Mislocalization of mitochondria and compromised renal function and oxidative stress resistance in Drosophila SesB mutants

Abstract

Mitochondria accumulate at sites of intense metabolic activity within cells, but the adaptive value of this placement is not clear. In Drosophila, sesB encodes the ubiquitous isoform of adenine nucleotide translocase (ANT, the mitochondrial inner membrane ATP/ADP exchanger); null alleles are lethal, whereas hypomorphic alleles display sensitivity to a range of stressors. In the adult renal tubule, which is densely packed with mitochondria and hence enriched for sesB, both hypomorphic alleles and RNA interference knockdowns cause the mitochondria to lose their highly polarized distribution in the tissue and to become rounded. Basal cytoplasmic and mitochondrial calcium levels are both increased, and neuropeptide calcium response compromised, with concomitant defects in fluid secretion. The remaining mitochondria in sesB mutants are overactive and maintain depleted cellular ATP levels while generating higher levels of hydrogen peroxide than normal. When sesB expression is knocked down in just tubule principal cells, the survival of the whole organism upon oxidative stress is reduced, implying a limiting role for the tubule in homeostatic response to stressors. The physiological impacts of defective ANT expression are thus widespread and diverse.

mitochondria are essential organelles for the function of all aerobic cells, as they produce ATP, buffer cytosolic calcium, and sequester apoptotic factors. Impairment of mitochondrial function, resulting in excitotoxic metabolic insufficiency, cell death, and oxidative damage, can contribute to severe tissue pathologies and diseases (36, 44). Mitochondria are also highly dynamic, and accumulate at sites with particularly high ATP consumption (3, 13, 17, 31, 37).

The mitochondrial adenine nucleotide translocase (ANT) is the most abundant protein of the inner mitochondrial membrane and a crucial component in the maintenance of cellular energy homeostasis, as well as in the formation of the mitochondrial permeability transition pore (2, 15). It catalyzes ADP/ATP exchange across the mitochondrial inner membrane (24) and thus has the potential to allow mitochondrial sensing of subcellular ATP:ADP poise. ANT is a target of oxidative damage, and functional inactivation has been associated with chronological age as well as physiological age (51). In mammals, the majority of species, including humans, possess three distinct genes for ANT (9, 25, 26, 28). Drosophila harbors two overlapping ANT genes, ANT1 (sesB), and ANT2, with 72% nucleotide identity with 78% amino acid sequence identity (52). Mutations affecting ANT1 in Drosophila were originally identified as a stress-sensitive strain of animals (sesB) that are conditionally paralytic in response to mechanical stress (18, 19). It was later discovered that sesB¹ is an adult viable hypomorphic allele of the gene encoding ANT1 and that null alleles are lethal (52). Mutations have been suggested to cause a low rate of ATP pumping into the cytoplasm and cause defects in vesicle recycling (40).

The role of ANTs in specific tissues within the whole organism can be studied in Drosophila using the GAL4 > UAS system (6). In this study, we used targeted over- or underexpression of sesB to investigate its importance in mitochondrial function in the Malpighian (renal) tubule, a simple transporting epithelium (12) with an extraordinarily density of mitochondria (48), and thus highly enriched for ANT transcripts. The Drosophila tubule provides a genetically tractable, metabolically highly active model epithelium for which multiple functional readouts are available (12). Here, we aimed to characterize the in vivo role of the sesB (ANT1) gene in cytosolic and mitochondrial Ca²⁺ signaling. In parallel, we assessed changes in mitochondrial morphology and distribution, both in the sesB¹ mutant and in flies in which RNA interference (RNAi) against sesB is expressed specifically in the Malpighian tubules, resulting in impairment of fluid secretion. Furthermore, manipulation of ANT1 in Malpighian tubule principal cells increased hydrogen peroxide (H₂O₂) production in both tubules and in isolated mitochondria and was sufficient to confer an organismal oxidative stress sensitive phenotype.

MATERIALS AND METHODS

Drosophila stocks and generation of transformants.

All strains were maintained on a standard Drosophila diet at 22°C, 55% humidity on a 12:12 h light-dark photoperiod. Wild-type flies were obtained from a Canton-S (CS) stock. The c42-GAL4 driver drives expression in tubule principal cells in the adult (7) and so is suitable for studies on acutely dissected adult tubules. The doubly homozygous c42-GAL4 > UAS-GFP_mt (c42mtGFP) flies, generated by classical recombination, express the fluorescent GFP reporter in principal cell mitochondria. To assess in vivo calcium signals, doubly homozygous c42-GAL4 > UAS-apoaequorin_cyto (c42aeq) or c42-GAL4 > UAS-apoaequorin_mt (c42mtaeq) flies were used, which express the apoaequorin luminescent calcium reporter in the cytosol or the mitochondria respectively, of the principal cells of the tubule main segment [upon which the diuretic neuropeptide Capa-1 acts (21, 42)]. Therefore, the doubly homozygous c42aeq, c42mtaeq flies were crossed with sesB mutant flies for cytosolic or mitochondrial calcium assays. For mitochondrial morphology and distribution in the principal cells of the tubule main segment, c42mtGFP flies were crossed with sesB mutant flies. The c710aeq flies express the apoaequorin transgene in the cytosol of the stellate cells of the tubule main segment. The doubly homozygous c710-GAL4 > UAS-apoaequorin_cyto (c710aeq) flies express the apoaequorin luminescent calcium reporter in the cytosol of the stellate cells of the tubule main segment (42). The ubiquitous Actin-GAL4 and hs-GAL4 lines were obtained from the Bloomington Stock Center (Bloomington, IN). c42-GAL4 also directs expression in a few other tissues in the adult fly (30). So to assess the impact of tubule-targeted sesB transgenes on the survival of whole flies under oxidative stress, we used our newly developed Urate Oxidase-GAL4 (UrO-GAL4) driver, based on the promoter region of a gene with expression utterly specific to principal cells of the main segment of third instar larval and adult tubules [expression only in the principal cells of the tubule main segment (46a)]. The use of the heat shock GAL4 driver (hs-GAL4), which drives the expression of the sesB RNAi transgene in all tubule cell types, enables us to use the diuretic neuropeptides Capa-1 (acting on principal cells) and Drosokinin (acting on stellate cells) sequentially in the same fluid transport assay to maximally stimulate the tubules.

As sesB is a recessive lethal locus, we used two classes of hypomorphic allele: heterozygotes for the sesB¹ mutant (obtained from the Bloomington Stock Center, Bloomington, IN) and UAS-controlled RNAi alleles. To generate constructs for heritable RNAi of sesB gene, an inverted repeat of a 441-base pair fragment of sesB was generated by PCR using the primers 5′CTTGGCTGCTGATACTGGCAAGGGTGG-3′and 5′-AGCACGAAAGCGCCACCAGTTCCTCTG-3′ and cloned as a tail-tail inverted repeat flanking the white intron into the P-element vector pWIZ (27). The construct was injected into w¹¹¹⁸ embryos by Bestgene, producing transgenic flies, in which hairpin ds-RNA could be expressed in cells of choice under control of the appropriate GAL4 driver lines (6). Validation of the sesB RNAi line was confirmed by quantitative (Q)-RT-PCR and Western blot analysis (Supplementary Figs. S1 and S2¹).

For overexpression studies, both open reading frames of the sesB gene (sesB-1 and sesB-2) were cloned using the Gateway system. The entry and destination vectors used were obtained from the Drosophila Gateway Vector collection (46). The primers listed were used in PCR for each of the sesB isoforms: 5′CACCATGGGCAAGGATTTCGATGCTGTT3′, 5′CAAGACCTTCTTGATCTCATCGTACAA3′ for sesB-1, and 5′CACCATGGGGAATATATCAGCATCCATA3′, 5′CAAGACCTTCTTGATCTCATCGTACAA3′ for sesB-2. Transgenic lines were generated using standard methods for P element-mediated germline transformation (BestGene). Independent transformants were isolated, and their chromosomal location determined according to standard genetic techniques. All transgenic lines were viable as homozygotes, thus facilitating crossing to produce flies in which the relevant reporter was targeted to cell-specific types using the appropriate GAL4 driver lines.

Calcium measurements using aequorin.

Luminometry experiments were carried out on live, intact tubules expressing a targeted transgene for either cytosolic or mitochondrially targeted apoaequorin as previously described (39, 48). For in vivo tubule experiments, flies of the following genotypes were used: c42aeq, c42mtaeq, c710aeq, UAS-sesB RNAi, and the resulting progeny from GAL4 > UAS crosses.

Immunocytochemistry.

The protocol used for immunohistochemistry was carried out as described previously (45) using mouse anti-GFP (1:1,000, Zymed), DAPI (1 μg/ml for 1 min, Sigma). An FITC-conjugated affinity-purified goat anti-mouse antibody (Jackson Immunologicals) was used in a dilution of 1:1,000 for visualization of the primary antiserum. The slides were viewed using a Zeiss 510 META confocal microscope.

Peptide antibody production.

Rabbit anti-peptide antibodies were raised against the sesB-2 epitope MGNISASITSQSKM (residues 1–14), sesB-1 and sesB-2 epitope RLAADTGKGGQREF (residues 155–158) by Genosphere Biotechnologies (Paris, France). Antisera were purified on a HiTrap NHS-activated HP column (Amersham Pharmacia Bio-tech, Buckinghamshire, UK) according to the manufacturer's instructions.

Renal fluid secretion assays.

Fluid transport assays were carried out as previously described (11) using live, intact tubules dissected from 7-day-old adults with the following genotypes: wild-type CS, sesB¹ mutant, hs-GAL4 > UAS-sesB RNAi. Fluid droplets were collected every 10 min, and the volumes of fluid were calculated. Basal rates of fluid secretion were monitored for 30 min, whereupon Capa-1 and/or Drosokinin, endogenous Drosophila neuropeptides that stimulate Ca²⁺ signaling and fluid transport (21, 47), was added at 10⁻⁷ M, and the secretion rate was then recorded for a further 30 min.

Oxidative stress survival experiments.

Experiments were carried out as previously described (33). Briefly, 50 ml vials were prepared containing 2 ml of solid medium composed of 1.3% low melting agarose, 1% sucrose, and 1% H₂O₂. Three- to five-day-old flies were placed in each vial in groups of 25–30. Fly lines (mutant sesB¹/Y and FM7a/Y, parental UO-GAL4, UAS-sesB-1, UAS-sesB-2, UAS-sesB RNAi; and progeny from GAL4 > UAS crosses) were maintained at 22°C, and numbers of surviving flies were counted at each time point. Survival curves were analyzed with the log-rank test (GraphPad Prism 4 software).

Organotypic imaging of mitochondria.

Live intact tubules were dissected from parental hs-GAL4, UAS-sesB RNAi, and progeny from GAL4 > UAS crosses allowed to adhere to glass-bottomed dishes coated with poly-l-lysine. Mitochondrial activity was assessed by treatment with the cationic and membrane-permeant mitochondrial dye, JC-1 (5 nM, Invitrogen), and imaged by confocal microscopy as previously described (48).

ATP assays.

ATP measurements were accomplished with a luciferin-based ATP Determination kit (Invitrogen). For each sample, 20 tubules were dissected and placed into 50 μl of Schneider's medium. Triplicate samples were generated for each genotype. All samples were sonicated, boiled for 10 min, and spun down at 5 000 g for 5 min at 4°C, and 25 μl of supernatant from each sample were used for the experiment. The assay was conducted using a Mithras LB 940 plate reader luminometer (Berthold technologies) according to manufacturer's instructions using ATP standards at final concentrations between 1 nM and 1 μM. Tubule samples were used at 1:9 (vol/vol) dilution in a final volume of 250 μl and assayed in duplicate. The ATP concentrations of tubule samples were calculated from the standard curve. Results are expressed as pmol ATP ± SE (n = 6).

Measurement of H₂O₂ generation from tubules and from isolated mitochondria.

For tubule assays, 10 pairs of rapidly dissected tubules from each genotype were added to 60 μl of 25 mM sodium phosphate buffer (pH 7.4), ruptured by brief sonication to release cellular contents, and spun down briefly at 5000 g at 4°C; 50 μl of the supernatant was used immediately in the Amplex Red assay (53).

Preparation of isolated mitochondria: Mitochondria from fly lines (UAS-sesB RNAi, hs-GAL4, and heat-shocked progeny from the cross) were isolated using the Pierce Mitochondria Isolation Kit for Tissue (Pierce Biotechnology, Rockford, IL). In brief, ∼100 adult flies were homogenized with 7–10 strokes of the homogenizer in a 0.3 mg/ml solution of trypsin in PBS on ice for 3 min. A quick refrigerated centrifugation was performed to pellet the sample and to remove the trypsin solution, and the pellet was then resuspended in PBS containing 4 mg/ml of BSA to quench the proteolytic activity of trypsin. The homogenate was treated with the kit reagents in the presence of Protease Inhibitor Cocktail (Sigma) and then subjected to a series of graded centrifugations. The mitochondrial pellet was suspended in mitochondria resuspension buffer [250 mM mannitol, 5 mM HEPES (pH 7.4), and 0.5 mM EDTA], and the protein content of the isolated mitochondria was determined by Bradford assay. Mitochondrial pellets (30 μg of protein) were used for the measurement of H₂O₂ production, described above.

Statistical analysis.

Data are presented as means ± SE. Significance of differences was assessed with Student's t-test (two-tailed) for unpaired samples, with significance taken as P < 0.05, marked graphically with an asterisk.

RESULTS AND DISCUSSION

ANT transcripts are tissue specifically expressed.

The Drosophila genes ANT1 (sesB) and ANT2 are duplicated in tandem and transcribed from a common promoter (52). Why are there two such similar genes? Previously ANT-lacZ transgenic flies were shown to express lacZ widely throughout the developmental stages (23); however, recent postgenomic resources allow a more refined view. FlyAtlas, a comprehensive microarray-based atlas of gene expression in multiple Drosophila tissues (8), reports distinct expression patterns for sesB and ANT2 genes (Table 1). SesB is ubiquitously expressed, though with depressed expression in the ovary and particularly testis. By contrast, ANT2 is virtually testis-specific.

Table 1.

Expression levels of sesB and ANT mRNA in larval and adult Drosophila tissues

Tissue	sesB mRNA Signal	Enrichment	ANT mRNA Signal	Enrichment
Brain	4,088 ± 85	1.00	32 ± 2	0.30
Head	4,886 ± 180	1.20	25 ± 2	0.20
Eye	6243 ± 86	1.48	35 ± 1	0.29
TAG	5,600 ± 177	1.30	32 ± 2	0.30
Salivary gland	9,963 ± 447	2.37	46 ± 5	0.39
Crop	5,047 ± 293	1.20	25 ± 2	0.20
Midgut	3,381 ± 87	0.80	65 ± 6	0.50
Tubule	5,904 ± 186	1.40	35 ± 3	0.30
Hindgut	5,264 ± 120	1.30	36 ± 2	0.30
Heart	5,486 ± 298	1.30	16 ± 0	0.14
Fat body	5,945 ± 874	1.41	29 ± 6	0.24
Ovary	2,743 ± 6	0.70	14 ± 0	0.10
Testis	467 ± 20	0.10	1,145 ± 86	9.40
MAG	3,570 ± 89	0.80	30 ± 2	0.30
Virgin spermatheca	5,595 ± 137	1.33	40 ± 3	0.33
Mated spermatheca	5,280 ± 285	1.25	44 ± 3	0.36
Adult carcass	7,528 ± 487	1.80	38 ± 7	0.30
Whole fly	4,210 ± 254		121 ± 5

Data were obtained from http://flyatlas.org (8). The signals represent mRNA signals globally normalized against all samples on all 4 chips for a tissue by Affymetrix GCOS software tissues (8). TAG, thoracicoabdominal ganglion; MAG, male accessory glands. Enrichment denotes enrichment of mRNA relative to whole fly.

The Drosophila sesB gene encodes two isoforms: sesB-1 and sesB-2 (which encodes an extra 13 amino acids at the NH₂ terminus). The Affymetrix probes used in FlyAtlas do not differentiate between these isoforms, but RT-PCR using primers designed to distinguish sesB-1 and sesB-2 transcripts confirmed that sesB-1 is ubiquitously expressed in the fly, while sesB-2 expression is not detected in tubules and seems to be almost specific to ovaries (Supplementary Fig. S3). These results suggest that ANT function is provided ubiquitously by sesB-1, though complemented with sesB-2 in ovary, while ANT2 is reserved for the functionally unique mitochondria of sperm. Within the tubule, confocal microscopy revealed mitochondrial localization of the sesB protein in principal cells of wild-type Malpighian tubule (Fig. 1A), which was identical to that seen with a UAS-mit-GFP reporter when driven in adult tubules using the c42-GAL4 driver (Fig. 1B).

Fig. 1.

Mitochondrial distribution in tubules from wild-type and sesB mutants. For orientation, arrows indicate apical mitochondria (yellow), basolateral mitochondria (white), principal cell nuclei (blue), and the tubule lumen (red) in A. A: immunocytochemistry using anti-sesB rabbit polyclonal antibody and anti-rabbit IgG-FITC conjugate reveal mitochondrial sesB protein localization in the main, fluid-transporting segment of the Malpighian tubule. B–F: expression of mitochondrial GFP reporter targeted to principal cells in adult tubules using the c42-GAL4 driver. B: control (c42mtGFP); C: SesB knockdown (c42mtGFP > sesB RNAi); D: SesB¹ mutants (c42mtGFP > sesB¹). In A–D, tubule diameter is taken as 30 μm. E–G: enlargement of basolateral plane, showing rod-like and branching normal mitochondria from c42mtGFP (E), compared with small globular mitochondria from c42mtGFP > sesB RNAi (F), and c42mtGFP > sesB¹ tubules (G). In E–G, scale bar represents 1 μm.

To study the effect of sesB on fly development, we assayed for viability after sesB-1, sesB-2 overexpressor, and sesB RNAi knockdown in all tissues using the binary GAL4 > UAS system and an Actin-GAL4 strain (6). Phenotypic characterization of lethality revealed that ubiquitous over-expression of sesB-1 and sesB-2 were both 2nd instar larval lethal, while RNAi of sesB is lethal at pupal stage (data not shown). This confirms that normal levels of sesB are essential for development and viability of the fly.

SesB mutants reveal striking morphology and localization defects.

Imaging of mitochondria in tubules overexpressing sesB RNAi transgene reveals striking differences in the subcellular localization and shape of mitochondria (Fig. 1, C and D) compared with the control (Fig. 1B). These differences are shown clearly by transverse section of the tubule main segment. In c42mtGFP control flies (Fig. 1B), mitochondria are positioned adjacent to the basolateral membrane and packed in the microvillar membrane, while in the sesB RNAi tubules (Fig. 1C), mitochondria are distributed throughout the cell, with short and globular morphology. In sesB¹ mutants (Fig. 1D) or tubules in which a sesB RNAi construct is driven (Fig. 1C), the pronounced apical and basal polarization of the mitochondria is lost, and they are distributed uniformly throughout the cytoplasm. In particular, the normally prominent threadlike brush border (Fig. 1B) is reduced and its mitochondrial density lower, in sesB mutants (Fig. 1, C and D). Furthermore, enlargement of basolateral plane shows that the mitochondria lose their normal threadlike appearance (Fig. 1E), and become short and globular (Fig. 1, F and G). This phenotype could be attributed to cytosolic calcium overload, which has been shown to alter mitochondrial morphology from elongated organelles into spherically shaped particles (4). Furthermore, alterations in mitochondrial morphology have been associated with mitochondrial metabolic state (14, 49) as well as apoptotic cell death (34).

SesB modulates both cellular and mitochondrial calcium levels.

We previously argued that mitochondrial placement within the cell, and apposition to major sites of ATP consumption, was critical to the neuroendocrine control by CAPA peptides (48). In particular, the insertion of a mitochondrion into each apical microvillus provides a structural basis for the spatial and temporal filtering of the intracellular calcium signal received by the apical mitochondria. As sesB knockdown results in loss of this intimate association, the long-term phase of mitochondrial calcium signal should be lost. Accordingly, the impact of sesB on calcium signaling was determined in intact transgenic tubules using the bioluminescent calcium reporter aequorin (42). The Drosophila neurohormone Capa-1 acts upon the principal cells of tubules via an IP₃-mediated pathway (38), producing a biphasic rise in cytosolic calcium (21) and eliciting calcium signals in the Golgi, peroxisomes (45), and mitochondria (48) via different calcium channels including TRP channels (10, 29). Transgenic fly lines homozygous for UAS-apoaequorin_cyto or UAS-apoaequorin_mt, and the GAL4 enhancer trap insertion c42, which drives expression in adult tubule principal cells (22, 42), were crossed with UAS-sesB RNAi lines. Targeting sesB RNAi to tubule principal cells resulted in significantly increased basal cytoplasmic [Ca²⁺]_i levels and abolished the Capa-1-induced calcium response (Fig. 2, A and C). Mitochondria are dynamic calcium stores (41) that track long-term changes in cytoplasmic calcium and are important in tubule function (10, 48); accordingly, sesB RNAi resulted in similarly increased basal [Ca²⁺]_mit levels and a markedly inhibited mitochondrial calcium response to Capa-1 (Fig. 2, B and C).

Fig. 2.

SesB knockdowns increase basal calcium levels and blocks neuropeptide stimulation. A, B: expression of sesB RNA interference (RNAi) was targeted to principal cells in adult tubules expressing either the cytosolic (A) or mitochondrial (B) targeted calcium reporter, apoaequorin. Tubules were stimulated with the neuropeptide Capa-1 (10⁻⁷ M, arrowed), and the typical calcium response shown in black. Data are expressed as mean [Ca²⁺] (nM) ± SE; n > 10, P < 0.05, compared with wild-type tubules. C: bar graphs represent cytosolic and mitochondrial basal [Ca²⁺] from tubule principal cells overexpressing targeted sesB RNAi, average [Ca²⁺] (nM); n = 6, where ***P < 0.0001, Student's t-test. D: expression of sesB RNAi was targeted to stellate cells in adult tubules expressing the cytosolic calcium reporter. Tubules were stimulated with the neuropeptide Drosophila leucokinin (Drosokinin, 10⁻⁷ M, arrowed). Overexpression of sesB RNAi resulted in a significantly decreased Drosokinin calcium response.

The neuropeptide Drosophila leucokinin (Drosokinin) acts upon the stellate cells and elicits an increase in cytoplasmic Ca²⁺ via the IP₃ pathway (38, 39). Targeting sesB RNAi to tubule stellate cells resulted in increased basal cytoplasmic [Ca²⁺]_i levels and in significantly decreased Drosokinin-induced calcium response (Fig. 2D). Normal sesB expression is thus required both for cellular calcium homeostasis and the neuropeptide responses of both principal and stellate cells.

Tubule principal cell-specific expression of sesB modulates ATP levels.

The importance of sesB in cellular energy homeostasis was demonstrated by assay of ATP levels in tubules using a luciferin-based ATP determination kit. Targeted overexpression of sesB-1 and sesB-2 in just tubule principal cells using the UO-GAL4 driver significantly increase the availability of ATP to the cell (Fig. 3A, black bars, P < 0.05 compared with parental lines). By contrast, RNAi against sesB expressed in the principal cells decreased tubule ATP levels significantly as predicted (Fig. 3B, black bar, P < 0.01 compared with parental lines). Although this reduction is not lethal, it might be expected to impact on tubule function.

Fig. 3.

Cellular ATP is modulated by sesB gene expression. Pools of 20 tubules were assayed for ATP content. Data are shown as means ± SE (n = 6). Significant difference *P < 0.05, **P < 0.01, and ***P < 0.001.

Effects of sesB gene on fluid transport.

Capa-1 acts to stimulate fluid production through [Ca²⁺]_mit by increasing ATP supply to the apical V-ATPase, which energizes secretion (48). With their reduced calcium response and lower levels of ATP, sesB mutants might be expected to show deficits in renal function. As predicted, Fig. 4A shows that viable hypomorphic alleles of sesB¹ display a significantly reduced rate in fluid transport. Furthermore, phenotypic analysis of c42 > UAS-sesB RNAi Malpighian tubules reveal these transport fluid at reduced rates compared with wild type. Figure 4B shows that reducing sesB expression in principal cells significantly decreases basal rates of fluid transport, as well as the physiological response to the diuretic neuropeptides Capa-1 and Drosokinin. The compromised calcium signaling and ATP supply in sesB hypomorphs thus feed through to downstream functional properties.

Fig. 4.

Impact of sesB mutations on tubule function. A: fluid transport by Drosophila renal tubule is significantly decreased in hypomorphic sesB¹ allele when the tubule is maximally stimulated by application of the neuropeptides Capa-1(10⁻⁷ M) and Drosokinin (10⁻⁷ M). *P < 0.05. B: tubules, in which UAS-sesB RNAi was driven by hs-GAL4, were stimulated at 30 min by the addition of Capa-1 and subsequently at 60 min with Drosokinin (arrows) cannot sustain high rates of stimulated fluid transport compared with wild type.

Does the mislocalization of mitochondria in sesB RNAi tubules cause the functional defects described or do the abnormalities in localization and function occurred in parallel? Malpighian tubules have intense metabolic demands, and the accepted paradigm is that mitochondria must be positioned properly within the cell to service regions of high ATP consumption. Only one study has shown that mitochondria move in and out of apical microvilli of insect tubules, Rhodnius prolixus (5), although our previous published work (48) showed mitochondrial activation in areas of high metabolic demand within the Drosophila Malpighian tubule. In this work, we demonstrate that the mitochondrial mislocalization and deficit of ATP supply is clearly manifested by fluid transport impairment in sesB mutants, suggesting that mislocalization contributes to functional physiological defects. However, it may also be that abnormalities in localization and function occur in parallel in the sesB mutants. For example, mitochondrial fission in cells is regulated by dynamin-related protein 1 (Drp-1). Dynamin activity appears to be regulated by the level of intracellular Ca²⁺ (16), and we have shown that calcium homeostasis is altered in sesB mutants, so it is possible that Ca²⁺ may directly activate Drp-1 and trigger mitochondrial fission.

Residual mitochondria in sesB mutants are overactive.

As well as altering placement, we observed reduced mitochondrial density in sesB mutants, suggesting that the remaining mitochondria might have to work harder to try to maintain ATP levels. Therefore, to investigate the potential effects on mitochondrial function by sesB RNAi transgene, the potential-sensitive dye JC-1 was used in live imaging experiments utilizing dissected tubules. A previous study has shown that, within renal principal cells, only basolaterally localized mitochondria are constitutively active, providing high levels of ATP for the function of basolateral ATPases (48). Figure 5B shows that targeted overexpression of sesB RNAi to tubule principal cells resulted in an increase in mitochondrial activation throughout the cell. It appears that, in sesB hypomorphs, the remaining mitochondria maintain tissue viability by increased activity, though at the potential risk of increased generation of reactive oxygen species (ROS).

Fig. 5.

Mitochondria are overactive in sesB mutants. A, B: imaging of live tubules using mitochondrial potential-sensitive dyes. Activated mitochondria, in which the inner membrane potential is hyperpolarized, accumulate red fluorescence with JC-1 (48). A: mitochondrial activity in tubules from hs-GAL4 flies. B: note the increase in the red fluorescence compared with A in basal surface of tubule principal cells from hs-GAL4 > UAS-sesB RNAi flies.

Does the low [ATP] generated by sesB RNAi tubules (Fig. 3) have an impact on H₂O₂ production? In mammalian mitochondria, low [ATP] can result in increased H₂O₂ production via different mechanisms at Complex 1, including a high NADH/NAD ratio (35). Indeed, H₂O₂ production increases sharply when SesB is knocked down with RNAi, both in whole tubules (Fig. 6A) and in mitochondria isolated from such tubules (Fig. 6B). However, given that current thinking suggests that in Drosophila mitochondria, O₂^·−/H₂O₂ production does not arise from NADH oxidation at Complex 1 (32), then this cannot provide a mechanistic link between the low ATP but high H₂O₂ observed in sesB RNAi tubules. An alternative explanation may lie with the role of glycerol 3-phosphate (G3P). Production of O₂^·−/H₂O₂ in Drosophila mitochondria has been shown to occur via action of glycerol 3-phosphate dehydrogenase (G3PDH), which converts G3P to dihydroxyacetone phosphate (DHAP) (32). The glycerol phosphate shuttle allows transport of electrons from cytosolic NADH to mitochondrial carriers of the oxidative phosphorylation pathway e.g., G3PDH. Assuming that this shuttle operates in flies, then if [ATP] is reduced, it is possible that glycolysis is increased, leading to increased [NADH], which in turn modulates activity of G3PDH, impacting on ubiquinone and also altering G3P/DHAP ratios. Data from Drosophila mitochondria show that production of O₂^·− arises from G3PDH activity and ubiquinone (32), on the cytoplasmic side of the inner membrane, suggesting that the glycerol phosphate shuttle may indeed result in increased O₂^·−/H₂O₂ production under conditions of low [ATP], as our data show.

Fig. 6.

Hydrogen peroxide (H₂O₂) production in tubules and in isolated mitochondria. A: detection of H₂O₂ in Malpighian tubules. Tubules from c42mtGFP > UAS-sesB RNAi and hs-GAL4 > UAS-sesB RNAi flies (heat-shocked) show increased H₂O₂ production compared with all tested genotypes. Data are shown as mean μM H₂O₂ ± SE (n = 6), where *P < 0.05, **P < 0.01, ***P < 0.001, Student's t-test. B: detection of H₂O₂ production in mitochondria. Mitochondria from heat-shocked hs-GAL4 > UAS-sesB RNAi flies show increased H₂O₂ production compared with parental lines. Data are shown as mean μM H₂O₂ ± SE (n = 3), where ***P < 0.001, Student's t-test.

Tubule knockdown of sesB sensitizes flies to oxidative stress.

Previous work has shown that sesB mutation causes detectable phenotypes such as stress sensitivity or reduced viability. In sesB mutants, higher levels of endogenous ROS generation might compromise organismal response to oxidative stress, and help to explain the observed viability phenotypes. Targeted overexpression of sesB-1 and sesB-2 in just tubule principal cells using the UO-GAL4 driver significantly increases survival upon oxidative stress challenge by H₂O₂ (Fig. 7A, red line, P < 0.0001 compared with parental lines). By contrast, RNAi against sesB expressed in the principal cells significantly decreases survival in H₂O₂ medium (Fig. 7B, black line, P = 0.0014 compared with sesB RNAi/WhCS and P = 0.0004 compared UO-GAL4 > WhCS parental lines). Furthermore, Fig. 7B shows that the male viable hypomorphic allele sesB¹ also displays a significantly reduced survival compared with FM7a siblings (red line, P < 0.0001 compared with parental line). Thus, a significant reduction in sesB expression in just tubule principal cells is deleterious for the whole fly, while overexpression of either sesB-1 or sesB-2 in tubule principal cells is sufficient to confer protection against H₂O₂-induced oxidative stress. This could be due to two reasons: cellular defense against oxidative stress could be dependent on cellular ATP levels, or the higher baseline ROS generation in sesB mutants may poise the organism to be more sensitive to additional external stress. Although we cannot rigorously distinguish these possibilities, the higher rates of ROS generation in overexpressors, and their protection against oxidative stress, argues for the former model.

Fig. 7.

Expression of sesBs exclusively in tubule principal cells modulates survival of the whole fly under oxidative stress. A, B: survival of flies after ingestion of H₂O₂. A: survival of flies in which sesB-1 and sesB-2 were targeted to tubule principal cells. B: survival of flies in which sesB RNAi was targeted to tubule principal cells and from male viable hypomorphic allele of sesB¹. Outcrossed parental lines are shown in gray (UAS-sesB-1, UAS-sesB-2, and UAS-sesB RNAi) or black (UO-GAL4 > WhCS) and were generated to control for genetic background of the sesB transgenics stocks. Data are expressed as numbers of surviving flies up to 150 h ± SE, n > 30 flies per vial, repeated 3 times for each genotype.

Conclusion

Normal levels of ANT activity are necessary for mitochondrial function and thus viability. The tissue selectivity of ANT expression observed suggests that tuning of ANT function is critical for organismal survival. In humans, mutations in ANT1 (the heart and skeletal muscle isoform) demonstrate a range of symptoms from myopathy (1) to autosomal dominant progressive external ophthalmoplegia (20). Drosophila is a useful model system for study of mitochondrial defects (43), usually in the context of neural function (50). In Drosophila, sesB mutants were first obtained in a screen for mechanical stress sensitivity (52), and mutants have been shown to have an impact on synaptic transmission (40). This study focused on an epithelium in which sesB is highly enriched and for which a range of quantitative readouts are available. The range of defects observed - in morphology, placement, H₂O₂, and ATP production, secretion, and oxidative stress sensitivity - expand our understanding of the important roles played by the ANT transporter family.

GRANTS

This work was funded by grants from the UK Biotechnology and Biological Sciences Research Council.

DISCLOSURES

No conflicts of interest are declared by the authors.