Mitochondria get a Parkin' ticket - Semantic Scholar

news and views 1. Cerletti, M., Shadrach, J. L., Jurga, S., Sherwood, R. & Wagers, A. J. Cold Spring Harb. Symp. Quant. Biol. 73, 317–322 (2008). 2. Uezumi, A., Fukada, S., Yamamoto, N., Takeda, S. & Tsuchida. K. Nature Cell Biol. 12, 143–152 (2010). 3. Joe, A. W. B. et al. Nature Cell Biol. 12, 153–163 (2010). 4. Carpenter, S. & Karpati, G. Pathology of Skeletal Muscle 2nd edn (Oxford University Press, 2001).

5. Vettor, R. et al. Am. J. Physiol. Endocrinol. Metab. 297, E987–E998 (2009). 6. Hu, E., Tontonoz, P. & Spiegelman, B. M. Proc. Natl Acad. Sci. USA 92, 9856–9860 (1995). 7. Yeow, K. et al. FEBS Lett. 506, 157–162 (2001). 8. Asakura, A., Komaki, M. & Rudnicki, M. Differentiation 68, 245–253 (2001). 9. Brack, A. S. et al. Science 317, 807–810 (2007).

10. Serrano, A. L., Baeza-Raja, B., Perdiguero, E., Jardi, M. & Munoz-Canoves, P. Cell Metab. 7, 33–44 (2008). 11. Naveiras, O. et al. Nature 460, 259–263 (2009). 12. Verma, S., Rajaratnam, J. H., Denton, J., Hoyland, J. A. & Byers, R. J. J. Clin. Pathol. 55, 693–698 (2002). 13. Plikus, M. V. et al. Nature 451, 340–344 (2008). 14. Rodeheffer, M. S., Birsoy, K. & Friedman, J. M. Cell 135, 240–249 (2008). 15. Tang, W. et al. Science 322, 583–586 (2008).

Mitochondria get a Parkin’ ticket Philipp Wild and Ivan Dikic Recent studies have revealed a prominent role of mitochondrial dysfunction in the development of one of the most common neurodegenerative disorders, Parkinson´s disease. The ubiquitin ligase Parkin and the protein kinase PINK1, whose mutations are associated with Parkinson´s disease, function in a pathway that links ubiquitylation with selective autophagy of damaged mitochondria. The pathogenesis of Parkinson´s disease (PD) is characterized by substantial loss of dopamine-containing neuronal cells and the formation of cytoplasmic inclusions (Lewy bodies) that accommodate aggregation-prone proteins1. Mutations in genes coding for the PTEN-induced putative kinase‑1 (PINK1) and the ubiquitin ligase Parkin have been associated with autosomal recessive forms of PD2. It has been reported previously that both proteins interact genetically and function within the same molecular pathway to influence mitochondrial dynamics3. On page 119 of this issue, Geisler et al. show that Parkin, in conjunction with PINK1, mediates selective autophagy of depolarized mitochondria4 (Fig. 1). Contrary to previous models proposing that loss of Parkin leads to enhanced cellular toxicity due to the accumulation of substrate proteins, this study suggests that mutant Parkin contributes to the onset of PD by a failure to trigger removal of dysfunctional mitochondria. Studies in Drosophila melanogaster have revealed that Parkin and PINK1 are directly involved in the regulation of mitochondrial functions — their mutants caused mitochondrial swelling in dopaminergic neurons and sperm5. Pioneering work has indicated that Parkin recruitment to mitochondria is required for the selective elimination of damaged mitochondria in mammalian cells6. Expression of Philipp Wild and Ivan Dikic are at the Goethe University Medical School, Institute of Biochemistry II and Frankfurt Institute for Molecular Life Sciences, Theodor-Stern-Kai 7, Frankfurt a.M., D-60590, Germany. e-mail: [email protected]

104

PD‑associated Parkin mutations affects its translocation to depolarized mitochondria and their subsequent clearance through the autophagy pathway. As these mutations disrupt the integrity of the E3 ligase RING domains, functional Parkin seems to be indispensable for proper mitophagy. Yet, how Parkin is recruited to depolarized mitochondria and the nature of its downstream signalling cascade remain unclear. Two recent reports demonstrated an important role for PINK1 in recruiting Parkin to the mitochondrial membrane4,7. PINK1 spans the outer mitochondrial membrane, with its kinase domain facing the cytosol. After mitochondrial depolarization, PINK1 interaction with Parkin is significantly enhanced4. Moreover, modulation of PINK1 expression or activity, either by siRNA-induced depletion or expression of kinase-deficient PINK1, reduces the relocation of Parkin to depolarized mitochondria and their turnover by mitophagy. Although PINK1 seems to function as an important sensor for mitochondrial damage upstream of Parkin, many issues related to their functional interactions remain unanswered. For example, how is the kinase activity of PINK1 regulated following mitochondrial damage, can PINK1 mediate phosphorylation of Parkin or other substrates, does this create a translocation signal for Parkin and/or does it stimulate the E3 ligase activity of Parkin? Such interplay represents an appealing mechanism to inducibly regulate the degradation of stressed mitochondria only upon the loss of membrane potential. Mitochondria are the principle sites of ATP production in aerobic cells. Yet, at the same time they are the major source of reactive oxygen

species (ROS), which damage cellular components, including the mitochondria. Various mechanisms, including autophagy, monitor and clear damaged mitochondria to maintain cellular homeostasis. Autophagy is characterized by the incorporation of cytoplasmic components into double-membrane vesicles (autophagosomes), which eventually fuse with lysosomes, mediating their degradation. Emerging evidence indicates that the engulfment of mitochondria by the autophagic machinery is not a random process. It engages autophagy receptors, molecules capable of binding to autophagy modifiers, MAP1LC3/ GABARAPs, as well as different substrates of autophagosomes8. Atg32, an outer mitochondrial membrane protein, has recently been described as a functional mitochondrial autophagy receptor in yeast9,10. Atg32 interacts with Atg8, the yeast homologue of MAP1LC3 and Atg11, a yeast adaptor protein for selective autophagy. A recently characterized mammalian mitophagy receptor, Nix/BNIP3L, binds to MAP1LC3/GABARAPs and mediates clearance of mitochondria upon depolarization and during erythrocyte differentiation and may represent the closest mammalian homologue to Atg32 (ref. 11). Geisler et al. now implicate, downstream of Parkin, the adaptor protein p62/ SQSTM1, which seems to require ubiquitylation as a targeting signal for mitophagy. Ubiquitin has been suggested to act as a signal for the selective autophagy of diverse cargos, including protein aggregates, ribosomes, peroxisomes and pathogens, in mammalian cells8. Earlier studies have shown that sperm mitochondria intensively decorated by ubiquitin following

nature cell biology VOLUME 12 | NUMBER 2 | FEBRUARY 2010 © 2010 Macmillan Publishers Limited. All rights reserved.

news and views Parkin Oxidative stress

Parkin

PINK1

PINK1

VDAC1

OMM

Phagophore

p62

Ub Ub Ub Ub

Parkin

p62

Ub Ub Ub Ub

LC3 p62 VDAC1

PINK1

VDAC1

Figure 1 Model of PINK1/Parkin-mediated mitophagy. In the absence of mitochondrial damage, E3 ligase Parkin is diffusely distributed throughout the cytosol. Upon mitochondrial membrane depolarization, PINK1, a kinase of the outer mitochondrial membrane (OMM), induces Parkin translocation to stressed mitochondria. Subsequently, Parkin mediates the formation of Lys 27-linked ubiquitin (Ub) chains on VDAC1. This leads to the recruitment of the autophagy receptor p62, which in turn, by binding to LC3, directs damaged mitochondria into forming autophagosomes (phagophore). Disruption of this sequential process at distinct steps may contribute to the development of PD due to accumulation of dysfunctional mitochondria and a concomitant increased cellular toxicity.

fertilization were cleared through autophagy 12. Geisler et al. found that Parkin colocalizes with ubiquitin chains at sites of condensed mitochondria after the loss of membrane potential, whereas pathogenic variants of Parkin cluster to mitochondria with a significant delay. The type of ubiquitin linkages depends on which of the seven lysines of ubiquitin serve as acceptor sites for chain elongation, and the authors showed that Parkin mediates the formation of both Lys 27 and Lys 63 linkages on substrates at depolarized mitochondria (Fig. 1). Importantly, different ubiquitin chains have specific consequences for the tagged substrates and are linked to diverse cellular functions13. The proteins that recognize and convert these signals into a functional response are responsible for this variety14. p62/SQSTM1 is one of these ubiquitin receptors, which is thought to bind ubiquitylated substrates through its ubiquitin-associated domain (UBA)8. The authors propose that p62/SQSTM1, which simultaneously binds ubiquitylated proteins and components of the autophagic machinery, functions downstream of Parkin. They showed that Parkin facilitates the recruitment of p62 to condensed mitochondria and subsequent p62dependent mitophagy. Conversely, p62 knockdown does not affect Parkin translocation to depolarized mitochondria, but is indispensable for mitochondrial clearance. The final cog in the wheel was the identification of a Parkin substrate that could mechanistically link ubiquitylation and autophagy. The authors

found that voltage-dependent anion channel 1 (VDAC1), a mitochondrial outer membrane protein, is ubiquitylated through Lys 27-linked chains, following mitochondrial membrane depolarization. So far, it is not clear whether the UBA domain of p62 can directly bind Lys 27 ubiquitin chains or whether the interaction is mediated through another ubiquitin receptor. The authors also showed that silencing of endogenous VDAC1 by siRNA disrupted the redistribution of Parkin to damaged mitochondria and prevented mitochondrial clearance. Both effects could be restored by re-transfection with wildtype VDAC1, indicating that the mitochondrial substrate VDAC1 is required for proper PINK1/ Parkin-directed mitophagy (Fig. 1). This latter result points to an important role for VDAC1 in the formation of this trimeric complex, in addition to being a Parkin substrate. This study provides insights into how functional mutations in two PD‑associated genes may contribute to the pathogenesis of PD by a failure to selectively clear damaged mitochondria (Fig. 1). A recent report also demonstrates the importance of PINK1 for Parkin translocation to mitochondria7. However, the model advanced by this group differs from the one proposed by Geisler et al.: they suggest that Parkin, in collaboration with PINK1, functions in the microtubule-dependent trafficking of depolarized mitochondria to lysosomes, rather than through the direct engagement of the autophagic machinery 7. Although both

models are not mutually exclusive, the observation by Geisler et al. that ubiquitylation of VDAC1 and p62 recruitment are prerequisites for PINK1/Parkin-mediated mitophagy argues for a more direct role of PINK1/Parkin in the engagement of the autophagy pathway. Although the present study contributes to a better understanding of the molecular mechanisms that potentially underlie the pathogenesis of PD, several open questions remain. For instance, it will be interesting to test whether there are other Parkin substrates in damaged mitochondria. Can VDAC1 constitute a key node at which autophagy and apoptosis pathways intersect? Ubiquitylation of VDAC1 and subsequent engulfment and degradation of damaged mitochondria might prevent the release of pro-apoptotic factors from mitochondria under physiological conditions. Moreover, it will be important to determine to which extent these findings can be generalized to other neuronal and non-neuronal cells, respectively. Several cell types, such as HeLa, lack detectable amounts of endogenous Parkin but nevertheless need to remove dysfunctional mitochondria to avoid increased ROS production. Given the redundancy of Parkin with other E3 ligases, it is appealing to consider the possibility of other ubiquitin ligases being involved in mitophagy in a cell type-specific manner. For instance, E3 ligases, such as MULAN, residing in or at the mitochondrial membrane, may regulate this


105

news and views process15. Notably, aberrant mitochondrial morphology observed in Parkin and PINK1 D. melanogaster mutants occurred in tissues with a high demand for energy, such as flight muscle, dopaminergic neurons and sperm. These cell types are more susceptible to oxidative stress and, hence, specific mechanisms may be required to tightly control the integrity of their mitochondria. Further studies addressing the contribution of selective autophagy

to mitochondrial quality control should also prove beneficial in developing new therapeutic approaches for the treatment of PD patients. 1. Shults, C. W. Proc. Natl Acad. Sci. USA 103, 1661–1668 (2006). 2. Kitada, T. et al. Nature 392, 605–608 (1998). 3. Poole, A. C. et al. Proc. Natl Acad. Sci. USA 105, 1638– 1643 (2008). 4. Geisler, S. et al. Nature Cell Biol. 12, 119-131 (2010). 5. Park, J., Kim Y. & Chung, J. Dis. Model Mech. 2, 336– 340 (2009). 6. Narendra, D., Tanaka, A., Suen & D. F., & Youle R. J. J. Cell Biol. 183, 795–803 (2008).

7. Vives-Bauza, C. et al. Proc. Natl Acad. Sci. USA 107, 378–383 (2010). 8. Kirkin, V., McEwan, D. G., Novak & Dikic, I. Mol. Cell 34, 259–269 (2009). 9. Okamato, K., Kondo-Okamato, N. & Ohsumi, Y. Dev. Cell 17, 87–97 (2009). 10. Kanki, T. et al. Dev. Cell 17, 98–109 (2009). 11. Novak, I. et al. EMBO Rep. 11, 45–51 (2010). 12. Sutovsky, P. et al. Nature 402, 371–372 (1999). 13. Ikeda, F. & Dikic, I. EMBO Rep. 9, 536–542 (2008). 14. Dikic, I., Wakatsuki, S. & Walters K. J. Nature Rev. Mol. Cell Biol. 10, 659–671 (2009). 15. Li, W. et al. PloS One 3, e1487 (2008).

Cohesin and Cdk1: an anaphase barricade Keith T. Jones Separation of sister chromatids at anaphase in metazoan cells requires only the cleavage of the kleisin subunit of centromeric cohesin, but efficient poleward movement of separated sisters requires the associated loss in Cdk1 activity. Activation of the anaphase-promoting complex/cyclosome ensures these events are coordinated. The ties holding sister chromatids together during mitosis must be sufficiently strong to provide counter tension to the pulling forces of microtubules attached to their kinetochores. This molecular tension begins during prometaphase so as to allow all sister chromatids to align on the equator of the metaphase spindle, a goal, known as congression, in which sister kinetochores are biorientated. At the transition from metaphase to anaphase, the counter forces holding sister chromatids capitulate, allowing microtubules to pull sisters apart. There must be a synchronous surrender in this tug of war, so that separation of all sisters is achieved at the same time. Similarly it is important for the microtubules not be confused en route; their cargo of separated sisters must now travel towards the spindle poles not the spindle equator. Synchronous sister release and efficient microtubule pulling forces help prevent lagging chromosomes, and so ensure correct chromosome numbers in the two daughter cells. Therefore the cohesive forces holding sister chromatids together, their synchronous dissipation at anaphase onset and their associated poleward movement are all essential features in preventing the formation of aneuploid daughter cells, which may go on Keith Jones is at the School of Biomedical Sciences, Faculty of Health, University of Newcastle, NSW 2308, Australia. e‑mail: [email protected]

106

to show compromised physiological function or a propensity for tumorigenicity. On page 185 of this issue, Oliveira et al.1 discover that cleavage of the kleisin component of cohesin (RAD21; Scc1/Mcd1 in budding yeast) and loss of Cdk1 activity are the key steps that drive the metaphase to anaphase transition in a higher eukaryote model system, the fruit fly Drosophila melanogaster. For a long time, the cohesin complex, which is composed of two ‘structural maintenance of chromosome’ (SMC) subunits, SMC1 and 3, and two non-SMC subunits, RAD21 and SCC3, has been a reasonable candidate to provide the counter tension that holds sister chromatids together at metaphase. In one model, it is thought to embrace sister chromatids in a ring, which would provide the necessary structure to hold sisters together 2. Moreover, RAD21 is a substrate of Separase, a protease activated at anaphase onset by the degradation of its inhibitory chaperone Securin. RAD21 cleavage would therefore break the ring-like structure, allowing sisters to be pulled apart, and this would be achieved in a timely manner by the controlled activation of Separase. Oliveira et al. used a fly line expressing RAD21, engineered with cleavage sites for the tobacco etch virus (TEV) protease, to test directly whether RAD21 cleavage, and so the cohesin complex, did indeed provide the counter tension to microtubules during metaphase. They could do this by

injecting the TEV protease, and in so doing immediately cleave RAD21. The early fly embryo is ideally suited for this experiment, as it is a syncytium with multiple mitotic spindles and no cell membranes to prevent diffusion of the injected protease. Metaphase-arrested embryos were needed to assess the effects of RAD21 cleavage, as it is only at this stage of the cell cycle that it can be ascertained whether cohesin complexes were providing the sole counter force to microtubules. This was achieved by inhibiting the anaphase-promoting complex/cyclosome (APC/C), an E3 ubiquitin ligase, whose activity at the onset of anaphase triggers polyubiquitylation, and immediate proteolysis, of Securin so freeing Separase to act on RAD21 (ref. 3). APC/C inhibition would normally be a feature of prometaphase at a time when sister congression had not yet been achieved, and would be mediated by activation of the spindle assembly checkpoint (SAC). The SAC is a collection of proteins that prevent APC/C activity through inhibition of CDC20, an essential APC/C component needed for triggering anaphase4. The SAC is normally satisfied only once all sister kinetochores are biorientated, occupied and under tension from microtubules, and as a consequence APC/CCDC20 can become active. Fly embryos with TEV-sensitive RAD21 were arrested at metaphase by one of two methods, either by overexpressing MAD2, a critical SAC member that can bind, and so sequester CDC20
