Physiology versus pathology in Parkinson's disease
Over the last few years, human genetics has provided a series of invaluable entry points to understanding the pathogenesis of Parkinson's disease (PD). Relative to dominantly inherited forms of PD caused by mutations in the proteins α-synuclein and leucine-rich repeat kinase 2 (LRRK2), recessive forms of PD do not seem to bear as close a resemblance to the sporadic disorder. Mutations in parkin produce a distinctive, early onset clinical syndrome with slower progression than the idiopathic disorder (1). Parkin functions as an E3 ubiquitin ligase, suggesting a role in targeting to the proteasome and protein turnover. Considering this biochemical activity, it is surprising that the pathology associated with mutations in parkin shows no Lewy bodies or obvious protein deposits. Indeed, ubiquitination by parkin may control other cellular processes such as endocytosis (2). Nonetheless, mutations in parkin do result in the loss of midbrain dopamine neurons. Mutations in the protein DJ-1 implicated in the response to oxidative stress also produce recessive parkinsonism (3, 4), but there is very little if any pathology available. Although the condition responds to dopamine replacement therapy, we do not know whether it actually involves a loss of dopamine neurons. Similarly, mutations in the PTEN-induced putative kinase 1 (PINK1), the subject of the paper by Kitada et al. in a recent issue of PNAS (5), cause a form of recessive parkinsonism (6), but its pathological basis also remains to be characterized. Nonetheless, a significant number of patients with later-onset parkinsonism that strongly resembles idiopathic PD have heterozygous, or occasionally homozygous, mutations in parkin, DJ-1, and PINK1 (7), supporting their relationship to the sporadic disorder.
Knockouts
The analysis of recessive parkinsonism has the advantage that it focuses on the loss of a normal function rather than on the gain of an abnormal function, which can be difficult to reproduce and understand. A number of groups therefore have disrupted the genes encoding parkin and DJ-1 in mice and other model organisms. Despite its predicted role in protein degradation, the loss of parkin in mice does not result in protein deposition or the degeneration of dopamine neurons (8, 9). On the other hand, parkin knockout mice do show physiological and behavioral deficits attributable to the nigrostriatal dopamine system (8), raising the possibility that these deficits somehow predispose to the degeneration observed in patients or at least result from the same pathogenic process. Similarly, the loss of DJ-1 in mice impairs dopamine release in the striatum, but without any reduction in the number of dopamine neurons, even in aged animals (10, 11). Because we know little about the pathology of DJ-1-associated PD, it remains possible that the knockout mouse may accurately model the human condition. Alternatively, defects in dopamine release may somehow predispose to dopamine cell degeneration in this case as well. Indeed, DJ-1-deficient animals are more sensitive to a number of neurotoxins (12–14), but again we do not know whether this results from a defect in dopamine release or is an independent consequence of DJ-1 loss.
Loss of parkin in mice does not result in protein deposition or the degeneration of dopamine neurons.
Recessive parkinsonism has proven to be particularly amenable to genetic analysis in Drosophila melanogaster. Loss of parkin in Drosophila produces severe mitochondrial abnormalities in the male germ line and in flight muscle, accompanied by degeneration that also may affect some of the dopamine neurons (15). Although parkin knockout mice show no degeneration of dopamine neurons or mitochondrial pathology, they do appear to exhibit a functional deficit in the respiratory chain (16). The identification of mitochondrial abnormalities is particularly significant because the neurotoxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) still provides what is in many ways the best animal model for PD and involves the inhibition of mitochondrial complex I (17). The high energy demands of flight muscle presumably make this tissue particularly vulnerable to a disturbance in oxidative phosphorylation.
The analysis of PINK1 has further supported a role for mitochondria in the pathogenesis of PD. First, PINK1 is a mitochondrially targeted kinase (6). Second, mutations in Drosophila PINK1 produce mitochondrial abnormalities very similar to those observed with mutations in parkin (18, 19). Further, the overexpression of parkin suppresses the effect of mutations in PINK1, indicating that parkin acts downstream of PINK1 in the same pathway.
PINK1
In a recent issue of PNAS, Jie Shen and colleagues (5) presented the first analysis of PINK1 deficiency in mammals. The phenotype strikingly resembles that observed in parkin and DJ-1 knockout mice, rather than in the Drosophila mutants. Similar to the DJ-1 knockout, PINK1-deficient mice show defects in evoked dopamine release measured by using striatal slices (5). However, these defects appear to differ, with inhibition of dopamine reuptake apparently normalizing the defect in release in the DJ-1 knockouts but not in the PINK1 (5, 10, 11). Because the measurements in this preparation reflect reuptake as well as release, the normalization of release by blocking reuptake in DJ-1-deficient mice suggests increased activity of the dopamine transporter, whereas the relative lack of effect in PINK1 knockouts suggests a primary defect in the release mechanism. Indeed, mice lacking PINK1 show less frequent as well as smaller individual release events from adrenal chromaffin cells (5). The authors further demonstrate defects in plasticity at the corticostriatal synapse that can be corrected with dopamine receptor agonists, dopamine supplementation, and amphetamine, supporting a specific, presynaptic defect in dopaminergic neurotransmission. However, the animals again show no loss of dopamine neurons, at least up to 9 months of age.
Based on the observations in parkin, DJ-1, and now PINK1 knockout mice, the authors hypothesize that PD may begin with a defect in dopamine release that progresses to frank degeneration. Obviously, clinical PD involves a deficiency in dopamine, but the question is whether an early, physiological defect in dopamine release actually contributes to the degeneration. In this regard, it is interesting that smoking protects against PD (20) and that nicotine stimulates dopamine release directly at striatal terminals (21, 22). By extension, the firing pattern of midbrain dopamine neurons may influence susceptibility to PD. Alternatively, the defect in dopamine release may simply precede degeneration without invoking a pathogenic role, or it may reflect a phenomenon entirely independent of the degeneration. Indeed, dopamine release by surviving cells has been postulated to increase with disease progression, compensating for the loss of other dopamine neurons but placing greater demands on those that remain (23).
Mitochondria
To understand the role of a defect in dopamine release on degeneration in PD, it will be very important to understand its relationship to mitochondrial dysfunction. In particular, the work in Drosophila shows that parkin and PINK1 have important roles in mitochondria that directly influence degeneration. The oxidation-dependent association of DJ-1 with mitochondria and the sensitization of DJ-1-deficient organisms to oxidative stress suggest that this protein also may function in mitochondria (4). Does the loss of these proteins then impair transmitter release directly due to energy failure? A recent study of synaptic transmission in a Drosophila mutant with reduced numbers of presynaptic mitochondria shows a moderate impairment in transmitter release after prolonged stimulation that apparently reflects reduced recruitment of a reserve synaptic vesicle pool due to the depletion of ATP (24). It is possible that the loss of PINK1 in mice impairs mitochondrial function, but the present study by Kitada et al. (5) reports no morphological change in mitochondria. If the defect in transmitter release is indeed secondary, as a result of mitochondrial dysfunction, which seems very likely given the strictly mitochondrial localization of PINK1, it would predict abnormalities in baseline transmitter release at other synapses. Like parkin and DJ-1, PINK1 is widely expressed. It therefore would be of great interest to explore possible defects in vesicle mobilization (24).
Although the degeneration observed in Drosophila flight muscle in the absence of PINK1 might seem to preclude a causative role for impaired transmitter release in degeneration, it is possible that transmitter release has a particularly important role in the survival of neurons. For example, the cycling of synaptic vesicles may help to eliminate toxins that otherwise would impair mitochondrial function. The vesicular monoamine transporter VMAT2, which loads dopamine into secretory vesicles for exocytotic release, protects against MPTP by sequestering the toxin away from mitochondria. It also may protect against the normal transmitter dopamine, which oxidizes easily and itself exhibits considerable toxicity (25). Defects in the release machinery thus may influence the cytosolic concentration of monoamines (26), contributing to mitochondrial dysfunction and degeneration. However, it will be important to understand how the same cellular process might be affected by DJ-1 acting at the level of reuptake and by PINK1 acting at the level of the release mechanism.
In summary, the work of Kitada et al. (5) provides yet another example of a defect in dopamine release in a recessive form of PD, raising the provocative hypothesis that this defect contributes to, rather than simply reflects, the degenerative process. Future work will need to characterize in more detail the effects of recessive PD genes on mitochondria in mammalian systems and to address the relationship between mitochondrial function and transmitter release. In addition, it will be very important to understand why the loss of genes that cause recessive PD in humans fails to produce degeneration in mice. Considering the increased sensitivity of DJ-1 mice in particular to oxidative stress, the production of disease may require a second hit, such as exposure to a toxin.
Acknowledgments
K.N. is supported by a Larry L. Hillblom Foundation Fellowship. R.H.E. is supported by the National Parkinson Foundation, the Michael J. Fox Foundation, the National Institute on Drug Abuse, the National Institute of Mental Health, and the National Institute of Neurological Disorders and Stroke.
Footnotes
- *To whom correspondence should be addressed. E-mail: robert.edwards{at}ucsf.edu
-
Author contributions: K.N. and R.H.E. wrote the paper.
-
The authors declare no conflict of interest.
-
See companion article on page 11441 in issue 27 of volume 104.
- © 2007 by The National Academy of Sciences of the USA
