TRAP1 rescues PINK1 loss-of-function phenotypes

Human Molecular Genetics, Jul 2013

PTEN-induced kinase 1 (PINK1) is a serine/threonine kinase that is localized to mitochondria. It protects cells from oxidative stress by suppressing mitochondrial cytochrome c release, thereby preventing cell death. Mutations in Pink1 cause early-onset Parkinson's disease (PD). Consistently, mitochondrial function is impaired in Pink1-linked PD patients and model systems. Previously, in vitro analysis implied that the protective effects of PINK1 depend on phosphorylation of the downstream factor, TNF receptor-associated protein 1 (TRAP1). Furthermore, TRAP1 has been shown to mitigate α-Synuclein-induced toxicity, linking α-Synuclein directly to mitochondrial dysfunction. These data suggest that TRAP1 seems to mediate protective effects on mitochondrial function in pathways that are affected in PD. Here we investigated the potential of TRAP1 to rescue dysfunction induced by either PINK1 or Parkin deficiency in vivo and in vitro. We show that overexpression of human TRAP1 is able to mitigate Pink1 but not parkin loss-of-function phenotypes in Drosophila. In addition, detrimental effects observed after RNAi-mediated silencing of complex I subunits were rescued by TRAP1 in Drosophila. Moreover, TRAP1 was able to rescue mitochondrial fragmentation and dysfunction upon siRNA-induced silencing of Pink1 but not parkin in human neuronal SH-SY5Y cells. Thus, our data suggest a functional role of TRAP1 in maintaining mitochondrial integrity downstream of PINK1 and complex I deficits but parallel to or upstream of Parkin.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

TRAP1 rescues PINK1 loss-of-function phenotypes

Li Zhang 2 Peter Karsten 2 Sabine Hamm 2 Joe H. Pogson 1 A. Kathrin Mu ller-Rischart 0 5 Nicole Exner 6 Christian Haass 4 5 6 Alexander J. Whitworth 1 Konstanze F. Winklhofer 0 4 5 Jo rg B. Schulz 2 3 Aaron Voigt 2 0 Department of Neurobiochemistry 1 MRC Centre for Developmental and Biomedical Genetics, and Department of Biomedical Sciences, University of Sheffield , Sheffield S10 2TN, UK 2 Department of Neurology, University Medical Center , RWTH Aachen, Aachen 52074, Germany 3 Ju lich Aachen Research Alliance (JARA) - Translational Brain Medicine , Aachen 52074, Germany 4 Munich Cluster for Systems Neurology (SyNergy) , Munich, Germany 5 German Center for Neurodegenerative Diseases (DZNE) , Munich, Germany 6 Department of Biochemistry, Adolf Butenandt Institute, Ludwig Maximilians University , Munich D-80336, Germany PTEN-induced kinase 1 (PINK1) is a serine/threonine kinase that is localized to mitochondria. It protects cells from oxidative stress by suppressing mitochondrial cytochrome c release, thereby preventing cell death. Mutations in Pink1 cause early-onset Parkinson's disease (PD). Consistently, mitochondrial function is impaired in Pink1-linked PD patients and model systems. Previously, in vitro analysis implied that the protective effects of PINK1 depend on phosphorylation of the downstream factor, TNF receptor-associated protein 1 (TRAP1). Furthermore, TRAP1 has been shown to mitigate a-Synuclein-induced toxicity, linking a-Synuclein directly to mitochondrial dysfunction. These data suggest that TRAP1 seems to mediate protective effects on mitochondrial function in pathways that are affected in PD. Here we investigated the potential of TRAP1 to rescue dysfunction induced by either PINK1 or Parkin deficiency in vivo and in vitro. We show that overexpression of human TRAP1 is able to mitigate Pink1 but not parkin loss-of-function phenotypes in Drosophila. In addition, detrimental effects observed after RNAi-mediated silencing of complex I subunits were rescued by TRAP1 in Drosophila. Moreover, TRAP1 was able to rescue mitochondrial fragmentation and dysfunction upon siRNA-induced silencing of Pink1 but not parkin in human neuronal SH-SY5Y cells. Thus, our data suggest a functional role of TRAP1 in maintaining mitochondrial integrity downstream of PINK1 and complex I deficits but parallel to or upstream of Parkin. - INTRODUCTION Parkinsons disease (PD) is a progressive neurodegenerative disorder characterized by the loss of dopaminergic (DA) neurons. Mitochondrial dysfunction is believed to play a major role in the etiology of PD. First evidence for this hypothesis was the observation that application of mitochondrial toxins inhibiting specifically complex I of the respiratory electron transport chain (ETC), like 1-methyl-4-phenyl1,2,3,6-tetrahydropyridine (MPTP) or rotenone causes PD-like symptoms in rodents and primates (reviewed in 1). Moreover, reduced function of complex I was reported in post-mortem brain material (substantia nigra and frontal cortex) derived from familial and sporadic PD patients (2,3). Although PD is idiopathic in most cases, several monogenic mutations have been identified to cause familial PD. Detailed analysis of these heritable PD variants and the functions of disease-linked genes provided valuable insights into disease mechanisms. For example, loss-of-function mutations in PTEN-induced kinase 1 (Pink1) and parkin are associated with autosomal recessive, early-onset PD (4,5). Pioneer work in Drosophila revealed that flies deficient for Pink1 or parkin display almost identical phenotypes. These phenotypes can be explained by depleted and dysfunctional mitochondria observed in Pink1- or parkin-deficient flies. Further analysis showed that PINK1 acts upstream of Parkin, suggesting that both proteins function in one pathway (6,7). Finally, data derived from work in cultured cells imply that PINK1 and Parkin are required for quality control and degradation of mitochondria by regulating mitophagy (8 11). Intracellular accumulations of a-Synuclein in so-called Lewy bodies (LBs) are the pathological hallmark of PD. Apart from being the main component of LBs, other findings also highlight the importance of a-Synuclein in the etiology of PD. Three point mutations in SCNA, the gene coding for a-Synuclein and duplication as well as triplication of the SCNA locus have been linked to PD (12 15). Interestingly, expression of the PD-associated mutant a-Synuclein[A53T] causes mitochondrial fragmentation in cultured human neurons. Parallel expression of wild-type PINK1, Parkin or DJ-1 but not of their PD-linked mutant variants rescued this phenotype (16). We have recently shown that the mitochondrial chaperone TNF receptor-associated protein 1 (TRAP1) rescues a-Synuclein[A53T]-induced mitochondrial fragmentation in SH-SY5Y cells (17). Moreover, regulation of TRAP1 expression levels modified the toxic effects of a-Synuclein[A53T] on DA neurons in Drosophila. Overexpression provided rescue, whereas reduction of TRAP1 levels enhanced the a-Synuclein[A53T]-dependent effects in flies and mammalian cells (17). TRAP1 is a mitochondrial chaperone and belongs to the heat shock protein 90 (HSP90) family. Originally, TRAP1 was identified to bind the TNF receptor in a yeast two-hybrid study and its expression was reported to be upregulated in certain cancers (18 20). Among the described functions, TRAP1 is thought to protect from apoptosis induced by diverse stressors like oxidants (21). In the context of PD, this function is of special interest as TRAP1 has been reported to be a downstream effector of PINK1 (21,22). It has been shown in various models that PINK1 can protect against cell death induced by oxidative stress. However, the mechanism underlying this activity has not been revealed unambiguously (23). Interestingly, TRAP1 is phosphorylated by PINK1 in vitro (21). This phosphorylation seems to be crucial in the context of protection from oxidative stress. A kinase-dead version of PINK1 or silencing of TRAP1 completely abolished the protective effects of PINK1 in cultured cells (21,22). Although these findings strongly suggest a fundamental role of TRAP1 in the etiology of PD, the mechanism of the protective effects of TRAP1 on mitochondrial function and integrity remains elusive. To investigate the role of TRAP1 in mitochondrial dysfunction in PD, we utilized in vivo and in vitro models to dissect the role of TRAP1 within the PINK1 and Parkin pathway. TRAP1 rescues morphological defects in Pink1-deficient flies Pink1-deficient flies (Pink1[B9] allele) display severe morphological defects like abnormal wing posture and indentation of the thorax. In addition, climbing and flight ability is strongly reduced in Pink1[B9] flies (6,7). TRAP1 is reported as being a downstream target of PINK1 and is required for the cytoprotective effects mediated by PINK1 in vitro (21). We therefore asked whether human TRAP1 might be able to mitigate Pink1 loss-of-function phenotypes in Drosophila. Ubiquitous (DaG) expression of human TRAP1 (DaG.TRAP1) ameliorated the abnormal wing posture observed in Pink1[B9] mutant flies (Fig. 1A and B). This rescue effect required a functional ATPase domain of TRAP1, since expression of an ATP-binding deficient variant of TRAP1 (TRAP1[D158N]) at similar mRNA and protein levels (Supplementary Material, Fig. S1A and B) did not provide any rescue ability with regard to wing posture defects in Pink1deficient flies (Fig. 1B). Interestingly, pan-neural (elav) expression of Pink1-RNAi recapitulated the abnormal wing posture phenotype. Concomitant pan-neural overexpression of human TRAP1 provided rescue, whereas parallel silencing of endogenous fly Trap1 showed a trend (although not significantly) to enhance the phenotype of Pink1-deficient flies (Fig. 1C, Supplementary Material, Fig. S1C and D). Next, we assayed the effect of TRAP1 on indentation of the thorax observed in Pink1[B9] flies. This phenotype was reduced in the presence of TRAP1 overexpression (Fig. 1D). In addition to the observed morphological defects, Pink1[B9] mutant flies displayed locomotor deficits like impaired flight and climbing abilities. We found that at least deficits in climbing ability were significantly ameliorated by ubiquitous TRAP1 expression in Pink1[B9] flies (Fig. 1E and F). TRAP1 restores mitochondrial defects in Pink1-deficient flies Pink1-deficient flies have a reduced content and disorganized structure of indirect flight muscles (6,7). Consequently, we tested whether ubiquitous TRAP1 expression might rescue muscle defects in Pink1[B9] mutant flies. In thorax sections of Pink1-deficient flies, we observed vacuolization of indirect flight muscle tissue. This vacuolization was not evident in flies with parallel ubiquitous expression of TRAP1 (Fig. 2A and B). Electron micrographs of indirect flight muscles derived from Pink1[B9] flies revealed a decline of mitochondria. The remaining mitochondria were enlarged and displayed fragmented cristae (Fig. 2C and D). These mitochondrial defects could also be reverted by overexpression of TRAP1 in Pink1deficient flies. To exert these rescue effects, a functional ATPbinding domain of TRAP1 was crucial, because overexpression of the ATP-binding deficient variant TRAP1[D158N] was not sufficient to restore muscle integrity and mitochondrial morphology compared with control flies (Fig. 2A D). The observed alterations in morphology and ultrastructure of mitochondria in Pink1[B9] mutant flies are indicative for a dysfunction of these organelles. The altered mitochondrial morphology observed in Pink1[B9] flies went along with a decline of mitochondrial mass as reflected in complex I abundance (Fig. 3A). The decline in mitochondria was accompanied by impaired mitochondrial function as indicated by a reduction of complex I activity (Fig. 3B), along with a reduction in ATP content in Pink1[B9] mutant flies (Fig. 3C). Interestingly, ubiquitous expression of human TRAP1[WT] but not of TRAP1[D158N] restored complex I abundance (Fig. 3A). Consistently, only TRAP1[WT] partially rescued complex I activity and ATP levels (Fig. 3B and C). TRAP1 attenuates decline of DA in Pink1-deficient flies We have recently shown that human TRAP1 is able to rescue a-Synuclein-induced decline of DA concentration in fly head lysates (17). Reduced DA levels are a major hallmark of PD and are observed in patients suffering from sporadic as well as Pink1-linked early-onset PD (24). Pink1 mutant flies display a decline in DA neurons (6,7). Thus we analyzed DA levels from Pink1-deficient flies with concomitant TRAP1 overexpression. Interestingly, no significant differences in DA levels were detected 10 days post-eclosion (d.p.e.). In contrast, 20 d.p.e. Pink1 mutant flies without TRAP1[WT] overexpression displayed a roughly 50% decline in DA content, which was not present in TRAP1[WT]-expressing flies (Fig. 3D). Thus, TRAP1 seems to specifically attenuate the decline in DA observed in Pink1 deficient flies. Taken together, ubiquitous expression of human TRAP1 significantly suppressed all signs of mitochondrial impairment in a Pink1[B9] background in Drosophila (Figs 2 and 3). Downregulation of complex I subunits is partially compensated by TRAP1 In PD patients, increased signs of oxidative stress and reduced activity of complex I are observed in affected brain regions (25 28). Moreover, inhibition of complex I is known to cause PD-like symptoms in rodents and primates (among others, reviewed in 29 31). We therefore asked whether the effects of TRAP1 on complex I activity and ATP concentration in Pink1[B9] flies (Fig. 3) were directly or indirectly linked to complex I activity. In flies, RNAi-mediated knockdown of NADH:ubiquinone oxidoreductases (complex I subunits) encoded by CG18624 and CG11455 is detrimental (32). Ubiquitous (DaG) silencing of CG18624 (NDUFB1 in humans) or CG11455 (NDUFS5 in humans) at 258C resulted in vital offspring. However, lethality was observed under more stringent conditions (298C). Interestingly, concomitant expression of TRAP1[WT] was able to rescue this lethality, whereas TRAP1[D158N] failed to rescue (Fig. 4A). In agreement with previous reports, we found that ubiquitous (DaG) silencing of CG18624 (using another RNAi construct) is lethal already at 258C and larvae exhibit reduced ATP levels (32). The decrease in ATP levels of larvae with RNAi-mediated silencing of CG18624 was restored to control levels by concomitant overexpression of TRAP1[WT] (Fig. 4B). Next, we analyzed whether TRAP1[WT] expression alone might enhance ATP levels. Compared with flies without TRAP1 expression or flies with TRAP1[D158N] expression, no significant increase in ATP levels was observed in TRAP1[WT]expressing flies (Fig. 4C). These data implicate that TRAP1 [WT] has a direct protective effect on the consequences of RNAi-mediated silencing of complex I subunits. TRAP1 does not rescue parkin loss-of-function phenotypes Previous data including our analysis suggest that TRAP1 functions downstream of PINK1. Another reported downstream factor of PINK1 is Parkin, an E3 ubiquitin ligase. Like mutations in Pink1, mutations in parkin cause autosomal recessive PD (4,5,33,34). Genetic studies revealed that overexpression of Parkin rescued phenotypes in Pink1-deficient flies (6,7). PINK1 and Parkin seem to act in concert to target nonfunctional mitochondria for mitophagy. As TRAP1 rescued Pink1 deficiency, we asked whether TRAP1 expression is also able to attenuate phenotypes in parkin loss-of-function flies. We found that TRAP1 mildly attenuated wing posture defects in parkin mutant flies (Fig. 5A). In contrast, other phenotypes observed in parkin-deficient flies like reduced climbing ability (Fig. 5B), reduced ATP levels in thorax (Fig. 5C), impaired mitochondrial integrity (Fig. 5D) and decline in levels of complex I subunits (Fig. 5E) were not rescued by TRAP1 overexpression. Although TRAP1 attenuated these loss-of-function phenotypes in Pink1[B9] mutant flies, TRAP1 expression was not sufficient to rescue effects induced by parkin loss-of-function. TRAP1 rescues mitochondrial defects in Pink1 but not in parkin-deficient SH-SY5Y cells Under normal conditions, roughly 25 30% of SH-SY5Y cells display a fragmented mitochondrial network (17,35 39). This phenomenon is based on the fact that 70 75% of the cells are in the interphase, showing a tubular network of mitochondria, whereas the remaining cells are close to or undergo cell division, which is accompanied by fragmentation of mitochondria (40). Accordingly, the observed degree of mitochondrial fragmentation at baseline is not related to mitochondrial dysfunction but rather to cell-cycle progression. Loss of Pink1 and/or parkin is known to alter the mitochondrial network, shifting mitochondria from a tubular appearance toward a highly fragmented mitochondrial network in vertebrate cells, independently of cell-cycle progression. Effective transient silencing of either Pink1 or parkin by siRNAi treatment strongly increased the percentage of cells displaying fragmented mitochondria (Fig. 6A C, Supplementary Material, Fig. S2A). In Pink1-silenced cells, overexpression of TRAP1 suppressed mitochondrial fragmentation to baseline levels. In contrast, TRAP1[D158N] and a cytoplasmic variant lacking the mitochondrial targeting sequence TRAP1[DMTS] failed to provide rescue (Fig. 6C, Supplementary Material, Fig. S2B). In cells with transient silencing of parkin, overexpression of TRAP1[WT] provided only mild rescue effects, whereas TRAP1[D158N] and TRAP1[DMTS] were not effective against mitochondrial fragmentation (Fig. 6C, Supplementary Material, Fig. S3). The protective effect of TRAP1[WT] was not only evident with respect to mitochondrial morphology. Silencing of Pink1 and parkin impaired mitochondrial function as indicated by a reduction of ATP concentration. Overexpression of TRAP1[WT] also restored ATP levels in si-Pink1-treated cells. This rescue effect was absent in si-parkin-treated cells (Fig. 6D). TRAP1[D158N] and TRAP1[DMTS] did not restore ATP concentration in si-Pink1 or si-parkin-treated cells. In this study, we analyzed the function of the mitochondrial chaperone TRAP1 within the PINK1 and Parkin pathway. We show that human TRAP1 is able to rescue phenotypes in Pink1 loss-of-function flies but has only minor effects on phenotypes of parkin-deficient flies. Flies lacking PINK1 display morphological abnormalities like abnormal wing posture, thorax indentation as well as locomotion deficits like reduction in climbing and flight abilities. Here we show that these phenotypes are significantly ameliorated upon expression of human TRAP1. The observed morphological abnormalities in Pink1-deficient flies can be ascribed to mitochondrial impairment observed in high-energy-demanding tissue like flight muscles. In flight musculature of Drosophilae lacking PINK1, swollen mitochondria exhibiting fragmented cristae are found (Fig. 2). The lack of functional mitochondria obviously coincides with reduced levels of complex I subunits, reduced complex I function and a decline in ATP (Fig. 3). These phenotypes also are at least partially rescued by overexpression of human TRAP1 in Pink1-deficient flies. Yet, this partial rescue of complex I activity and ATP levels might not be sufficient to restore flight ability because of the enormous energy demand in the flight musculature. Moreover, our data strongly suggest that the protective effects of TRAP1 require the proteins inherent ATP binding (and probably ATP hydrolysis), as ATP-binding deficient TRAP1[D158N] did not provide rescue activity in any condition tested. Mutations in Pink1 are rescued by overexpression of Parkin (6,7,10,41,42), suggesting that both gene products act in one pathway. This was first shown in flies, where loss-of-function mutations in Pink1 and parkin cause very similar/identical phenotypes. As overexpression of PINK1 failed to rescue parkin mutations, PINK1 was placed upstream of Parkin (6,7). Interestingly, our data imply that human TRAP1 failed to rescue parkin-deficient flies. Thus, TRAP1 has to be placed either upstream of Parkin or TRAP1 might act independently of Parkin in a parallel pathway downstream of PINK1. PINK1 and Parkin are required for mitochondrial quality control and targeting of dysfunctional mitochondria for mitophagy (43 45). Under normal conditions, mitochondria constantly undergo fission and fusion cycles. This process is used to maintain functional mitochondria, separating damaged portions of the ETC in one daughter mitochondrion (unhealthy) and retaining all functional ETC components in the other (healthy) daughter mitochondrion (Fig. 7). The healthy mitochondrion is fully functional and might fuse with other healthy mitochondria, whereas unhealthy mitochondria are identified by quality control mechanisms and targeted for degradation via mitophagy. The key to this degradation process seems to be the breakdown of the mitochondrial membrane potential. In healthy mitochondria, PINK1 is usually cleaved by the inner mitochondrial membrane protease PARL (46 49). Processed PINK1 is subsequently efficiently degraded by the proteasome (46). In case of impaired membrane potential, PINK1 accumulates on mitochondria (49 51). This accumulation of PINK1 results in recruitment of Parkin from the cytosol to damaged mitochondria by an unknown mechanism. It is discussed controversially whether this process involves phosphorylation of Parkin by PINK1 (49,52 57). Once Parkin localizes to damaged mitochondria, subsequent ubiquitination of target proteins in the outer membrane directs damaged mitochondria toward mitophagy. This process may involve tight regulation of mitochondrial fission and fusion. TRAP1 is most likely not involved in this process, as we did not observe any alterations in mitochondrial morphology after overexpression of TRAP1 (17). TRAP1 probably also does not directly influence mitochondrial quality control via direct interaction with the PINK1 and Parkin pathway. We showed that expression of human TRAP1 in flies compensated for the detrimental effects of RNAimediated silencing of complex I subunits (Fig. 4). Thus, TRAP1 seems to be beneficial for complex I integrity and/or maintaining a functional ETC in general (Fig. 7). There are several putative sites where TRAP1 might act to protect mitochondria (Fig. 7). Further analysis is needed to reveal the detailed molecular mechanism of TRAP1 function(s). Nevertheless, our hypothesis is strongly supported by recent data showing that expression of the yeast complex I homolog Ndi1p rescued Pink1 deficiency but not parkin deficiency (32). Similar effects have been reported for overexpression of the PD-linked genes DJ-1 and HtrA2/Omi as well as the nicotinamide adenine dinucleotide-dependent protein deacetylase (Sir2), the transcription factor FOXO and the electron carrier vitamin K2 (58 61), all of them rescuing phenotypes in Pink1- but not parkin-deficient flies. These data place complex I function either upstream of Parkin or suggest an alternative parallel pathway independent of Parkin. Again, the finding that overexpression of yeast complex I homolog Ndi1p rescues mutations in Pink1 but not parkin is supporting this assumption (32). Very recent data by Costa et al. (62) provide independent confirmation of TRAP1 function within the PINK1 and Parkin pathway. In their manuscript, it is convincingly shown that fly TRAP1 (dTRAP1) rescues phenotypes in Pink1- but not parkindeficient flies. In addition, Costa et al. carefully analyzed dTrap1 loss-of-function phenotypes. dTrap1-deficient flies were vital and did not display any morphological abnormalities or obvious phenotypes. Detailed analysis of dTrap1deficient flies, however, revealed a mild reduction in longevity, a decrease in motor performance and reduced mitochondrial function. In addition, dTrap1-deficient flies were sensitive to exogenous stressors (heat) and ETC inhibitors (like rotenone and antimycin). Conversely, dTrap1 overexpression protected flies from these toxins (62). These findings are in line with our previous data, showing the protective effects of TRAP1 toward ETC inhibitors in cultured cells (17). A functional ETC causes polarization of the mitochondrial membrane. As depolarization of mitochondria is required to initiate mitophagy, TRAP1 might negatively regulate mitophagy by maintaining the ETC in a functional state. Restored ATP levels and mitochondrial integrity in Pink1-deficient flies and SH-SY5Y cells by TRAP1 strongly imply functional electron transport and membrane potential of mitochondria. In this scenario, Parkin is placed downstream of PINK1 and TRAP1. Upon mitochondrial damage, unequal fission results in healthy and dysfunctional, depolarized mitochondria. Dysfunctional, depolarized mitochondria are efficiently removed by mitophagy. Accumulation of dysfunctional and depolarized mitochondria results in increased ROS production. Moreover, potential fusion with healthy mitochondria will have detrimental effects on the entire mitochondrial pool (Fig. 7). An increase in dysfunctional mitochondria is obvious in TRAP1 and even more prominent in PINK1 deficiency. In our model, fast degradation of these dysfunctional mitochondria in a Parkin-dependent manner reduces detrimental effects on overall mitochondrial integrity. This is why increased Parkin levels rescue PINK1 and TRAP1 deficiency. Reversely, this explains why neither PINK1 nor TRAP1 provides robust rescue effects on Parkin deficiency (Figs 5 and 6). Although an excess of both proteins will keep mitochondria in a healthy state, appearing dysfunctional/depolarized mitochondria are not targeted to mitophagy due to a lack of Parkin. The presence of dysfunctional mitochondria with the abovementioned detrimental side effects cannot be compensated for by PINK1 or TRAP1. Besides a linear pathway in which PINK1 and Parkin control mitophagy, both proteins seem to exert additional functions. These functions are likely to be more important in the context of the observed TRAP1 rescue activity. The fact that Parkin rescues the phenotype of Pink1-deficient flies already implies the existence of alternative pathways, as the absence of PINK1 abolishes recruitment of Parkin to mitochondria and thus Parkin-dependent mitophagy (11,51,54,63,64). Our data strongly suggest that TRAP1 has a fundamental role in PD pathology. TRAP1 is involved in maintaining mitochondrial polarization most likely through stabilization of complex I. Moreover, we showed that TRAP1 is able to ameliorate complex I deficiency (this study) and toxicity induced by a-Synuclein (17). Thus, enhancing TRAP1 activity might be a reasonable approach in future therapies of PD. MATERIALS AND METHODS Fly stocks used in this study were obtained from Bloomington Drosophila Stock Center: w[],Pink1[B9]/FM7i, P{w+mC ActGFP} (BL34749), P{w+mW.hsGawB}elavC155 (BL458, referred to in text as elav) and w[]; P{w[+mW.hs] GAL4-da.G32}UH1 (BL5460, DaG in text). Flies expressing short hairpin RNA to facilitate gene silencing via RNAi were either from the Vienna Drosophila RNAi Center (VDRC)w[1118];;P{GD14981}v30033 (white-RNAi in text), w[1118];;P{GD11336}v21860/CyO (Pink1-RNAi in text), w[1118],P{GD1556}v9271 (vasa-RNAi in text)or from the National Institute of Genetics (NIG-fly, Japan): 3152R-1 (Trap1-RNAi in text). Transgenic flies carrying UAS-hTRAP1 (wild-type and D158N) on second and third chromosomes were generated by BestGene, Inc. using sitedirected integration (on second location 28E7, strain 9723; on third location 76A2, strain 9732). Unless otherwise noted, flies were raised on standard cornmeal medium at 258C. Negative geotaxis Negative geotaxis (climbing) analysis was performed 5 d.p.e. with flies raised at 258C and shifted to 298C after eclosion. Groups of 10 flies per vial (2.5 cm diameter) were gently tapped to the bottom and the number of flies crossing a line at 8 cm height within a time period of 10 s was scored. Each analysis was repeated 10 times with 60 s resting interval. The number of male flies tested per genotype was n . 150 for Pink1[B9] rescue and n . 60 for park[25] rescue experiments. Wing posture Wing posture defects in Pink1[B9] (n . 200) and park[25] (n . 60) flies were scored 5 d.p.e. with flies raised at 258C and shifted to 298C after eclosion. In case of pan-neural (elav) induction of RNAi, wing posture defects were scored 20 d.p.e. (n . 100). Flight assays Flight assays were performed as previously described (65). The number of flies per genotype was n . 100. Thorax indentations Thorax indentations were assessed by visual inspection of young adult flies, and the presence of indentations was scored regardless of severity or number. Histology Thoraces were prepared from 5-day-old adult flies (raised at 258C and adults shifted to 298C) and treated as previously described (65). Semi-thin sections were stained with Toluidine blue, whereas ultra-thin sections were examined using a transmission electron microscope (FEI tecnai G2 Spirit, 120 kV). Fly head DA assay Fly head DA assay was performed as described previously (17). Three fly heads flash-frozen in liquid nitrogen (raised at 258C and adults shifted to 298C) were homogenized (Precellys 24 homogenizer) in 100 ml homogenization buffer (0.1 M perchloric acid). For HPLC analysis, 20 ml of supernatant from each sample was used (Dionex Ultimate 3000; running buffer: 57 mM citric acid, 43 mM sodium acetate, 0.1 mM EDTA, 1 mM octane sulfonic acid, 20% methanol, pH 4.3). Samples were separated on a chromatographic column (Dionex Acclaim C18, 5 mm, 2.1 150 mm2 column, at 258C), and DA was electrochemically detected on a graphite electrode (Dionex ED50 electrochemical detector with following conditions: disposable carbon electrode at 0.8 V, flow rate 0.2 ml/min). DA (Sigma-Aldrich) standards of 0.1, 0.25 and 0.4 mM were used for creation of a standard curve. The Chromeleon 6.6 software was used for HPLC data analysis. Mitochondria Mitochondria were isolated from 30 fresh 5-day-old flies ( 35 mg tissue), using Mitochondria Isolation Kit (MITOISO1, SIGMA), according to the manufacturers instructions. The final mitochondrial pellet was re-suspended in 60 ml of 10 mM Tris HCl, pH 7.6, and diluted to a 1:10 solution for complex I activity measurement. Mitochondrial complex I activity Mitochondrial complex I activity was analyzed using 2 ml from mitochondrial extraction (see above), 4 ml of 10 mM NADH and 194 ml of incubation buffer (70 ml of 50 g/l lipidfree BSA, 4 ml of 17.5 mM decyl-ubiquinon, 0.6 ml of 0.1 M DCIP, 1 ml of 1 mM antimycin-A, 250 ml of 0.1 M, pH 7.5, potassiumphosphate buffer, add H2O to 1 ml, freshly made). This solution was incubated for 1 min and measured at 378C every 30 s for 10 min at 600 nm wavelength. Activity of the complex I is expressed per milligram protein (mU/mg protein): mU NetA/min d/1, where NetA, the rate of absorbance change per minute; d 100, dilution factor of the reaction solution; 1 0.0191 mM 1 cm21, extinction coefficient of NADH at 600 nm and pH 8.1. Complex I activity was normalized to protein concentration of lysates (measured using DC Protein Assay, Bio-Rad). Each lysate was measured in triplicate. Results from triplicate measurements were depicted as fold change compared with control (PINK1[B9]/+;DaG). The number of independent repetitions was n 5 per genotype. Complex I subunit Complex I subunit levels were detected in western blot analysis using anti-NDUFS3 (ab14711, Abcam, 1:1000). RNA interference on SH-SY5Y cells Cells were reverse-transfected with stealth siRNA (Invitrogen) using Lipofectamine RNAiMAX (Invitrogen). Fresh medium was added 24 h after transfection. After additional 24 h, cells were transfected with the indicated DNA constructs using Lipofectamine/Plus (Invitrogen). Cells were fixed with 3.7% PFA for 10 min and permeabilized by incubation with 0.2% Triton for 10 min at room temperature. Mitochondria were visualized by staining with an anti-Tom20 pAb (1:1000, Santa Cruz). Assessment of mitochondrial morphology SH-SY5Y cells were reversely transfected with the indicated stealth siRNA (Invitrogen) using Lipofectamine RNAiMax (Invitrogen) according to the manufacturers instructions. Twenty-four hours after transfection, fresh medium was added. The indicated DNA constructs and mitoGFP were transfected another 24 h later using Lipofectamine/Plus. One day after transfection, cells were fixed with 3.7% paraformaldehyde in PBS for 10 min at room temperature. After permeabilization with 0.2% Triton X-100 in PBS, cells were blocked with 5% goat serum in 0.2% Triton X-100 in PBS for 1 h. Fixed cells were incubated with an antibody against TRAP1 diluted in blocking solution for another hour at room temperature. After washing with PBS, the cells were incubated with secondary antibody for 45 min at room temperature. After extensive washing, the coverslips were mounted on glass slides and analyzed by fluorescence microscopy. For quantification, the mitochondrial morphology was determined in a blinded manner. Cells displaying an intact network of tubular mitochondria were classified as tubular. When this network was disrupted and mitochondria appeared predominantly spherical or rod-like, they were classified as fragmented. Quantifications were based on triplicates of at least three independent experiments. Shown is the percentage of transfected cells with fragmented mitochondria (n . 2500 cells). Fluorescence microscopy was performed using Zeiss Axio Imager.A2 as described (16). Confocal images were obtained with an inverted laser scanning confocal microscope (Zeiss Axiovert 200 M). Cellular ATP measurement Cellular ATP measurement was performed using ATP Bioluminescence Assay Kit HS II (Roche Applied Science) according to the manufacturers instructions. Lysates of fly thoraces (flies raised at 258C and adults shifted to 298C) were freshly prepared as described in Liu et al. (66) and subsequently used for ATP measurement. SH-SY5Y cells were transfected as indicated for ATP measurement. Forty-eight hours post-transfection, medium was replaced by DMEM supplied with 3 mM glucose for 3 h. Cells were washed in PBS and harvested by scraping. Bioluminescence was measured with a Mithras LB 940 luminometer (Berthold Technologies) and normalized to total protein levels. Real-time RT PCR Isolation of total cellular RNA, cDNA synthesis and real-time RT PCR were performed as described (67). Quantification was performed with the 7500 Fast Real Time System (Applied Biosystems) based on triplicates per primer set for each RNA sample. Statistical analysis of RT PCR data is based on at least three independent experiments. RNA expression was normalized with respect to endogenous reference genes: human b-actin; Drosophila ribosomal protein 49 (rp49). Relative expression was calculated for each gene using the delta delta cycle threshold method. The following forward (for) and reverse (rev) primers were used to analyze mRNA abundance of respective human genes: b-actin for: 5-TGGACTTCGAGCAAGAGA-3, b-actin rev: 5-AGGAAGGAAGGCTGGAAGAG-3, parkin for: 5-CGA CCC TCA ACT TGG CTA CT-3, parkin rev: 5-GAC ACA CTC CTC TGC ACC ATA C-3, Pink1 for: 5-CCA ACA GGC TCA CAG AGA AG-3, Pink1 rev: 5-AGC GTT TCA CAC TCC AGG TT-3. The following forward (for) and reverse (rev) primers were used to analyze mRNA abundance of respective genes in flies: rp49 for: 5-TCG GAT CGA TAT GCT AAG CTG TCG CAC-3, rp49 rev: 5-AGG CGA CCG TTG GGG TTG GTG AG-3, dTrap1 for: 5-AGG CAG AGT CAC CGA TCC-3, dTrap1 rev: 5-TGA TGC CTG CTT GGT CTC-3, hTrap1 for: 5-TCG CTG GAA AAC TCC TTG-3, hTrap1 rev: 5-GAG GAC ATT CCC CTG AAC CT-3. TRAP1[DMTS] TRAP1[DMTS] lacks the first 67 amino acids containing the potential mitochondrial localization sequence. The corresponding TRAP1 cDNA sequence (NM_016292.2) was PCR-amplified, introducing a new start codon at amino acid position 67 using the following primers: for: 5-ATG GAG GAA CCC CTG CAC TCG AT-3, rev: 5-TCA GTG TCG CTC CAG GGC CTT G-3. After cloning the obtained PCR product into pCRII-TOPO vector (Invitrogen), the TRAP1[DMTS] cDNA was subcloned into pcDNA3.1+ (Invitrogen) using KpnI and XhoI restriction sites. SUPPLEMENTARY MATERIAL Supplementary Material is available at HMG online. ACKNOWLEDGEMENTS The authors thank Bjo rn Falkenburger for excellent support regarding statistical analysis. Conflict of Interest statement. None declared. This work is/was funded by the Bundesministerium fu r Bildung und Forschung (Nationales Genomforschungsnetz (NGFN+): 1GS08137 to A.V. and J.B.S., 1GS08139 to K.F.W. and C.H. and the Kompetenznetz Degenerative Demenzen (KNDD) to J.B.S. (01GI0703/01GI1005C). The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript. A.J.W. is supported by funding from the Medical Research Council (G070091), Wellcome Trust (WT089698) and European Union FP7 (MEFOPA). C.H. is supported by a Forschungsprofessur provided by the LMU Excellence Program.

This is a preview of a remote PDF:

Li Zhang, Peter Karsten, Sabine Hamm, Joe H. Pogson, A. Kathrin Müller-Rischart, Nicole Exner, Christian Haass, Alexander J. Whitworth, Konstanze F. Winklhofer, Jörg B. Schulz, Aaron Voigt. TRAP1 rescues PINK1 loss-of-function phenotypes, Human Molecular Genetics, 2013, 2829-2841, DOI: 10.1093/hmg/ddt132