|Home | About | Journals | Submit | Contact Us | Français|
The gene for Parkin, an E3 ubiquitin ligase, is mutated in some familial forms of Parkinson's disease, a severe neurodegenerative disorder. A homozygous mutant of the Drosophila ortholog of human parkin is viable but results in severe motoric impairment including an inability to fly, female and male sterility, and a decreased life span. We show here that a double mutant of the genes for Parkin and the metal-responsive transcription factor 1 (MTF-1) is not viable. MTF-1, which is conserved from insects to mammals, is a key regulator of heavy metal homeostasis and detoxification and plays additional roles in other stress conditions, notably oxidative stress. In contrast to the synthetic lethality of the double mutant, elevated expression of MTF-1 dramatically ameliorates the parkin mutant phenotype, as evidenced by a prolonged life span, motoric improvement including short flight episodes, and female fertility. At the cellular level, muscle and mitochondrial structures are substantially improved. A beneficial effect is also seen with a transgene encoding human MTF-1. We propose that Parkin and MTF-1 provide complementary functions in metal homeostasis, oxidative stress and other cellular stress responses. Our findings also raise the possibility that MTF-1 gene polymorphisms in humans could affect the severity of Parkinson's disease.
Parkinson's disease (PD) is the second most prevalent progressive neurodegenerative disorder and the most common age-related movement disorder (10, 13, 43, 59). Many molecular aspects of PD pathogenesis still need to be clarified. Extensive studies point to oxidative stress as a major contributor to the disease (28). Besides the gene for Parkin, an E3 ubiquitin ligase, four other genes, PINK1, DJ1, UCHL1, and α-synuclein, have been implicated in rare, early-onset, familial forms of PD, whereas LRRK2 is predominantly responsible for late-onset PD (20, 57, 70). Much effort has gone into the development of animal models of PD, including models in the fly Drosophila melanogaster. In the studies presented here, we use a strain in which the ortholog of the human parkin gene has been disrupted by the insertion of a P-element transposon into the coding region (24, 48).
In mammals, the proteins PINK1 and Parkin cooperate to ensure proper quality control of mitochondria, and Parkin is particularly important for autophagy of faulty mitochondria (reviewed in references 8 and 73). In agreement with this notion, Parkin deficient flies suffer from mitochondrial malfunction (24, 45, 48), which distorts muscle structure and causes severe locomotor defects and an inability to fly (24, 48). Furthermore, both male and female parkin-null Drosophila mutants are sterile (52), exhibit an increased sensitivity to multiple stresses (including oxidative stress), and have a reduced life span (23, 48).
Maintenance of metal homeostasis is an essential requirement for the proper functioning of all organisms. An adequate supply of essential trace metals, such as copper and zinc, is important, whereas an excess can be highly toxic. Alterations in copper homeostasis due to mutations in copper transporters cause Wilson's and Menkes disease (11, 31, 65). An imbalance in trace metal levels has also been implicated in neurodegenerative disorders such as Parkinson's and Alzheimer's disease, as well as in senescence processes (42, 53, 54). To investigate the possible interplay of Parkin function with metal homeostasis, we modulated the concentration of the metal responsive transcription factor 1 (MTF-1) in parkin mutant Drosophila. MTF-1 is conserved in evolution, and its homologs have been characterized in humans (7, 36, 44), mice (25, 51, 69), fish (3, 9), and Drosophila (17, 60, 75). MTF-1, also referred to as metal response element binding transcription factor 1, is a zinc-finger protein that regulates transcription of its target genes by binding to DNA sequence motifs known as metal response elements (MREs), which are typically located proximal to the transcription start site (12, 27, 32, 41, 64, 68). The majority of MTF-1 preferentially localizes to the cytoplasm in quiescent, nonstressed cells but translocates to the nucleus upon heavy metal load and a number of other stressful conditions (34, 58, 62).
Apart from counteracting the effects of heavy metal load, MTF-1 also induces transcription of metallothionein genes in response to oxidative stress and infection (2, 21, 22). Metallothioneins (MTs), are small, cysteine-rich, metal-binding proteins with a major role in metal homeostasis and detoxification (30, 46). MTs occur in all eukaryotes, as well as in some prokaryotes. Heavy metals such as zinc, copper, and cadmium are complexed by the cysteine sulfhydryl groups, which can also exert antioxidant function (2, 29). In this context it is noteworthy that in a mouse PD model, dopaminergic (DA) neurons of an MT-knockout mutant are more vulnerable to L-DOPA toxicity than neurons from mice with wild-type MT (15, 39). This suggests that MTs play a protective role against DA-quinone-induced neurotoxicity. There are more than 10 functional metallothionein genes in humans, four in the mouse (50, 71) and at least four in Drosophila, termed MtnA to MtnD (17, 40). The Drosophila MTs are involved to different degrees in the defense against heavy metal stress. MtnA is the most important under copper load, while MtnB preferentially binds cadmium and protects against cadmium intoxication. MtnC and MtnD, despite sharing 67% amino acid identity with MtnB, have only a minor role in protection against heavy metals, at least when MtnA and MtnB are present (16, 18). The MtnA and MtnB genes (also referred to as Mtn and Mto, respectively) are differentially regulated during development (61). In addition to metallothioneins, MTF-1 also regulates, in Drosophila, the expression of ferritins, the copper importer Ctr1B, the zinc exporter ZnT35C, glutathione S-transferase, and an ABC transporter (60, 63, 74).
MTF-1 proteins of human and Drosophila are highly similar in their DNA-binding zinc finger region but quite divergent outside of it. Nevertheless, they can largely complement each other in the protection against metal stress (6, 75). A major difference between mammals and Drosophila is that metallothionein genes in mammals are mainly induced by zinc and cadmium, whereas in Drosophila they are best induced by copper and cadmium (18, 75, 76). Moreover, disruption of the MTF-1 gene in the mouse results in embryonic lethality (25), which is not the case for fly mutants, which are viable and fertile. However, the fly mutants do display sensitivity to cadmium, zinc, and copper load, as manifested by a reduced life span on heavy-metal-supplemented food, and also cannot tolerate copper starvation (17, 60). The Drosophila allele MTF-1140-1R carrying a 4.1-kb deletion of the coding region has the strongest phenotype and is considered a null mutation (17). We have used this allele for experiments requiring an MTF-1 loss of function.
A link between cellular heavy metal handling and the parkin mutant phenotype was suggested by our previous finding of a partial rescue of parkin mutant flies upon zinc supplementation or chelation of redox-active metals (55, 56). In the present study we set out to investigate the interaction between metal homeostasis and Parkin function or, more specifically, MTF-1 and Parkin. We found that parkin mutants combined with a knockout of MTF-1 are not viable, a genetic constellation termed synthetic lethality. Parkin mutant Drosophila suffer from oxidative stress as a result of heightened reactive oxygen species (ROS) production. A strong ubiquitous MTF-1 expression dramatically ameliorates the parkin mutant phenotype: our results show that MTF-1 decreases oxidative stress, normalizes the concentration of essential trace metals, increases the frequency of development to adulthood, restores female fertility, improves muscle/mitochondrial morphology and locomotion, and considerably extends the life spans of parkin mutant flies.
One liter of standard fly food was composed of 55 g of cornmeal, 10 g of wheat flour, 100 g of yeast, 75 g of glucose, 8 g of agar, and 15 ml of the antifungal agent Nipagin (15% in ethanol; Brenutag Schweizerhall AG, Basel, Switzerland). For experiments involving survival, development, eclosion frequency, ROS measurements, real-time PCR, and transmission electron microscopy (TEM) muscle analysis, several conditions were tested, namely, normal food (NF), NF supplemented with zinc chloride (4 mM) or N-acetylcysteine (15 mM) or bathocuproine disulfonate (BCS; copper chelator) (0.3 mM), and bathophenanthroline sulfonated sodium salt (BPS; iron [copper] chelator) (0.1 mM). All flies were maintained at 25°C on a 12/12-h light-dark cycle.
UAS-MTF-1 flies were generated by using the full-length MTF-1 cDNA cloned into the pUAST vector. tub-MTF-1 constructs were made by cloning the MTF-1 cDNA under the control of the constitutive α-tubulin promoter. Both constructs were injected into the w1118 fly strain along with the p(Δ2-3) helper plasmid, and transformants were selected based on the eye color (red/orange). The MTF-1140-1R-null allele, generated by homologous recombination, was characterized previously (17). The UAS-MTF-1, MT (tub-MtnA), MTF-1140-1R strains were generated by recombination in our laboratory. We combined (i) w; +; tub-MtnA/TM6B (tub-MtnA), (ii) w; tub-MTF-1/CyO; + (tub-MTF-1), and (iii) w; Actin-Gal4; UAS-MTF-1/TM6B (Actin-Gal4; UAS-MTF-1) flies with w; +; park25/25 (park25/25) flies by recombination. Unless specified otherwise, w; +; park25/TM6B, w+ flies (heterozygous parkin mutants hereafter referred to as park25/+ flies) were used as controls for the w; +; park25/25 flies (homozygous parkin mutants hereafter referred to as park25/25 flies). park25/25 is a null mutation of the Drosophila parkin gene (24). The TM6B balancer in the control flies was confirmed to have no effect on any of the experiments performed (by removing it).
For life span experiments, 1- to 2-day-old flies (20 per vial) maintained at 25°C on a 12/12-h light-dark cycle were examined for each genotype at least in triplicate. Surviving flies were transferred to fresh food vials every 2 days and counted daily. In the experiment with w; Actin-Gal4; UAS-MTF-1, park 25/TM6B flies, the survivors in 23 parallel independent sets were counted at regular time intervals. In each life span assay testing different conditions, the conditions of park25/+ and park25/25 control flies raised on NF were the same. The variations in the average life span of control flies in different experiments can be attributed to subtle experimental variations. The metal chelator concentrations selected for ROS determination had no significant effect on feeding behavior of the flies (55). Fertility was assayed by placing single parkin mutant males with three to four virgin yw females and by placing single virgin parkin mutant females with two yw males. Vials were checked 3 to 7 days later for the presence of larvae and eclosing adults. For the analysis of the eclosion frequency, the number of days allowed for egg laying and the parent population were the same for all vials of NF or zinc (Zn)-supplemented food, and progeny flies were counted at the same time.
A climbing assay was performed as described previously (47). Flies of each of the four genotypes, (i) park25/+, (ii) park25/25, (iii) tub-MTF-1, park25/+, and (iv) tub-MTF-1, park25/25, were anesthetized with CO2 and individually counted and placed in food vials 24 h before the assays were performed to enable a full recovery from the effects of CO2. Ten flies were placed in an empty 110-by 27-mm vial; a horizontal line was drawn 100 mm above the bottom of the vial, and another identical vial was used as a cover to provide more mobile space. After the flies had acclimated for 10 min at room temperature, each genotype was assayed in triplicate for five trials per set per genotype. The procedure involved gently tapping the flies (on a soft surface) down to the bottom of the vial. The flies were given 30 s to climb the vial, and the number of flies that crossed the 100-mm mark each time was recorded. These values were then averaged, and a group mean and standard error was calculated. The mean values of various fly groups were statistically compared using an unpaired Student t test. The study was repeatedly performed with the same group of flies on every alternate day up to 10 days in an isolation room at 25°C and 60 to 70% humidity under standard lighting conditions. Preliminary studies indicated no significant difference in the outcome of the climbing assays performed in normal light or red light conditions.
The Drosophila metallothionein MtnA promoter (−446 to +74) was cloned from genomic DNA using the primer pair 5′-CGG GAT CCA GGT ATG GGC TAT TTA GGC C-3′ and 5′-GGG ATG GCC CCA AAG GAT CTG-3′ in a pCasper4-derived vector carrying EYFP-coding sequence and the simian virus 40 poly(A) site. Details were reported previously (6, 18). Transgenics of MtnA-EYFP combined with parkin heterozygous or homozygous knockouts were made. Both fly types were then frozen at the same age, and photographs of EYFP expression were taken with a Leica MZ FLIII fluorescence stereomicroscope and a Nikon CoolPix 950 digital camera (Leica, Heidelberg, Germany) at an exposure of 730 ms.
Total RNA was purified from adult Drosophila tissue by using the Nucleospin RNA II protocol (Macherey-Nagel) and eluted in 60 μl of RNase-free water. cDNA was prepared by using the Transcriptor high-fidelity cDNA synthesis kit from Roche. The cDNA obtained was further purified by using an AM 1906 Ambion DNA free kit and used for analysis by real-time PCR on the Tecan Genesis 200/8 robot using the Eurogenentec Mesa Green qPCR Mastermix Plus for SYBR assays. The quantitative-PCR (qPCR) run was performed on an Applied Biosystems apparatus (ABI Prism SDS 7900 HT) in a 384-well format with a reaction volume of 10 μl. ΔΔCT values were calculated by subtracting the ΔCT calibrator value from the ΔCT sample value, and ΔCT values were calculated by subtracting the ΔCT endogenous control value from the ΔCT target gene/calibrator value. The normalization strategy used has been described in Vandesompele et al. (67). All of the fold change values are normalized to the respective park25/+ values on NF. The housekeeping genes used were actin5c, TBP, and GAPDH. Two sets of primer sequences were used for each of the transcripts quantified: for parkin, the first primer set was 5′-AAG ATC ATA TTT GCC GGT AAG GAA-3′ and 5′-CGC TTT GCT GAC CCA AGT C-3′, which amplify a 73-bp fragment only from the parkin heterozygous control flies, and the second set was 5′-CAA AGC CCT GTC CAA AAT GC-3′ and 5′-GCG CGT GTG CAG ACC AT-3′; for MTF-1, the first primer set was 5′-TGT CCG GCT GCG ATA AGG-3′ and 5′-GCC ATT GTG CAG ACG AAG GT-3′, which amplify a 68-bp fragment from wild-type MTF-1-containing flies, and the second set was 5′-GCA TTC AAC ACG CGC TAC A-3′ and 5′-ACA GTT GAA CGT CTC GCC ATT-3′; for MtnB, the first primer set was 5′-TTG CAA GGG TTG TGG AAC AAA-3′ and 5′-TGC AGG CGC AGT TGT CC-3′, which amplify a 65-bp fragment and the second set was 5′-AAG TCG AGA AAT AGA TAC ATA CAA GAT GGT-3′ and 5′-CGC ACT TTT GGG CCG AG-3′; and for foi, the first primer set was 5′-GTG GCT GCG GGT CTG TTC-3′ and 5′-TTT GTG CGA GGC CGA GAT-3′, which amplify a 69-bp fragment, and the second set was 5′-TGG CGA TGC CCT ACT TCA C-3′ and 5′-TGA TCA TCC CCC GCT CAT-3′.
Fresh dichlorofluorescein diacetate (2,7-DCFH-DA) from Invitrogen/Molecular Probes catalog no. C369 was mixed with dimethyl sulfoxide (DMSO) to make a 1 mM stock solution (2,7-DCFH-DA/DMSO). A 40 μM working solution was prepared in HEPES buffer (30 mM). A total of 50 heads of frozen adult flies of a specific genotype (same age) were removed and collected in an Eppendorf tube. This was done in triplicate for each condition. Each sample was then homogenized using cold protein homogenization buffer (a 1:3 [wt/vol] dilution of 0.32 mM sucrose, 20 mM HEPES, 1 mM MgCl2, and 0.5 mM phenylmethylsulfonyl fluoride [PMSF] protease inhibitor at pH 7.4) and centrifugation for 20 min at 4°C and 20,000 × g. The protein content of the supernatant was determined by using the Bio-Rad diagnostics kit, and a final concentration of 0.4 mg/ml was used as the standard for the ROS assay of each genotype tested. The fluorescence intensity (emission acquisition) was monitored for 45 min after the sample (20 μl) was kept in a cuvette (1.5 by 1.5 mm) in the fluorimeter immediately after the addition of the DCFH-DA dye (2 μl) to prevent loss of signal due to fading of fluorescence. Excitation of dye was at 485 nm and emission at 520 nm. The curve area of fluorescence intensity, which was recorded every 5 min (for 45 min) in the range of 500 to 600 nm was integrated, and the total area was used for comparison, with the final result obtained in counts per second. The fluorimeter was standardized using 0.05% H2O2, a positive ROS-generating species.
Ovaries from female parents were dissected in Grace's insect medium (1×; Gibco/Invitrogen) at room temperature (RT). The dissected ovaries were immediately fixed in 4% paraformaldehyde, 0.2% Triton X-100 dissolved in Grace's medium for 20 min without shaking, The fixative was washed three times with phosphate-buffered saline (PBS) plus 0.5% Triton (PBST) for 10 min each. The samples were then labeled for 1 h with fluorescently labeled phalloidin (Phalloidin-Alexa 568; Molecular Probes) and diluted 1:200 in PBST at room temperature. This was followed by three washes with PBST for 10 min each and a second labeling with Toto (nuclear stain) diluted 1:1,000 in PBST at room temperature. The samples were washed three times with PBST for 10 min each time, followed by two washes with PBS for 10 min each, and then embedded in Vectashield (mounting medium) overnight at −20°C. Ovarioles were dissected from the ovaries and mounted on glass slides. Pictures were taken using a confocal lens at 20× magnification.
Female flies, (i) park25/+, (ii) tub-MTF-1, park25/+, and (iii) yw and MTF-1 knockout controls, were allowed to lay eggs on NF or metal-supplemented food (100 μM cadmium sulfate, 500 μM copper sulfate, 500 μM ammonium ferric citrate, and 4 mM zinc chloride) for 4 days and removed afterwards. The resulting progeny were collected at regular intervals and frozen. This procedure was repeated until the required number of 50 flies was obtained in triplicate for each genotype. Each sample set of frozen flies was then subjected to homogenization using cold protein homogenization buffer (0.32 mM sucrose, 20 mM HEPES, 1 mM MgCl2, 0.5 mM PMSF protease inhibitor at pH 7.4), and the samples were normalized for protein content. A final concentration of 1 mg/ml was prepared by diluting the samples in 0.2 M HNO3 to obtain a total assay volume of 1 ml. A highly sensitive flame atomic absorption spectrophotometer (FAAS; GTA-120/PSD-120, Varian Australia, Pty., Ltd., Mulgrave, Victoria, Australia) was used to detect the metal content in each genotype assessed. Cd and Zn concentrations were recorded by the same flame. Likewise, Cu and Fe concentrations were measured together.
Dissected thoraces of 2-day-old anesthetized adult flies (park25/+, park25/25, and Act-Gal4; UAS-MTF-1, park25/25) were kept in ice-cold fixative (2.5% glutaraldehyde-0.1 M sodium cacodylate buffer adjusted to 328 mOsm/liter with sucrose [pH 7.4]) for 4 h at 4°C. Postfixation was performed with 1% OsO4 and 0.1 M sodium cacodylate (pH 7.4) for 2 h at 4°C, and the sample was washed overnight with 0.1 M sodium cacodylate (pH 7.4) before going through a series of progressive dehydration steps in a graded ethanol series of 70, 80, and 96% alcohol for 10 min each time, followed by three 10-min washes in 100% alcohol and a final 20-min wash in propylene oxide. The sample was treated with a 1:1 propylene oxide-Epon (Epon 812) mix for 2 h and then embedded in Epon overnight at 70°C. Blocks of thoraces were trimmed, and semithin sections of dorsal longitudinal muscles were stained with toluidine blue dye, which labels nucleic acids, thus staining both the nuclei and the cytoplasm (5). Ultrathin sections of 70-nm thickness were made with an ultramicrotome for the selected sections. For TEM, sections were contrast stained on the grid first with 2% uranyl acetate and then with 2.5% lead citrate (Reynold's), each for 20 min at room temperature. The sections were inspected with a Philips CM100 TEM with a GATAN Orius camera.
JMP software (SAS Institute) was used for statistical evaluations. Life span (survival) assays were analyzed with the Kaplan-Meier log rank statistical test. Brain ROS levels and qPCR results were compared by one-way analysis of variance. The results are expressed as mean ± the standard deviation.
The metal-responsive transcription factor 1 (MTF-1) is a key regulator of metal homeostasis in Drosophila (17, 60, 75). Based on our findings of a connection between the parkin knockout and trace metal status (55), we tested the effect of an MTF-1 knockout in a parkin mutant background. The result was clear-cut, in that no surviving double mutant was ever observed in 50 independent crosses. This “synthetic lethality,” observed at the pupal stage, was rescued by a cDNA transgene of MTF-1 driven by the constitutive tubulin enhancer/promoter. In a genetic cross between parents heterozygous for the parkin and MTF-1 recombined deletions (Fig. 1), 69 of 217 progeny (32%) were parkin and MTF-1 homozygous knockouts also expressing an MTF-1 transgene (statistical expectation, 40%). This rescue by an MTF-1 cDNA transgene confirms the absence of any secondary hits as the cause of the observed lethality. Furthermore, it also excludes the possibility that an intronic open reading frame located within the MTF-1 gene (D. Steiger, K. Steiner, and W. Schaffner, unpublished data) is responsible for the effect and not MTF-1 itself. To find out what condition could overcome the pupal lethality, we maintained the heterozygous parkin, MTF-1 parent flies on N-acetylcysteine (NAC), which is a precursor to glutathione, an established antioxidant. Since in either condition the lack of a functional parkin gene or of an MTF-1 gene increases the ROS level (see below), we reasoned that keeping a lower ROS level by other means might also overcome the synthetic lethality of the double mutants. After testing different concentrations of NAC, an optimal supplement of 15 mM was chosen for the main experiment. Indeed, raising the progeny under this condition resulted in a substantial rescue of the synthetic lethality: of the total progeny, 19% were homozygous parkin and MTF-1 double mutants (statistical expectation, 25%) (Table 1). Other antioxidants tested included ascorbate, zinc, and metal chelators of copper and iron (BCS and BPS, respectively), which we had shown before to positively influence the parkin mutant phenotype (55, 56). However, none of these was able to rescue synthetic lethality.
We examined the life span of Drosophila MTF-1-overexpressing transgenic lines (tub-MTF-1 and Act-Gal4; UAS-MTF-1) in a park25/25 background. These experiments illustrated that an elevated expression of MTF-1 from the ubiquitously active tubulin enhancer/promoter prolonged the life span of parkin mutants significantly, from a median of 7 days for the mutants alone to 21 days (Fig. 2A). In this experiment, the maximal life span was extended from 12 to 41 days by the MTF-1 transgene (Fig. 2A). The stronger combination with actin-Gal4 driving UAS-MTF-1 revealed a similar effect: in an experiment with 23 independent replicas, 10% of the mutant animals were still alive at day 34 (Fig. 2B). Overexpression of MTF-1 in control and wild-type flies (park25/+ or park+/+) did not increase their normal life span (data not shown). Also, in an independent study, MTF-1 overexpression did not extend life span of flies kept on standard food (4). Elevated MTF-1 expression not only prolonged the life span of parkin mutant adult flies but also enhanced survival during development. In a genetic cross involving parkin heterozygous parents, only 2.5% of the eclosing progeny flies were homozygous parkin mutants (statistical expectation, 25%). In comparison, the same cross but also expressing an MTF-1 transgene resulted in a 15% eclosion frequency (Table 2).
Strikingly, female fertility and fecundity was completely rescued by MTF-1. When crossed with wild-type males, parkin mutant females with the tubulin-driven MTF-1 transgene produced the same number of progeny as a cross of wild-type males and females. Drosophila gonad formation requires a complex morphogenetic process (35, 37). As in the majority of metazoans, Drosophila oogenesis occurs within the ovarian follicles in which germ line cells develop in close proximity to specialized somatic cells (Fig. 3A to A″). Parkin mutant females lack the proper spatiotemporal development in the germarium and thus have stunted ovaries with few mature oocytes, which fail to get fertilized (Fig. 3B to B). The restoration of female fertility by strong MTF-1 expression was also evident at the morphological level: dissected ovaries showed a normalized structure with follicles formed in the germarium and mature stages in the posterior regions of the ovariole, with several oocytes ready for fertilization (Fig. 3C to C″). In contrast, the sterility phenotype of parkin mutant males, which is due to defective spermatogenesis at the individualization step (24), was not rescued. This suggests that Parkin is particularly important for male fertility.
Strong MTF-1 expression dramatically improved the climbing ability of parkin mutant flies (Fig. 4). What is more, they generally moved around fast, responded by running away when physically perturbed, jumped, and occasionally displayed short flight episodes (data not shown). To investigate the effect of elevated MTF-1 expression on muscle morphology, we examined the ultrastructure of the indirect flight muscle in heterozygous control flies, parkin mutants, and the MTF-1 transgenic flies. Cross-thoracic sections of control adults analyzed by TEM revealed well-organized muscle fibers in parallel stripes with a regular M- and Z-line banding pattern and darkly stained, electron-dense mitochondria with regularly packed cristae (Fig. 5A to C). In contrast, age-matched parkin mutants had abnormal muscle structure with large vacuoles, a reduced muscle content with mostly irregular arrangement and enlarged mitochondria with disintegrated cristae (Fig. 5D to F). The MTF-1-overexpressing parkin mutant flies displayed a clear rescue effect in that muscle fiber structure was more regular with less prominent vacuoles; moreover, mitochondria had more densely packed cristae with considerably lesser signs of disintegration in comparison to parkin mutants (Fig. 5G to I).
Owing to their high cysteine content, metallothioneins can act as antioxidants, in addition to their obvious role as metal chelators (33, 38). Basal and induced levels of metallothionein expression depend on the transcription factor MTF-1 (17, 51, 74). To find out whether metallothionein genes could be induced by MTF-1 in a parkin mutant background, we used a reporter line in which enhanced yellow fluorescent protein (EYFP) is driven by the promoter of MtnA, the most highly expressed Drosophila metallothionein gene. Compared to heterozygous controls, even in the absence of any heavy metal load the basal expression of this reporter was increased, likely due to elevated oxidative stress in the parkin mutant flies (Fig. 6C and D). This metal-independent upregulation of the MtnA promoter, was strictly dependent on MTF-1, since no trace of fluorescence could be detected in flies lacking MTF-1 (data not shown). Reverse transcription-PCR (RT-PCR) of parkin mutants also revealed elevated transcript levels of MTF-1 (2-fold) (Fig. 7A) and the embryo-enriched metallothionein MtnB (6-fold) (Fig. 7B) compared to park25/+ flies. MTF-1 overexpression in park25/25 flies was achieved from the tubulin promoter or indirectly with the stronger UAS-Act-Gal4 system; both induced a >200-fold increase in MtnB transcripts (Fig. 7B). Conversely, the level of parkin transcripts was increased in the MTF-1 knockout flies (Fig. 7C). This pattern of regulation can be explained by a partial redundancy of Parkin and MTF-1, where one is upregulated to compensate for the loss of the other. The synthetic lethality of a combined knockout of parkin and MTF-1 genes mentioned above is in line with this hypothesis. Another experiment revealed that the transcript levels of the zinc importer foi (37) in parkin mutants were enhanced 3-fold by elevated MTF-1 expression, which may contribute to the normalized structure of ovaries and rescued fertility of female parkin mutants (Fig. 7D). This hypothesis is further supported by the increased level of zinc ions found in tub-MTF-1; park25/25 flies (discussed below, see also Fig. 8B).
Parkin mutant flies display high levels of ROS indicative of intrinsic oxidative stress (23, 72). In agreement with MTF-1 having an antioxidant function, we observed a substantial decrease of ROS levels in the heads of parkin mutants expressing an MTF-1 transgene: ROS dropped to approximately half of the levels in park25/25 flies (Fig. 8A). Previously, we had observed that limiting the availability of redox-active metals, achieved by supplementing the food with chelators for copper and iron (BCS and BPS, respectively), also increased the life span of parkin mutant flies (55). Furthermore, the w; tub-MTF-1; park25/25 flies raised on the metal-chelator-supplemented food displayed a somewhat lower ROS level (Fig. 8A). ROS levels were not significantly changed in heterozygous control flies, either upon ubiquitous MTF-1 overexpression or following dietary intake of metal chelators. The MTF-1 knockout flies showed the highest levels of ROS, probably due to the reduced expression of MTF-1-dependent antioxidant genes such as metallothioneins (Fig. 8A). Metals such as zinc and redox active copper and iron are required in trace amounts for several structural and biological processes in organisms (66). park25/25 flies display not only reduced basal levels of zinc (see also reference 56) but also of copper and iron in comparison to control flies (Fig. 8B). Tubulin-driven MTF-1 expression in park25/25 restores the basal level of these metals (Fig. 8B). The concentrations of cadmium are generally low since it is a nonessential, toxic metal. Upon supplementing fly food with metals, the levels became quite similar in all three genotypes tested (Fig. 8C). This is particularly important in the case of zinc supplementation, which we had previously shown to improve the condition of parkin mutant Drosophila (56).
Human MTF-1 has been expressed in Drosophila and shown to largely, but not completely, rescue the metal sensitivity of Drosophila lacking its endogenous MTF-1 (6). We therefore tested the effect of actin-Gal4-driven hMTF-1 expression in a parkin mutant background. Indeed, the life span of parkin mutants was increased from a median of 7 days to 19 days, which is close to the 21 days obtained with elevated Drosophila MTF-1 expression (Fig. 9). Other rescue effects paralleled those observed with elevated Drosophila MTF-1 expression (see above) but were less pronounced (data not shown).
Metallothionein genes are the major targets of MTF-1; in Drosophila, the metallothionein A gene (MtnA) shows the strongest expression. Thus, we also tested whether the overexpression of MtnA in a parkin mutant background exerted a beneficial effect similar to that of MTF-1 overexpression. To this end, we crossed inter se three independent lines with tubulin-driven MtnA overexpression in a parkin heterozygous background to raise parkin homozygous knockouts. The median life span of the parkin mutants was extended up to 17 days (Fig. 9), but other rescue effects associated with elevated MTF-1 expression (discussed in results) were not observed.
We show here that metal-responsive transcription factor 1 (MTF-1) plays a crucial role in modulating the severity of a Parkin loss-of-function phenotype. On the one hand, the combined loss of Parkin and MTF-1 is not viable, i.e., displays synthetic lethality. On the other hand, elevated expression of MTF-1 dramatically improves the condition of parkin mutant flies: there is an overall extension of life span, females regain normal fertility, and the motoric abilities of flies improve to the point that they can walk fast and even display short episodes of flight. The latter is noteworthy since flight muscles have a high energy consumption that depends on robust mitochondrial function. At the histological level, the degenerated mitochondria characteristically seen in parkin mutants are rescued to a more regular, electron-dense structure following MTF-1 overexpression. In addition, flight muscles have an improved myofibrillar arrangement. In earlier studies, mitochondrial and muscle degeneration observed in parkin mutant flies were proposed to be a result of excessive oxidative stress (24, 48). Mitochondrial malfunction can indeed result in an increased susceptibility to oxygen radical damage, and mitochondrion-associated increase in ROS production has been implicated in Parkinson's disease (19, 26, 73). In line with this concept, the major target genes of MTF-1 are metallothioneins (MTs), which encode small, cysteine-rich proteins that can scavenge heavy metals, notably the redox-active copper, and ROS. The elevated basal level of MTF-1 and MT transcripts in parkin mutant flies can thus be seen as a compensatory attempt to counteract enhanced ROS levels (24). In accordance with such a scenario is the rescue effect seen upon food supplementation with the antioxidant N-acetylcysteine, which, however, falls short of the dramatic effect seen with elevated MTF-1 expression. Besides metallothioneins and a number of other stress-associated genes, ferritin genes are a major target of MTF-1 in Drosophila (74). Ferritins are well-characterized iron-binding proteins that keep iron in a soluble and nontoxic form in the cell. Thus, ROS production by redox-active iron might be lowered via the upregulation of ferritin levels. The multiple targets of MTF-1 help to explain why overexpression of metallothionein alone was less effective than MTF-1 overexpression in improving the condition of parkin mutants. In a separate series of experiments we found that strong expression of an ortholog, the human MTF-1, also increased life span, rescued female fertility, and improved locomotion ability, although the effects were less pronounced than with Drosophila's own MTF-1, presumably due to the evolutionary distance between mammals and insects. This is in line with previous findings from our laboratory that mammalian and Drosophila MTF-1 transgenes are able to largely, but not completely, compensate for each other's absence (6).
Although we observed remarkable improvements following elevated MTF-1 expression, a complete rescue of parkin mutants, including male fertility and sustained flight ability, was only observed with a parkin transgene. Furthermore, it has to be pointed out that although MTF-1 knockout flies display high ROS levels, they show no signs of Parkinson's disease-like symptoms. Together, these findings are consistent with the idea that oxidative stress is an important but not the sole culprit in PD etiology (1, 14, 49). Nevertheless, the dramatic effect of MTF-1 on a parkin loss-of-function mutation underscores the importance of this transcriptional regulator in cellular stress response. Furthermore, it is tempting to speculate that polymorphisms of the MTF-1 gene in humans will affect the onset and/or the severity of Parkinson's disease.
We thank Jessica Greene and Leo Pallanck (University of Washington) for providing the park25/25 flies, Gery Barmettler and Theresa Bruggmann (Center for Microscopy and Image analysis) for TEM analysis of the muscle sections, Martin Moser for real-time analysis, Eva Freisinger and Tamara Huber (Inorganic Chemistry, University of Zürich [UZH]) for FAAS experiments, Ivan Ostojic (FMI Basel) for statistical analyses, and Ben Schuler (Department of Biochemistry, UZH) for providing the fluorimeter for ROS analysis. We also thank Till Strassen for maintenance of the fly stocks and George Hausmann for critical reading of the manuscript.
This study was supported by Swiss National Science Foundation grant 31003A-113993 and the University of Zürich.
Published ahead of print on 7 March 2011.
†The authors have paid a fee to allow immediate free access to this article.