Genetic analysis of dTSPO, an outer mitochondrial membrane protein, reveals its functions in apoptosis, longevity, and Aβ42-induced neurodegeneration

The outer mitochondrial membrane (OMM) protein, the translocator protein 18 kDa (TSPO), formerly named the peripheral benzodiazepine receptor (PBR), has been proposed to participate in the pathogenesis of neurodegenerative diseases. To clarify the TSPO function, we identified the Drosophila homolog, CG2789/dTSPO, and studied the effects of its inactivation by P-element insertion, RNAi knockdown, and inhibition by ligands (PK11195, Ro5-4864). Inhibition of dTSPO inhibited wing disk apoptosis in response to γ-irradiation or H2O2 exposure, as well as extended male fly lifespan and inhibited Aβ42-induced neurodegeneration in association with decreased caspase activation. Therefore, dTSPO is an essential mediator of apoptosis in Drosophila and plays a central role in controlling longevity and neurodegenerative disease, making it a promising drug target.

Translocator protein 18 kDa is of particular interest to neurodegeneration, being abundantly expressed in glial cells recruited and activated during neuro-inflammation. Thus, TSPO intensity is increased in Alzheimer's disease (AD), stroke, and multiple sclerosis (MS), making imaging TSPO ligands an important system for diagnosing neurodegenerative diseases (Venneti et al., 2006;Lavisse et al., 2012). A single TSPO amino acid substitution (Ala147Thr) has been associated with human adult separation anxiety (Costa et al., 2009) and TSPO expression in neurons has been implicated in modulating long-term potentiation and learning (Tokuda et al., 2010). Hence, TSPO may be an important target for treating neurological diseases (Veiga et al., 2007). TSPO has also been implicated in the pathogenesis of heart disease (Bird et al., 2010;Schaller et al., 2010), atherosclerosis (Bird et al., 2010), and inflammatory bowel disease . In addition, TSPO expression is increased in a variety of cancers. Consequently, TSPO ligands have been reported to have therapeutic effects on certain tumors through modulation of cellular proliferation and apoptosis (Furre et al., 2005;Fafalios et al., 2009;Zheng et al., 2011).
Although TSPO ligands have been widely applied in clinical imaging and therapeutics, the function of TSPO in the biology of the cell and mitochondrion are still poorly understood. To address this deficiency, we have analyzed the physiological consequences of inactivation of the tspo gene in Drosophila. Analysis of dTSPO-deficient Drosophila has confirmed that dTSPO plays a central role in the regulation of apoptosis and that its modulation can extend lifespan and ameliorate the toxicity of Ab42 over-expression.

Drosophila has a TSPO which can be inactivated
The tspo gene is highly conserved from bacteria to humans. Therefore, we were able to identify the Drosophila tspo homologue, CG2789/ dTSPO, using a BLAST search of the Drosophila protein sequences for homology with the human TSPO protein sequence. The alignment of the Drosophila TSPO polypeptide with those of human and other species is shown in Fig. S1. The mitochondrial localization of the dTSPO protein was confirmed by the co-localization of MitoTracker Deep Red which is resistant to fixation in cultured Drosophila S2 cells together with the synthetic fluorescent TSPO affinity probe N-(5-Isothiocyanato-2-phenylindol-3-ylglyoxyl)-N′-(7-nitrobenz-2-oxa-1,3diazol-4-yl)-1,6-diaminohexane (named compound 18) (Taliani et al., 2010) (Fig. 1A).
Disruption of the Drosophila OMM with digitonin removed Voltagedependent anion channel (VDAC, also named porin), the marker of OMM, together with dTSPO. The inner mitochondrial membrane (IMM) localized cytochrome c oxidase subunit I (COX-I) protein was not affected. These results confirm the OMM location of dTSPO (Fig. 1B).
A review of the Drosophila mutant repository revealed a strain in which a P-element had inserted into the tspo gene, tspo [EY00814]. We also depleted the Drosophila TSPO by expression of dsRNA (RNAi) homologous to the dTSPO mRNA regulated by the Gal4/UAS system in various fly tissues. Western blot analysis confirmed that the tspo Pelement inactivation (tspo À/À) and whole body dsRNA dTSPO knockdown flies lacked dTSPO protein, while another OMM protein, porin (VDAC), was unchanged (Fig. 1C). Still, both tspo À/À mutant flies and dsRNA whole body knockdown flies were viable and grossly normal and had comparable developmental timing as wild type flies (Fig. S2).

Inactivation of dTSPO protects cells from apoptosis
If dTSPO participates in the intrinsic pathway of apoptosis in Drosophila, its inactivation should inhibit cell death. Using irradiation-induced apoptosis in Drosophila 3rd-instar larvae tissues (Wichmann et al., 2006) (Fig. 2), we found that c-ray at 30 Gray followed by 3 h recovery stimulated apoptosis in wing disc cells detected by TUNEL staining ( Fig. 2A,C). Quantification of the ratio of area of TUNEL positive pixels versus whole wing disc pixels revealed that apoptosis induction was drastically suppressed in tspo À/À and whole-body dTSPO knockdown flies, both male and female (Fig. 2B,D). Similarly, cells isolated from 3rd-instar larval brains exposed to H 2 O 2 and monitored for apoptosis by Propidium Iodide (PI)/FITC-Annexin V double staining and flow cytometry (Fig. 3A) as well as by caspase 3/ 7 activity (Fig. 3B) revealed significant suppression of apoptosis in tspo À/À fly cells. Therefore, inactivation of dTSPO inhibited apoptosis in flies of both genders.
As resistance to various stresses is frequently associated with longevity, the sensitivity to oxidative stress induced by H 2 O 2 and metabolic stress induced by starvation were analyzed in wild type and tspo À/À flies. Male tspo À/À flies were more resistant to H 2 O 2 than wild-type flies, though female flies were not ( Fig. 4D) (With log rank test for male, P = 0.0015; for female, P = 0.0932). dTSPO deletion did not affect sensitivity to starvation and heat stress, in neither male nor female flies (Fig. S3).
To determine the relation of dTSPO expression to aging, we analyzed the dTSPO mRNA and protein levels at 22°C throughout adult life. The dTSPO mRNA increased continuously from 0 to 30 days after eclosion (DAE) (Fig. S4A). This was paralleled by a progressive increase in dTSPO protein from 30 to 60 DAE. Interestingly, dTSPO protein levels were elevated immediately after eclosion, suggesting that dTSPO may be carried over from the larval or pupal stage (Fig. S4B).

Inactivation of dTSPO imparts resistance to neurodegenerative disease
As aging is the leading risk factor for AD and apoptosis of neurons has been reported in AD (Castro et al., 2010), we investigated whether inactivation of dTSPO could ameliorate the toxic effects of overexpression of the human AD-associated amyloid peptide, Ab42, transcribed from the neuron-specific promoter (elav > Ab42). Ab42 expression in the fly brain induces neurodegeneration, shortens lifespan ( Fig. 5A, Table S2), and results in the disintegration of the brain tissue (Iijima et al., 2004). Both the systemic partial reduction of brain dTSPO in tspo +/À flies (elav>Ab42, tspo +/À) and the neuronal-specific dsRNA knockdown of dTSPO (elav>Ab42, dTSPO-RNAi) reduced the Ab42-induced toxicity and restored the normal lifespan ( Fig. 5A, Table  S2), even though knockdown of dTSPO in neurons (elav>dTSPO-RNAi versus elavGal4 control) did not extend lifespan of wild type flies ( Fig. 5A, Table S2). Despite the male-specific enhanced longevity and oxidative stress resistance in dTSPO-depleted flies, the depletion of dTSPO on Ab42-expressing flies was comparably protective for both male and female flies.
To determine the physical basis for the protective effect of dTSPO depletion on Ab42-toxicity, the tissue integrity was monitored using the number of vacuole-like lesions in brain slices of aged flies. Both partial systemic inactivation (tspo +/À) and neuronal knockdown of dTSPO reduced the numbers of brain lesions (Fig. 5B,C).
To assess why Ab42-induced brain vacuolation was reduced by dTSPO depletion, we determined the neuronal activity of caspase 3/7 in Ab42 over-expressing flies, caspase activation being associated with Ab42 toxicity (Castro et al., 2010;Rohn, 2010). While neuronal expression of human Ab42 increased caspase 3/7 activity by 20% in fly heads by 20 DAE, systemic partial inactivation of dTSPO (tspo +/À) and neuronspecific knockdown of dTSPO returned the caspase activity to that of normal brain tissue (Fig. 5D).
To determine whether dTSPO depletion reduced Ab42 toxicity by inhibiting Ab42 expression, the level of Ab42 mRNA was monitored by quantitative RT-PCR. In Ab42 over-expressing flies, inactivation of dTSPO did not diminish Ab42 mRNA levels in neurons. Hence, the rescue effects of dTSPO depletion are not caused by altered Gal4/UAS system or altered synthesis/stability of Ab42 mRNA (Fig. S5).

The dTSPO is necessary to sustain mitochondrial function
To determine whether dTSPO was important in sustaining mitochondrial function, mitochondrial respiration, OXPHOS complex specific activities, or mitochondrial ATP production and mitochondrial aconitase inactivation these parameters were assessed at 2-5 DAE and 7-10 DAE flies. While the 2-5 DAE flies showed minimal changes in OXPHOS (Fig. S5), the 7-10 DAE flies exhibited reduced mitochondrial respiration, OXPHOS enzyme activities, and increased mitochondrial oxidative stress ( Fig. 6).
Mitochondria isolated from flies at 2-5 DAE had a basal respiration rate using site I substrates (pyruvate and malate) that was mildly increased in the absence of ADP (state IV), but the ADP-stimulated respiration rate (state III) was unaffected (Fig. S6A). This resulted in a modest decreased Respiration Control Ratio (RCR) (state III/state IV respiration rate) in association with a slightly reduced membrane potential (Fig. S6B). The specific activities of the electron transport chain complexes I, II, III and IV were not significantly reduced ( Fig. S6C) nor was mitochondrial oxidative stress ( Fig. S6D) significantly increased in tspo +/À or À/À 2-5 DAE flies.
By contrast, mitochondria isolated from 7 to 10 DAE dTSPO-deficient (tspo À/À) flies resulted in a reduction in both basal and ADP-stimulated respiration using site I substrates, resulting in a constant RCR (Fig. 6A). This was associated with a striking reduction in complexes I specific activity in both the tspo +/À and À/À flies and substantial reductions in complex II and IV activities in tspo À/À flies (Fig. 6B). Despite this obvious reduction of OXPHOS, the tspo +/À and À/À flies did not have reduced ATP levels relative to controls (Fig. S7). Finally, the 7-10 DAE tspo À/À Drosophila exhibited a marked reduction in mitochondrial aconitase activity revealing chronic mitochondrial oxidative damage (Fig. 6C), though the level of malondialdehyde (MDA), a by-product of lipid peroxidation measured in TBARS assay (thiobarbituric acid reactive substances), in older male and female mutant flies was not significantly increased relative to wild-type flies. Hence, the increased mitochondrial ROS seen in mutant flies did not increase ROS-generated macromolecular damage (Fig. S8).

Discussion
The TSPO has been proposed as an important OMM factor for controlling apoptotic cell death. In mammals, there are 2 tspo gene paralogs, tspo1 and tspo2, having slightly different functions . In Drosophila, by contrast, we have found only one tspo gene (CG2789/dTSPO). Therefore, unlike mammalian systems, inactivation of the Drosophila tspo gene has permitted direct examination of TSPO function. This has revealed that dTSPO inactivation in Drosophila inhibits apoptosis and mitochondrial bioenergetics while ameliorating neurodegenerative disease and extending lifespan.
Genetic inactivation of dTSPO reduced apoptotic cell death triggered by c-ray in living animals and by chemical inducers (H 2 O 2 ) in isolated brain cells. Hence, our results confirmed that Drosophila mitochondria still play an essential role in apoptosis and that this is mediated by the dTSPO located in the OMM.
While apoptosis plays a significant role in development in many species, depletion of dTSPO did not affect the development or gross morphology of Drosophila. This suggests that the signals and/or The systemic knockdown of dTSPO extended lifespan even in the face of the severe inhibition of the OXPHOS enzymes and chronically elevated mitochondrial oxidative stress. This places mitochondrial dTSPO-mediated cell death as essential for translating mitochondrial dysfunction into reduced longevity. In mammals, apoptosis can be initiated by activation of the mPTP through oxidative stress (Wallace & Fan, 2009). As mammalian apoptosis has been implicated in regulating longevity (Biteau et al., 2011;Raffaello & Rizzuto, 2011;Rufini et al., 2013), loss of cells due to the activation of the intrinsic pathway of apoptosis mediated through TSPO would seem to be central to longevity. The fact that Drosophila lifespan was not extended by the neuronal-specific knockdown of dTSPO in the presence of increased mitochondrial oxidative stress, but lifespan was extended with systemic inactivation of dTSPO suggests that the dTSPO knockdown effect is not neuron-specific. This contrasts with reports that over-expression of the antioxidant enzyme Cu/Zn superoxide dismutase (Sod1) in Drosophila motor-neurons does extend lifespan (Parkes et al., 1998).
Neuronally expressed human Ab42 shortened the lifespan of Drosophila in association with increased brain cell apoptosis and both of these effects of Ab42 were ameliorated by dTSPO deficiency in both male and female flies. Therefore, TSPO-associated apoptosis is an important factor in mediating neuronal cell death and thus neurodegenerative disease, even though neuronal dTSPO deficiency did not extend lifespan.
The differential effect of oxidative stress on male versus female flies indicates that sex differences modulate the sensitivity to oxidative stress and aging. This may be analogous to our observation that in mammalian cells a portion of the estrogen receptor protein is located in the mitochondrial matrix and when activated by 17b-estradiol rapidly increases the specific activity of mitochondrial Mn superoxide dismutase (Pedram et al., 2006). It is possible that similar sex-specific factors may mitigate some of the deleterious consequences of oxidative stress in females. However, these sex-specific factors may not exist in Drosophila brain, as the Ab42 neurotoxicity was comparably restored by dTSPO depletion in both males and females.
While it has been reported that TSPO probes colocalized with glial markers and not neuronal markers (Venneti et al., 2008), the protection for Ab42-mediated neurodegeneration when dTSPO was specifically knockeddown in fly neurons, argues that dTSPO is both expressed in neurons and important for their integrity. Our data thus support the observation that mammalian TSPO functions in hippocampal neurons and affects long-term potentiation and learning (Tokuda et al., 2010).
If the mPTP is involved in mediating the intrinsic pathway of apoptosis in Drosophila, as it is in vertebrates (Li et al., 2007), then our data imply that the TSPO may be an important OMM component of the mPTP. If this is the case, then it may be interacting with ATP synthase dimers that have recently been proposed to form the IMM component of the mPTP (Giorgio et al., 2013). In Drosophila, the Ca 2+ efflux mediated by the fly mPTP counterpart is associated with tetracaine-and thiol-sensitive IMM depolarization but is not modulated by Ca 2+ via cyclophilin D nor is it associated with mitochondrial swelling or cytochrome c release (von Stockum et al., 2011). Hence, the mechanism by which dTSPO depletion is coupled to inhibition of apoptosis still requires further clarification.
Similarly, the mechanism by which dTSPO depletion imparts protection from neurodegenerative disease and extension of lifespan in the face of OXPHOS dysfunction merits further investigation. Under the conventional view of the mitochondrial free-radical theory of aging, aging is caused by damage to macromolecules by mitochondrial dysfunction-triggered ROS over-production. However, this 'theory' remains contested due to observations such as those in this manuscript where mitochondrial aconitase inactivation suggested increased mitochondrial ROS production yet lifespan was extended and MDA levels were not increased. Possibly, the increased mitochondrial ROS (A) (B) (C) Fig. 6 Effect of dTSPO depletion on 7-10 DAE Drosophila mitochondrial OXPHOS. Mitochondria were isolated from whole body homogenates of tspo +/+, tspo +/À, and tspo À/À flies (mixture of equal number of males and females). (A) Mitochondrial oxygen consumption rate when metabolizing site I (pyruvate + malate) substrates in the absence (state 4) or presence (State 3) of ADP. N = 3 for each genotype. (B) Mitochondrial OXPHOS complexes I to V specific activities normalized using citrate synthase activity. N = 3 for each enzyme value. (C) Mitochondrial aconitase activity in male and female tspo +/+ versus À/À flies expressed as the percentage ratio of the endogenous activity divided by the Fe 2+ -recovered activity. N = 3 for each assay. *P < 0.05, ** P < 0.01, *** P < 0.001. production resulting from dTSPO depletion activated antioxidant defences mitigating macromolecular damage through induction of hormesis via JNK pathway and thus extended lifespan. In Drosophila, the JNK pathway has been shown to be activated by ROS and activated JNK can extend lifespan (Biteau et al., 2011). While JNK can also sensitize cells to apoptosis, this effect might be blocked by the dTSPO deficiency. The net result would be protection from neurodegenerative disease and extension of lifespan.
In summary, we have found that dTSPO is an OMM protein that is essential for Drosophila apoptosis and that elimination of dTSPO inhibits apoptosis, inhibits neurodegeneration, and extends lifespan. Given the high species conservation of TSPO, the role of TSPO in apoptosis must be both ancient and conserved. Hence, pharmacological modulation of TSPO may be a productive approach for treating degenerative diseases.

Experimental procedures Production of dTSPO polyclonal antibody
A 15 amino acid peptide spanning from N164 to S178 of CG2789/ dTSPO (NH 2 -CNPEKEQAPKDEEKPS-COOH) was used to produce a Drosophila polyclonal antibody. Peptide synthesis, conjugation, antiserum production, ELISA screening and affinity purification were done by Covance, Inc. (Princeton, NJ, USA).

Fly stocks
Drosophila were raised on standard cornmeal medium at 22°C (used for qRT-PCR and western blot of dTSPO in aged flies) or 25°C (used in all other assays). The tspo[EY00814] strain, obtained from the Bloomington Drosophila Stock Center, has a P-element insertion in the 3′ regulatory region of tspo gene. The UAS-dTSPO-RNAi stock was obtained from Vienna Drosophila RNAi Center (VDRC) and contained a transgene which can be transcribed into a dsRNA that targets the dTSPO mRNA. Pan-neuronal driver elav-Gal4 and UAS-human WT Ab42 were kindly provided by Dr. Yi Zhong in Cold Spring Harbor Laboratory and Tsinghua University, China. The drivers of UAS-Dcr2; actin-Gal4/CyO, and da-Gal4 were also obtained from Bloomington. The strains were all backcrossed to w 1118 background.

Longevity and oxidative stress/starvation/heat stress resistance assays
To monitor longevity on standard food, fruit flies were collected at 2-3 DAE by brief CO 2 anesthesia and 20 flies placed in each vial with standard cornmeal agar medium. For female studies, only virgin flies were collected. Lifespan was determined at 25°C and approximately 50% humidity with 12 h/12 h light/dark cycle. The number of dead flies was counted every 4 days and the surviving flies transferred to fresh cornmeal agar medium.
To monitor oxidative stress and starvation resistance, 3-6 DAE flies were maintained on standard cornmeal medium after eclosion and then transferred to vials containing filter papers previously soaked with 4 w/v % sucrose (control group), or 5 w/v% H 2 O 2 in 4 w/v% sucrose (pH = 7.2) (oxidative stress group), or water without H 2 O 2 , sucrose (starvation group). The number of dead flies was recorded every few hours and the flies were transferred to new vials with new filter papers every 24 h. To monitor heat stress resistance, flies were maintained on standard cornmeal medium after eclosion and incubated in 37°C. The number of dead flies was counted and the survival rate was calculated.
For longevity and oxidative stress assays, Kaplan-Meier statistics were used to determine the median lifespan/survival period, and log rank test was used to calculate P value to determine statistical significance. At least three independent measurements were performed for each experiment.

Quantification of neurodegeneration
Fly heads were fixed in standard Bouin's Fixative, embed in paraffin blocks, and sectioned. Sections were placed on slides and examined by fluorescent microscopy using the Rhodamine channel. The brain tissue fluoresced as red due to the endogenous fluorescence of the 'white' gene product. The appearance of nonfluorescent 'black vacuoles' within the brain indicated regions of neurodegeneration. To quantify neurodegeneration, the images of the sections were captured and the number of vacuoles counted.

Caspase 3/7 activity
Isolated 3rd-instar larval brain cells treated with H 2 O 2 , or isolated fly heads were homogenized firmly in Homogenization Buffer (225 mM mannitol, 75 mM sucrose, 10 mM MOPS, 1 mM EGTA, pH 7.2) on ice, then centrifuged at 300 g for 5 min. The supernatant was collected and an equal volume of reaction buffer (ApoONE kit, promega, Madison, WI, USA) was added in 96-well plate wells. The plate was shaken gently for 5 min, then incubated in dark for 15 h. Fluorescence was measured with fluorescent spectrophotometer (NOVOstar, BMG Labtech, Ortenberg, Germany), with the excitation at 499 nm and emission at 521 nm.

c-irradiation and TUNEL staining
The 3rd-instar larvae were subjected to c-irradiation (30 Gray), allowed to recover for 3 h at room temperature, and their wing discs isolated for TUNEL staining. Wing discs were dissected and fixed for 20 min at room temperature in 4% PFA in PBS. The samples were then washed three times in PBT buffer (0.1% Tween-20 in PBS) for 10 min per wash, incubated in equilibration buffer (ApopTag kit; Millipore) for 1 h, and incubated again in reaction buffer (TdT enzyme; ratio 7:3; ApopTag kit) at 37°C overnight. On the next day, the TdT reaction mix was replaced with stop buffer (diluted 1:34 in dH 2 O; ApopTag kit) and incubated at 37°C for 3-4 h. The samples were washed three times, 5 min per wash, blocked in blocking solution (PBS, 0.3% Triton-X 100, and 5% normal goat serum) at room temperature for 1 h, and incubated with anti-digoxigenin antibody solution (diluted 47:53 in blocking solution; ApopTag kit) overnight in the dark at 4°C. On the following day, the samples were washed four times in PBS, 20 min per wash, and imaged. Apoptosis was quantified by measuring the area of the TUNEL pixels and dividing it by the area of the total disc using ImageJ software (National Institutes of Health).

Flow cytometry
Brains were dissected from 40-50 3rd-instar larvae flies and incubated in 1 9 Trypsin-EDTA (GIBCO) at room temperature with gently agitated using 200 lL micropipette tips every 15-20 min until no visible tissue fragments could be seen. The cell suspension was centrifuged at 800 g for 5 min, the pellet washed twice with PBS, and resuspended with serum-free Schneider's Drosophila Medium for apoptosis assay. After staining, the cell suspension was analyzed by Flow Cytometry (Accuri C6 Personal Flow Cytometer, BD Biosciences, San Jose, CA, USA).
Cell apoptosis was assessed by flow cytometry following Propidium Iodide (PI) /FITC-Annexin V double staining assay. FITC-Annexin V negative cells together with Propidium Iodide stained nuclei were excluded and dead versus healthy cells assessed by FITC-Annexin V staining. The cells were incubated with H 2 O 2 for 3 h, and then spun down and washed with cold PBS. The pellet was resuspended with 1 9 binding buffer (FITC-Annexin V Apoptosis Detection Kit I, BD Pharmingen, Franklin Lakes, NJ, USA). The density of the cells was adjusted to 1 9 10 6 per mL and 100 lL of cell suspension was transferred to a 1.5-mL Eppendorf tube. The cells were then loaded with 5 lL FITC-Annexin V and 10 lL PI for 15 min at RT in the dark, 400 lL of Binding Buffer was added to each tube, and the relative fluorescence was assessed by flow cytometry.

ATP content
Adenosine tri-phosphate content was determined by using an ATP Determination Kit (Molecular Probes, Grand Island, NY, USA). Five 7-10 DAE flies were homogenized on ice in 100 lL of water in 1.5-mL Eppendorf tube and the supernatant collected by centrifugation at 18 000 g for 5 min. This extract was diluted 1000-fold and 20 lL assayed in 96-well plates, the ATP content normalized to total protein.

TBARS assay
Malondialdehyde content was determined by using the TBARS Assay Kit (Cayman Chemical, Ann Arbor, MI, USA). Twenty flies of 20-23 DAE were homogenized on ice in 250 lL of RIPA buffer in 1.5-mL Eppendorf tube and the supernatant collected by centrifugation at 1600 g for 10 min. This extract was assayed in 96-well plates without dilution. The MDA content was normalized to total protein.

Mitochondrial biochemistry
Mitochondrial isolation Unless otherwise indicated, mitochondria were isolated from 20-30, 2-5 DAE, whole flies. The flies were gently crushed in a 1.5-mL Eppendolf tube with 10 strokes of a fitted pestle in 1 mL Homogenization Buffer at 4°C. The extracts were filtered through eight layers of cheesecloth and then centrifuged at 300 g for 5 min. The mitochondria were collected from the supernatant by centrifugation at 6000 g for 10 min. For respiration and membrane potential assays, the mitochondrial pellet was resuspended in 0.5 mL of Respiration Buffer (225 mM mannitol, 75 mM sucrose, 10 mM KCl, 10 mM TrisHCl and 5 mM KH 2 PO 4 , pH 7.2); For enzymatic assay of OXPHOS complexes, citrate synthase, and aconitase, the pellet was resuspended in 0.5 mL of hypotonic medium (25 mM K 2 HPO 4 , 5 mM MgCl2, pH 7.2) and the mitochondrial membranes disrupted by 2-3 liquid nitrogen freeze-thaw cycles.

Mitochondrial respiration
Respiration assays were performed in 1 mL of respiration buffer using a Clark-electrode. Oxygen consumption rates without (state IV) and with (state III) ADP were recorded (Tong et al., 2007). Respiration dTSPO regulates apoptosis and aging, R. Lin et al. 515 rates were normalized to Bradford protein content, corrected for BSA content.
Mitochondrial membrane potential Mitochondrial membrane potential was measured by the mitochondrial uptake of 20 lM TMRM in a reaction containing 25 mM succinate. The decline in the buffer TMRM concentration was monitored at 575 nm with fluorescent spectrophotometer (NOVOstar, BMG Labtech). Relative membrane potentials were reported for tspo À/À, tspo +/À mitochondria relative to control +/+ mitochondria assayed in parallel wells.
OXPHOS complexes, citrate synthase, and aconitase assays Oxidative phosphorylation complex assays were performed in 200 lL reactions in 96 well plate, monitored with a plate reader (SpectraMax Paradigm, Molecular Devices, Sunnyvale, CA, USA). The relative activities were normalized to the citrate synthase activity. Chemicals were from Sigma-Aldrich unless otherwise specified.
Complex IV activity: Complex IV (cytochrome c oxidase) activity was determined by the oxidation of reduced cytochrome c, monitored at 550 nm in reaction solution (50 mM HEPES, 1 mM DDM, 50 lM cytochrome c/reduced, pH 7.2).
Complex V activity: Complex V (F1-ATP synthase) activity was measured as the rate of hydrolysis of ATP, generated by the conversion of phosphoenolpyruvate to pyruvate by pyruvate kinase (PK), linked to the reduction of pyruvate to lactate by lactate dehydrogenase (LDH). The reaction buffer was (40 mM TrisHCl, 20 lM EGTA, 0.2 mM NADH, 2.5 mM PEP, 25 lg mL À1 Antimycin A, 50 mM MgCl2, 0.5 mg mL À1 LDH, 0.5 mg mL À1 PK, 2.5 mM ATP, pH 8.0), with the rate being monitored by the oxidation of NADH at 340 nm. The reliance on proton pumping was confirmed by demonstrating the oligomycin (2 lM) sensitivity of the reaction.
Aconitase activity and reactivation. Aconitase activity was assayed using the Aconitase Assay Kit (Cayman Chemistry Co.). Endogenous activity was determined in the initial extract and after reactivation with ferrous ammonium sulfate, the difference was used to indicate the extent of ROS inactivation. We used the percentage ratio of endogenous/reactivated total as the% aconitase inactivated by ROS.

Supporting Information
Additional Supporting Information may be found in the online version of this article at the publisher's web-site.

Fig. S1
Multiple sequences alignment of TSPO from various species aligned using CLUSTW. Amino acids on a black background are identical. Those on a gray background are similar for side chain hydrophobicity.

Fig. S2
The duration of lifecycle in wild type and tspo À/À flies. The number of days from crossing to eclosion was recorded, for each genotype, n = 10~40. Each bar = mean AE SEM.

Fig. S3
The sensitivity to starvation and heat stress in wild type and tspo À/À flies. (A) The male (upper) and female (lower) flies of 3-6 DAE were fed with water containing 4 w/v% sucrose (control group) or without sucrose (starvation group). Numbers of dead flies were counted and percent surviving calculated. N = 10-30 flies for each curve. (B) The male 3-6 DAE flies fed with regular food were incubated in 37°C, the rate of survival was monitored after 15-h incubation. N = 6 for each group, indicating six vials of flies.   the absence (state 4) or presence (state 3) of ADP, N = 8 for each genotype. (B) Relative mitochondrial membrane potential assessed by TMRM uptake in mitochondria, N = 3 per genotype. (C) Mitochondrial OXPHOS complex I-IV activities normalized to citrate synthase activity, N = 4 repetitions per assay. (D) Mitochondrial aconitase activity in male and female tspo +/+ and À/À flies, expressed as the ratio of endogenous activity divided by total activity following aconitase following reactivation with Fe 2+ . N = 3 for each genotype. For all data above, *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. S7
Effect of dTSPO depletion on ATP content. ATP content in whole body tissue of 7-10 DAE flies was similar in tspo À/À flies compared with tspo +/+ wild type controls. For each genotype, n = 3.

Fig. S8
Effect of dTSPO depletion on MDA level in TBARS assay. MDA level in whole body tissue of 20-23 DAE flies was similar in tspo À/À flies compared with tspo +/+ wild-type controls. For each genotype, n = 3. Table S1. Median, maximum lifespan and statistical analysis of dTSPO mutant, knockdown flies and wild-type flies treated with PK11195 and Ro5-4864. Table S2. Median, maximum lifespan and statistical analysis of Ab42expressing flies with modification of dTSPO inactivation. dTSPO regulates apoptosis and aging, R. Lin et al.