the suitability of using Drosoph-ila as a model system for studying INCL (Glaser et al.2003). Both the Ppt1 gene and Ppt1 enzyme activity areuniformly expressed among different tissues of the flyand throughout all stages of development in a patterntypical of housekeeping genes. A small deletion, Df(1)446-20, was generated as part of the analysis. Df(1)446-20removes Ppt1 along with three neighboring genes. Flieshomozygous for Df(1)446-20 and completely lackingPpt1 enzyme activity are viable and fertile.

In this report, two new point-mutant alleles of Ppt1,along with Df(1)446-20, were used to characterize phe-notypes caused by the loss of Ppt1. Ppt1-deficient fliesaccumulate autofluorescent storage material, a pheno-type characteristic of all NCLs. The osmiophilic deposits

2380 A. J. Hickey et al.

have a highly laminar morphology more similar todeposits seen in Tay–Sachs disease than to the morphol-ogy of GRODs seen in INCL, and the deposits accumu-late almost exclusively in the adult CNS. In addition, thelife span of Ppt1-deficient flies is reduced by �30%, butunlike in human INCL, the life-span reduction is not theconsequence of progressive neurodegeneration. Ppt1-deficient flies show no evidence of alterations in brainmorphology or of apoptotic cell death of CNS neurons,even at extreme ages.

MATERIALS AND METHODS

Drosophila strains and genetics: Flies were maintained oncornmeal-brewer’s yeast-glucose medium at 23� and 55%relative humidity. Creation of the Df(1)446-20 deficiency andthe UAS:DmPpt18.1 cDNA transgene have been described pre-viously (Glaser et al. 2003; Korey and MacDonald 2003).Flies containing the P{Act5C-Gal4}17bFO1 and P{GAL4-elav.L}3transgenes were obtained from the Bloomington Stock Cen-ter. To demonstrate rescue of the aberrant storage phenotypeby the Ppt1 cDNA transgene, Df(1)446-20; UAS:DmPpt8.1/1;P{GAL4-elav.L}3/1 and Df(1)446-20; UAS:DmPpt8.1/1; H2/1flies were created by crossing Df(1)446-20/Y; CyO/1; P{GAL4-elav.L}3/H2 males to Df(1)446-20; UAS:DmPpt8.1/CyO females.

Electron microscopy: Brains and other tissues of interestwere dissected from flies of appropriate genotype and fixedin 6.5% glutaraldehyde in 0.1 m sodium phosphate buffer atpH 7.4 for 3–3.5 hr. The tissue was then washed 2 3 10 min inphosphate buffer and 2 3 10 min in 0.1 m sodium cacodylatebuffer at pH 7.4, postfixed in 1% osmium in cacodylate bufferfor 1 hr, washed 1 3 5 min in water, prestained in 2% uranylacetate for 30 min, washed in water again for 5 min, dehy-drated in an acetone series, and embedded in Epon 812/Araldite. Sections (0.08–0.25 mm thick) of each sample werestained with uranyl acetate and Reynold’s lead and wereobserved using a Zeiss 910 transmission electron microscope.Two to 5 animals of each genotype were examined, and 40–50cells were examined at random from each animal to identifyabnormal ultrastructural deposits. The ages of adult flies usedfor analysis were 1–3 weeks or as indicated in each figure legend.

Isolation of point mutations in Ppt1: Male w1118 flies weremutagenized by feeding them 20 mm ethyl methanesulfonatein 1% sucrose using standard protocols (Grigliatti 1986).Mutagenized males were mated en masse to virgin por1575 oss/FM6, w1 females. Single */FM6 females were recovered, placedindividually into vials, and allowed to mate with FM6/Y males.Progeny from crosses that produced viable */Y male progenywere identified, and a single */Y male was collected from eachvial and assayed for Ppt1 enzyme activity using the fluorogenicsubstrate 4-methylumbelliferyl-6-thiopalmitoyl-b-d-glucoside(4-MU-6S-palm-b-glc). The enzyme assay was modified fromour previous study (Glaser et al. 2003). Briefly, the head wasdissected from each */Yfly and placed into the well of a 96-wellplate. Each well contained 20 ml water plus 10 ml of substrate(0.375 mg/ml 4-MU-6S-palm-b-glc, 0.2 m Na-phosphate/0.1 m

citrate, pH 4.0, 15 mm DTT, 0.09% BSA, and 5 units/mlb-glucosidase). The heads were then homogenized in the 4-MU-6S-palm-b-glc substrate using a 96-well pestle. Plates wereincubated at 30� for 2 hr after which 100 ml of stop buffer wasadded to each well, and the amount of fluorescence at 460 nmwas measured. A total of 2200 viable mutagenized chromo-somes were screened, and two independent strains were foundthat had no detectable Ppt1 enzyme activity. The Ppt1 gene ineach strain was amplified from genomic DNA by PCR using

Deep Vent Polymerase (New England Biolabs, Beverly, MA).Both strands of the gene from each strain were sequenced, anda single GC-to-AT transition mutation was found in each strainproducing the S77F and A179T mutations of Ppt1.RNA interference: To construct a foldback RNAi construct

of Ppt1, we designed two sets of PCR primers to amplifyidentical fragments encompassing the first 800 bp of the Ppt1coding region. One fragment was 8 bp longer than the other toproduce an unpaired region for the hairpin turn in the RNA.The first amplified fragment was flanked by EcoRI and BamHIrestriction sites. The primers used were: 59-CCGGAATTCCAATGATTTCTATCTGCTGT-39 and 59-CGCGGATCCGTAAAGGGCTGGATGACTTT-39. The second amplified fragmentwas flanked by XbaI and BamHI restriction sites. The primersused were: 59-TGCTCTAGAGCAATGATTTCTATCTGCTGT-39and 59-CGCGGATCCTGCTCTCAGTAAAGGGCTGG-39. ThePCR fragments were amplified and cloned using the ZeroBlunt TOPO TA kit (Invitrogen, San Diego). The two frag-ments were digested with XbaI/BamHI and EcoRI/BamHI, asappropriate, and ligated into a 1600-bp concatamer. The con-catamer was then cut with XbaI and EcoRI and cloned intothe pUAST transformation vector. The pUAST:Ppt1-RNAi con-struct was injected into Drosophila embryos, and transformantswere obtained using standard methods (Rubin and Spradling1982). Two independent insertions of the pUAST:Ppt1-RNAiconstruct were recombined onto a single second chromo-some to generateUAS:Ppt1-RNAi8/7/CyO flies.UAS:Ppt1-RNAi8/7/CyO females were crossed to y1 w*; P{Act5C-GAL4}17bFO1/TM6B males to generate the UAS:Ppt1-RNAi8/7/1; P{Act5C-GAL4}17bFO1/1 progeny used for EM analysis.Analysis and quantitation of autofluorescence: Tissues from

animals of the appropriate genotype were dissected in salineand mounted directly in Vectashield (Vector Laboratories,Burlingame, CA). Tissues were examined and photographed onan Axioscope-2 Plus deconvolution microscope using a fluo-rescein filter set (Zeiss, Thornwood, NY). All images areflattened 2-mm Z-stacks collected using Openlab software(Improvision). Pixel intensities were distributed across a rangeof 0–127 arbitrary intensity units. To quantitate the amount ofautofluorescence in the brain, the total number of pixels in aflattened Z-stack image of a whole brain was determined andparsed into 10-unit increments of increasing fluorescenceintensity. All brain preparations used for quantitation wereimaged using the same exposure time and intensity. In addi-tion, the optic lobes were removed from brains used for quan-titation to reduce variability caused by differences in the amountof optic lobe tissue recovered during dissection of the brainfrom the head capsule.Viability curves: Approximately 300 female flies of appro-

priate genotype were collected at 0–2 days of age and weredistributed into vials with �30 females and �5 males per vial.The flies were maintained at 23� and were passaged twice perweek. The number of dead females was counted at each passage.The data were analyzed using the Kaplan–Meier survival analysispackage in StatView 5.0. P-values were calculated using theMantel–Cox log rank statistic.

RESULTS

Loss of Ppt1 causes accumulation of aberrant stor-age material: To determine whether aberrant storagematerial accumulates in flies lacking Ppt1, we dissectedbrains from Df(1)446-20 homozygous female flies andexamined them by electron microscopy. Abnormal de-posits were readily apparent (Figure 1). Unlike theirregular granular deposits (GRODs) seen in human

Phenotypes of Ppt1-Deficient Drosophila 2381

Figure 1.—Df(1)446-20 flies accumulate abnormal storage material. Tissues from Df(1)446-20 flies were analyzed by electronmicroscopy (EM). Osmiophilic laminar deposits were detected in the brains of 5-day (A) and 35-day (B and C) posteclosion fe-males. Deposits in the brains of 14-day posteclosion males were larger and less compacted (D). Deposits reminiscent of curvilinearand rectilinear deposits were detected in 35-day adult thoracic ganglion (E). Rare deposits were detected in the brains of third-instar larvae (F), and less sharply defined deposits were also seen in cells of 35-day adult gut (G). Brains from wild-type flies lackedabnormal deposits, but they did have abundant vacuoles containing fine granular material (H). Enlargements of selected depositsare shown to the right in A, B, F, G, and H. Abnormal storage deposits are indicated by arrows. Mitochondria are indicated byarrowheads. Nuclei are indicated by the letter N. Bars, 1 mm.


INCL, the deposits in Ppt1-null flies were fairly homog-enous in structure, being predominantly spherical inshape and composed of tens to hundreds of concentriclayers of material, each layer being 4–12 nm in thick-ness, and often surrounding a granular core from whichthe laminar material seemed to emanate (Figure 1, Aand B, see also Figure 3, B–D). The deposits are similarin structure to the membranous cytoplasmic bodies ob-served in human Tay–Sachs and Sandhoff diseases, aswell as to deposits observed in flies mutant for the lyso-somal sugar transporter benchwarmer and in flies mutantfor dnpc1a, the fly homolog of human NPC1 that ismutated in Nieman–Pick type C disease (Dermaut et al.2005; Huang et al. 2005; see later discussion). The lam-inar deposits were strongly osmiophilic and were foundin the cytoplasm, typically near the nucleus, in �90% ofthe cell bodies in any given field of view. No depositswere detected in sections from the neuropil regionsof the brain. Individual deposits averaged �0.5 mm indiameter in 5-day-old animals up to 2 mm in 35-day-oldanimals, indicating that the deposits grow in size throughthe accumulation of additional layers of storage mate-rial as the fly ages (compare Figure 1A and 1B). Depositsconsisting of what appear to be aggregates of individualcomponents became more common as the flies aged(Figure 1C). Storage material in the brains of Df(1)446-20 males formed deposits that were consistently largerand more often aggregated and that contained moreprominent gaps and spaces between the layers of ma-terial, as compared to deposits in females (Figure 1D).Laminar deposits like those found in the brain were alsofound in the thoracic ganglion; more rarely, depositswere also detected that had a morphology reminiscentof the curvilinear and rectilinear profiles observed inthe classic late infantile and variant late infantile formsof NCL (Figure 1E). There were no abnormalities in themorphology of other cellular organelles, and the laminardeposits did not appear to be derived from structurallyabnormal mitochondria, as has been reported for Ppt1mutants in Caenorhabditis elegans (arrowheads in Figure1; Porter et al. 2005).

Laminar deposits were also found in some tissuesoutside the adult CNS, including third-instar larval brainand adult gut epithelium (Figure 1, F and G). Comparedto the adult brain, however, these tissues had deposits thatwere smaller, more diffuse, less osmiophilic, and muchless abundant, particularly in larval brain. It is unclearwhether the deposits seen in the larval brain are precur-sors of the deposits seen in adults. No abnormal storagematerial was found in muscle or fat body cells in the adult(data not shown). Finally, no abnormal deposits weredetected in the brains of w1118 or Oregon R control fliesor Df(1)446-20 heterozygotes (Figure 1H and data notshown). Vesicles containing uniform granular contentswere common in brain cells of young wild-type flies butwere not detected in the Df (1)446-20 mutant cells(compare Figure 1H to Figure 1, A–G).

Loss of Ppt1 is necessary and sufficient for accu-mulation of abnormal storage material: The Df(1)446-20 deletion removes three genes in addition to Ppt1(Glaser et al. 2003), so it was necessary to demonstratethat the loss of the Ppt1gene was what specifically causedaccumulation of the abnormal storage material detectedby EM. Initial attempts to rescue the storage materialphenotype in Dp(1)446-20 flies using a genomic clonefrom the Ppt1 locus were unsuccessful. Ppt1 expressionwas negligible no matter where the clone was inserted inthe genome despite inclusion of significant amountsof flanking sequence (our unpublished observation).To circumvent this problem, we used Ppt1 cDNA cloneUAS:DmPpt18.1 to express Ppt1 enzyme activity in Df(1)446-20 flies. UAS:DmPpt18.1 was created for previous studiesand can be controlled using the GAL4/UAS system(Brand and Perrimon 1993; Korey and MacDonald

2003). Pan-neural expression of UAS:DmPpt18.1 was in-duced in Df(1)446-20 flies using an elav:GAL4 driver.The level of Ppt1 enzyme activity measured in extracts ofwhole heads from these flies was 71.8% of the levelobserved in heads from wild-type flies (data not shown).This means that enzyme levels specifically in the brainwere about threefold higher than those of wild type.This is because the brain contributes only about one-quarter of the total Ppt1 enzyme activity that is measuredin a whole-head extract, as estimated by comparingenzyme levels in extracts of dissected brains vs. thoseof whole heads in wild-type animals (our unpublishedobservation). This level of neuron-specific overexpres-sion of Ppt1 did not produce any obvious phenotypes inyoung adult flies, but did reduce viability in animalsolder than 2–3 weeks of age (data not shown). This wasnot unexpected, since overexpression of Ppt1 is knownto be deleterious (Korey and MacDonald 2003).

Brains from Df(1)446-20 flies expressing UAS:DmPpt18.1

were isolated and examined by EM. No laminar de-posits or aberrant storage material of any type weredetected, whereas laminar deposits were readily detectedin brain cells in sibling controls (Figure 2). This resultdemonstrates that deletion of Ppt1 is necessary forthe accumulation of aberrant storage material seen inDf(1)446-20 flies.

To demonstrate that mutation to Ppt1 is sufficient tocause accumulation of abnormal storage material, wegenerated point mutations in the Ppt1 gene using ethylmethanesulfonate mutagenesis and analyzed the mu-tants by EM. Individual male flies hemizygous for mu-tagenizedX chromosomes were screened for Ppt1 enzymeactivity using the fluorogenic substrate 4-methylumbel-liferyl-6-thiopalmitoyl-b-d-glucoside (see materials and

methods). Two mutant alleles of Ppt1were recovered thatproduced no detectable enzyme activity capable of me-tabolizing the methylumbelliferyl substrate. The mutantalleles were then cloned from genomic DNA by PCRand sequenced. Each allele was found to contain a singlepoint mutation that changes a conserved amino acid,


specifically, serine 77 to phenylalanine (S77F) and alanine179 to threonine (A179T). The analogous residues inhuman PPT1 are serine 69 and alanine 171. The A179Tmutation is located within the distal end of the substrate-binding pocket (Figure 3A). The alanine side chain likelycomes in contact with fatty acid substrates, suggesting thatthe alanine-to-threonine mutation could interfere withsubstrate binding. The S77F mutation is located at thebottom of an indentation on a side of the enzyme separatefrom the substrate-binding pocket and the active siteresidues (Figure 3A). The change of serine 77 to a bulkyphenylalanine could reduce Ppt1 stability. In this regard, it

is interesting to note that the side chain of serine 77 isimmediately adjacent to a conserved vicinal disulfidebridge that may be important in Ppt1 structure. Finally,females were created that were trans-heterozygous foreach pairwise combination of the Ppt1 point mutationsand the Df(1)446-20 deletion, specifically, Df(1)446-20/Ppt1S77F, Df(1)446-20/Ppt1A179T, and Ppt1S77F/Ppt1A179T. Brainsfrom these females were isolated and analyzed by EM forthe presence of abnormal storage material. Laminardeposits identical in structure to those observed inDf(1)446-20 animals were found in all three genotypes(Figure 3, B–D). This result provides compelling geneticevidence that it is solely the loss of Ppt1 enzyme activity,and not background mutations, that causes the observedaccumulation of abnormal storage material.

RNAi-mediated knockdown of Ppt1 expression wasalso used to demonstrate that reduction of Ppt1 enzymeactivity causes accumulation of abnormal storage mate-rial. A Ppt1 RNAi transgene that consists of an invertedrepeat of the first 800 bp of the Ppt1 cDNA separated by

Figure 2.—Neuron-specific expression of a Ppt1 cDNA res-cues the abnormal storage material phenotype. Brains fromfemale flies of genotypes Df(1)446-20; UAS:DmPpt18.1/1;P{GAL4-elav.L}3/1 (A) and from sibling control flies of geno-type Df(1)446-20; UAS:DmPpt18.1/1; H2/1 (B) were analyzedby EM. No laminar deposits were detected in Df(1)446-20 fliesexpressing the Ppt1 cDNA, but deposits were readily detectedin sibling controls lacking the GAL4 driver. Abnormal storagedeposits are indicated by arrows. Nuclei are indicated by theletter N. Bars, 1 mm.

Figure 3.—Abnormal storage material accumulates in thebrains of females with point mutations in Ppt1. (A) The loca-tions of the S77F and A179T point mutations of Ppt1 areshown in the crystal structure of bovine PPT1 (Bellizziet al. 2000). The protein backbone is shown in yellow. Palmi-tate in the binding pocket of the enzyme is shown in red. Theactive site of the enzyme is located at the top of the palmitate,in the middle of the enzyme in the orientation displayed. Ser-ine 77 and alanine 179 are shown in purple in spacing-fillingmode, with alanine 179 located at the bottom of the bindingpocket and serine 77 located in an indentation at the top ofthe molecule. Brains from female flies of genotypes Df(1)446-20/Ppt1S77F (B), Df(1)446-20/Ppt1A179T (C), and Ppt1S77F/Ppt1A179T (D) were examined by EM. Laminar deposits identi-cal to those observed in Df(1)446-20 animals were found in allthree trans-heterozygous genotypes.


a short spacer was constructed and inserted into theDrosophila germline (seematerials andmethods). Thetransgene can be controlled using the Gal4/UAS system.The strongest suppression of Ppt1 enzyme activity wasobserved when expression of the RNAi transgene wasdriven using the tissue-general Act5C:Gal4 driver. Levelsof Ppt1 enzyme activity assayed in extracts of femaleheads varied from fly to fly, ranging from 0 to 10% ofwild-type levels (data not shown). Brains were isolatedfrom the heads of such Ppt1 knockdown flies and thenassayed by EM for the presence of abnormal storagematerial. Laminar deposits like those seen in bothDf(1)446-20 and Ppt1 point mutant flies were observed(Figure 4). In comparison to flies containing a mutationin the Ppt1 gene, the RNAi flies had deposits that wereless abundant, being found in �30% of the cells in agiven field of view instead of the 90% seen in Ppt1 nullflies. Also, the deposits seen in RNAi flies most often hadnotably fewer laminar layers, although deposits with alarger number of layers were occasionally observed (Figure4). This result further supports the conclusion that adeficiency of Ppt1 enzyme activity results in the accu-mulation of abnormal storage material in the form oflaminar cytoplasmic deposits.

Loss of Ppt1 causes accumulation of CNS-specificautofluorescent inclusions: Tissues were dissected fromDf(1)446-20 homozygous female flies and were exam-ined by fluorescence microscopy to determine whetherabnormal storage material in Ppt1-mutant flies could bedetected as autofluorescent inclusions. Adult Df(1)446-20 flies clearly accumulated autofluorescent inclusionsin the CNS (Figure 5). All animals, including Drosoph-ila, accumulate autofluorescent material called lipofus-cin as a normal consequence of aging (Porta 2002).The autofluorescent inclusions detected in Df(1)446-20flies, however, were more abundant and larger in sizethan the lipofuscin that accumulated in age-matchedwild-type control flies (Figure 5A). The autofluorescentinclusions, which fluoresced across a wide range ofwavelengths, were found in the adult CNS, including thebrain and ventral ganglia, but not in the larval CNS, orin any other adult tissue examined, or in flies heterozy-gous for the Df(1)446-20 deletion (Figure 5, A and B,and data not shown). Since the autofluorescent inclu-sions were observed in young adults but not in larvae,they must begin forming during metaphorphosis. Toaddress the timing of their appearance, we quantifiedautofluorescent inclusions at various times during meta-morphosis and in adults at different ages. By 36 hr afterpupation, autofluorescent inclusions were detected inthe pupal brain, but no difference was observed betweenDf(1)446-20 and wild-type flies (Figure 6). By 72 hr afterpupation, however, autofluorescent inclusions in theDf(1)446-20 brains were both significantly more abun-dant and significantly brighter than the lipofuscin foundin wild-type brains, and this difference persisted in adultflies (Figure 6).

To demonstrate that mutation to Ppt1 is sufficient tocause autofluorescent inclusions, we isolated brainsfrom Df(1)446-20/Ppt1S77F, Df(1)446-20/Ppt1A179T, andPpt1S77F/Ppt1A179T females and analyzed them for auto-fluorescence. Autofluorescent inclusions were substan-tially more abundant in all three trans-heterozygousgenotypes than in wild-type controls, supporting theconclusion that loss of Ppt1 activity causes the accumu-lation of autofluorescent inclusions (Figure 7). Surpris-ingly, pan-neural expression of the Ppt1 cDNA did notrescue the autofluorescence phenotype, nor did tissue-general expression of the Ppt1 RNAi transgene pheno-copy the autofluorescence. Autofluorescence was rescued,however, in Ppt1-deficient males containingDp(1;Y)578,a Y chromosome containing a duplication of the 8Aregion of the X chromosome where Ppt1 is located (datanot shown). The presence of autofluorescent inclusionsin three different trans-heterozygous females, each hav-ing two independently derived Ppt1 mutations, com-bined with rescue by Dp(1;Y)578 provides strong geneticevidence that Ppt1 deficiency is the cause of the auto-fluorescence phenotype. The relationship between theautofluorescent inclusions and the laminar bodies seenby EM is unclear. While they could be the same deposits,

Figure 4.—RNAi-mediated knockdown of Ppt1 causes ac-cumulation of abnormal storage material. Brains from fliesof genotype UAS:DmPpt1-RNAi8/7/1; P{Act5C-Gal4}17bFO1/1were examined by EM. The majority of abnormal storage de-posits had fewer laminar layers than did deposits in Ppt1mutant flies (A), although deposits with a larger numbersof layers were also observed (B).


it is also possible that they are different types of depositsindependently produced by loss of Ppt1.

Ppt1-mutant flies have a reduced life span but noobvious neurodegeneration: Ppt1-mutant flies are viable

and fertile, but their life span is reduced significantlycompared to that of w1118 controls. The median lifespans of Df(1)446-20/Ppt1S77F, Df(1)446-20/Ppt1A179T, andPpt1A179T/Ppt1S77F females were reduced 42, 37, and 20%,

Figure 5.—Adult Df(1)446-20homozygous flies accumulate au-tofluorescent inclusions in theCNS. (A) Flattened optical Z-stacksof Df(1)446-20 or w1118 brains fromfemale third-instar larvae (3L),1-day-old adult females (1 d), 29-day-old adult females (29 d), and58-day-old adult females (58 d)were obtained by deconvolutionlight microscopy. (B) Autofluores-cent inclusions were also detectedin the thoracic ganglion ofDf(1)446-20 flies, but not in non-CNS tis-sues, including fat body, gut, orovary. The solidly white tissues inthe fat body sample of Df(1)446-20 and in the ovary preparationsare autofluorescent tissues, likethe chorion of developing eggs,and differ from the punctate auto-fluorescent inclusions found inthe CNS.

Figure 6.—Autofluorescent inclusions form during metamorphosis and increase in number and intensity with age. Brains wereisolated at various times postpupation (hours) and posteclosion (days), and flattened optical Z-stacks like those shown in Figure 5were obtained. The amount of autofluorescence in each brain was quantitated by counting the number of pixels in the image in10-unit increments of increasing pixel intensity. Results for brains from female Df(1)446-20 (shaded lines) and w1118 (solid lines)are shown. Each mean and standard deviation value was calculated from three to five independent samples.


respectively (Figure 8 and Table 1). Seeing a life spanreduction in all three trans-heterozygous genotypes isconsistent with mutation to the Ppt1 gene causing theobserved reductions. Effects of genetic background,however, cannot be completely excluded, particularlygiven the modest effects. The greater reduction indeficiency heterozygotes compared to point mutantheterozygotes could indicate that the point mutants ofPpt1 are not null alleles, at least for Ppt1 functionsrelated to reduction in life span. Alternatively, back-ground mutations on the Df(1)446-20 chromosome couldreduce life span dominantly beyond the reductioncaused by loss of Ppt1; however, a reduction in life spanin Df(1)446-20 heterozygotes has not been observed.The median life spans of male flies hemizygous for

Df(1)446-20, Ppt1S77F, or Ppt1A179T were reduced 32, 29,and 21%, respectively, compared to those of w1118 controls(P , 0.001; data not shown). While this result is con-sistent with the idea that loss of Ppt1 reduces viability inadult males, hemizygosity of the male X chromosomeprevents us from drawing strong conclusions. Finally, wewere unable to determine if pan-neural expression ofthe UAS:DmPpt18.1 cDNA can rescue the reduction in lifespan. elav-driven expression of the cDNA causes a re-duction in viability on its own, likely due to overexpres-sion of Ppt1 enzyme activity, which is known to bedeleterious (Korey and MacDonald 2003). The reduc-tion in viability caused by Ppt1 overexpression precludesthe ability to detect rescue of the reduced life spancaused by Ppt1 deficiency.

To determine whether Ppt1-mutant flies might alsohave a neurodegeneration phenotype that correlateswith the reduction in life span, the gross morphology ofbrains isolated from Ppt1-mutant flies was assessed inhemotoxylin/eosin-stained paraffin sections and in wholebrains from Ppt1-mutant animals containing GFP markergenes that highlight specific brain structures such as themushroom bodies. No consistent alteration of brainmorphology was observed in Ppt1-mutant flies, andno overt evidence of neurodegeneration was detected,such as vacuolization, brain shrinkage, or a reduction inthe number of nuclei (data not shown). Furthermore,we found no evidence of elevated levels of apoptosis inflies as old as 60 days as assayed by terminal deoxynu-cleotidyltransferase-mediated nick end labeling or byimmunolocalization of H2A.X phosphorylation (datanot shown; Madigan et al. 2002). There was also no evi-dence for degradation of ommatidial structure, as assayedby the deep pseudopupal technique (data not shown).These results suggest that loss of Ppt1 in flies does notcause neurodegeneration and that neurodegenerationdoes not cause the observed reduction in life span.

Figure 7.—Mutation to Ppt1 issufficient to cause autofluores-cence. Brains from female fliesof genotypes Df(1)446-20/Ppt1S77F

(A), Df(1)446-20/Ppt1A179T (B),Ppt1S77F/Ppt1A179T (C), and w1118

(D) were examined by deconvolu-tion light microscopy. Autofluo-rescent inclusions like thoseobserved in Df(1)446-20 animalswere readily detected in all threetrans-heterozygous genotypes.

Figure 8.—Ppt1 deficiency reduces adult viability. Survivalcurves of Df(1)446-20/Ppt1S77F (triangles), Df(1)446-20/Ppt1A179T (squares), Ppt1A179T/Ppt1S77F (diamonds), and w1118

(circles) are shown. The number of flies that survived was de-termined every 3–4 days for the length of the experiment.Mean, median, and maximum life spans and P- and n-valuescalculated from these curves are shown in Table 1.


DISCUSSION

Ppt1-mutant flies accumulate abnormal storage ma-terial: Mutation of the Drosophila Ppt1 gene causes theage-dependent accumulation of abnormal storage ma-terial, a characteristic feature of all NCLs, and of lyso-somal storage diseases in general. The storage materialthat accumulates in the brains of Ppt1-mutant flies, how-ever, has a very different morphology from that of theGROD that accumulates in human INCL. The depositsare highly laminar in structure, most similar in mor-phology to the membranous cytoplasmic bodies thataccumulate in Tay–Sachs and Sandhoff diseases (seeimages in Itoh et al. 1984; Perl 2001). In addition, thecurvilinear deposits detected in the thoracic ganglionare most similar to the curvilinear profiles characteristicof late infantile NCL (Figure 1E). It is not unknown forhuman NCL patients to have deposits of differing mor-phologies in different cell types. Most interestingly, thereis even a case of adult-onset NCL in which laminar de-posits similar in structure to those reported here wereintermingled with, and continuous with, curvilinear de-posits in neurons of the patient’s spinal anterior horn(Iseki et al. 1987). In addition to being morphologicallydifferent, the storage material in Ppt1-mutant flies mayalso be biochemically different from the deposits seenin INCL. The storage material in Ppt1-mutant flies couldnot be detected using a variety of lipophilic stains orperiodic acid Schiff, stains that readily detect the storagematerial produced in all NCLs, including INCL (ourunpublished observations). The differences in morphol-ogy and biochemical composition between the storagematerial produced in Ppt1-mutant flies and the GRODseen in INCL patients suggest that the loss of Ppt1 hasdifferent biochemical consequences in flies and inmammals. Such differences could arise from differenceseither in the substrates of Ppt1 or in more general as-pects of lysosome function and homeostasis that in-directly affect the downstream consequences of Ppt1loss. The latter possibility is supported by the observa-tion that the upregulation of other lysosomal enzymes,typically seen in NCLs, does not appear to occur in Ppt1-mutant flies. Levels of b-glucuronidase and hexosamin-idase A are significantly elevated in INCL patients andin Ppt1-knockout mice, but they are unchanged inPpt1-mutant flies (our unpublished observation; Prasad

and Pullarkat 1996; Griffey et al. 2004). Finally, thegeneral idea that Ppt1-mutant phenotypes vary amongorganisms because of species-specific differences inPpt1 substrates and/or lysosomal catabolic pathways isalso supported by the phenotype of Ppt1 mutants in C.elegans. Loss of Ppt1 in C. elegans causes mitochondrialpathology without accumulation of aberrant storagematerial, a phenotype very different from the pheno-types observed in either flies or mammals and suggestiveof unique functions for Ppt1 in worms (Porter et al.2005).

No evidence of a neurodegeneration phenotype wasfound in Ppt1-mutant flies. Gross brain morphology andrates of apoptotic cell death in the brain were unalteredeven in aged Ppt1 mutants. The reason why Ppt1 de-ficiency leads to neural cell death in mammals but notin flies is unknown. It is possible that flies simply do notlive long enough for some pathological byproduct orprocess caused by the loss of Ppt1 to reach toxic levels.Alternatively, Ppt1 might have an essential function inmammalian neurons that is lacking in flies. For exam-ple, PPT1 may localize to synaptic vesicles rather than tolysosomes in cultured mouse neurons, suggesting thatPPT1 plays an important role in turnover of palmitoy-lated signaling molecules at the synapse (Heinonen

et al. 2000; Lehtovirta et al. 2001; Ahtiainen et al.2003; see, however, Virmani et al. 2005). The subcellularlocalization of Ppt1 in Drosophila is not known; how-ever, if this neural-specific function of Ppt1 were lackingor nonessential in flies, then loss of Ppt1 might be with-out consequence for synaptic function. While we havegenerated Ppt1-specific antibodies to address this ques-tion, we have thus far been unable to detect endoge-nous Ppt1 on Western blots or by immunocytology. Theabsence of an obvious neurodegeneration phenotypedoes not preclude the possibility of more subtle abnor-malities in neural function. Ppt1-mutant flies have ashortened life span, suggesting that loss of Ppt1 haspathological consequences sufficient to cause an age-dependent reduction in viability (Figure 8 and Table 1).In this context, it is interesting to note that abnormalstorage material accumulates almost exclusively in theCNS, suggesting that fly neurons, like human neurons,are preferentially sensitive to Ppt1 loss (Figures 1 and 5).Thus, it would not be surprising if the observed re-duction in viability of Ppt1-mutant flies is a consequence

TABLE 1

Ppt1-deficient flies have reduced life span

Genotype Mean (%D) Median (%D) Max (%D) P-value n

w1118 77 83 104 NA 335Df(1)446-20/Ppt1S77F 51 (�34) 48 (�42) 97 (�7) ,0.0001 318Df(1)446-20/Ppt1A179T 55 (�29) 52 (�37) 104 (0) ,0.0001 286Ppt1A179T/Ppt1S77F 65 (�16) 66 (�20) 104 (0) ,0.0001 291


of pathology specifically in the CNS as opposed to othertissues. A primary focus of future studies will be to de-termine whether Ppt1-mutant flies have phenotypes in-dicative of age-dependent loss of neural function.

Drosophila NCL models: Mutation to two lysosomalgenes of Drosophila, in addition to Ppt1, causes diseasephenotypes associated with NCLs. Cathepsin D (cathD)encodes a lysosomal aspartic proteinase that, when mu-tated in sheep, causes ovine NCL. Ovine NCL exhibitsphenotypes like those of human NCLs, including ac-cumulation of autofluorescent storage material andneurodegeneration. cathD-mutant flies accumulate au-tofluorescent storage material and show an age-dependentincrease in the number of apoptotic cells in the brain,although the increased cell death does not produce areduction in life span (Myllykangas et al. 2005). Thefact that mutations in two different NCL-related homo-logs in Drosophila, Ppt1 and cathD, both recapitulate atleast some phenotypes associated with human diseasesuggests that the molecular mechanisms that cause thosephenotypes in NCLs may be conserved in flies. TheDrosophila benchwarmer gene (bnch; allelic to the genespinster) encodes a lysosomal transmembrane carbohy-drate transporter that, when mutated, causes accumu-lation of autofluorescent storage material as well asneurodegeneration. The latter effect is likely a conse-quence, at least in part, of disrupted endosome-to-lysosome transport pathways (Nakano et al. 2001; Sweeney

and Davis 2002; Dermaut et al. 2005). Human homo-logs of bnch are not known to be NCL genes but aremutated in lysosomal storage diseases caused by de-fective efflux of lysosomal substrates, particularly asoccurs in sialic acid storage disease (discussed in Dermaut

et al. 2005). The abnormal deposits that accumulate inPpt1-mutant flies are similar in structure to a class ofdeposits seen in bnch mutants (Dermaut et al. 2005).The fact that mutations in three different lysosomalproteins in Drosophila—Ppt1, cathD, and bnch—all causethe accumulation of abnormal autofluorescent storagematerial suggests that at least some consequences ofperturbation to lysosomal function are conserved be-tween flies and humans. Storage material accumulationalone, however, is clearly not sufficient to cause neuro-degeneration, as evidenced by the lack of neurode-generation and minimal cell death in Ppt1 and cathDmutants, respectively (this report; Myllykangas et al.2005). A lack of correlation between the severity ofstorage material accumulation and the degree of neu-rodegeneration has also been noted in human NCLdisease and has been used to suggest that accumulationof storage material and neurodegeneration are separatemanifestations caused by the loss of particular CLNgenes (see discussion in Oswald et al. 2005).

A primary focus for future studies will be the iden-tification of genes that can dominantly modify the phe-notypes produced by Ppt1 mutations. Identification ofsecond-site modifier genes may provide insight into the

normal functions of Ppt1, as well as into how the loss ofthose functions could contribute to disease. In addition,determination of whether such loci also modify pheno-types caused by mutations in cathD or bnch will help usto differentiate the modifiers specific for Ppt1-relatedfunctions from the modifiers that affect phenotypes com-mon to lysosomal storage disease mutations in general.

We acknowledge technical support by the Wadsworth Center’sElectron Microscopy and DAI Light Microscopy Core Facilities, andwe thank the Bloomington Drosophila Stock Center for fly stocks andthe anonymous reviewers for helpful comments. This work wassupported by National Institutes of Health grants NS44572 to R.L.G.and NS33648 to M.E.M.

LITERATURE CITED

Ahtiainen, L., O. P. Van Diggelen, A. Jalanko and O. Kopra,2003 Palmitoyl protein thioesterase 1 is targeted to the axonsin neurons. J. Comp. Neurol. 455: 368–377.

Bellizzi, J. J. I., J. Widom, C. Kemp, J. Y. Lu, A. K. Das et al.,2000 The crystal structure of palmitoyl protein thioesterase1 and the molecular basis of infantile neuronal ceroid lipofusci-nosis. Proc. Natl. Acad. Sci. USA 97: 4573–4578.

Bilen, J., and N. M. Bonini, 2005 Drosophila as a model of humanneurodegenerative disease. Annu. Rev. Genet. 39: 153–171.

Bonini, N. M., and M. E. Fortini, 2003 Human neurodegenerative dis-ease modeling using Drosophila. Annu. Rev. Neurosci. 26: 627–656.

Brand, A. H., and N. Perrimon, 1993 Targeted gene expression as ameans of altering cell fates and generating dominant pheno-types. Development 118: 401–415.

Camp, L. A., and S. L. Hofmann, 1993 Purification and properties ofa palmitoyl-protein thioesterase that cleaves palmitate fromH-Ras. J. Biol. Chem. 268: 22566–22574.

Camp, L. A., L. A. Verkruyse, S. J. Afendis, C. A. Slaughter and S. L.Hofmann, 1994 Molecular cloning and expression of palmitoyl-protein thioesterase. J. Biol. Chem. 269: 23212–23219.

Cooper, J. D., 2003 Progress towards understanding the neurobiol-ogy of Batten disease or neuronal ceroid lipofuscinosis. Curr.Opin. Neurol. 16: 121–128.

Dawson, G., S. A. Dawson, C. Marinzi and P. E. Dawson,2002 Anti-tumor promoting effects of palmityol:protein thioes-terase inhibitors against a human neurotumor cell line. CancerLett. 187: 163–168.

Dermaut, B., K. K. Norga, A. Kania, P. Verstreken, P. Hongling

et al., 2005 Aberrant lysosomal carbohydrate storage accompa-nies endocytic defects and neurodegeneration in Drosophilabenchwarmer. J. Cell Biol. 170(1): 127–139.

Driscoll, M., and B. Gerstbrein, 2003 Dying for a cause: inverte-brate genetics takes on human neurodegeneration. Nat. Rev.Genet. 4: 181–194.

Ezaki, J., M. Takeda-Ezaki, M. Koike, Y. Ohsawa, H. Taka et al.,2003 Characterization of Cln3p, the gene product responsiblefor juvenile neuronal ceroid lipofuscinosis, as a lysosomal inte-gral membrane glycoprotein. J. Neurochem. 87: 1296–1308.

Gao, H., R.-M. N. Boustany, J. A. Espinola, S. L. Cotman, L. Srinidhiet al., 2002 Mutations in a novel CLN6-encoded transmem-brane protein cause variant neuronal ceroid lipofuscinosis inman and mouse. Am. J. Hum. Genet. 70(2): 324–335.

Glaser, R. L., A. J. Hickey, H. L. Chotkowski and Q. Chu-LaGraff,2003 Characterization of Drosophila palmitoyl-protein thioester-ase 1. Gene 312: 271–279.

Goebel, H. H., and K. E. Wisniewski, 2004 Current state of clinical andmorphological features in human NCL. Brain Pathol. 14: 61–69.

Goswami, R., A. J. Kilkus, T. Han, S. A. Dawson and G. Dawson,2005 Differential regulation of ceramide in lipid-rich microdo-mains (rafts): antagonistic role of palmitoyl:protein thioesteraseand neutral sphingomyelinase 2. J. Neurosci. Res. 81: 208–217.

Griffey, M., E. Bible, C. Vogler, B. Levy, P. Gupta et al., 2004 Adeno-associated virus 2-mediated gene therapy decreases autofluores-cent storage material and increases brain mass in a murine modelof infantile neuronal ceroid lipofuscinosis. Neurobiol. Dis. 16:360–369.


Grigliatti, T., 1986 Mutagenesis, pp. 39–58 in Drosophila: A Practical Ap-proach, edited by D. B. Roberts. IRL Press, Oxford/Washington, DC.

Gupta, P., A. A. Soyombo, A. Atashband, K. E. Wisniewski, J. M.Shelton et al., 2001 Disruption of PPT1 or PPT2 causes neuro-nal ceroid lipofuscinosis in knockout mice. Proc. Natl. Acad. Sci.USA 98: 13566–13571.

Haltia, M., 2003 The neuronal ceroid-lipofuscinoses. J. Neuro-path. Exp. Neurol. 62(1): 1–13.

Haltia, M., J. Rapola, P. Santavuori and A. Keranen, 1973 In-fantile type of so-called neuronal ceroid lipfuscinosis. Part 2.Morphological and biochemical studies. J. Neurol. Sci. 18: 269–285.

Heinonen, O., A. Kyttala, E. Lehmus, T. Paunio, L. Peltonen et al.,2000 Expression of palmitoyl protein thioesterase in neurons.Mol. Genet. Metab. 69: 123–129.

Hofmann, S. L., A. K. Das, J.-Y. Lu and A. A. Soyombo, 2001 Posi-tional candidate gene cloning of CLN1, pp. 69–92 in Batten Dis-ease: Diagnosis, Treatment, and Research, edited by K. E. Wisniewski

and N. Zhong. Academic Press, San Diego.Huang, X., K. Suyama, J. Buchanan, A. J. Zhu and M. P. Scott,

2005 A Drosophila model of the Niemann-Pick type C lysosomestorage disease: dnpc1a is required for molting and sterol homeo-stasis. Development 132: 5115–5124.

International Batten Disease Consortium, 1995 Isolation of anovel gene underlying Batten Disease, CLN3. Cell 82: 949–957.

Iseki, E., N. Amano, S. Yokoi, Y. Yamada, K. Suzuki et al., 1987 Acase of adult neuronal ceroid-lipofuscinosis with the appearanceof membranous cytoplasmic bodies localized in the spinal ante-rior horn. Acta Neuropathol. 72: 362–368.

Isosomppi, J., O. Heinonen, J. O. Hiltunen, N. D. Greene, J. Vesaet al., 1999 Developmental expression of palmitoyl proteinthiesterase in normal mice. Dev. Brain Res. 118: 1–11.

Itoh, H., J. Tanaka, Y. Morihana and T. Tamaki, 1984 The finestructure of cytoplasmic inclusions in brain and other visceralorgans of Sandhoff disease. Brain Dev. 6: 467–474.

Korey, C. A., and M. E. MacDonald, 2003 An over-expressionsystem for characterizing Ppt1 function in Drosophila. BMCNeurosci. 4: 30.

Lehtovirta, M., A. Kyttala, E.-L. Eskelinen, M. Hess, O. Heinonen

et al., 2001 Palmitoyl protein thioesterase (PPT) localizes intosynaptosomes and synaptic vesicles in neurons: implications forinfantile neuronal ceroid lipofuscinosis (INCL). Hum. Mol.Genet. 10: 69–75.

Lu, J. Y., L. A. Verkruyse and S. L. Hofmann, 1996 Lipid thioestersderived from acylated proteins accumulate in infantile neuronalceroid lipofuscinosis: correction of the defect in lymphoblasts byrecombinant palmitoyl-protein thioesterase. Proc. Natl. Acad.Sci. USA 93: 10046–10050.

Lu, J.-Y., L. A. Verkruyse and S. L. Hofman, 2002 The effects oflysosomotropic agents on normal and INCL cells provide furtherevidence for the lysosomal nature of palmitoyl-protein thioester-ase function. Biochim. Biophys. Acta 1583: 35–44.

Madigan, J. P., H. L. Chotkowski and R. L. Glaser, 2002 DNAdouble-strand break-induced phosphorylation of Drosophila his-tone variant H2Av helps prevent radiation-induced apoptosis.Nucleic Acids Res. 30(17): 3698–3705.

Marsh, J. L., and L. M. Thompson, 2004 Can flies help humanstreat neurodegenerative disease? BioEssays 26(5): 485–496.

Mitchison, H. M., S. L. Hofmann, C. H. Becerra, P. B. Munroe,B. D. Lake et al., 1998 Mutations in the palmitoyl-protein thio-esterase gene (PPT; CLN1) causing juvenile neuronal ceroidlipofuscinosis with granular osmiophilic deposits. Hum. Mol.Genet. 7: 291–297.

Mole, S. E., 2004 The genetic spectrum of human neuronal ceroidlipofuscinoses. Brain Pathol. 14: 70–76.

Myllykangas, L., J. Tyynela, A. Page-McCaw, G. M. Rubin, M. J.Haltia et al., 2005 Cathepsin D-deficient Drosophila recapitu-late the key features of neuronal ceroid lipofuscinosis. Neuro-biol. Dis. 19(1–2): 194–199.

Nakano, Y., K. Fujitani, J. Kurihara, J. Ragan, K. Usui-Aoki et al.,2001 Mutations in the novel membrane protein spinster inter-fere with programmed cell death and cause neural degenerationin Drosophila melanogaster. Mol. Cell. Biol. 21(11): 3775–3788.

Oswald, M. J., D. N. Palmer, G. W. Kay, S. J. A. Shemilt, P. Rezaieet al., 2005 Glial activation spreads from specific cerebral foci

and precedes neurodegeneration in presymptomatic ovine neu-ronal ceroid lipofuscinosis (CLN6). Neurobiol. Dis. 20: 49–63.

Perl, D. P., 2001 Barney Sachs and the history of the neuropatholog-ical description of Tay-Sachs Disease, pp. 11–24 in Tay-SachsDisease, edited by R. J. Desnick and M. M. Kaback. AcademicPress, San Diego.

Porta, E. A., 2002 Pigments in aging: an overview. Ann. NY Acad.Sci. 959: 57–65.

Porter, M. Y., M. Turmaine and S. E. Mole, 2005 Indentificationand characterization of Caenorhabditis elegans palmitoyl proteinthioesterase 1. J. Neurosci. Res. 79: 836–848.

Prasad, V. V., and R. K. Pullarkat, 1996 Brain lysosomal hydrolasesin neuronal ceroid-lipofuscinoses. Mol. Chem. Neuropathol. 29:169–179.

Ranta, S., Y. Zhang, B. Ross, L. Lonka, E. Takkunen et al., 1999 Theneuronal ceroid lipofuscinosis in human EPMR and mnd mutantmice are associated with mutations in CLN8. Nature 23: 233–236.

Ranta, S., M. Topcu, S. Tegelberg, H. Tan, A. Ustubutun et al.,2004 Variant late infantile neuronal ceroid lipofuscinosis in asubset of Turkish patients is allelic to Northern epilepsy. Hum.Mutat. 23: 300–305.

Rubin, G. M., and A. C. Spradling, 1982 Genetic transformation ofDrosophila with transposable element vectors. Science 218: 348–353.

Santavuori, P., M. Haltia, J. Rapola and C. Raitta, 1973 In-fantile type of so-called neuronal ceroid lipofuscinosis. Part 1.A clinical study of 15 patients. J. Neurol. Sci. 18: 257–267.

Savukoski, M., T. Klockars, V. Holmberg, P. Santavuori, E. S.Lander et al., 1998 CLN5, a novel gene encoding a putativetransmembrane protein mutated in Finnish variant late infantileneuronal ceroid lipofuscinosis. Nat. Genet. 19: 286–288.

Schulz, A., S. Dhar, S. Rylova, G. Dbaibo, J. Alroy et al.,2004 Impaired cell adhesion and apoptosis in a novel CLN9Batten disease variant. Ann. Neurol. 56: 342–350.

Sleat, D. E., R. J. Donnelly, H. Lackland, C.-G. Li, I. Sohar et al.,1997 Association of mutations in a lysosomal protein with clas-sical late-infantile neuronal ceroid lipofuscinosis. Science 277:1802–1805.

Suopanki, J., J. Tyynela, M. Baumann and M. Haltia, 1999 Palmitoyl-protein thioesterase, an enzyme implicated in neurodegenera-tion, is localized in neurons and is developmentally regulated inrat brain. Neurosci. Lett. 265: 53–56.

Sweeney, S. T., and G. W. Davis, 2002 Unrestricted synaptic growthin spinster—a late endosomal protein implicated in TGF-B-mediatedsynaptic growth regulation. Neuron 36: 403–416.

Tyynela, J., D. N. Palmer, M. Baumann and M. Haltia, 1993 Stor-age of saposins A and D in infantile neuronal ceroid lipofuscino-sis. FEBS Lett. 330: 8–12.

Vesa, J., E. Hellsten, L. A. Verkruyse, L. A. Camp, J. Rapola et al.,1995 Mutations in the palmitoyl protein thioesterase gene causinginfantile neuronal ceroid lipofuscinosis. Nature 376: 584–587.

Virmani, T., P. Gupta, X. Liu, E. T. Kavalali and S. L. Hofmann,2005 Progressively reduced synaptic vesicle pool size in cul-tured neurons derived from neuronal ceroid lipofuscinosis-1knockout mice. Neurobiol. Dis. 20: 314–323.

Wheeler, R. B., J. D. Sharp, R. A. Schultz, J. M. Joslin, R. E. Williams

et al., 2002 The gene mutated in variant late-infantile neuronalceroid lipfuscinosis (CLN6) and in nclf mutant mice encodes anovel predicted transmembrane protein. Am. J. Hum. Genet.70: 537–542.

Wisniewski, K. E., 2005 Neuronal ceroid-lipofuscinoses. Gene-Reviews (http://www.genetests.org.).

Wisniewski, K. E., E. Kida, A. A. Golabek, W. Kaczmarski, F. Connellet al., 2001 Neuronal ceroid lipofuscinoses: classification anddiagnosis, pp. 1–34 in Batten Disease: Diagnosis, Treatment, andResearch, edited by K. E. Wisniewski and N. Zhong. AcademicPress, San Diego.

Wolfgang, W. J., T. W. Miller, J. M. Webster, J. S. Huston, L. M.Thompson et al., 2005 Suppression of Huntington’s disease pa-thology in Drosophila by human single-chain Fv antibodies. Proc.Natl. Acad. Sci. USA 102(32): 11563–11568.

Zhang, Z., Y.-C. Lee, S.-J. Kim, M. S. Choi, P.-C. Tsai et al.,2006 Palmitoyl-protein thioesterase-1 deficiency mediates theactivation of the unfolded protein response and neuronal apo-ptosis in INCL. Hum. Mol. Genet. 15(2): 337–346.

Communicating editor: T. H. Eickbush


palmitoyl-protein thioesterase 1 deﬁciency in drosophila ... · are in vivosubstrates of ppt1 are...

Documents