Defects in PINK-PRKN-PARK7/DJ-1-dependent mitophagy and autosomal recessive Parkinson disease

KEYWORDS:

• DJ-1
• mitochondria
• mitophagy
• neurons
• optineurin
• PARK7
• parkin
• parkinson’s disease
• PINK1
• PRKN

Parkinson disease (PD) comes in many clinical varieties. One subtype of PD is characterized by autosomal recessive inheritance, young age at onset, a relatively pure motor phenotype and very slow disease progression. Known causes of this subtype of PD are loss-of-function mutations in (in order of decreasing prevalence) the PRKN, PINK1 and PARK7/DJ-1 genes. Compelling evidence from many laboratories has revealed that PINK1 and PRKN selectively label damaged mitochondria for autophagic destruction (mitophagy). PINK1 sets this quality-control cascade in motion by accumulating on damaged mitochondria and phosphorylating ubiquitin that is present in small amounts on the outer mitochondrial membrane (OMM). Phospho-ubiquitin is a high-affinity receptor for PRKN. Binding of PRKN to phospho-ubiquitin in combination with PINK1-mediated PRKN phosphorylation potently activates the E3 ubiquitin ligase activity of PRKN. PRKN-mediated ubiquitination of OMM proteins induces recruitment of the ubiquitin-binding autophagy receptor OPTN (optineurin) and the downstream autophagic machinery.

Compared with PINK1 and PRKN, the precise molecular function of PARK7/DJ-1 has remained more enigmatic. Whereas PRKN possesses the typical domain structure of a RING-between-RING E3 ubiquitin ligase and PINK1 contains a serine/threonine protein kinase domain, the sequence and structure of PARK7 do not have any features that are strongly predictive of a specific biochemical function. Another difficulty is that patients with PARK7 mutations are exceedingly rare, which has limited the availability of patient-derived cells with PARK7 mutations and has forced investigators to rely on less disease-relevant models.

Recently, we measured PINK1-PRKN-mediated mitophagy in skin fibroblasts and iPSC-derived neurons from a PD patient with a homozygous PARK7 mutation (p.Pro158del) [1]. This mutation dramatically impairs the stability of PARK7, leading to an almost complete loss of the protein. PINK1-PRKN-mediated mitophagy, triggered by exposure to mitochondria-depolarizing agents, is severely impaired in the PARK7 mutant fibroblasts and neurons, and the magnitude of this mitophagy defect is similar to that observed in cells from PD patients with PINK1 or PRKN mutations. Expression of wild-type PARK7 in the PARK7 mutant fibroblasts and neurons rescues mitophagy. Conversely, knockdown of PARK7 in wild-type cells disrupts PINK1-PRKN-mediated mitophagy. Activation of PINK1 and PRKN on depolarized mitochondria is preserved in the PARK7 mutant cells, but recruitment of OPTN to depolarized mitochondria is impaired. Strikingly, nonselective, starvation-induced autophagy, which does not require OPTN or other autophagy receptors, is intact in PARK7 mutant cells as well as in PINK1 and PRKN mutant cells. Importantly, confocal microscopy, structured illumination microscopy and western blotting of subcellular fractions demonstrated that endogenous wild-type PARK7 in healthy control fibroblasts and dopaminergic neurons translocates to depolarized mitochondria, where it colocalizes with PRKN, phospho-ubiquitin and LC3. The time course of translocation of PARK7 to depolarized mitochondria is very similar to that of OPTN, but clearly slower than that of PRKN. PARK7 translocation to depolarized mitochondria depends on PINK1 and PRKN, as it is abolished in PINK1 and PRKN mutant cells. Overexpression of PARK7 does not rescue the mitophagy defect of PINK1 or PRKN mutant cells, indicating that PARK7 is unable to compensate for upstream loss of PINK1 kinase activity or PRKN E3 ubiquitin ligase activity.

OPTN co-immunoprecipitates with PARK7 both in basal conditions and after mitochondrial depolarization. Proximity ligation assay (PLA) experiments confirm that endogenous PARK7 and OPTN are in intimate proximity and show that a substantial proportion of the PARK7-OPTN PLA dots relocate from the cytosol to mitochondria after mitochondrial depolarization. This finding suggests that PARK7 and OPTN are in a complex in the cytosol in basal conditions and that some of these complexes are recruited to damaged mitochondria. We artificially targeted PARK7 to the OMM by fusing the N-terminal transmembrane segment of the OMM protein TOMM20 to PARK7. Interestingly, expression of this OMM-attached PARK7 construct is sufficient to recruit OPTN to mitochondria in the absence of mitochondria-depolarizing agents, even in PINK1 and PRKN mutant cells. Thus, PARK7 artificially anchored to the OMM is able to draw OPTN to mitochondria independently of PINK1-PRKN pathway activation. In summary, PARK7 plays an essential role in PINK1-PRKN-mediated mitophagy at a position downstream of PINK1 and PRKN. PARK7 binds with OPTN, comigrates with OPTN to ubiquitinated mitochondria, and promotes mitochondrial OPTN accumulation.

Several important questions remain to be addressed. OPTN is a scaffold protein with multiple protein interaction domains, and it is still unknown which domain of OPTN mediates the (direct or indirect) interaction with PARK7. It is also unclear how interaction with PARK7 exactly promotes mitochondrial accumulation of OPTN. Previous work has shown that PARK7 can counteract oxidative stress by virtue of the H₂O₂-scavenging properties of its C106 residue. Conceivably, binding to PARK7 might safeguard OPTN against oxidation in the highly oxidative microenvironment surrounding damaged mitochondria. However, a PARK7 construct in which C106 is replaced by alanine rescues mitophagy in PARK7 mutant cells as efficiently as wild-type PARK7, indicating that the function of PARK7 in mitophagy does not require C106 oxidation. Also, the reactive oxygen species scavenger Mito-TEMPO does not mitigate the mitophagy defect of PARK7 mutant cells, implying that the mitophagy defect is not caused by oxidative stress.

Multiple enzymatic functions have previously been ascribed to PARK7. PARK7 has been reported to function as a protease, an aldehyde-adduct hydrolase, a glyoxalase or as an enzyme that prevents damage caused by the glycolytic intermediate 1,3-bisphosphoglycerate. Whether PARK7 exerts any of these enzymatic activities during or after its mitochondrial translocation and how this might contribute to mitophagy, is unclear. Alternatively, PARK7 could promote mitochondrial accumulation of its binding partner OPTN non-enzymatically, e.g., by interacting with an OMM protein.

It should be kept in mind that many divergent, mitophagy-independent functions have also been described for PINK1, PRKN and PARK7. Whether PINK1, PRKN and PARK7 mutations cause PD via mitophagy disruption or via mitophagy-independent mechanisms, is still a matter of debate. Nevertheless, the convergence of the disruptive effects of PINK1, PRKN and PARK7 mutations on mitophagy is unlikely to be a coincidence and suggests that this pathway is fundamentally important in the pathogenesis of autosomal recessive PD.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by FWO grant G072821N to W.V., D.I. is supported by a PhD Fellowship of the Research Foundation Flanders (FWO).