Background
Parkinson's disease (PD), affecting 1% of the population over 65, is one of the most common neurodegenerative disorders characterized with selective loss of dopaminergic neurons in the substantia nigra [
1‐
3]. Although molecular mechanism of disease is not fully known, studies indicate that mitochondrial dysfunction and oxidative stress could play a role in neuronal loss [
4]. Three mutated genes –
Parkin, DJ-1, and
PINK1 – implicated in mitochondria and oxidative stress-related survival pathways are typically present in the brain with Parkinson's disease (PD) with apparent autosomal recessive inheritance. Two other genes – α
-Synuclein and
LRRK2 – are present in an autosomal dominant pattern and are associated with prominent intracellular protein inclusions [
5,
6]. A number of pathogenic mutations identified in the
DJ-1 gene are linked to early onset familial Parkinson's disease (PD) [
7‐
11]. Mechanistically, the DJ-1 protein has been reported to be sensitive to oxidative stress and it may function as an anti-oxidant protein [
12‐
15].
DJ-1 is a small protein composing 189 amino acids and it is expressed in variety of tissues including brain [
16,
17]. Although its precise biochemical function is not fully known, recent studies indicate that it has been involved in many diverse biological processes including protease, chaperone and antioxidant activities [
13,
18‐
20]. Crystal structure and biochemical data show that DJ-1 forms dimers (37 KD) [
18,
20‐
23]. In addition to dimer formation, DJ-1 also tends to form high molecular weight (HMW) complexes (250–700 kD) during normal condition [
21,
24‐
26] It is possible that DJ-1 and other DJ-1 interacting proteins might form these HMW complexes. Although presence of α-synuclein is controversial [
21,
24,
27], Parkin has been associated with DJ-1 HMW complex [
21] However, whether the DJ-1 HMW complex remains intact in brains with sporadic PD is not clear. Regulation of the formation and distribution of the DJ-1 HMW complexes remains largely obscure at this stage. In this study, we analyzed the levels of DJ-1 HMW complex as well as the DJ-1 protein in the substantia nigra, temporal cortex and mitochondria fraction from the rapidly autopsied temporal cortex from the brains of PD patients by native and denatured Western blots. We found that the total DJ-1 protein was significantly reduced in the substantia nigra and the DJ-1 HMW complex in the cortex mitochondria is also significantly decreased in brains of sporadic PD patients. These results suggest abnormal DJ-1 expression levels and DJ-1 protein structure changes may contribute to PD pathogenesis.
Discussion
DJ-1 is the third gene that has been linked to PD. Mutations in the
DJ-1 gene cause early onset PD with autosomal recessive inheritance [
7‐
11]. In this study, we have analyzed its levels in the more common sporadic form of PD. We analyzed HMW DJ-1 complex and total DJ-1 protein levels in PD patients. We found significantly lower total DJ-1 levels in the PD substantia nigra region, indicating reduction in DJ-1 protein levels might be related to the substantia nigra pathology in PD patients. In addition, we have shown that DJ-1 is localized in mitochondria of both neuronal and glial cells supporting the role of DJ-1 in mitochondrial function. When we compared the levels of total and HMW complex of DJ-1, we have discovered significantly decreased levels of HMW complex of DJ-1 in mitochondrial preparation of the cortex from PD brains.
We have found that total level of DJ-1 in the substantia nigra is decreased in PD compared to health controls (Fig.
1). However, other groups have reported either no change or increased levels of DJ-1 in PD brain [
25,
35‐
37]. One possibility of the discrepancy with previous groups is that they have not investigated DJ-1 specifically in the substantia nigra from PD brains. For instance, we could not observe any change in DJ-1 protein levels in the PD frontal cortex. Second, post-mortem interval (PMI) is shorter (< 2.5 hrs) and more consistent between PD and NPC in our study compared to other groups [
25,
35‐
37], which are critical factors. Third, we have analyzed a larger sample group (PD n = 18, NPC n = 9) compared to others. Variation and range within each group from smaller number of samples might affect the accuracy of the conclusion drawn from the data.
Oxidative stress has been well-documented in the substantia nigra in PD [
12‐
15]. Recent studies have shown that over-expression of wild type DJ-1 in cultured human dopaminergic cells protected the cells from death induced by hydrogen peroxide and 6-hydroxydopamine, while over-expressing the L166P mutant DJ-1 had no such protective effect [
34]; knocking down endogenous DJ-1 rendered cells susceptible to oxidative damage [
34]. Also, wild type DJ-1 inhibited the mutant A53T α-synuclein-induced protein aggregation and cytotoxicity by increasing the expression of heat shock protein 70 [
34]. Accordingly, our studies have shown that total levels of DJ-1 were significantly lower in the substantia nigra region of PD than that of NPC, implying that decreased DJ-1 levels might account for the loss of anti-oxidation capability in the substantia nigra region.
A recent study has also shown increased oxidative damage in the PD cortical region [
32], but little has been reported on the subject. We proposed that DJ-1 HMW complex could account for this anti-oxidative stress function. However, since we did not observe any significant changes in the total DJ-1 levels in the cortex region, this suggests that the subcellular location of DJ-1, rather than the total amount of DJ-1, may be more important for its biological role. Recent studies in DJ-1 subcellular localization have identified an endogenous pool of DJ-1 in the mitochondrial matrix and inter-membrane space [
26,
38]. Therefore, DJ-1 is an integral mitochondrial protein that may have important functions in regulating mitochondrial physiology [
26], consistent to our result that displayed co-localization of DJ-1 and mitochondria. Moreover, in response to oxidative stress, DJ-1 is translocated from cytosol to mitochondria and nucleus. DJ-1 tagged with mitochondrial signal confers more neuro-protection against oxidative stress when compared to WT or nuclear tagged DJ-1 in cell assays [
39]. Thus, mitochondrial localization of DJ-1 is important to exert cell-protective effect in oxidative stress condition. Interestingly, we found that the DJ-1 HMW complex in the mitochondria fraction extracted from the PD cortical region is significantly lower than NPC. In DJ-1 HMW complex, Parkin, an E3-ubiquitin ligase, has been found [
21]. Moreover, oxidative stress promotes association of Parkin and DJ-1[
25]. Therefore, reduced levels of HMW complex in mitochondria isolated from PD cortex, may suggest reduced interaction or dissociation of DJ-1 with Parkin or possibly with other proteins as well. Additionally, recent studies also have shown that the subunit of mitochondrial complex I in the PD cortex region was oxidatively damaged and its function was impaired [
40], implying that HMW DJ-1 complex might play a role in protecting the cortex mitochondria from oxidative stress.
Although total DJ-1 levels were reduced in substantia nigra, we could not observe any change in DJ-1 HMW complex levels in sustantia nigra. On the other hand, in cortex we could not observe any change in levels of DJ-1 HMW complex in the cortex of PD patients as well, but when we examined mitochondria enriched preparation from the cortex we observed a reduction in DJ-1 complex. It could be that DJ-1 levels or distribution of DJ-1 between HMW and dimers in mitochondria might be important for DJ-1 function. Therefore, it would be also interesting to examine the HMW and total DJ-1 levels in the substantia nigra mitochondria. However, due to the large amount of tissue required to prepare the mitochondria fraction, while relatively small amount of the substantia nigra tissue was available, mitochondrial fractions from substantia nigra has not yet been examined until we collect enough tissue for this study in the future.
It is worthwhile to note that large variations were seen with each group, despite the fact that all samples were taken from the same brain region. It is practically impossible to obtain a sample set from patients in the same cohort with identical genetic backgrounds, which would reduce the variation. In order to clearly distinguish PD and NPC, we categorized our patients based on pathological definitions. The inherent heterogeneity of the samples meant that we could not rule out other genetic factors that could affect DJ-1 levels. Future studies are anticipated to analyze DJ-1 HMW complex and total DJ-1 in the mitochondria fraction of substantia nigra, when a sufficient amount of brain tissue is available for biochemical analysis, and to investigate the possible mechanisms involved.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
HN analyzed the results and performed the statistical analysis, and revised the manuscript PH participated in acquisition of data and performed immunohistochemistry and immunocytochemistry experiments. Both HN and PH did trouble shooting to ensure experimental results reliable. TB and LS involved in collection of tissue used in this study. WX and YS designed experiments. YS supervised HN and PH in experiments and participated in preparation of the manuscript. All authors read and approved the final manuscript.