PTEN-Induced Kinase 1 (PINK1)
The discovery of PTEN-induced kinase 1 (PINK1) was first described in 2001 by Valente et al. as a novel locus for autosomal recessive Parkinson’s Disease (PD). Termed PARK6, the new locus was identified in a genome-wide homozygosity screen performed in a Sicilian family with multiple PD-affected members (Valente et al. 2001, 2004). Localized to the mitochondria, PINK1 was the first nuclear-encoded mitochondrial protein to be implicated in PD pathogenesis and strongly suggests the role of mitochondrial pathomechanisms in PD (Valente et al. 2004). The precise mechanism by which PINK1 contributes to mitochondrial dysfunction, however, remains unclear as there remains a significant gap in knowledge in our understanding of the role of PINK1 in human disease.
First identified as a serine/threonine kinase with reported homology to the Ca2+/calmodulin family, PINK1 does not fall within any of the previously identified kinase families as determined by sequence similarity of their kinase domains (Manning et al. 2002), indicating the PINK1 substrates may be unique. Studies examining the cellular role of PINK1 have shown that the serine-threonine kinase protects cells from oxidative stress and, thus, is proposed to serve an important role in protecting the cell from mitochondrial crisis (Jendrach et al. 2009). Recent publications have suggested several intriguing hypothesis addressing the cellular functions of PINK1. This review will highlight the role of PINK1 in mitochondrial fission and fusion, in clearance of damaged mitochondria through mitophagy, and in mitochondrial trafficking. Additionally, insight gained from studying PINK1 in model organism or through the identification of PINK1 specific substrates will also be discussed.
Localization and Structure
Located on chromosome 1p35–p36, PINK1 is composed of three main domains: an N-terminal mitochondrial targeting sequence (MTS), a small transmembrane domain (TM), and a highly conserved serine/threonine kinase domain that has been reported to exhibit autophosphorylation activity in vitro (Zhou et al. 2008; Thomas and Cookson 2009). Analysis of PINK1’s kinase domain led to the recognition of highly conserved subdomains involved in either anchoring and orienting the ATP molecule (subdomains I–IV) or binding of the peptide substrate (subdomains VIa–XI) (Silvestri et al. 2005). PINK1 is made up of eight exons spanning 1.8 kB, corresponding to an ubiquitously expressed 581 amino acid protein. The highest expression levels are observed in the heart, skeletal muscle, and testis (Silvestri et al. 2005), all mitochondrial dense tissue. In the brain, PINK1 expression is primarily neuronal and localized to the hippocampus, Substantia nigra, and the cerebellar Purkinje cells (Thomas and Cookson 2009). The 581 amino acids (63 kDa protein) of PINK1 are further processed upon mitochondrial import to generate two mature isoforms, 54 and 45 kDa, respectively (Deas et al. 2009).
It is well established that PINK1 localizes to the mitochondria as the N-terminal mitochondrial target sequence is sufficient for mitochondrial import of PINK1 (Valente et al. 2001; Silvestri et al. 2005). Specific subcellular localization of PINK1, however, remains somewhat controversial as it has been reported to be located on the outer mitochondrial membrane (OMM), the mitochondrial intermembrane space (IMS), the inner mitochondrial membrane (IMM), and even the cytoplasm (Thomas and Cookson 2009). Recent work on the topology of PINK1 has concluded that PINK1 localization depends on the transmembrane domain, which embeds itself into the outer mitochondrial membrane leaving the kinase domain exposed to the cytosol (Zhou et al. 2008). This would imply that PINK1 specific substrates are located in the cytosol or localized to the outer mitochondrial membrane. Interestingly, two of the most widely reported interactors or substrates of PINK1, TRAP1, and Omi/HtrA2 reside within the mitochondria. Therefore, PINK1 interactors (TRAP1 and Omi/HtrA2) and PINK1 localization need to be rectified through additional mechanistic studies.
Substrates and Interactors of PINK1
Identified as a serine/threonine kinase, PINK1 has been shown to exhibit autophosphorylation activity in vitro. However, despite robust phosphorylation activity, the in vivo substrates remain unknown making it difficult to ascertain the biochemical function and physiological substrates of PINK1 (Geisler et al. 2010). One of the first specific PINK1 substrates to be identified was the mitochondrial chaperone tumor necrosis factor type 1 receptor associated protein 1 (TRAP1). Through a combination of both in vitro and in vivo assays, Pridgeon et al. showed that the phosphorylation of TRAP1 by PINK1 played a critical role in protecting cells from oxidative-stress-induced cell death (Pridgeon et al. 2007). The protective action of PINK1 through suppression of cytochrome c release was impaired by PD-linked mutations including G309D, L347D, and W437X, which resulted in a reduced capability for PINK1 to phosphorylate TRAP1. Reported to be a direct substrate for PINK1, this work by Pridgeon et al. provided some of the first insight into the mechanism by which loss of PINK1 leads to neurodegeneration.
Further confirmation that the kinase activity of PINK1 is required for its in vivo function came with the identification of the PD relevant substrate, Parkin. Parkin, also known as PARK2, is one of the most commonly affected PD genes and encodes for an E3 ubiquitin ligase that is also associated with autosomal recessive forms of PD. Shown to mediate both classical and nonclassical ubiquitin linkages, Parkin has been implicated in both protein degradation and inclusion-body formation (Kim et al. 2008). The first indication that Parkin and PINK1 may be linked in a common pathway came from initial work in Drosophila, where it was determined that Parkin and PINK1 function in a linear pathway with PINK1 upstream of Parkin (described below). However, the mechanism by which the two proteins interacted together remained unknown until Kim et al. elegantly performed studies in both mammalian and Drosophila systems, that they clearly demonstrated that PINK1’s kinase activity regulates the translocation of Parkin to the mitochondria (Kim et al. 2008). Given the role of Parkin in protein degradation and PINK1’s proposed role in mitochondrial dynamics, the identification of Parkin as a PINK1 specific substrate has important implications into PD-related mechanisms in which PINK1 and Parkin work together to provide a mitochondrial quality control system. This was further confirmed by Geisler et al., demonstrating that PINK1 kinase activity and its mitochondrial localization sequence are required for the recruitment of Parkin to depolarized mitochondria. Parkin localization results in selective degradation of the damaged mitochondria through selective autophagy of the mitochondria, a process termed mitophagy (Geisler et al. 2010).
Supporting a role of PINK1 in protecting cells from mitochondrial dysfunction, Plun-Favreau et al. reported that the serine protease Omi/HtrA2 is an interactor of PINK1 (Plun-Favreau et al. 2007). Loss of function of Omi/HtrA2 has been reported in the literature to result in a neurodegenerative disorder with a similar phenotype to PD. Classically, Omi/HtrA2 is thought to be released from mitochondria during apoptosis upon permeabilization of the outer mitochondrial membrane. Once in the cytosol, the serine protease binds to inhibitor of apoptosis (IAP) proteins and thus contributes to induction of apoptosis (Plun-Favreau et al. 2007). While proposed to be a proapoptotic protein, the parkinsonian neurodegenerative phenotype observed in Omi/HtrA2 knockout mice has lead to the hypothesis that the serine protease may also serve to protect mitochondrial homeostasis. Further work indicated that PINK1 and Omi/HtrA2 are both components of the same regulatory pathway involved in the maintenance of mitochondria homeostasis. They provided evidence that p38-mediated phosphorylation of Omi/HtrA2 occurs in a PINK1-dependent manner (Plun-Favreau et al. 2007). The authors concluded that PINK1-dependent phosphorylation of HtrA2 may modulate the protease’s proteolytic activity and thereby serve to provide a mechanism by which PINK1 phosphorylation of HtrA2 results in increased resistance of cells to mitochondrial stress. It remains unclear, however, whether Omi/HtrA2 is a direct substrate of PINK1 and whether differences in cell viability due to PINK1 inactivation are indirectly affecting other kinases that are responsible for direct phosphorylation of Omi/HtrA2 activity.
Investigating PINK1 Function Using Model Organisms
Several model organisms have been used to help researchers elucidate the roles of PINK1. Specifically, given the role of PINK1 in PD, many of the studies in model organisms have been focused toward determining the mechanism by which loss of PINK1 contributes to the pathogenesis of PD. Using C. elegans as a model organism, loss of function PINK1 mutations result in mutant worms demonstrating reduced mitochondrial cristae lengths (Samann et al. 2009). Additionally, worms harboring the mutant form of PINK1 also display an increased sensitivity to the reactive oxygen species catalyst, paraquat, which could be rescued upon expression of transgenic PINK1. The reported mitochondrial phenotype coinciding with an observable sensitivity to oxidative stress supports findings in other model organisms and the hypothesis that PINK1 functions to preserve mitochondrial homeostasis in the cell (Samann et al. 2009).
A second model organism that has been used extensively to gain insight into the function of PINK1 in the cell has been Drosophila. Interestingly, these fruit fly models for PD resemble many of the characteristics of the human form of PD including the loss of dopaminergic (DA) neurons along with the presentation of various locomotive defects (Park et al. 2009). In 2006, three papers came out reporting that loss of PINK1, either by loss of function mutations or through the removal of the Drosophila PINK1 homologue results in male sterility, indirect flight muscle, and DA neuronal degeneration accompanied by locomotive defects. Additionally, loss of PINK1 resulted in defects in mitochondrial morphology (increased mitochondrial cristae fragmentation and mitochondrial swelling observed in tissues with high energy demands) and demonstrated an increased sensitivity to multiple stressors including oxidative stress (Park et al. 2009).
Interestingly, Drosophila PINK1 fly mutants phenocopy almost all of the observed phenotypes of the Parkin knockout, including indirect flight muscle/DA neuronal degeneration, impaired flight, and slow climbing ability. Subsequent genetic studies probing for the mechanism behind the phenotype similarity observed with Parkin and PINK1 mutants resulted in the discovery of an epistatic relationship between the two genes. Further studies revealed that PINK1 and Parkin are linked in a linear pathway involved in protecting the integrity and function of the mitochondria with Parkin acting downstream of PINK1 (Park et al. 2009). Together, these two models strongly implicate the role of mitochondrial dysfunction in PINK1 or Parkin-related PD pathogenesis.
To further elucidate the physiological role of PINK1 in mammalian cells and to gain insight into the disease mechanism of PINK1’s involvement in PD pathogenesis, several labs have worked to understand novel phenotypes of PINK1-deficient mice. PINK1−/− mice have functional defects in the dopaminergic system, including a decrease in dopamine release in striatal slices. This functional defect could be rescued by treatment with dopamine receptor agonists or other agents that increase dopamine release (Kitada et al. 2007). These results suggest a critical role for PINK1 in providing a functional link between mitochondria and regulation of dopamine release and thus may provide insight into the pathogenesis of PINK1 in playing a role in causing PD-associated dopaminergic dysfunction. Unlike studies in both Drosophila and C. elegans models, loss of murine PINK1 resulted in no gross changes in the structure of mitochondria; however, the authors did report that the overall size of mitochondria increased in the PINK1-deficient mice compared to control (Gautier et al. 2008). While no gross changes in mitochondrial structure were observed, significant defects in mitochondrial respiration were detected in complexes I, II, III, and IV. At 3–4 months of age, this respiration defect was specific for the striatum; however, mitochondrial respiration activity was decreased in other areas, including the cerebral cortex, at 2 years of age compared to control. Similar mitochondrial defects in other regions outside of the striatum could also be induced following induction of ROS formation through the treatment of H2O2 or mild heat shock, suggesting that ROS accumulation as a consequence of aging may exasperate the mitochondrial dysfunction (Gautier et al. 2008).
Mitochondrial Function of PINK1
Mitochondria are dynamic organelles that require active fission and fusion to maintain a distinct morphology, not only to maintain mitochondrial homeostasis but also to maintain changing cellular energy demands (Thomas and Cookson 2009). Work performed in Drosophila and mouse models have found that loss of PINK1 results in enlarged/swollen mitochondria, suggesting a potential role for PINK1 in regulating mitochondrial morphology. The distinct morphology of the mitochondria is tightly regulated through two opposing process: mitochondrial fission and fusion. These two processes have also been shown to play important roles in maintaining mitochondrial maintenance and function, along with regulating the stabilization of the mitochondrial genome in response to the changing cellular environment (Okamoto and Shaw 2005). This regulation will be described briefly below, but for a more detailed review behind the mechanism of mitochondrial fission and fusion, readers are encouraged to refer to Chan et al. for a comprehensive overview of the mechanism of fission and fusion (Chan 2006).
Capable of undergoing several fission/fusion events, mitochondria are frequently changing shape in response to their surrounding cellular energy demands. Conserved dynamin-related GTPases are responsible for maintaining the balance between mitochondrial fission and fusion. Two highly conserved proteins, the outer mitochondrial membrane GTPases Mitofusins (MFN 1 or MFN 2) and the inner membrane GTPase OPA1, are responsible for controlling mitochondrial fission. Mitofusins are large GTPases embedded in the outer mitochondrial membrane (OMM) that initiate interaction between the two mitochondria. Fusion of the inner mitochondrial membrane (IMM) is predicted to be regulated by OPA1, a dynamin-related GTPase that localizes to the inner mitochondrial space (IMS) and associates with the IMM (Chan 2006).
Mitochondrial fission is controlled by two proteins: Drp1 and hFIS1. The small protein, hFIS1, uniformly localizes to the OMM and is responsible for recruiting Drp1 to sites of intended mitochondrial fission. Drp1 is a large dynamin-related GTPases found in the cytosol that localizes to the mitochondria to initiate mitochondrial fission. Current theories suggest that upon recruitment to the mitochondria, Drp1 forms a ring around the mitochondria and acts as a mechanochemical enzyme that uses GTP hydrolysis to drive mitochondrial constriction and fission. Interestingly, Drp1 has been reported to be recruited to the mitochondria in the absence of hFIS1, and thus hFIS1 is not required for Drp1 localization to the mitochondria (Chan 2006).
While it has been proposed that PINK1’s function in the cell is to regulate mitochondrial dynamics, the exact mechanism of PINK1’s involvement with mitochondrial fission and fusion remains to be elucidated. Knockdown of PINK1 in Drosophila models have demonstrated that loss of PINK1 results in elongated mitochondria, while studies in C. elegans resulted in heavily fragmented mitochondria. Studies in mammalian cell models have been even more controversial where loss of PINK1 has been reported to result in either excess fusion, excess fission, or result in no effect on mitochondrial morphology, depending on the cell model and the method of knockdown (Thomas and Cookson 2009; Jendrach et al. 2009). Mitochondrial fission and fusion are important processes which allow the mitochondria to control the structure and distribution of the mitochondria network. Given the proposed role of mitochondrial defects in the pathology of PD, a better understanding of the involvement of PINK1 in mitochondrial biology and homeostasis may provide important insight into the specific pathogenic pathways that leads to mitochondrial dysfunction in PD.
PINK1 and Parkin have previously been reported to interact genetically and function within the same linear molecular pathway (Park et al. 2009). The mechanistic details of how a serine/threonine kinase and an E3 ligase function to protect mitochondrial dynamics has remain largely unknown, until recently, with the discovery that PINK1 may play a direct role in the elimination of mitochondria through interaction with Parkin (Geisler et al. 2010). In an elegant set of experiments, Geisler et al. demonstrated that PINK1’s kinase activity and its mitochondrial localization sequences are both required to translocate Parkin to depolarized mitochondria. From there, Parkin subsequently mediates the formation of two distinct polyubiquitin chains, and degradation of the damaged organelle is further promoted through the addition of the autophagic adaptor p62 (SQSTM1), whose presence is essential for the clearance of mitochondria (Geisler et al. 2010).
Following mitochondria depolarization, indicative of mitochondrial damage, PINK1 interaction with Parkin is significantly enhanced. Furthermore, modulations of PINK1 expression levels either through siRNA knockdown or expression of kinase deficient PINK1 results in a reduction of Parkin translocation to depolarized mitochondria and thus the subsequent degradation of the damaged mitochondria. Therefore, PINK1 appears to serve as an important sensor for mitochondria damage upstream of Parkin (Geisler et al. 2010). Many questions still remain to further elucidate the recently identified novel role for PINK1’s involvement in mitophagy. In particular, it is still unclear how the kinase activity of PINK1 is regulated following mitochondrial damage and whether PINK1 can mediate phosphorylation of Parkin to stimulate its E3 ligase activity. Given the role of both Parkin and PINK1 in autosomal recessive PD, this model provides an attractive hypothesis to provide a common pathomechanism in which loss of either protein leads to the onset of hereditary PD.
Mitochondrial trafficking is a critical process to ensure proper distribution of the mitochondria within the cell. Mitochondrial transport occurs in both anterograde and retrograde directions within the cell and is controlled by a combination of cytoskeleton and mitochondrial proteins, including the atypical GTPase Miro and the adapter protein Milton (Weihofen et al. 2009). Miro and Milton complex on the outer mitochondrial membrane, facing the cytosol where they link to the heavy chain of a kinesin to drive anterograde transport of the mitochondria along microtubules within the cell. Work published by Weihofen et al. showed that PINK1 forms a multiprotein complex through interactions with the two cytoplasmic proteins Miro and Milton, and this suggests a novel role for PINK1 in mitochondrial trafficking (Weihofen et al. 2009). This model depends on the topology report of Zhou et al., as it requires the kinase to be anchored to the OMM with its kinase domain facing the cytoplasm to ensure interactions with Miro and Milton. Mitochondrial transport is also tightly regulated by mitochondrial dynamics, including fission and fusion (Okamoto and Shaw 2005). PINK1, therefore, in addition to direct interactions with Miro and Milton, may also regulate mitochondrial transport through its roles in mitochondrial fission and fusion. The proposed role of PINK1 in regulating mitochondrial transport provides yet another means by which PINK1 can protect the cell from oxidative stress or other forms of mitochondrial crisis.
PINK1 and Parkinson’s Disease
Parkinson’s disease is a progressive neurodegenerative disorder in which dysregulation of mitochondrial structure and function has come to the forefront as one of the central factors in the pathogenesis of PD (Chu 2010). Characterized by a loss in dopaminergic neurons in the Substantia nigra, as well as the presence of ubiquitin-positive Lewy neuritis and Lewy bodies, PD is progressive disease with no current therapy to slow progression of the disease. Understanding the molecular pathways that result in PD pathology remains an important goal to facilitate development of more effective therapeutic strategies. While the majority of PD cases appear to be sporadic, there is a significant amount of PD cases that are caused by specific genetic defects linked to familial forms of PD, which have provided important insight into the pathology underlying PD. Loss of function of PINK1 has been implicated in autosomal recessive PD as most of the identified pathogenic mutations occur within the kinase domain and result in decreased kinase activity.
Neurons are highly specialized cells that have very specific energy demands, and thus proper mitochondrial function and distribution play a critical role in maintaining neuronal function. As previously discussed, mitochondrial fission and fusion play a critical role in both mitochondrial homeostasis and trafficking (Okamoto and Shaw 2005). Therefore, modulation of the fission and fusion process due to PINK1 mutants found in autosomal recessive PD may result in an inability to properly traffic mitochondria along the length of the neuron to the synapse. This would result in an energy deficit throughout the neuron and could culminate in the degeneration of the DA neurons. Additionally, given the interaction between PINK1 and characterized mitochondrial trafficking proteins, Miro and Milton, loss of PINK1 may have a more direct means to modulate mitochondrial trafficking (Weihofen et al. 2009).
PINK1 also appears to play a critical role in maintaining mitochondria function as it may serve as an important role in ensuring that the mitochondria population remains maximally functional (Geisler et al. 2010). Given recent insight into the potential role of PINK1 and Parkin in mitophagy, as described in the section above, PINK1 may also be directly involved in the clearance of damaged mitochondria. Loss of PINK1 may therefore result in an inability to segregate and remove damaged mitochondria, thus leading to accumulation of mitochondrial dysfunction within the neuron (Geisler et al. 2010). The inability to maintain energetic demands of the neuron would therefore result in neuronal degradation and may provide a molecular pathway to explain the mitochondrial dysfunction observed as one of the classic pathologies of PD (Thomas and Cookson 2009; Chu 2010). It will be interesting to see if the molecular pathways by which Parkin mutations lead to PD are the same as PINK1 given the linear epistatic relationship between the two proteins.
Several recent publications have implicated PINK1 in playing a role in regulating mitochondrial morphology, mitochondrial homeostasis, and mitochondrial trafficking. Insights from the Parkinson’s field have also demonstrated that PINK1 appears to function to protect the cell from stresses, such as oxidative stress, that would otherwise compromise the mitochondria and lead to cellular distress or death. Given the role of the mitochondria in cellular metabolism, it could be proposed that altered mitochondrial bioenergetics through the disruption of PINK1 function may play an important role in PD pathology. However, there remains much to be done to further understand the role of PINK1 in PD. More mechanistic studies focusing on the localization and processing of PINK1 remain to be elucidated. Most likely, the reason behind the controversy related to the localization of PINK1 is due in some part to the inability to detect PINK1 endogenously with commercially available antibodies. Therefore, researchers are often forced to tag the protein or overexpress it to such a level that is no longer biologically relevant. Similarly, PINK1 interactors and substrates remain largely undetermined, probably somewhat due for the same reason as it is difficult to study protein interactions and kinase activity at the endogenous single-cell level. Likewise, the exact mechanism by which PINK1 regulates mitochondrial dynamics is beginning to become clear; however, more work is required to further identify the specific substrates and interactors of this serine/threonine kinase. Clearly, PINK1 is involved in several pathways aimed at maintain mitochondrial homeostasis. A better understanding of how PINK1 is modified or processed, along with an awareness of how these changes alter the function of the kinase both under normal and stressed conditions, will provide important clues into the mechanism by which PINK1 protects cells from mitochondrial crisis.