In the aftermath of a recent spate of gene discoveries for Parkinson disease, some of the prizes scientists were chasing at the 35th Annual Conference of the Society for Neuroscience, held last month in Washington, D.C., were the identification of parkin substrates and the characterization of the latest gene linked to PD.
The latter, called leucine-rich repeat kinase 2 (LRRK2), is now under heavy scrutiny in several laboratories. For example, Veerle Baekelandt, at KU Leuven, Belgium, reported that LRRK2 mRNA is widely expressed throughout the cerebral cortex of the mouse, as well as in the amygdala and the striatum, an area of the brain that is particularly relevant to Parkinson disease. Weiping Gai and colleagues from Flinders University, Adelaide, Australia, and elsewhere, backed up this finding. These researchers reported immunohistochemical localization of LRRK2 in tissue samples taken from PD, Alzheimer disease (AD), dementia with Lewy bodies, and multiple system atrophy patients. Gai and colleagues found that LRRK2 and α-synuclein colocalized in intraneuronal inclusions, in LB disorders. They also reported that LRRK2 occurs in neurofibrillary tangles in AD samples, though not in plaques. On the other hand, Andrew Grover and colleagues at the Sun Health Research Institute in Sun City, Arizona, reported that LRRK2 antibodies recognize human microglia and that LRRK2-positive microglia are associated with pigmented neurons of the substantia nigra, those same neurons that degenerate in PD.
As for parkin substrates, it appears that we have to hold our breath a bit longer before we find out exactly how this E3 ubiquitin ligase contributes to PD progression. Hints abound, but breakthroughs are not in sight yet. However, Wanli Smith, Peter Pei, and colleagues, working at Christopher Ross’s lab at Johns Hopkins University, Baltimore, Maryland, in collaboration with Ted Dawson’s group, reported a connection between parkin and LRRK2. First, they showed that LRRK2 itself appeared to be predominantly cytoplasmic. When they used antibodies to detect the protein, they found punctate labeling indicative of protein aggregation in a small number of cells (HEK-293 SH-SY5Y, or primary neurons) transfected with the LRRK2 gene. This aggregation increased when parkin or synphilin was coexpressed with LRRK2, Smith reported. He also revealed that in immunoprecipitation studies, LRRK2 and parkin or synphilin (but not DJ-1 or α-synuclein) bound tightly. The researchers also found that expression of parkin led to increased ubiquitination of the aggregates, though not of LRRK2. Mutant LRRK2 caused neurodegeneration in SH-SY5Y cells and in primary neurons. All told, the findings suggest that LRRK2 interacts with parkin and synphilin, suggesting a possible pathogenic pathway, and that mutant LRRK2 is toxic via a gain-of-function mechanism. This work will appear this month in PNAS, Smith et al., 200, in press.
Work from Ted Dawson's and Valina Dawson’s own labs at Johns Hopkins suggest that the aminoacyl-tRNA synthetase cofactor, p38/JTV-1, may be a bona fide parkin substrate involved. Hanseok Ko reported that of all the proposed substrates, only p38 levels were elevated in the brains of both young and old parkin-null mice (see also Ko et al., 2005). Ko reported that p38 is also elevated in brains of patients with familial and idiopathic PD, and that while normal parkin can prevent p38-induced apoptosis, the PD-associated R42P parkin mutant cannot. His poster also revealed that overexpression of p38 in mouse brain leads to loss of dopaminergic neurons in the substantia nigra.
James Palacino, working with Michael Schlossmacher at Brigham and Women’s Hospital, Boston, reported a potential mechanism whereby p38 may mediate parkin toxicity. Palacino found that p38 levels are about threefold higher in brain samples taken from people with autosomal recessive juvenile parkinsonism (AR-JP) than in control samples, though with a single anti-mouse p38 antibody he failed to detect any changes in expression of the protein in brain samples from parkin-null mice. What would be the consequences of elevated p38/JTV1? Because p38 has no ubiquitin E3 ligase activity of its own, Palacino wondered if it cooperates with parkin to regulate the degradation of far-upstream signal element (FUSE) binding proteins, or FBPs. These FUSE binding proteins, of which there are three family members, regulate the expression of c-myc and other genes, and their degradation is known to depend on p38/JTV1.
Palacino looked to see if parkin deficiency affects FBP levels. He reported that both FBP1 and 2, also known as KSRP (KH-type splicing-regulatory protein) are elevated in brain samples taken from both AR-JP patients and parkin-null mice. Next, he checked to see how parkin deficiency might affect transcripts that may be degraded by FBPs. First, he measured steady-state levels of mRNA coding for a slate of about 11 proteins that he previously found were differentially expressed in parkin-deficient mice (see ARF related news story). He reported that none of these levels were altered in parkin-null animals. Next, he turned his attention to the protein tyrosine kinase Src because Src mRNA is known to be degraded in an FBP2-dependent manner. Palacino found that when he transfected parkin into M17 cells, which on their own have little or no endogenous parkin, both FBP1 and FBP2 decreased and that this was accompanied by a significant increase in Src mRNA levels. Also, he reported that Src mRNA levels were over fivefold lower in the brains of parkin-null mice. The data suggest that parkin leads to a reduction in the amount of FBPs, which in turn leads to an increase in the steady-state levels of Src mRNA. Src kinase activity was also much lower in parkin-null animals, he reported.
How these findings relate to PD pathology is unclear, presently. But Palacino pointed out that Src regulates mitochondrial respiration. This process, of course, has been linked time and again to PD through a variety of mechanisms, such as apoptosis, generation of reactive oxygen species, and the toxicity of PD-inducing chemicals such as MPTP. In fact, Src regulates the phosphorylation of a subunit of cytochrome c oxidase, the last redox protein in the mitochondrial respiratory chain, causing an increase in electron transfer. Palacino reported that cytochrome c oxidase activity is low in parkin-deficient M17 cells, but that it doubles when he expresses small amounts of exogenous parkin in the cells. Taken together, the evidence forges a molecular link among parkin, p38/JTV1, and mitochondrial fitness.
Another parkin substrate that has drawn some attention is Pael-R, or parkin-associated endothelin receptor-like receptor. Pael-R was initially identified by Ryosuke Takahashi, RIKEN, in Saitama, Japan, and colleagues as a parkin binding partner. It can induce cell death when overexpression leads to accumulation of misfolded protein. This G-protein-coupled receptor also accumulates in the brains of people with autosomal recessive juvenile onset AR-JP (see ARF related news story). Now, Takahashi and colleagues have developed a Pael-R-based transgenic mouse model that recapitulates one of the major features of PD, namely, the loss of catecholaminergic neurons.
Takahashi and colleagues crossed transgenic mice expressing Pael-R under the regulation of the PDGF promoter with parkin-null animals. Takahashi reported that while parkin-null mice appeared normal, animals that also had one or both copies of the Pael-R transgene exhibited age-related changes similar to those seen in PD patients. At 6 months, for example, the numbers of tyrosine hydroxylase (TH)-positive neurons are similar in all animals, but by 12 months, there is a dose-dependent loss of TH neurons in both the substantia nigra and locus coeruleus of the Pael-R Tg/parkin-/- animals. At 24 months, this loss is more pronounced, and again is greater in the homozygotes than in the hemizygotes. Using immunohistochemistry, the scientists also found that the number of dopaminergic terminals is reduced in the striatum, a major projection site for dopaminergic neurons. Both TH and dopamine transporter (DAT) reactivity tissue is lost in striata from two-year-old transgenics, though the number of striatal neurons does not change as visualized by staining with the neuronal marker NeuN. In addition, Takahashi reported a huge increase (almost fivefold) in GFAP immunoreactivity in the striatum; this indicates rampant gliosis, another pathological feature of PD. In summary, the mice might make a good model for studying dopaminergic degeneration and the role of parkin/Pael-R in that process. It will be interesting to see what behavioral data emerge from studies on these animals.
An entirely different twist on parkin came from the laboratory of Henry Querfurth at St. Elizabeth’s Medical Center in Boston. Charles Moussa’s poster described how tagged wild-type parkin on a lentiviral vector managed to protect neuroblastoma and muscle cells from death induced by overexpression of APP or the β-secretase substrate C-100. At least in this system, Aβ appears to be a parkin substrate. Specifically, parkin ubiquitinated Aβ and reduced its level inside the cells, the researchers suggest.
And last but not least, what about ubiquitination without degradation? Several posters pointed to a curious, parkin-catalyzed ubiquitination that does not affect protein turnover. Darren Moore, from Ted Dawson’s lab, presented evidence that parkin is involved in the multiple mono-ubiquitination of heat-shock protein 70 (Hsp70). Through mass spectroscopy, Moore identified Hsp70 as a parkin binding partner. Mutation in the RING2 domain, or C-terminal end of parkin, abolishes this interaction. Moore reported that the mono-ubiquitination of Hsp70 occurs both in-vitro and in cultured cells. However, inhibiting the proteasome has no effect on the turnover of the ubiquitinated Hsp70, suggesting that the ubiquitination is not necessarily a first step in degradation. Similarly, James Olzmann, from Lian Li’s lab at Emory University in Atlanta, reported that parkin polyubiquitinates DJ-1, which has also been implicated in familial PD. But Olzmann reported that it was specifically the L166P mutant of DJ-1 that gets ubiquitinated, not wild-type protein. In Olzmann’s hands, too, inhibiting the proteasome had no effect on levels of the ubiquitinated L166P DJ-1, indicating that the ubiquitination does not affect degradation of the protein, at least by the proteasome. What role these ubiquitination reactions have, if any, in vivo remains to be determined, but proteasome degradation does not appear to be it.—Tom Fagan.
- Loss of Parkin in Mammals Takes Steam Out of Mitochondria
- New Substrate for Parkin Links Disease to ER Stress
- Ko HS, von Coelln R, Sriram SR, Kim SW, Chung KK, Pletnikova O, Troncoso J, Johnson B, Saffary R, Goh EL, Song H, Park BJ, Kim MJ, Kim S, Dawson VL, Dawson TM. Accumulation of the authentic parkin substrate aminoacyl-tRNA synthetase cofactor, p38/JTV-1, leads to catecholaminergic cell death. J Neurosci. 2005 Aug 31;25(35):7968-78. PubMed.
No Available Further Reading