Pink1 (PTEN induced putative kinase 1) encodes a serine-threonine kinase which has been linked to autosomal recessive and some sporadic forms of Parkinson disease [1-3]. Targeted to the mitochondria, Pink1 is thought to be involved in mitochondrial protection by preventing oxidative damage and maintaining mitochondrial integrity, where loss of function of Pink1, in humans and in Drosophila melanogaster, show substantial mitochondrial defects in sensitive tissues [4-8]. It is becoming increasingly apparent that protection during cell stress is due to the involvement of Pink1 in mitochondrial fission/fusion events [1,9]. This involvement implicates Pink1 as a key regulator of fission/fusion, acting upstream of the E3 ubiquitin ligase, parkin, to maintain proper mitochondrial integrity and function [4,6,10,11]. In this role, recruitment of parkin to the mitochondria by Pink1 results in the ubiquitination of various mitochondrial proteins, promoting fission and mitophagy [12-15]. In contrast, studies have found that loss of parkin or Pink1 function can also result in increased fission, promoting mitophagy [16,17]. Although the fission/fusion decision is not fully understood, results do highlight the importance of the Pink1/parkin pathway in maintaining mitochondrial homeostasis.

The Pink1/parkin pathway has been linked to components of the insulin receptor (INR) pathway, including: interaction of PTEN with Pink1 [18] and DJ-1 [19,20], an indirect interaction with Akt through parkin [21], an interaction with Akt through DJ-1 [22,23], and transacttivation of Pink1 by Foxo [24,25]. It has also been suggested that Pink1 may activate Foxo indirectly through Sir2 [26]. This is thought to be a protective mechanism, where Foxo activation results in the transcription of genes such as SOD2 and Thor. In addition to genes that promote stress resistance, under conditions of oxidative stress or starvation, Foxo transcription factors may also target genes that promote cell cycle arrest and apoptosis [27,28]. Overexprssion of Foxo has been linked to neurotoxicity [29,30] and overexpression in the developing Drosophila eye results in a characteristic phenotype with reductions in cell number and area [31]. Genetic expression studies using the fly eye have been enormously successful in the study of neurodegeneration. This is due to the conservation of key signaling pathways between humans and Drosophila, and the ease of quantifying degeneration of photoreceptor neurons associated with each Drosophila ommatidium. To further investigate the roles of Pink1/parkin, we have performed expression studies to determine the effects of Pink1 and parkin on the Foxo-induced phenotype of developmental defects in the Drosophila eye. We hypothesized that through an interaction with the INR pathway, or through mitochondrial protective effects, Pink1 and parkin would be capable of alleviating the detrimental effects of Foxo overexpression. In contrast, our findings show that reduction of Pink1 is able to suppress the effects of Foxo overexpression, where co-overexpression of Foxo with Pink1 or parkin results in an increased severity of the Foxo-induced phenotype. These findings suggest a complex role for the Pink1/parkin pathway in cell fate decisions.

Under cell stress conditions, Foxo transcription factors are activated and target genes that promote cell survival and/or apoptosis [27,28]. The transactivation of Pink1 by Foxo [24,25] suggests that there may be recruitment of the Pink1/parkin pathway to help maintain mitochondrial homeostasis during cell stress. Acting in this protective role, we hypothesized that Pink1 and parkin may alleviate the Foxo-induced phenotype of developmental defects in the Drosophila eye. This would presumably be

due to the regulation of mitochondrial fission/fusion events and through mitophagy to degrade dysfunctional mitochondrial fragments, maintaining the overall mitochondria health of the cell [1,9]. In contrast, our results indicate that the Pink1/parkin pathway may be involved in aspects of cell fate other than protection. Our findings show that co-overexpression of Pink1 or parkin results in an increased severity of the Foxo-induced phenotype, and that a reduction in Pink1 is able to improve on the phenotype. This suggests that there may be a more complex role for the Pink1/parkin pathway under cell stress conditions.

Many transcriptional targets of Foxo have been identified, including molecules involved in metabolism, oxidative stress resistance, cell cycle arrest and apoptosis [27,28]. The Foxo-induced phenotype of developmental defects in the Drosophila eye is likely due to the transcription of pro-apoptotic gene targets. Drosophila studies link phosphorylation of Foxo to neurodegeneration, and have identified the pro-apoptotic Hid gene as one responsible target, where overexpression of Hid causes dramatic eye degeneration [29,30,37]. In contrast, Foxo has also been shown to prevent mitochondrial dysfunction and neurodegeneration, and is proposed to function downstream of Pink1 [26]. In this protective role, Foxo is thought to act through targets including the mitochondrial superoxide dismutase SOD2, a gene involved in stress resistance. With the ability to promote cell survival or apoptosis, changes in Foxo activity may be the mechanism behind the effects of Pink1 and parkin on the Foxo-induced phenotype. Co-overexpression of Pink1 or parkin may affect Foxo activity to increase the transcription of pro-apoptotic targets, thus increasing the severity of the phenotype. Improvement of the Foxoinduced phenotype seen with reductions in Pink1 may also indicate a change in Foxo activity, suggesting that the relationship between Pink1 and Foxo is more complex than Pink1 indirectly activating Foxo downstream. It is also possible that Pink1 and parkin are acting outside of the INR pathway to affect cell survival. The influence of Pink1 and parkin on mitofission/mitofusion events is not fully understood, and some mechanisms involved are closely tied to apoptosis. It may be that under certain conditions these mechanisms are utilized to promote apoptosis instead of cell protection. One example would be the ubiquitination of VDAC by parkin [14]. VDAC is a major component of the permeability transition pore (PTP), and is involved with mitochondrial outer membrane permeabilization (MOMP) through interacttions with pro-apoptotic Bcl-2 proteins [38]. With both the PTP and MOMP implicated as initiators of apoptosis, mitofission events triggered by ubiquitination of VDAC must be controlled so to prevent release of apoptotic factors from the mitochondria. Compounding factors, such as the effects of Hid in the Foxo-induced phenotype, may result in overwhelming instability during increases in Pink1 or parkin expression, making this degree of control impossible. In this instance, Pink1 and parkin may actively participate in the initiation of apoptosis, a novel role for the Pink1/parkin pathway that warrants further investigation.

Expression of the constitutively active version of Foxo (Foxo^AA) with co-overexpression of Pink1, parkin or Pink1^RNAi seems to indicate that the Pink1/parkin effect on the Foxo-induced phenotype is independent of Akt signalling, supporting the idea that Pink1 and parkin may be acting outside of the INR pathway. In contrast, the change in significance when expressing Foxo^AA in the Pink1^B9 mutant background suggests that there is Akt involvement. Interestingly, this may indicate that there is a role for Pink1 in the cell that is independent of its kinase function, and that this additional role is somehow involved in the Akt signalling pathway. In this respect, the apparent rescuing effect of the Foxo-induced phenoltype during decreases in Pink1 expression would be partially due to the decrease in kinase activity, and partially due to the presence of the Pink1 protein. Future studies looking into an additional role for Pink1, apart from its kinase function, may yield new interactions and targets in the Pink1/parkin pathway.

In conclusion, our results show that Pink1 and parkin are able to increase the effects of Foxo in Drosophila, highlighting a possible role for the Pink1/parkin pathway in cell death. In addition, the constitutively active version of Foxo allows us to exclude a general requirement for Akt during increased expression of Pink1 or parkin, however, suggests that there may be an additional role for Pink1 apart from its kinase function. Further studies looking at the effect of Pink1 and/or parkin on Foxo activity, and the role of the Pink1/parkin pathway in mitochondrial fission/fusion events, may uncover underlying mechanisms that mediate a shift towards apoptosis. Moreover, it is likely that the Pink1/parkin pathway is involved in various aspects of cell fate decisions, contrary to the view that Pink1 and parkin act exclusively as protective proteins.

This research was funded by the Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Grant and Parkinson Society Canada Friedman Pilot Project Grant to BES. We thank the family of Jerry Friedman for their generosity. Student support was received from the Department of Graduate Studies at Memorial University to AMT. We thank Dr. Liqiu Men (Bio-molecular Imaging Cluster, Memorial University of Newfoundland) for help with SEM.