|Home | About | Journals | Submit | Contact Us | Français|
Oxidative damage to cell macromolecules by reactive oxygen species is associated with numerous diseases and aging. In Drosophila, RNAi-mediated silencing of the mitochondrial antioxidant manganese superoxide dismutase (SOD2) throughout the body dramatically reduces life span, accelerates senescence of locomotor function, and enhances sensitivity to applied oxidative stress. Here, we show that Sod2 knock-down in the musculature alone is sufficient to cause the shortened life span and accelerated locomotor declines observed with knock-down of Sod2 throughout the body, indicating that Sod2 deficiency in muscle is central to these phenotypes. Knock-down of Sod2 in the muscle also increased caspase activity (a marker for apoptosis) and caused a mitochondrial pathology characterized by swollen mitochondria, decreased mitochondrial content and reduced ATP levels. These findings indicate that Sod2 plays a crucial role in the musculature in Drosophila and that the consequences of Sod2 loss in this tissue extend to the viability of the organism as a whole.
The electron transport chain (ETC) in mitochondria is the primary source of reactive oxygen species (ROS) in eukaryotes, with 0.1% or more of oxygen entering the chain being univalently reduced to superoxide [1, 2]. Superoxide escaping from the ETC and other ROS derived from superoxide represent a significant oxidative threat to mitochondria and cells as a whole . Superoxide dismutase (SOD, EC 126.96.36.199) enzymes are antioxidants that protect cells from oxidative stress by catalyzing the dismutation of superoxide to oxygen and hydrogen peroxide. Hydrogen peroxide is subsequently reduced to water by catalase (EC 188.8.131.52) or a number of peroxidases .
The fruit fly, Drosophila melanogaster, possesses the three major eukaryotic SOD isoforms: SOD1 found predominantly in the cytosol, SOD2 in the mitochondrial matrix, and SOD3 in the extracellular space . The localization of SOD2 to the mitochondrial matrix places it in close proximity to the ETC, suggesting that this enzyme has an important role in protecting cells from superoxide produced by mitochondrial respiration. Accordingly, complete loss of SOD2 activity in mice results in a number of pathological conditions including dilated cardiomyopathy, accumulation of lipid in liver and skeletal muscle, motor disturbances with neurodegeneration (in a mixed C57BL/6 and 129Sv background), and a very short life span [6–8]. At the cellular level, Sod2 mutant mice exhibit increased oxidative damage to lipid in addition to genomic and mitochondrial DNA [9, 10], decreased mitochondrial respiration rates, deficits in ETC activity reminiscent of mitochondrial disease  and elevated hepatic apoptosis . In Drosophila, knock-out or knock-down of Sod2 throughout the body also severely shortens life span [12–15]. Furthermore, Sod2 deficient flies exhibit accelerated declines in olfactory and locomotor behavior with age in addition to neurodegeneration associated with elevated apoptosis [14, 16]. All of these studies underscore the importance of SOD2 in protecting cells and tissues from oxidative damage.
In this study we investigated whether Sod2 functions within specific tissues to support normal life span and locomotor function. We find that expression of Sod2 RNAi transgenes in the musculature, but not the nervous system, dramatically shortens life span and accelerates loss of locomotor behavior across age. Flies with knock-down of Sod2 in muscle exhibit mitochondrial pathology, reduced ATP content and elevated caspase activity, supporting the hypothesis that mitochondrial dysfunction in the musculature is central to the shortened life span and accelerated decline in locomotor function in these animals.
Flies were reared to adulthood and housed for all aging studies at 25°C and 55% relative humidity under a 12 hour light–dark cycle on a sugar: yeast: cornmeal: agar medium (10%: 2%: 3.3%: 1% w/v) supplemented with 0.2% Tegosept (Sigma Chemical Co., St. Louis, MO) and active yeast. Sod2 was knocked-down by using the Gal4/UAS system  to express Sod2 inverted repeat (Sod2-IR) transgenes previously described . The UAS-Sod2-IR24, UAS-Sod2-IR15 and daughterless-Gal4 (da-Gal4) lines were provided by John Phillips (University of Guelph). Other fly lines were from the following sources: Mef2-Gal4, Sunita Gupta Kramer (Rutgers); Appl-Gal4; Lawrence Goldstein (University of California, San Diego); D42-Gal4, Gabrielle Bouilanne (University of Toronto); GMH5-Gal4, R. J. Wessells (University of Michigan); 24B-Gal4, repo-Gal4 and elav-Gal4, Drosophila Stock Center (Bloomington, IN); 91Y-Gal4 and 188Y-Gal4, J. Douglas Armstrong (University of Edinburgh); UAS-Shibirets1, Ron Davis (Baylor College of Medicine), Flies with a second chromosome harboring the UAS-lacZ and UAS-Sod2-IR24 transgenes were generated by recombination and confirmed by PCR using the primers 5′-GAACACGTCGCTAAGCGAAAGC-3′ and 5′-GCGAAATTTTACGGGCCACGAAC-3′ to detect UAS-Sod2-IR and the primers 5′-CTGACGGGCTCCAGGAGTC-3′ and 5′-CTACCGGATTGATGGTAGTGGTC-3′ to detect UAS-lacZ.
Groups of 25 adult males (0–3 days old) per genotype were collected under brief CO2 anesthesia and homogenized in extraction buffer (50 mM potassium phosphate/0.1 mM EDTA/2% Triton-X-100, pH 7.8) on ice. To rupture mitochondria, samples were sonicated for 20 sec and then incubated at 4°C for 45 min. Ruptured mitochondrial samples were centrifuged at 16,000 × g for 15 min at 4°C and the resulting supernatant was stored at 4°C. Protein concentration was determined using the Lowry method (DC Protein Assay, Bio-Rad). Samples containing equal amounts of protein were electrophoresed by Discontinuous Native PAGE (4% stacking gel (pH 6.8), 20% resolving gel (pH 8.8)) in sample buffer (0.5 M Tris-HCl/50% glycerol/0.01% bromophenol blue) at 80–100 V. SOD activity was measured colorimetrically using a modified version of an “in-gel” assay previously described . Briefly, gels containing electrophoresed protein samples were incubated in 2.5 mM nitroblue tetrazolium and 50 mM potassium phosphate buffer for 20 min in the dark under gentle agitation. Gels were washed briefly in 50 mM phosphate buffer and then incubated in 28 mM N,N,N′,N′-tetramethylethylenediamine/28 μM riboflavin/50 mM potassium phosphate for 15 min in the dark with gentle agitation. Following a brief wash in 50 mM potassium phosphate buffer, gels were exposed to white light for ~15 min for full color development. In-gel SOD activity was quantified by densitometry using Alpha Imager software (Alpha Innotech Corp., San Leandro, CA).
Adult males (1–3-days old, 100–200 flies per genotype) were collected under brief CO2 anesthesia and transferred to fresh food vials at a density of 25 flies per vial. Flies were transferred to fresh food vials every 2–4 days. Dead and surviving flies were counted during each transfer to fresh food.
Groups of 25 male flies (1–3 days old) were collected under brief CO2 anesthesia and allowed to recover at least 18 hours at 25°C and 55% relative humidity prior to assay. Negative geotaxis (bang-induced climbing) was measured in Rapid Iterative Negative Geotaxis (RING) assays as previously described . Flies were transferred to a RING apparatus and rested for 1 min. Negative geotaxis was initiated by sharply rapping the apparatus on a table three times in rapid succession. The flies’ positions in the tubes were captured in digital images taken 4 sec after initiating the behavior. Data (distance climbed for each fly) were extracted from the images using Scion Image as described . The performance of flies in a single vial in 5 consecutive trials was averaged to generate a single datum. Five vials of flies per genotype were tested to generate N=5 for each genotype in individual experiments. After each RING test, flies were transferred to fresh food vials and housed until the next assessment of negative geotaxis.
Determination of decline-time 50 (DT50, time required for negative geotaxis to decline to 50% of the youngest flies) was performed as described previously . DT50 values, interpolated from second-order curve fits using Prism 4.02 (GraphPad Software, San Diego, CA), were determined for each vial of flies individually and then compiled for each genotype.
Temperature-sensitive Shibire (Shits1) was expressed using the Gal4/UAS system . Shits1 encodes a dominant-negative dynamin that blocks neurotransmission at chemical synapses at 31°C [20, 21]. Adult flies expressing Shits1 (containing a Gal4 driver and two UAS-Shits1 transgenes) and control flies (containing the UAS-Shits1 transgenes but no Gal4 driver) were collected and allowed to recover as above. Negative geotaxis was first assessed in RING assays performed at 25°C. Flies were allowed to recover for 1 hour, transferred to 31°C for 30 minutes, and then tested again in RING assays at the elevated (i.e. restrictive) temperature. The effect of Shits1 expression on RING was quantitated as percent negative geotaxis remaining at the restrictive temperature [(negative geotaxis at 31°C) ÷ (negative geotaxis at 25°C) × 100%] for each genotype.
Gal4 expression was assessed by using a UAS-lacZ reporter gene in a Sod2 knock-down background. Flies with a double-transgenic UAS-lacZ, UAS-Sod2-IR24 second chromosome (generated by recombination) were mated to Gal4 driver stocks to produce progeny with concurrent expression of lacZ and knock-down of Sod2.
The spatial expression patterns of da-Gal4, Mef2-Gal4 and 24B-Gal4 across age were assessed by β-galactosidase histology. Briefly, adult flies were collected under brief CO2 anesthesia, embedded in Tissue-Tek O.C.T. (Sakura Finetak USA, Inc.), transferred in embedding medium to a specimen holder and frozen at −40°C. Whole-body tissue sections (15μM) were cut using a Hacker Bright Cryostat (model OTC5000, GMI, Ramsey, MN) at −20°C. Sections were transferred to slides, allowed to air dry for 1 hour at room temperature and then fixed in 2% glutaraldehyde for 20 min. Slides were washed twice for 5 min in phosphate-buffered saline (PBS) and placed at 37°C until dry. β-galactosidase (EC 184.108.40.206) activity was visualized by adding pre-warmed X-gal substrate to the slides for 30 min. Slides were washed twice for 5 min in PBS and coverslips mounted in 70% glycerol in PBS. Digital images were obtained using a Zeiss Axioplan-2 microscope, Axiocam CCD camera, and Axiovision software (Carl Zeiss, Germany).
Quantitative assessment of whole-body lacZ expression across age was performed by spectrophotometric assessment of β-galactosidase activity as previously described . Three adult males were homogenized in extraction buffer (50 mM potassium phosphate, 1 mM MgCl2, 0.5 μg/ml leupeptin, 0.5 μg/ml aprotenin, 0.7 μg/ml pepstatin A, pH 7.2) at the indicated ages. Homogenates were centrifuged at 16,000 × g for 5 min. β-galactosidase activity measurements (i.e. change in absorbance at 562 nM over 5 min in a Ultrospec 2000 spectrophotometer (Pharmacia Biotech, Piscataway, NJ)) in the resulting supernatants were initiated by addition of 100 μM chlorophenol red-β-D-galactopyranoside. β-galactosidase activity was expressed as the absorbance change per minute per milligram of homogenate protein. β-galactosidase activity in control flies containing either the UAS or Gal4 transgenes alone was negligible.
Thoraces were dissected from adult male flies under anesthesia and immediately fixed in 1% glutaraldehyde/2% paraformaldehyde/0.2 M sodium cacodylate (pH 7.4), osmicated (1% osmium tetroxide), stained (1% uranyl acetate) and dehydrated in an ethanol series with microwave assistance. Fixed and dehydrated samples were briefly incubated in propylene oxide and then embedded in propylene oxide/Embed 812 resin. Thin sections (80–100 nm) were transferred to copper grids and visualized using a JEOL 1010 transmission electron microscope (JEOL USA Inc., Peabody, MA). Mitochondrial content in representative images was quantitated as the fraction of total image area occupied by mitochondria. Four to eight representative low magnification images were analyzed for each genotype at each age using Scion Image software resulting in a total image field of 0.65–1.3 mm2 analyzed.
ATP content was assessed using the ATP Bioluminescent Assay Kit (Sigma Chemical Co., St. Louis, MO). For each sample, five thoraces were dissected from adult males under CO2 anesthesia, homogenized in 6 M guanine hydrochloride and heated at 95°C for 5 min to eliminate endogenous ATPase activity. Homogenates were centrifuged at 16,000 × g for 15 min at room temperature and the resulting supernatant was diluted 10-fold in sample buffer (0.2 M glycine, 50 mM MgCl2, 4 mM EDTA, pH 7.4). Luminescence was measured following addition of ATP Assay Kit reagents according to the manufacturer’s instructions in a Wallac 1420 Victor V plate reader (Perkin-Elmer, Waltham, MA). ATP concentrations of thorax homogenates were interpolated from standard curves using ATP standards.
Caspase activity was assessed as the hydrolysis of the peptide substrate Ac-asp-glu-val-asp-7-amido-4-methylcoumarin (Ac-DEVD-AMC) essentially as described . Briefly, two thoraces were homogenized in lysis buffer (50 mM Hepes, pH 7.5/100 mM NaCl/1 mM EDTA/0.1% CHAPS/10% sucrose/5 mM DTT/0.5% Triton X-100/4% glycerol) and centrifuged for 5min at 13,000 × g at 4°C to generate a supernatant. The supernatant was incubated with 25 mM Ac-DEVD-AMC for 1 h at 27°C and fluorescence of AMC product was determined using a Spectra Max Fluorescent Microplate Reader (Molecular Devices, Sunnyvale, CA) with excitation and emission set at 360 nm and 460 nm, respectively. Protease-specific caspase product was determined by subtracting fluorescence of a duplicate sample pre-treated with caspase inhibitor Ac-DEVD-CHO (2.5 mM) for 15 min from each sample.
Statistical treatment of data (one- and two-way ANOVA, Bonferroni’s multiple comparison tests, Tukey’s HSD multiple comparison tests, log-rank tests) were performed with JMP 5.01 (SAS Institute, Cary, NC, USA) and Prism 4.02 (GraphPad Software, San Diego, CA, USA).
Knock-down of Sod2 in Drosophila via RNA interference (RNAi) using GAL4 to ubiquitously express a UAS-Sod2-inverted repeat (UAS-Sod2-IR) transgene substantially reduces SOD2 enzymatic activity, shortens life span and elicits hallmarks of mitochondrial dysfunction that are consistent with oxidative damage . We confirmed that ubiquitous GAL4-driven expression of Sod2-IR knocked-down SOD2 (but not SOD1) activity (Fig. 1A) and decreased the life span of males (Fig. 1B) and females . Additionally, ubiquitous expression of Sod2-IR caused a rapid age-dependent loss of negative geotaxis (i.e. bang-induced climbing, a locomotor behavior ) across age in males (Fig. 1C) and females . Interestingly, negative geotaxis in 1-day-old Sod2 knock-down and control flies was indistinguishable, indicating that newly emerged flies with ubiquitous expression of Sod2-IR are not globally sick. This finding is consistent with a previous report showing that ubiquitous Sod2 knock-down does not affect eclosion of adult flies possibly because oxygen metabolism or ROS signaling are fundamentally different during development and adulthood .
We reasoned that expression of Sod2-IR in the nervous system might accelerate age-related impairment in negative geotaxis and shorten life span. Tyrosine hydroxylase-Gal4 (TH-Gal4) expresses in dopaminergic neurons required for negative geotaxis , making it a useful tool for exploring the consequences of Sod2-IR expression in regions of the nervous system relevant to the behavior. To identify additional Gal4 lines for neurons involved in negative geotaxis, we assessed negative geotaxis in flies with expression of Shits1 (a temperature-sensitive dominant-negative dynamin that blocks neurotransmission [20, 26, 27] via a number of different Gal4 drivers. Flies with pan-neural expression of Shits1 via 188Y-Gal4, 91Y-Gal4  and Appl-Gal4  were largely paralyzed and had little or no negative geotaxis behavior remaining at the restrictive temperature (31°C), consistent with inhibition of neurotransmission throughout the nervous system (Fig. 2A). Expression of Shits1 by the spatially restricted Gal4 drivers D42 (motorneurons,  and OK307 (giant fiber neurons, ) also caused substantial defects in negative geotaxis at the restrictive temperature (Fig. 2A). In contrast, control flies with the UAS-Shits1 transgene and no Gal4 driver (Shits1/+) or flies with expression of Shits1 in mushroom body, olfactory, gustatory, central complex or eye neurons via a number of other Gal4 drivers had normal negative geotaxis at the restrictive temperature (Fig. 2B). Together, these studies indicate that 91Y, 188Y, Appl, D42, OK307 and TH express Gal4 in neurons required for negative geotaxis.
Expression of Sod2-IR via the pan-neuronal Gal4 drivers 188Y, 91Y and Appl caused modest but statistically discernable impairments in negative geotaxis across age compared to control flies harboring either the Gal4 or Sod2-IR transgene alone (Fig 3A–C). Expression of Sod2-IR using another pan-neuronal Gal4 driver, elav-Gal4 , likewise caused a subtle but statistically significant defect in negative geotaxis with age (Fig. 3D). Expression of Sod2-IR via 188Y-Gal4 and elav-Gal4 also decreased life span modestly (Fig. 3E and F). Additionally, expression of Sod2-IRin motor (D42-Gal4), dopaminergic (TH-Gal4) or giant fiber (OK307-Gal4 and C17-Gal4, ) neurons required for negative geotaxis or throughout the glia (repo-Gal4, ) caused subtle impairments in negative geotaxis across age and/or marginally reduced life span (Fig. 4). The detrimental effects of nervous system Sod2-IR expression on negative geotaxis and life span, however, were substantially smaller than those observed following ubiquitous Sod2-IR expression (Table).
In stark contrast to the modest effects of Sod2-IR expression in the nervous system, expression of two independent Sod2-IR transgenes (Sod2-IR15 and Sod2-IR24) in the musculature using Mef2-Gal4  dramatically reduced life span (Fig. 5A) and greatly accelerated the age-related impairment in negative geotaxis (Fig. 5B) compared to control flies with either the Gal4 driver alone or the Sod2-IR transgene alone. Sod2-IR expression in the musculature via 24B-Gal4  also shortened life span substantially (Fig. 5C) and hastened the age-related loss of locomotor function (Fig. 5D). Expression of Sod2-IR via Mef2-Gal4 or 24B-Gal4 reduced SOD2 activity in whole body extracts without significantly impacting the activity of SOD1 (Fig. 6), confirming that expression of Sod2-IR knocks-down the function of Sod2. Interestingly, the life span and locomotor effects of Sod2-IR expression in the musculature via Mef2-Gal4 and 24B-Gal4 were comparable to or possibly more severe than with the ubiquitous drivers Actin-Gal4 and da-Gal4 (Table).
We expressed Sod2-IR in cardiac muscle to determine whether Sod2 function in this tissue might be important for life span and locomotor behavior across age. Median life span was reduced by ~11% (Fig. 5E) and age-related loss of locomotor behavior was significantly accelerated (Fig. 5F) by expression of Sod2-IR driven by the cardiac Gal4 line GMH5 . The life span and locomotor effects of GMH5-driven Sod2-IR expression, however, were considerably less severe than observed with Mef2-Gal4 and 24B-Gal4 (Table).
We examined lacZ reporter expression in whole-body adult cryosections to characterize the expression patterns for da-Gal4, Mef2-Gal4 and 24B-Gal4 (Fig. 7A, C and E). Gal4drivers often exhibit age-dependent patterns of expression [22, 37] and could, in principle, be sensitive to the enhanced oxidative stress in flies with knock-down of Sod2. We therefore assessed da-Gal4, Mef2-Gal4 and 24B-GAL4 expression patterns across the shortened adult life span of flies that concurrently expressed UAS-lacZ and UAS-Sod2-IR24 transgenes. As expected [13, 29], da-Gal4 drove expression of lacZ throughout the body (Fig. 7A). Mef2-Gal4 drove expression mainly in the thoracic muscle (Fig. 7C) while 24B-Gal4 drove lacZ expression in thoracic muscle, abdominal muscle and fat body (Fig. 7E and inset for d1). This is largely consistent with the muscle-specific expression patterns previously reported for these two Gal4 drivers [34, 38–40]. At the level of resolution possible with this approach, we did not observe a substantive change in lacZ expression across the short life span of Sod2 knock-down flies driven by the three Gal4 drivers (Fig. 7A, C and E). The spatial patterns of expression in da-Gal4, Mef2-Gal4 and 24B-Gal4 drivers overlap in the musculature, but do not appreciably change with age in short-lived Sod2 knock-down animals.
We quantified lacZ reporter activity in whole body extracts of Sod2 knock-down animals throughout development and adulthood to characterize the temporal expression patterns of the da-Gal4, Mef2-Gal4 and 24B-Gal4 drivers. Whole body lacZ reporter activity was dynamic across age for all three drivers. Expression of lacZ driven by da-Gal4 was easily detectable during late larval and pupal stages, increased to a peak of expression at day 7 of adulthood, and then declined by day 11 of adulthood (close to the maximum life span of flies expressing Sod2-IR ubiquitously) to a level similar to that found in larvae and pupae (Fig. 7B). Mef2-Gal4 drove minimal lacZ expression during larval and pupal stages, but thereafter drove very strong expression in adults that increased out to the near-maximal life span of 7 days in flies expressing Sod2-IR via Mef2-Gal4 (Fig. 7D). 24B-Gal4 expressed lacZ at moderate levels during development with peaks of expression at days 1 and 7 of adulthood (Fig. 7F). Although the three Gal4 drivers have distinct temporal patterns of expression, they share the common feature of robust expression during adulthood.
To address whether normal expression of SOD2 in muscles is limiting for life span and age-related locomotor impairment, we assessed survival and negative geotaxis across age in flies overexpressing Sod2 using the Gal4 drivers Mef2 and 24B. Overexpression of Sod2 in the musculature did not extend life span or ameliorate age-related locomotor impairment (data not shown). In fact, Sod2 overexpression driven by Mef2-Gal4 resulted in a slight decrease in life span and acceleration of age-related locomotor impairment.
To gain insight into the cellular consequences of knocking-down Sod2, we examined the ultrastructure of indirect flight muscle (IFM) by transmission electron microscopy (TEM). IFM ultrastructure was indistinguishable at 1-day (Fig. 8A and C) and 7-days (Fig. 8B and D) of age in control flies. The IFM in control animals at both ages had a regular arrangement of myofibrils interspersed by rows of densely-packed mitochondria. The IFM from flies expressing Sod2-IR in the musculature via 24B-Gal4 (Fig. 8E) or Mef2-Gal4 (Fig. 8G) also had largely normal ultrastructure at 1-day of age, although there were occasional clusters of swollen mitochondria (Fig. 8I, J). The swollen mitochondria in 1-day old Sod2 knock-down flies exhibited a rarified arrangement of cristae and in some cases were associated with lysosomes (Fig. 8I, arrow), indicating that the damaged mitochondria were being removed by autophagy. By 7 days of age, the IFM in flies with Sod2 knock-down in muscle had an obvious reduction in mitochondrial content yet they retained normal myofibril ultrastructure (Fig. 8F and H). Quantitative assessment of TEM images revealed that while mitochondrial content was maintained in control flies out to 7 days of age, it was reduced by 32–49% in Sod2 knock-down animals during this same time (Fig. 9A).
Ubiquitous disruption of Sod2 function through mutations in mice  and flies  or RNAi-mediated knock-down of Sod2 in flies  causes oxidative damage to and decreased function of mitochondrial enzymes important in energy metabolism. We measured ATP content in extracts from Sod2 knock-down flies to determine whether decreased SOD2 activity in the musculature impaired mitochondrial function. We focused on ATP content in the thorax because this tissue is predominantly composed of muscle in Drosophila. ATP content increased significantly from day 1 to day 7 of adulthood in control flies harboring a Sod2-IR transgene alone, 24B-Gal4 alone, or Mef2-Gal4 alone (Fig. 9B) as expected from previous reports . ATP content at 1 day of age was indistinguishable in control flies and in flies with knock-down of Sod2 in the muscle (Fig. 9B). By 7-days of age, however, Sod2 knock-down flies had a 25–30% reduction in thoracic ATP content compared to age-matched controls (Fig. 9B).
Decreased SOD2 activity causes an elevation in apoptosis in mouse liver, retina and heart [9, 42–44] and also in the fly brain . We therefore determined whether knock-down of Sod2 in the musculature altered caspase activity (a marker of apoptosis) and, if so, whether caspase activity changed with age in Sod2 knock-down flies. As expected from previous studies , control flies had low levels of caspase activity in thoraces out to 4 days of age (Fig. 9C). In Sod2 knock-down flies, caspase activity was significantly elevated at all ages examined.
SOD2 is a critical antioxidant localized to mitochondria . Whole-body loss of Sod2 function through RNAi-mediated silencing or knock-out of the endogenous Sod2 gene, dramatically shortens life span and causes behavioral defects in flies [12–15], and mice [6–8]. Here, we used tissue-specific expression of Sod2-IR transgenes in Drosophila to address whether these effects on life span and behavior are due to loss of Sod2 function throughout the body or within specific tissues. We focused on the nervous system and the musculature due to their high metabolic activity with attendant generation of ROS .
Expression of Sod2-IR transgenes throughout the nervous system, in specific subsets of neurons required for negative geotaxis behavior, or in glia modestly shortened life span and accelerated age-related loss of negative geotaxis. The effects of nervous system expression of Sod2-IR, however, were considerably less severe than observed in flies with whole-body Sod2-IR expression (Table). While other interpretations are possible, these data suggest that the nervous system may not be a key site of Sod2 function within the context of life span and locomotor function across age in flies.
Consistent with the possibility that tissues outside of the nervous system are important for Sod2 function, expression of Sod2-IR in the musculature via two different Gal4 lines (Mef2 and 24B) dramatically shortened life span and accelerated the age-dependent loss of locomotor behavior (Table). The effect of Sod2-IR expression via these two Gal4 drivers was equal to or even greater than the effect of ubiquitous Sod2 knock-down (Table). Importantly, we confirmed that Mef2-Gal4 and 24B-Gal4 drive expression in the thoracic flight muscle and found that both were expressed throughout the short life span of Sod2 knock-down flies. Together, these data indicate that Sod2 functions within the musculature to protect flies from premature death and loss of locomotor behavior.
Previous studies point toward Sod2 being important for normal muscle function and integrity. For example, Sod2 knock-out mice exhibit reduced activity of ETC components and lipid accumulation in skeletal muscle [7, 11]. Additionally, Sod2 knock-out mice display a dilated cardiomyopathy [6, 7]. These studies, however, do not explicitly address the importance of Sod2 function in the muscle for whole organism vitality and vigor. Our studies extend these previous reports by demonstrating that normal life span and locomotor function depend on expression of Sod2 within the musculature in Drosophila.
The effects of Sod2 knock-down in the musculature led us to propose that overexpression of this enzyme in this tissue might extend life span or preserve locomotor behavior across age. Overexpression of Sod2 in muscle, however, had no effect or possibly even negative effects on these two measures. One possible interpretation of these data is that Sod2 in Drosophila muscle is not limiting for life span or locomotor senescence. An extrapolation of our findings is that the life span-extending effect of ubiquitous Sod2 overexpression  might be due to increased SOD2 activity in a tissue other than or in addition to the musculature. Our data, however, are also consistent with other interpretations. For example, increased longevity was obtained via ubiquitous Sod2 overexpression conditionally during adulthood only  whereas in the GAL4 system used here, Sod2 overexpression occurs during development and adulthood. Hence, it is possible that overexpression of Sod2 during development has negative physiological consequences that counteract any positive effects of Sod2 overexpression on aging during adulthood. Additionally, it is possible that overexpression of Sod2 beyond a level providing maximal benefit might have negative consequences on aging. Since ROS can act as important signaling molecules  in addition to driving oxidative damage, any negative outcomes associated with Sod2 overexpression could be due to inhibition of ROS signaling. Additional experiments are required to address these possibilities.
Sod2 knock-out mice exhibit dilated cardiomyopathy, a severe degenerative state thought to be an important contributor to the perinatal mortality in these animals [6, 7]. Interestingly, patients with hereditary hemochromatosis that carry a Sod2 allele associated with reduced enzymatic activity were found to be at a higher risk for developing cardiomyopathy . Although expression of Sod2-IR in Drosophila cardiac muscle shortened life span and accelerated the loss of locomotor function across age, these effects were rather modest compared to knocking-down Sod2 ubiquitously or in the flight muscle (Table). Hence in flies, Sod2 function within cardiac muscle is important for normal life span and locomotor function across age, but other muscles such as thoracic flight muscle are likely to be more important sites of SOD2 activity.
Sod2 knock-out mice display swollen mitochondria in nervous and muscle tissue [6, 43, 48], although this phenotype is not universally observed . Knock-down of Sod2 in the musculature of flies caused mitochondrial swelling in 1-day-old animals that was followed by a decrease in both mitochondrial and ATP content by 7 days of age. Loss of Sod2 in muscle, therefore, causes a progressive decrease in mitochondrial functional capacity that manifests during adulthood in Drosophila.
There is a strong link between oxidative stress and apoptosis . For example, exposure of cells or intact animals to pro-oxidant chemicals (i.e. exogenous oxidative stress) induces apoptosis [23, 50, 51]. Oxidative stress derived from endogenous sources can also activate apoptosis. Sod2 knock-out mice have elevated apoptosis in liver, retina and heart [9, 42, 44, 52] and Sod2 null flies have increased apoptosis in the central brain . Our studies indicate that knock-down of Sod2 in the musculature is sufficient to induce a large increase in apoptosis in Drosophila, which could be an important mechanism related to the short life span and possibly other phenotypes in Sod2 knock-down flies. Additionally, our studies support the notion that induction of apoptosis by knock-down of Sod2 can occur in a tissue autonomous fashion and therefore does not require loss of Sod2 throughout the body.
This work was supported by grant AG024259 from the National Institutes for Aging (M.G.) and grant MOP79519 from the Institute of Aging of the Canadian Institutes of Health Research (L.S.).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.