Evidence that neither PINK1 nor DJ-1 protect against neuronal loss in mice comes from knockdown and knockout experiments. In 2005, researchers led by Jie Shen at Brigham and Women’s Hospital, Boston, reported that DJ-1 knockouts do have dopaminergic deficits and are hypoactive in open field tests. However, the animals show no obvious signs of neuronal loss, at least up to the age of 12 months (see ARF related news story). Since Parkinson disease is a progressive disorder, one might expect neuronal losses to become evident as these animals age further. Not so, it seems. In the May 29 Molecular Neurodegeneration, Shen and Hiroo Yamaguchi report follow-up studies on these DJ-1 knockout mice. They show that even at the ripe old age of 24-27 months, the animals show no obvious dopaminergic neuron losses in the substantia nigra, or evidence of α-synuclein-containing Lewy bodies, another hallmark of PD. The results suggest that DJ-1 may be essential for dopaminergic transmission but not for the survival of dopaminergic neurons.

A similar pattern seems to be emerging for PINK1. Shen and colleagues turned their expertise to the study of that gene as well, and in the June 11 PNAS online, they report that no dopaminergic neuronal loss is evident in PINK1 knockout animals up to 9 months old. First author Tohru Kitada and colleagues generated mice with a deletion of exons 4-7 of the PINK1 gene, which leads to a truncated and unstable PINK1 mRNA. They tested the animals at 2-3 months of age, and also at 8-9 months for tyrosine hydroxylase-positive neurons in the SN and also for striatal dopamine levels (SN dopaminergic neurons project their axons into the striatum). They found that both were normal. This finding is supported by an earlier study led by Xu Gang Xia at Thomas Jefferson University, Philadelphia, and Zuoshang Xu at University of Massachusetts Medical Center in Worcester. Writing in the March 5 International Journal of Biological Sciences, first author Hongxia Zhou and colleagues reported that conditionally silencing PINK1 in mouse brain by using RNAi has no effect on dopaminergic neurons in SN of mice up to 1 year old. The animals also performed normally on a rotarod test of motor function.

PINK1 and DJ-1 knockout mice now join parkin-deficient animals in failing to recapitulate dopaminergic neuronal losses, presumed to be a key pathologic lesion in PD. Is this purely a reflection of mouse neurobiology, or does it suggest that in some Parkinson cases there is no neuronal loss? “That’s a good question,” wrote Mark Cookson, National Institutes of Health, in an e-mail interview. “We don’t know in the case of PINK1 or DJ-1 as no convincing recessive mutation cases have been autopsied. We do know that all three have loss of 18F-DOPA PET signal in the striatum, which would be consistent with a presynaptic deficit. It is formally possible that there is a dopamine deficit that is not cell loss, but it seems unlikely, and we know that is not the case for parkin,” he wrote. In the case of parkin mutations there is good autopsy data showing neuronal loss in patients (see Hayashi et al., 2000).

Basis for Dopaminergic Loss
While DJ-1 and PINK1 knockouts have no overt neuronal losses, they do have dopaminergic defects that seem consistent with the presynaptic deficits seen in Parkinson patients. From their earlier work, Shen and colleagues concluded that dopaminergic overflow, the balance between dopamine (DA) release and uptake, is reduced in DJ-1-negative striatal slices (see ARF related news story). In addition, they found that long-term depression (LTD) in corticostriatal medium spiny neurons, a dopamine D2 receptor-mediated phenomenon, is compromised in DJ-1-negative slices. Now, Shen and colleagues reveal similar effects of PINK1 loss. In their PNAS paper, Kitada and colleagues report that DA overflow is reduced in PINK1-negative striatum and that this seems to be due to poor dopamine release, since the DA re-uptake blocker nomifensine, which greatly enhances DA signals in control striatum, had very little effect in striatal slices from PINK1-negative animals. Poor DA release also explains why PINK1-negative striatal slices are compromised in LTD and LTP, a D1 and D2 receptor mediated process, and why they can both be rescued by appropriate DA receptor agonists. “Thus the observed deficits in corticostriatal LTP and LTD are most compatible with a specific defect in presynaptic dopaminergic function,” write Kitada and colleagues.

Exactly what kind of defect remains to be seen, but much work points to a mitochondrial role for PINK1 (see ARF related news story and ARF news story) and for DJ-1 (see ARF related news story). “If PINK1 is indeed a mitochondrial kinase, our results would suggest a possible functional link between mitochondria and the regulation of DA exocytosis,” write Kitada and colleagues.

Some clues as to how PINK1 facilitates the mitochondria/DA connection are provided in the June 19 PLoS Biology. Lian Li and colleagues at Emory University School of Medicine, Atlanta, Georgia, report that PINK1 protects cells against oxidative stress by phosphorylating the mitochondrial chaperone TRAP1, also known as heat shock protein 75 (Hsp75). Joint first authors Julia Pridgeon, James Olzmann, and colleagues found that PINK1 binds to and colocalizes with Hsp75 in mitochondria. Furthermore, they showed that PINK1 phosphorylates Hsp75 in vitro and in PC12 cells. The researchers found that the kinase activity was increased when the cells were challenged with hydrogen peroxide, suggesting that PINK1 helps protect against oxidative stress. Interestingly, when the researchers tested PINK1 carrying the PD mutations G309D and L347P, they found that the Hsp75 kinase activity of the protein was abolished. In the case of a third mutation, W437X, the activity was reduced by about 30 percent in cells and by about 50 percent in response to peroxide.

What effect might phosphorylation of Hsp75 have? Mitochondria are infamous for kick-starting the apoptotic cascade, so the authors looked to see if PINK1 may regulate release of mitochondrial cytochrome c, a crucial and early step in apoptosis. Pridgeon and colleagues found that overexpression of PINK1 in PC12 cells suppresses cytochrome c release and prevents apoptosis in response to hydrogen peroxide. Significantly, in cells expressing the PD PINK1 mutants, cytochrome c release was greater than in cells expressing wild-type protein. PINK1’s ability to protect against peroxide-induced apoptosis was also compromised in cells when Hsp75 was knocked down by RNA interference, suggesting that the mitochondrial chaperone mediates the protection afforded by PINK1. Similarly, when the researchers knocked down PINK1 with siRNAs, Hsp75 phosphorylation, both constitutive and peroxide-induced, was reduced. “These results, together with the in vitro phosphorylation data, provide compelling evidence that TRAP1 [aka Hsp75] is a bona fide substrate for PINK1 kinase,” write the authors.

DJ-1 has also been linked to protection against oxidative stress. Most recently, researchers in Germany led by Philipp Kahle at the Ludwig Maximilians University of Munich, reported that PD mutations in the DJ-1 gene destabilize the protein and hamper protection against peroxide. In the March 1 Journal of Biological Chemistry online, first author Karin Gorner and colleagues reported that the L166P mutation, which appears to produce the severest phenotype, destabilized the C-terminal helix-kink-helix motif which is essential for DJ-1 stability. This in turn may scupper signal transduction pathways which help cells respond to and survive oxidative stress. Gorner and colleagues found that wild-type, but not mutant DJ-1, can stimulate signaling through the Akt kinase and block apoptosis.

How do these in vitro cellular findings relate to Parkinson pathophysiology? Shen and colleagues found, for example, that DJ-1-negative mice showed no obvious signs of oxidative damage and no cell loss. But as they point out, mice are kept in well-controlled environments and do not live as long as we do. It is possible that DJ-1 and PINK1 are needed to protect against cumulative oxidative stress in people. Whether Parkinson patients with PINK1 or DJ-1 mutations may respond well to antioxidant therapy remains to be seen. “This is possible, but I would strongly caution against making the assumption that because there is no neurodegeneration in mice, that there is none in humans. Maybe it’s more restricted to the nigra or to a nigral subregion, or maybe there is a lot of dysfunction prior to [neuronal] death in the patients and the rate of real cell loss is slow,” wrote Cookson.—Tom Fagan.

References:
Pridgeon JW, Olzmann JA, Chin L-S, Li L. PINK1 protects against oxidative stress by phosphorylating mitochondrial chaperone TRAP1. PLoS Biology. 2007. July;5:e172. Abstract

Kitada T, Pisani A, Porter DR, Yamaguchi H, Tscherter A, Martella G, Bonsi P, Zhang C, Pothos EN, Shen J. Impaired dopamine release and synaptic plasticity in the striatum of PINK1-deficient mice. PNAS early edition. June 14, 2007. Abstract

Yamaguchi H, Shen J. Absence of dopaminergic neuronal degeneration and oxidative damage in aged DJ-1-deficient mice. Mol. Neurodegen. May 29 online publication. Abstract

Gorner K, Holtorf E, Waak J, Pham T-T, Vogt-Weisenhorn DM, Wurst W, Haass C, Kahle PJ. Structural determinants of the C-terminal helix-kink-helix motif essential for protein stability and survival promoting activity of DJ-1. J. Biol. Chem. 2007, May 4;282:13680-13691. Abstract

Zhou H, Falkenburger BH, Schulz JB, Tieu K, Xu Z, Xia XG. Silencing of the Pink1 gene expression by conditional RNAi does not induce dopaminergic neuron death in mice. Int. J. Biol. Sci. 2007, March 5;3:242-250. Abstract