A central goal for many of us working on the molecular biology of Parkinson disease (PD) has been to identify a central pathway linking the protein products of genes known to cause this disease when mutated. Two papers recently published in Nature describe a very convincing connection between two proteins unequivocally linked to parkinsonism (Clark et al., 2006; Park et al., 2006). The two studies show that knockouts in the Drosophila PINK1 and parkin homologues both result in an overlapping phenotype of degeneration that prominently includes mitochondrial dysfunction. Furthermore, parkin overexpression can rescue the PINK1 knockout phenotype, but not vice versa, suggesting that these two proteins are in a common pathway with parkin downstream of PINK1. These findings show, in a paradigm similar to that seen in the molecular genetics of AD, that identifying different genetic causes of disease may lead to the elucidation of a common molecular pathogenesis.

These data also add...  Read more

A central goal for many of us working on the molecular biology of Parkinson disease (PD) has been to identify a central pathway linking the protein products of genes known to cause this disease when mutated. Two papers recently published in Nature describe a very convincing connection between two proteins unequivocally linked to parkinsonism (Clark et al., 2006; Park et al., 2006). The two studies show that knockouts in the Drosophila PINK1 and parkin homologues both result in an overlapping phenotype of degeneration that prominently includes mitochondrial dysfunction. Furthermore, parkin overexpression can rescue the PINK1 knockout phenotype, but not vice versa, suggesting that these two proteins are in a common pathway with parkin downstream of PINK1. These findings show, in a paradigm similar to that seen in the molecular genetics of AD, that identifying different genetic causes of disease may lead to the elucidation of a common molecular pathogenesis.

These data also add credence to the idea that mitochondrial dysfunction is central to several forms of PD/parkinsonism. From a genetic perspective, there has been recent support for this, ranging from evidence that polymerase G mutations, which presumably lead to increased mutation of the mitochondrial genome, may be a cause of early onset PD (Davidzon et al., 2006), to the observation that somatic deletion mutations in the mitochondrial genome are more common in neurons of the substantia nigra compared to several other neuronal populations (Bender et al., 2006; Kraytsberg et al., 2006). While the latter data explicitly suggest that these mutations are not the cause of PD, as they are found both in individuals who died of PD and neurological controls, one wonders whether they contribute to the preferential vulnerability of nigrostriatal dopaminergic neurons.

There are two unresolved issues about this work. The first is one of conservation of function across species. It is interesting that the Drosophila PINK1 and parkin models have a much more extreme phenotype than do human patients. In fact, the muscle loss in the fly models seems at least superficially reminiscent of some of the mitochondrial disorders where muscle and optic nerve atrophy are common problems. However, human parkin and PINK1 patients have a more restricted area of damage, namely nigral neurons, without extensive damage to other systems, as far as we know. Why is there such a dramatic species difference? One possibility is that the two sets of genes are not perfect functional homologues. Both kinases and E3 ligases are large gene families, and it may be a simplification to infer that they have the same targets in mammals and invertebrates (see previous ARF comment). Though it may be much simpler and very useful to reconstruct pathways in an organism with lesser redundancy, it does mean that we have to apply some caution in using these models to predict human pathways when we don’t understand the intervening steps. One interesting experiment would be to see if human parkin rescued the fly phenotypes; if so, this would imply that the two proteins are strict functional orthologues. Showing that E3 ligase-deficient parkin mutations fail to do so would also be important in elucidating how this enzyme affects mitochondria, which is currently hard to reconcile mechanistically—although there are suggestions that parkin has a ubiquitin-independent function important for maintaining mitochondrial DNA integrity (see below).

A second caveat about the PINK1-parkin relationship is that at face value there are two lipid membranes between the two proteins because PINK1 is in the mitochondrial matrix, whilst parkin is predominantly cytosolic. This does not diminish the impact of the current observations, but it needs to be resolved. One possible explanation is that PINK1 is not strictly mitochondrial, or parkin is not strictly cytosolic. There is experimental support for both possibilities. In transient transfections, with all the attendant caveats about overexpression and modification of proteins, some mature PINK1 ends up in the cytosol (Beilina et al., 2004). Parkin has been associated with the outer mitochondrial membrane (Darios et al., 2003) and more recently reported to be within mitochondria in proliferating cells, although not in quiescent cells such as neurons (Kuroda et al., 2006). Therefore, it is possible that the two proteins communicate by shuttling between mitochondria and cytoplasm. An alternative explanation is that a shared interactor or substrate of the two enzymes mediates the communication between the mitochondrion and the cytosol. Detailed studies of where the two proteins, and their substrates, are found in the brain of Drosophila and vertebrates are therefore warranted.

View all comments by Mark Cookson
View all comments by Andrew Singleton

The parkin-PINK Connection: Bridging Together Proteasome and Mitochondria
The identification over the past few years of several genes causative of autosomal dominant and recessive forms of parkinson disease has represented a true revolution in PD research. A huge body of research has been performed aimed at understanding the physiological role of proteins encoded by PD genes and the mechanisms by which mutated proteins lead to neurodegeneration. These findings have elucidated two major pathways related to neuronal cell death, namely mitochondrial dysfunction and impairment of the ubiquitin-proteasome system (UPS). While PINK1 and DJ-1 are stress-related proteins exerting a protective role against mitochondrial dysfunction and oxidative stress, other proteins, such as parkin and UCH-L1, play crucial roles for the integrity of the UPS. In this light, it looked as if there were two well distinct mechanisms, possibly leading to a common final pathway of neuronal cell death. Studies on α-synuclein, another PD-related protein, have shown that its pathogenic forms can induce...  Read more

The parkin-PINK Connection: Bridging Together Proteasome and Mitochondria
The identification over the past few years of several genes causative of autosomal dominant and recessive forms of parkinson disease has represented a true revolution in PD research. A huge body of research has been performed aimed at understanding the physiological role of proteins encoded by PD genes and the mechanisms by which mutated proteins lead to neurodegeneration. These findings have elucidated two major pathways related to neuronal cell death, namely mitochondrial dysfunction and impairment of the ubiquitin-proteasome system (UPS). While PINK1 and DJ-1 are stress-related proteins exerting a protective role against mitochondrial dysfunction and oxidative stress, other proteins, such as parkin and UCH-L1, play crucial roles for the integrity of the UPS. In this light, it looked as if there were two well distinct mechanisms, possibly leading to a common final pathway of neuronal cell death. Studies on α-synuclein, another PD-related protein, have shown that its pathogenic forms can induce both mitochondrial and UPS impairment, thus building the first link between these two pathways. Additional evidence in support of a mitochondrial-UPS connection came from parkin animal models (such as mouse and Drosophila), surprisingly showing signs of mitochondrial damage and increased oxidative stress.

Now, three papers, two published in Nature (Clark et al., 2006; Park et al., 2006) and one in PNAS (Lu et al., 2006), independently reported that PINK1 knockout Drosophila models develop a mitochondrial phenotype remarkably similar to parkin knockouts, and—most interestingly—that parkin overexpression can rescue this phenotype. These findings unequivocally put PINK1 and parkin within the same pathway, with parkin acting downstream of PINK1, and definitely establish a tight connection between mitochondria and the UPS.

What could be the function of parkin within mitochondria? Is this related to its ubiquitin-ligase activity, or does it represent a distinct, still unknown function of this protein? Previous studies had already provided hints of a possible mitochondrial role of parkin. For instance, Moore et al. showed an interaction between parkin and DJ-1 under oxidative stress conditions, while Darios and coworkers demonstrated a protective role of parkin against mitochondrial swelling and cytochrome c release in a model of mitochondrial-related cell death induced by the toxin ceramide. However, how parkin can exert its protective function within the mitochondria is still obscure. On the other hand, while it is now established that PINK1 is imported and processed within the mitochondria, it is still unclear whether a proportion of the mature protein could be exported back to the cytosol, as suggested by Beilina and coworkers. Thus, it appears more and more clear that, instead of two (or more) distinct pathways, we are facing a complex scenario where PD-related proteins can variably interact at different cellular levels. To understand this scenario and all its players is now the most enthralling challenge of PD research.

References:
Clark IE, Dodson MW, Jiang C, Cao JH, Huh JR, Seol JH, Yoo SJ, Hay BA, Guo M. Drosophila pink1 is required for mitochondrial function and interacts genetically with parkin. Nature. 2006 May 3. Abstract

Moore DJ, Zhang L, Troncoso J, Lee MK, Hattori N, Mizuno Y, Dawson TM, Dawson VL. Association of DJ-1 and parkin mediated by pathogenic DJ-1 mutations and oxidative stress. Hum Mol Genet. 2005 Jan 1;14(1):71-84. Abstract

Darios F, Corti O, Lücking CB, Hampe C, Muriel MP, Abbas N, Gu WJ, Hirsch EC, Rooney T, Ruberg M, Brice A. parkin prevents mitochondrial swelling and cytochrome c release in mitochondria-dependent cell death. Hum Mol Genet. 2003 Mar 1;12(5):517-26. Abstract

Beilina A, van der Brug M, Ahmad R, Kesavapany S, Miller DW, Petsko GA, Cookson MR. Mutations in PTEN-induced putative kinase 1 associated with recessive parkinsonism have differential effects on protein stability. Proc Natl Acad Sci U S A. 2005 Apr 19;102(16):5703-8. Abstract

View all comments by Enza Maria Valente