The Parkinson's disease-associated protein DJ-1 plays a positive nonmitochondrial role in endocytosis in Dictyostelium cells

ABSTRACT The loss of function of DJ-1 caused by mutations in DJ1 causes a form of familial Parkinson's disease (PD). However, the role of DJ-1 in healthy and in PD cells is poorly understood. Even its subcellular localization in mammalian cells is uncertain, with both cytosolic and mitochondrial locations having been reported. We show here that DJ-1 is normally located in the cytoplasm in healthy Dictyostelium discoideum cells. With its unique life cycle, straightforward genotype-phenotype relationships, experimental accessibility and genetic tractability, D. discoideum offers an attractive model to investigate the roles of PD-associated genes. Furthermore, the study of mitochondrial biology, mitochondrial genome transcription and AMP-activated protein kinase-mediated cytopathologies in mitochondrial dysfunction have been well developed in this organism. Unlike mammalian systems, Dictyostelium mitochondrial dysfunction causes a reproducible and readily assayed array of aberrant phenotypes: defective phototaxis, impaired growth, normal rates of endocytosis and characteristic defects in multicellular morphogenesis. This makes it possible to study whether the underlying cytopathological mechanisms of familial PD involve mitochondrial dysfunction. DJ-1 has a single homologue in the Dictyostelium genome. By regulating the expression level of DJ-1 in D. discoideum, we show here that in unstressed cells, DJ-1 is required for normal rates of endocytic nutrient uptake (phagocytosis and, to a lesser extent, pinocytosis) and thus growth. Reduced expression of DJ-1 had no effect on phototaxis in the multicellular migratory ‘slug’ stage of the life cycle, but resulted in thickened stalks in the final fruiting bodies. This pattern of phenotypes is distinct from that observed in Dictyostelium to result from mitochondrial dyfunction. Direct measurement of mitochondrial respiratory function in intact cells revealed that DJ-1 knockdown stimulates whereas DJ-1 overexpression inhibits mitochondrial activity. Together, our results suggest positive roles for DJ-1 in endocytic pathways and loss-of-function cytopathologies that are not associated with impaired mitochondrial function.

a: Circular map of construct pPROF688 with insertion of DJ-1 fragment (DA) into pDNeo2 A fragment of the DJ-1 gene (75-479 bp) was cloned into the SacI and BamHI restriction enzyme sites of pDNeo2 in the antisense orientation. Neo R : G418 resistance which is controlled by the actin-15 promoter and terminator (A15-P and A15-T). AmpR: ampicillin resistance; lac Z operon: for blue/white screening. Ori: origin of DNA replication; The DA expression cloning cassette is regulated by the actin-6 promoter (A6-P) and the actin-8 terminator (A8-T).
b: Circular map of construct pPROF690 with replacement of Tet cassette with full DJ-1 in pPROF267 The Tet cassette in pPROF267 was replaced by the entire DJ-1 gene in the sense orientation using ClaI and XhoI restriction enzyme sites. G418 resistance cassette (Neo R ) in this construct is regulated by the A6-P promoter and the CPI-T terminator. The other gene cassettes designations are shown in Supplementary Figure 2a.
c: Circular map of construct pPROF693 with insertion of DJ-1 without the stop codon (DGFP) into pA15GFP The DGFP was cloned into the ClaI site of pA15GFP in the sense orientation. The DGFP:GFP gene is flanked by a functional promoter and terminator (A15-P and A8-T). The other gene cassettes designations are shown in Supplementary Figure 2b . Blue fluorescence -DAPI (4',6-diamidino-2-phenylindole) staining of nuclei was not performed as DAPI is not cell permeant. The DJ-1:GFP fusion protein was found throughout the cell in the cytoplasm with no enrichment in the mitochondria. Images were taken on an Olympus BX61 fluorescence microscope. Both wild type and transformed live cells in suspension showed significant levels of membrane blebbing under the conditions used. We do not know the reason for this but it did not affect the localization of DJ-1. After background subtraction and contrast adjusment, the image was deconvolved using the Olympus CellSens Dimension 1.16 software.  Supplementary Figure S5 Supplementary Figure S5. Expression of GFP in a transformant expressing GFP-tagged DJ-1. Western blot of protein from AX2 (negative control) and HPF1245 expressing GFP-tagged DJ-1 (52 kDa -arrowed). Crude protein lysates were separated by SDS-PAGE on a 12% polyacrylamide gel and transferred to a Hybond TM -P+ nylon (Amersham) membrane before washing, blocking and detection using Enhanced Chemifluorescence with anti-GFP-rabbit IgG fraction (Alexa Fluor 488 conjugate) diluted 1/500 in blocking buffer. No GFP could be detected in the parental AX2 strain. Shown here are the relationships between maximal respiration and the individual contributions by (a) Complex II activity, comprised of (b) the component that is coupled to Complex III and (c) the component coupled to the Dictyostelium alternative oxidase (AOX). The proton-motive force generating portion of the maximum uncoupled respiration is the sum of Complex I and Complex II/III contributions (d) and provides energy for ATP synthesis and other mitochondrial processes such as protein, ion and metabolite transport. In intact cells the contributions to basal respiration include (e) the proton "leak" ("consumption" of mitochondrial respiratory energy by processes other than ATP synthesis eg. protein import) and (f) direct oxygen consumption by other cellular enzymes (oxidases and oxygenases).

Supplementary Figure S6
If a given respiratory complex were functionally impaired in the DJ-1 antisense-inhibited or overexpression strains, it would contribute a different (smaller) fraction of respiratory O2 consumption and so lie on a different regression line. Multiple regression analysis showed that this was not the case and that the relative contributions of individual respiratory complexes were unaffected by changes in the level of DJ-1 expression.
Each point represents the average from 2 to 17 independent experiments on a given strain, each experiment involving 4 technical replicates (separate wells in the assay plate). All regressions were highly significant (significance probabilities shown in each panel margin).