Parkin overexpression during aging reduces proteotoxicity, alters mitochondrial dynamics, and extends lifespan

Adult-Onset Induction of Parkin Increases Longevity.

To determine whether Parkin can modulate organismal aging, we first examined the impact of Parkin overexpression on Drosophila lifespan. Studies of the genetics of aging and lifespan determination are prone to confounding effects because of uncontrolled differences in genetic background between test and control lines (35). To confront this issue, we have used the mifepristone (RU486) inducible-GAL4 system [annotated P[Switch] or Gene-Switch (36, 37)] throughout this study. This system eliminates genetic background effects because all flies share the same genetic background and only differ with respect to the presence of the inducing agent (RU486) or diluent (ethanol) in the food. Firstly, we used the ubiquitous daughterless–Gene-Switch (daGS) driver line to activate a Upstream Activating Sequence (UAS)-parkin transgene created by Greene et al. (19) during both development and adulthood. Western blot analysis showed that Parkin is significantly overexpressed in daGS>UAS-parkin flies upon RU486 treatment during development and the adult lifespan compared with uninduced flies (Fig. 1A and Fig. S1 A and B). RU486 treatment had no effect on Parkin expression levels in control flies (Fig. S1 C and D). Ubiquitous up-regulation of Parkin with this driver line resulted in as much as a 28% increase in both the mean and also the maximum lifespan in female flies (Fig. 1B and Fig. S2A) and smaller positive effects in male flies (Fig. S2 B and C). Interestingly, lower levels of Parkin induction are sufficient to extend lifespan (Fig. S2 D and E), whereas increased doses of RU486 produced diminishing returns in terms of lifespan extension (Fig. S2 F and G). No major longevity effects were observed in female or male control flies exposed to RU486 (Fig. S2 H–K). Next, we took advantage of the inducible nature of the daGS driver to determine the temporal requirements for Parkin-mediated lifespan extension. Whereas induction of the parkin transgene exclusively during development had no major impact on adult longevity (Fig. S2L), daGS>UAS-parkin flies were significantly longer-lived when exposed to RU486 exclusively in the adult stage (Fig. 1C); again, no major effects on longevity were observed in control flies exposed to RU486 using these paradigms (Fig. S2 M and N). Together, our data demonstrate a role for Parkin in prolonging lifespan in adult Drosophila.

Toward an initial exploration of the mechanisms underlying Parkin-mediated lifespan extension, we examined a number of behavioral and physiological parameters in long-lived Parkin flies. Because overexpression of Parkin during both development and adulthood resulted in maximum extension of lifespan, this paradigm was used in all further experiments. Long-lived Parkin flies displayed normal food consumption (Fig. 1D and Fig. S3A) and an increase in spontaneous physical activity in young flies (Fig. S3 B–D). Moreover, although interventions that extend lifespan are often associated with reproductive tradeoffs (38), Parkin overexpression resulted in an increase in fecundity in young flies (Fig. S3E). An increased capacity to withstand extrinsic stress is a hallmark of many longevity interventions (39, 40). Interestingly, Parkin-mediated longevity was not associated with increased resistance to hyperoxia (85% oxygen) (Fig. S3F). However, long-lived Parkin flies showed improved survival when maintained on an agar-only diet to induce starvation (Fig. 1E). Feeding RU486 to control flies (daGS>w¹¹¹⁸) did not alter feeding behavior (Fig. S3G), physical activity (Fig. S3 H–J), fecundity (Fig. S3K), hyperoxia resistance (Fig. S3L), or starvation resistance (Fig. S3M). Taken together, our data indicate that Parkin-mediated lifespan extension is not associated with reduced feeding behavior, motor activity, or a reproductive tradeoff.

Parkin Overexpression Counteracts Age-Onset Proteotoxicity.

Parkin functions as an E3 ubiquitin ligase to mediate attachment of ubiquitin monomers or chains to substrate proteins (12). To better understand Parkin-mediated lifespan extension, we set out to examine the relationships between parkin gene activity, polyubiquitination, and aging. K48-linked polyubiquitin predominantly, but not exclusively, targets proteins for proteasomal degradation, whereas K63-linked polyubiquitin regulates protein function, subcellular localization, or protein–protein interactions (11, 41). Firstly, we examined levels of K48-linked polyubiquitination in long-lived Parkin flies and controls during aging. Whereas no differences were observed in young flies, chronic overexpression of Parkin resulted in a significant increase in K48-linked polyubiquitination in aged flies (Fig. 2A and quantification in Fig. S4A). Feeding RU486 to control flies (daGS>w¹¹¹⁸) did not alter the levels of K48-linked polyubiquitination during aging (Fig. S4 B and C). Interestingly, however, no significant difference in the levels of K63-linked polyubiquitin chains were observed in long-lived Parkin flies (Fig. S4D).

To determine whether Parkin-mediated lifespan extension was associated with reduced proteotoxicity during aging, we characterized the age-related deposition of protein aggregates by immunofluorescence. In agreement with previous findings (7), we observed that Drosophila flight muscles accumulate aggregates of ubiquitinated proteins during aging (compare Fig. 2 B and D). The cumulative area of these aggregates increases substantially with age (compare Fig. 2 F and H and see quantification in Fig. 2 J and K), consistent with a loss of proteostasis. Importantly, long-lived Parkin flies displayed reduced levels of protein aggregates during aging (compare Fig. 2 D and E) and also a decrease in the size of individual aggregates (Fig. 2 H–K). Feeding RU486 to control flies (daGS>w¹¹¹⁸) did not alter the accumulation of protein aggregates during aging (Fig. S4 E–G). In a complementary approach, we examined the levels of insoluble ubiquitinated proteins in whole bodies of long-lived Parkin flies and controls during aging. Consistent with the immunofluorescence data, we observed that long-lived Parkin flies display reduced levels of insoluble ubiquitinated proteins during aging (Fig. 2L and quantification in Fig. S4H). Feeding RU486 to control flies (daGS>w¹¹¹⁸) did not alter the levels of insoluble ubiquitinated proteins during aging (Fig. S4 I and J). Taken together, our findings indicate that up-regulation of Parkin preserves protein homeostasis during aging, which is associated with reduced levels of proteotoxicity in aged flies.

Parkin Overexpression Modulates Mitochondrial Dynamics and Activity During Aging.

In recent years, considerable attention has been focused on the role of Parkin in mediating the elimination of dysfunctional mitochondria via mitophagy (16–18). Although the details of how this happens remain uncertain, a growing body of data supports the idea that the Parkin-mediated polyubiquitination of mitochondrial outer-membrane (OM) proteins is important for mitophagy (17). Indeed, proteomic studies have shown that the translocation of Parkin to dysfunctional mitochondria is associated with a broad activation of the UPS, resulting in the K48-linked polyubiquitylation and degradation of many OM proteins (27). Studies in Drosophila and mammalian cells have shown that Mfn represents an evolutionarily conserved mitochondrial substrate of Parkin (23, 25, 42). These findings have led to the proposal that the ubiquitin-mediated turnover of Mfns would then inhibit mitochondrial fusion and, thus, promote the segregation of dysfunctional mitochondria and facilitate their clearance by mitophagy (16–18, 23, 25, 43).

Because mitochondrial dysfunction is a hallmark of aging (2), we set out to examine the relationships between parkin gene activity, aging, and mitochondrial dynamics. In the first place, we sought to explore the effects of aging on the steady-state abundance of the dMfn. Using antibodies against dMfn (23), we examined dMfn levels in control and long-lived Parkin flies during aging. In doing so, we observed a significant increase in dMfn levels in aged control flies relative to young control flies (Fig. 3A and quantification in Fig. S5A). Interestingly, long-lived Parkin flies displayed significantly reduced levels of dMfn at both time points. Next, we examined whether Parkin-mediated changes in dMfn levels were associated with altered mitochondrial morphology during aging. Drosophila adult flight muscle is an ideal system to study alterations in mitochondrial structure and function (44–46). In this tissue, we observed mitochondrial fragmentation (i.e., smaller and rounder mitochondria), as visualized by electron microscopy (EM), in young and aged Parkin-overexpressing flies (compare Fig. 3 B, b and D, d with Fig. 3 C, c and E, e, respectively, and quantification in Fig. 3F). These findings are consistent with a model whereby Parkin-mediated turnover of dMfn inhibits mitochondrial fusion and, thereby, results in an increased segregation of individual mitochondria from the mitochondrial network. Feeding RU486 to control flies (daGS>w¹¹¹⁸) did not alter dMfn levels during aging (Fig. S5B) or mitochondrial morphology (Fig. S5 C–G).

Next, we sought to determine whether Parkin-mediated changes in mitochondrial dynamics were functionally significant. To do so, we assayed a number of markers of mitochondrial activity in long-lived Parkin flies and controls during aging. Firstly, we assayed the enzymatic activity of citrate synthase, a key enzyme in the Krebs cycle and a widely used mitochondrial marker. Indeed, Parkin overexpression conferred a significant increase in citrate synthase activity both in young and aged flies (Fig. 3G). Furthermore, long-lived Parkin flies showed an increase in the activity of mitochondrial respiratory complexes I (Fig. 3H) and complex II (Fig. 3I) in young flies. The peroxisome proliferator-activated receptor-γ coactivator (PGC)-1 family of transcriptional coactivators are key regulators of mitochondrial activity in mammals (47) and Drosophila (48, 49). Hence, we examined the expression of dPGC-1 in long-lived Parkin flies and controls. Interestingly, we observed that Parkin-mediated changes in mitochondrial activity were associated with increased dPGC-1 gene expression in young flies (Fig. 3J). Feeding RU486 to control flies (daGS>w¹¹¹⁸) did not alter markers of mitochondrial activity during aging (Fig. S5 H–K). Taken together, our findings show that Parkin-mediated longevity is associated with alterations in mitochondrial dynamics and activity during aging.

Parkin Overexpression in Adult Neurons Extends Longevity.

To better understand the relationships between parkin gene activity, chronic maintenance of the nervous system and organismal aging, we examined the impact of neuron-specific overexpression of Parkin on Drosophila lifespan. To do so, we used the pan-neuronal Elav–Gene-Switch (ElavGS) driver line to increase parkin gene expression specifically in neurons. Firstly, we validated that there was an increase in parkin gene activity in heads of ElavGS>UAS-parkin flies upon RU486 treatment (Fig. S6 A and B). Our previous results with the ubiquitous daGS driver line showed that increasing Parkin levels exclusively in adult tissues can extend lifespan. Therefore, we tested the effects of both constitutive and adult-onset neuronal overexpression of Parkin to refine our understanding of the spatiotemporal requirements of Parkin-mediated lifespan extension. Induced neuronal up-regulation of Parkin, both constitutively (Fig. 4A) and in adult neurons (Fig. 4B), increased both mean and maximum lifespan in female flies and produced no major effect in male flies (Fig. S6C). Other doses of RU486 gave diminishing returns in lifespan extension and higher doses shortened lifespan (Fig. S6D). Again, no positive effects on longevity were observed in control flies exposed to RU486 using this paradigm (Fig. S6 E–G). Similar to our findings with ubiquitous overexpression, we observed that neuron-specific overexpression of Parkin did not reduce food consumption (Fig. 4C and Fig. S7A) or physical activity (Fig. S7 B–D) and increased fecundity in young flies (Fig. S7E). Interestingly, neuron-specific Parkin overexpression was sufficient to improve survival under starvation conditions (Fig. 4D). Feeding RU486 to control flies did not alter feeding (Fig. S7F), starvation resistance (Fig. S7G), physical activity (Fig. S7 H–J), or fecundity (Fig. S7K).

Neuronal Overexpression of Parkin Reduces Proteotoxicity and Enhances Mitochondrial Activity in the Aging Brain.

To better understand neuronal Parkin-mediated lifespan extension, we examined a number of markers of Parkin activity in heads, which are greatly enriched for neuronal tissue. Firstly, we examined levels of K48-linked polyubiquitination in heads of long-lived neuronal Parkin flies and controls during aging. Whereas no differences were observed in young flies, neuronal overexpression of Parkin resulted in a significant increase in K48-linked polyubiquitination in heads of aged flies (Fig. S8A). To determine whether targeted activation of Parkin impacts protein homeostasis in the aging brain, we examined the levels of protein aggregates in aged brains of neuronal Parkin-overexpressing flies and controls. Long-lived flies with neuronal Parkin overexpression displayed reduced levels of protein aggregation in the aged brain (Fig. 4E and Fig. S8C). Feeding RU486 to control flies did not alter levels of protein aggregation in the aged brain (Fig. S8 B and B′). Together, these data indicate that neuronal Parkin overexpression can protect against proteotoxicity in the aging brain.

To determine whether Parkin overexpression can impact mitochondrial metabolism in aging neurons, we examined a number of markers of mitochondrial activity in heads. Interestingly, neuronal overexpression of Parkin was associated with increased citrate synthase activity (Fig. 4F) and dPGC-1 gene expression (Fig. S8D) in heads of young flies and an increase in respiratory complex I activity in heads of aged flies (Fig. 4G). Feeding RU486 to control flies (ElavGS>w¹¹¹⁸) did not alter markers of mitochondrial activity in heads during aging (Fig. S8 E–I). Taken together, these findings indicate that an increase in neuronal Parkin expression protects against proteotoxicity and enhances mitochondrial activity in the aging brain. Moreover, these Parkin-mediated changes in the aging brain result in an increase in health span at the organismal level.

Lifespan Analysis and Physiological Assays.

Lifespan analysis and other physiological assays were carried out as described previously (48). See SI Materials and Methods for details.

Immunostaining and EM.

Hemithoraces or brains were immunostained and analyzed as described in ref. 7. Standard procedure for EM analysis was carried out as described previously (44, 45). See SI Materials and Methods for details.

Molecular Biology.

Protocols for Western blot, analysis of detergent-insoluble fractions, mitochondria purifications, citrate synthase assay, and complex I and II activity assays used in this study can be found in SI Materials and Methods.