The PINK1/Parkin pathway regulates ... - Semantic Scholar

Human Molecular Genetics, 2011, Vol. 20, No. 16 doi:10.1093/hmg/ddr235 Advance Access published on May 25, 2011

3227–3240

The PINK1/Parkin pathway regulates mitochondrial dynamics and function in mammalian hippocampal and dopaminergic neurons Wendou Yu 1, Yaping Sun 2, Su Guo 2 and Bingwei Lu 1,∗ 1

Department of Pathology, Stanford University School of Medicine, Stanford, CA 94305, USA and 2Department of Biopharmaceutical Sciences, Programs in Biological Sciences and Human Genetics, University of California, San Francisco, CA 94143, USA Received April 14, 2011; Revised and Accepted May 19, 2011

PTEN-induced putative kinase 1 (PINK1) and Parkin act in a common pathway to regulate mitochondrial dynamics, the involvement of which in the pathogenesis of Parkinson’s disease (PD) is increasingly being appreciated. However, how the PINK1/Parkin pathway influences mitochondrial function is not well understood, and the exact role of this pathway in controlling mitochondrial dynamics remains controversial. Here we used mammalian primary neurons to examine the function of the PINK1/Parkin pathway in regulating mitochondrial dynamics and function. In rat hippocampal neurons, PINK1 or Parkin overexpression resulted in increased mitochondrial number, smaller mitochondrial size and reduced mitochondrial occupancy of neuronal processes, suggesting that the balance of mitochondrial fission/fusion dynamics is tipped toward more fission. Conversely, inactivation of PINK1 resulted in elongated mitochondria, indicating that the balance of mitochondrial fission/fusion dynamics is tipped toward more fusion. Furthermore, overexpression of the fission protein Drp1 (dynamin-related protein 1) or knocking down of the fusion protein OPA1 (optical atrophy 1) suppressed PINK1 RNAi-induced mitochondrial morphological defect, and overexpression of PINK1 or Parkin suppressed the elongated mitochondria phenotype caused by Drp1 RNAi. Functionally, PINK1 knockdown and overexpression had opposite effects on dendritic spine formation and neuronal vulnerability to excitotoxicity. Finally, we found that PINK1/Parkin similarly influenced mitochondrial dynamics in rat midbrain dopaminergic neurons. These results, together with previous findings in Drosophila dopaminergic neurons, indicate that the PINK1/Parkin pathway plays conserved roles in regulating neuronal mitochondrial dynamics and function.

INTRODUCTION Parkinson’s disease (PD) is a movement disorder featuring bradykinesia, progressive rigidity, tremor and loss of postural stability, which have been attributed to the loss of dopaminergic neurons in the substantia nigra, although neurons in other brain regions are also affected. A total of 13 loci (PARK1-13) have been associated with rare forms of familial PD (FPD). The mechanisms of action of the FPD genes appear diverse and are not well understood. Model organisms have been instrumental in elucidating FPD gene function (1,2). Studies in Drosophila have shown that Parkin (PARK2) and

PTEN-induced putative kinase 1 (PINK1) (PARK6) work in a common pathway, with Parkin acting downstream of PINK1 (3 – 5). Further genetic studies in Drosophila have revealed that the PINK1/Parkin pathway regulates mitochondrial morphology by tipping the balance of mitochondrial fission/fusion dynamics toward more fission in dopaminergic neurons (6) and muscle cells (6 – 9). However, studies in mammalian cells have generated divergent results. In human cervical carcinoma HeLa cell line and neuroblastoma cell lines with dopaminergic feature (M17 and SH-SY5Y), PINK1 overexpression increased the interconnectivity of mitochondria, whereas small hairpin RNA

∗

To whom correspondence should be addressed at: Department of Pathology, Stanford University School of Medicine, R270 Edwards Building, Stanford, CA 94305, USA. Tel: +1 6507231828; Fax: +1 6504986616; Email: [email protected]

# The Author 2011. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]

3228

Human Molecular Genetics, 2011, Vol. 20, No. 16

(shRNA)-mediated knockdown increased mitochondrial fragmentation (10– 13). On the other hand, cultured fibroblasts from patients carrying Parkin mutations contained longer and more branched mitochondria than controls (14). Furthermore, in monkey COS-7 cells, PINK1 overexpression resulted in more fragmented mitochondria, whereas PINK1 RNAi led to longer and more tubular mitochondrial morphology, which could be rescued by the overexpression of fission proteins hFis1 or Drp1 (dynamin-related protein 1) (6). It should be noted that these results are derived from either transformed cancer cell lines, or fibroblasts that are actively dividing and lack the characteristics of postmitotic neurons. Although it is possible that the PINK1/Parkin pathway may exert differential effects depending on the cell type and culture condition, it is important to examine its role in mammalian neurons, the cell types that are in high demand for energy and are the primary targets in many neurodegenerative diseases (2). In this study, we used primary cultures of rat hippocampal and dopaminergic neurons to study the role of the PINK1/Parkin pathway in regulating mitochondrial dynamics. Our results suggest that the PINK1/Parkin pathway plays a conserved role in tipping the balance of mitochondrial fission/fusion dynamics toward more fission.

RESULTS PINK1 overexpression increases mitochondrial fragmentation, whereas its knockdown has opposite effects in rat hippocampal neurons To study mitochondrial morphology, we used the EGFP or DsRed reporter fused to the mitochondrial-targeting sequence of cytochrome c oxidase (Mito-EGFP or Mito-DsRed) to label mitochondria. Unlike the more interconnected mitochondrial network described for HeLa or SH-SY5Y cell lines, mitochondria in the processes of cultured rat hippocampal neurons showed both punctate and thin, elongated tubular morphology, with gaps between adjacent mitochondrial units (Figs 1A and 2A, control panels). The easily observable morphology of mitochondria and their heterogeneity in this system thus provide a simple and ideal model to study mitochondrial dynamics. To examine the effect of PINK1 on mitochondrial morphology, rat hippocampal neurons were co-transfected with Mito-EGFP and wild-type human PINK1 (PINK1-wt). PINK1-wt clearly increased mitochondrial fragmentation: mitochondria became more punctate and shorter, and there were wider gaps between adjacent mitochondria (Fig. 1A and B). A kinase-dead form of PINK1, PINK1 (K219M), and a PD-associated mutant, PINK1 (L347P) (15), were also tested. In contrast to PINK1-wt, these PINK1 mutants increased mitochondrial interconnectivity upon overexpression, resulting in fewer gaps between the longer mitochondrial units (Fig. 1A and B). This result suggests that these PINK1 mutants may act in a dominant-negative fashion, consistent with results from earlier studies (16). It further supports that PINK1 affects mitochondrial morphology in a kinase-activity-dependent manner. We characterized mitochondrial morphology in detail using three parameters: mitochondrial index (summed mitochondrial length divided by

dendrite length), mitochondrial number (total number of mitochondrial segment per 100 mm dendrite length) and mitochondrial size (mitochondrial segment length in micrometer). It is apparent that PINK1 or Parkin overexpression resulted in increased mitochondrial number, smaller mitochondrial size and reduced mitochondrial index, whereas PINK1 mutant had opposite effects (Fig. 1B; Mito index: control 63.3 + 2.7%; PINK1-wt 36.2 + 2.7%; PINK1 K219M 76.9 + 2.6%; PINK1 L347P 77.6 + 2.4%. Mito number: control 12.8 + 1.1; PINK1-wt 27.1 + 1.2; PINK1 K219M 6.4 + 0.6; PINK1 L347P 6.0 + 0.5. Mito size: control 5.3 + 0.5; PINK1-wt 1.5 + 0.1; PINK1 K219M 14.3 + 1.4; PINK1 L347P 14.9 + 1.4; n ¼ 20). Together, these results support that PINK1 tips the balance of mitochondrial fission/fusion dynamics toward more fission in mammalian hippocampal neurons, consistent with findings in Drosophila dopaminergic neurons (6) and flight muscle (6 – 9). To test the effect of PINK1 deficiency on mitochondrial morphology, we used a lentiviral vector that expresses PINK1 shRNA driven by the H1 promoter, and Mito-EGFP driven by the ubiquitin promoter. With this lentiviral construct, we could achieve simultaneous knocking down of PINK1 and observing mitochondrial morphology with Mito-EGFP in the same neuron. A luciferase RNAi construct served as a control. To test the efficiency of RNAi knockdown, PINK1 protein expression was detected in transfected neurons with a rabbit anti-PINK1 antibody. As shown in Figure 1A, neurons with PINK1 shRNA expression as indicated with the co-expressed Mito-EGFP reporter showed decreased PINK1 level, compared with neighboring neurons without PINK1 shRNA expression (double arrow heads). Notably, neurons expressing PINK1 shRNA showed more elongated mitochondria in a continuous network, compared with neurons transfected with control shRNA or with Mito-EGFP alone. Measurements of mitochondrial index, mitochondrial number and mitochondrial size all showed significant differences (Fig. 1B; Mito index: control 63.3 + 2.7%; Luciferase RNAi 66.9 + 2.1%; PINK1 RNAi 79.0 + 2.2%. Mito number: control 12.8 + 1.1; Luciferase RNAi 14.4 + 1.0; PINK1 RNAi 9.3 + 1.2. Mito size: control 5.3 + 0.5; Luciferase RNAi 4.6 + 0.3; PINK1 RNAi 14.1 + 2.3; n ¼ 20). These loss of function results further support the notion that PINK1 tips the balance of mitochondrial fission/fusion dynamics toward more fission in mammalian hippocampal neurons. Parkin overexpression also results in more fragmented mitochondria and it suppresses PINK1 RNAi effects in rat hippocampal neurons We next examined the relationship between PINK1 and Parkin in regulating mitochondrial morphology in hippocampal neurons. As shown in Figure 2A and C, overexpression of myc-tagged Parkin-wt resulted in more fragmented mitochondria, compared with Mito-EGFP transfection alone. The effect was so dramatic that it was rare to detect tubular mitochondria in Parkin-wt overexpression neurons. Parkin is an E3 ubiquitin ligase and the C-terminal RING finger domain plays a crucial role in mediating its function, as a Parkin mutant (W453Stop) having a C-terminal 13-amino acid truncation was found to be pathogenic (17). We made a mutant form of Parkin (Parkin-


3229

Figure 1. Effects of PINK1 overexpression and knockdown on mitochondrial morphology. (A) Cultured rat hippocampal neurons were transfected with Mito-EGFP and HA-tagged PINK1-wt or PINK1 (K219M), and PINK1 (L347P), or with control luciferase shRNA and PINK1 shRNA expressed from the lentiviral vector. Immunofluorescence analysis was performed with chicken anti-EGFP and rat anti-HA antibodies or rabbit anti-PINK1 antibodies. Arrowheads in the control panels show fragmented mitochondria in a neuron transfected with PINK1-wt. Arrows in PINK1-L347P panels show a neuron with no PINK1-L347P expression. Double arrowheads in PINK1 RNAi panels show PINK1 expression in a neuron not transfected with PINK1 shRNA. (B) Quantification of Mito index, Mito number and Mito size in dendrites from neurons transfected with the constructs is shown in (A) (∗ P , 0.05; ∗∗ P , 0.01; ∗∗∗ P , 0.001 in Student’s t-test). Scale bar, 15 mm in lower magnification panels, 5 mm in higher magnification panels in this and the subsequent figures.

mut) by reversing the coding sequence of the last 21 amino acids (Fig. 2B). Unlike Parkin-wt, Parkin-mut lost the ability to induce increased mitochondrial number and decreased mitochondrial size and mitochondrial index (Fig. 2A and C; Mito index: control 70.8 + 3.0%; Parkin-wt 33.1 + 2.0%; Parkinmut 72.7 + 2.5%. Mito number: control 15.7 + 1.1; Parkin-wt 31.0 + 2.3; Parkin-mut 15.0 + 1.1. Mito size: control 4.6 + 0.4; Parkin-wt 1.3 + 0.1; Parkin-mut 5.0 + 0.3; n ¼ 20). We conclude that, like PINK1, Parkin also tips the balance of mitochondrial fission/fusion dynamics toward more fission and that this effect requires the integrity of the C-terminal RING finger domain and presumably the E3 ubiquitin ligase activity of Parkin. Next, we asked whether Parkin overexpression could rescue PINK1 RNAi effect, as is the case in Drosophila studies. In PINK1 RNAi neurons, added Parkin-wt expression led to more fragmented mitochondria, compared with neighboring neurons transfected with PINK1 shRNA alone, which consistently showed more elongated and fused mitochondria (Fig. 2A, arrows; Fig. 2C; Mito index: control 70.8 + 3.0%; Parkin-wt 33.1 + 2.0%; PINK1 RNAi 79.0 + 2.2%; PINK1 RNAi + Parkin-wt 38.9 + 2.1%. Mito number: control 15.7 + 1.1; Parkin-wt 31.0 + 2.3; PINK1 RNAi 9.3 + 1.2; PINK1 RNAi + Parkin-wt 27.6 + 1.2. Mito size: control 4.6 + 0.4; Parkin-wt 1.3 + 0.1; PINK1 RNAi 14.1 + 2.3; PINK1 RNAi + Parkin-wt 1.5 + 0.1; n ¼ 20). The

fragmentation of mitochondria in PINK1 RNAi/Parkin-wt overexpression background was similar to that seen in neurons overexpressing Parkin-wt alone, suggesting that the effect of Parkin-wt overexpression is epistatic to that of PINK1 deficiency. These results are consistent with the notion that PINK1 and Parkin work in the same pathway to tip the balance of mitochondrial fission/fusion dynamics toward more fission, with Parkin acting downstream of PINK1. To further explore the role of PINK1 in regulating mitochondrial fission/fusion dynamics, we next used a photoactivation assay to examine the effects of altered PINK1 activities on mitochondrial fusion. Rat hippocampal neurons were transfected with Mito-Dendra2. Dendra2 is a photoconvertible fluorescent protein that can be irreversibly converted from green fluorescence (maximum excitation and emission at 486 and 505 nm, respectively) to red fluorescence (maximum excitation and emission at 558 and 575 nm, respectively). With Mito-Dendra2 and live imaging, we were able to track lateral diffusion of photo-converted mitochondrial fluorescence, a reflection of mitochondrial fusion. To quantify mitochondrial fusion activity, we measured the relative diffusion rate of the photo-converted fluorescent signals (dendritic length occupied by the photo-converted signals at the time of imaging/dendritic length occupied by the photoconverted signals right after photo-conversion) in control

3230


Figure 2. Parkin overexpression leads to mitochondrial fragmentation and it suppresses PINK1 RNAi effect. (A) Cultured hippocampal neurons were transfected with Mito-EGFP and myc-tagged Parkin-wt, Parkin-mut or Parkin-wt plus PINK1 RNAi. Immunofluorescence was performed with chicken anti-EGFP and mouse anti-myc antibodies. Neurons marked with arrowheads in the Parkin-mut panels are control neurons transfected with Mito-EGFP only. In the PINK1 RNAi + Parkin panels, neurons were transfected with the lentiviral vector expressing PINK1 shRNA, mito-EGFP and myc-tagged Parkin-wt. Arrows show mitochondrial morphology in a neuron transfected with PINK1 shRNA only. (B) A diagram showing the human Parkin constructs used. (C) Quantification of Mito index, Mito number and Mito size in the dendrites of neurons transfected with the constructs is shown in (A)(∗∗∗ P , 0.001 in Student’s t-test).

neurons and neurons with PINK1 overexpressed or knocked down. As shown in Supplementary Material, Fig. S1, compared with control neurons, neurons transfected with wild-type PINK1 showed reduced diffusion of photo-converted signals. In contrast, neurons transfected with PINK1 RNAi showed increased diffusion of photo-converted signals (relative diffusion rates: control 2.79 + 0.26; PINK1-wt 1.02 + 0.07; PINK1 RNAi 4.31 + 0.40; n ¼ 6). These results suggest that the effect of PINK1 in tipping the balance of mitochondrial fission/fusion dynamics toward more fission is conferred at least in part through its inhibition of fusion.

PINK1 and Parkin functionally interact with Drp1 in rat hippocampal neurons Drp1 is a GTPase that directly promotes mitochondrial fission in mammalian cells (18). To test whether Drp1 genetically interacts with PINK1, neurons were transfected with Drp1-wt or a dominant-negative Drp1 K38A (Drp1-KA), with or without PINK1 shRNA co-transfection. As expected, Drp1-wt overexpression caused mitochondrial fragmentation, whereas Drp1-KA overexpression resulted in mitochondrial elongation, indicating that the balance of mitochondrial


3231

Figure 3. Wild-type Drp1 but not the dominant-negative Drp1K38A promotes mitochondrial fragmentation and suppresses PINK1 RNAi effect. (A) Cultured hippocampal neurons were transfected with Mito-EGFP and HA-tagged monkey Drp1-wt or Drp1 K38A (Drp1-KA). Immunofluorescence was performed with chicken anti-EGFP and rat anti-HA antibodies. Overexpression of Drp1-wt resulted in mitochondrial fragmentation, whereas Drp1 KA led to mitochondrial elongation. In the PINK1 RNAi panels, neurons were co-transfected with PINK1 shRNA lentiviral vector and HA-tagged Drp1-wt or Drp1-KA. (B) Quantification of Mito index, Mito number and Mito size in the dendrites of neurons transfected with the constructs is shown in (A)(∗∗∗ P , 0.001 in Student’s t-test).

fission/fusion dynamics was tipped toward more fission or fusion, respectively (Fig. 3A and B). In PINK1 RNAi neurons, co-expression of Drp1-wt led to mitochondrial fragmentation, whereas mitochondria with elongated and fused morphology persisted in Drp1-KA co-expression neurons (Mito index: control 67.1 + 2.3%; Drp1-wt 46.3 + 2.2%; Drp1 KA 77.6 + 1.9%; Drp1-wt + PINK1 RNAi, 44.8 + 1.8%; Drp1 KA + PINK1 RNAi; 78.6 + 1.6%. Mito number: control 14.4 + 1.4; Drp1-wt 19.9 + 1.3; Drp1 KA 6.5 + 0.5; Drp1-wt + PINK1 RNAi 19.9 + 1.3; Drp1 KA + PINK1 RNAi 5.6 + 0.5. Mito size: control 4.8 + 0.4; Drp1-wt 2.3 + 0.2; Drp1 KA 13.3 + 1.1; Drp1-wt + PINK1 RNAi 2.5 + 0.1; Drp1 KA + PINK1 RNAi 16.7 + 1.9; n ¼ 20). A lentiviral construct engineered to simultaneously express Drp1 shRNA and Mito-DsRed was then used to further examine the effect of Drp1 loss-of-function on mitochondrial morphology. Drp1 RNAi resulted in mitochondrial elongation as expected (Fig. 4A, arrows). Either PINK1 or Parkin overexpression suppressed Drp1 RNAi-induced mitochondrial elongation (Fig. 4A and B; Mito index: control 67.5 + 3.1%; Drp1 RNAi 77.8 + 2.1%; PINK1 + Drp1 RNAi 60.9 + 2.1%; Parkin + Drp1 RNAi 59.7 + 2.4%. Mito number: control 14.7 + 1.1; Drp1 RNAi 5.5 + 0.5; PINK1 + Drp1 RNAi 13.0 + 1.0; Parkin + Drp1 RNAi 15.3 + 1.0. Mito size: control 4.6 + 0.4; Drp1 RNAi 16.7 +

2.2; PINK1 + Drp1 RNAi 5.0 + 0.4; Parkin + Drp1 RNAi 4.1 + 0.3; n ¼ 22). Although a linear genetic relationship between PINK1/Parkin and Drp1 was not evident from these results, their functional genetic interactions do support a role for PINK1/Parkin in mitochondrial dynamics regulation. PINK1 functionally interacts with optical atrophy 1 in rat hippocampal neurons Optical atrophy 1 (OPA1) is a dynamin-like GTPase localized to mitochondrial inner membrane where it directly promotes mitochondrial fusion (19). We used shRNA to knockdown OPA1 function and examined the effects on mitochondrial morphology. Hippocampal neurons were transfected with Mito-EGFP or lentiviral constructs expressing OPA1 shRNA and Mito-DsRed on the same vector (Fig. 4C). In neurons transfected with Mito-EGFP alone (Fig. 4C, arrow heads), mitochondria were well separated and showed the mixed punctate and tubular morphology seen in control neurons (Figs 1A and 2A). In neurons transfected with both OPA1 shRNA and Mito-EGFP, as judged by the expression of both Mito-DsRed and Mito-EGFP, an increase in the number of punctate, smaller-sized mitochondria and a decrease in mitochondrial size and mitochondrial index were observed (Fig. 4C and D). This result is consistent with the pro-fusion

3232


Figure 4. PINK1 and Parkin interact with mitochondrial fission/fusion molecules Drp1 and OPA1. (A) PINK1 or Parkin overexpression can block the Drp1 RNAi effect. Cultured rat hippocampal neurons were transfected with Drp1 RNAi (Mito-DsRed) and PINK1 or Parkin constructs. Immunofluorescence was performed with a chicken anti-EGFP antibody. Arrows show fused mitochondria in a neuron transfected with Drp1 RNAi only. (B) Quantification of Mito index, Mito number and Mito size in the dendrites of neurons transfected with the constructs is shown in (A) (∗∗ P , 0.01; ∗∗∗ P , 0.001 in Student’s t-test). (C) OPA1 RNAi promoted mitochondrial fragmentation and blocked the PINK1 RNAi effect. Arrowheads show normal mitochondrial morphology in a neuron transfected with Mito-EGFP only. (D) Quantification of Mito index, Mito number and Mito size in the dendrites of neurons transfected with the indicated vectors is shown in (C) (∗∗∗ P , 0.001 in Student’s t-test).

function of OPA1. To test whether OPA1 RNAi could rescue the PINK1 RNAi effect, neurons were co-transfected with PINK1 shRNA and OPA1 shRNA that were monitored with the co-expressed Mito-EGFP and Mito-DsRed reporters, respectively. As shown before, in neurons transfected with PINK1 shRNA alone, mitochondria appeared elongated and continuous (Fig. 1A). However, in neurons transfected with both PINK1 shRNA and OPA1 shRNA, mitochondria morphology was restored to normal (Fig. 4C, compare with Figs 1A and 2A control panels; Mito index: control 64.9 +

1.5%; OPA1 RNAi 38.8 + 1.7%; PINK1 RNAi + OPA1 RNAi 62.9 + 1.7%. Mito number: control 13.8 + 0.7; OPA1 RNAi 28.2 + 1.2; PINK1 RNAi + OPA1 RNAi 11.4 + 0.5. Mito size: control 4.8 + 0.2; OPA1 RNAi 1.6 + 0.1; PINK1 RNAi + OPA1 RNAi 5.7 + 0.2; n ¼ 26). This result further supports that PINK1 tips the balance of mitochondrial fission/fusion dynamics toward more fission in mammalian hippocampal neurons. It is also consistent with the notion that mitochondrial morphology is determined by balanced fission/fusion activities. Reducing fission and fusion


3233

Figure 5. PINK1 and Parkin interact with Mfn1 in rat hippocampal neurons. (A) Mitochondrial morphology in neurons transfected with Mfn1-Myc and EGFP, or co-transfected with Parkin-EGFP and Mfn1-Myc or co-transfected with PINK1-HA and Mfn-Myc. (B) Quantification of Mito index, Mito number and Mito size in the dendrites of the transfected neurons is shown in (A)(∗∗∗ P , 0.001 in Student’s t-test).

simultaneously could have little effect on mitochondrial morphology.

PINK1 and Parkin functionally interact with Mfn1 in rat hippocampal neurons Mitofusins (Mfns) are dynamin-like GTPases localized to the outer membrane of mitochondria where they directly promote mitochondrial fusion (20). There are two Mfns in mammalian cells: Mfn1 and Mfn2, with Mfn1 showing general expression and Mfn2 more restricted expression (21,22). To test whether the PINK1/Parkin pathway genetically interacts with the Mfn in mammalian neurons, myc-tagged mouse Mfn1 was expressed in hippocampal neurons together with Mito-EGP. Mfn1 was localized to mitochondria as judged by the perfect co-localization of Mito-EGFP and myc staining (Fig. 5A). When either Mito-EGFP or Mfn-myc was used as a mitochondrial marker to quantify Mito index, Mito number or Mito size, there was no statistical difference in the measurements of these parameters (Mito index: Mito-EGFP 85.8 + 1.8%; Mfn1-myc 82.6 + 1.7%; P ¼ 0.20. Mito number: Mito-EGFP 6.2 + 0.5; Mfn1-myc 6.3 + 0.5; P ¼ 0.83. Mito size: Mito-EGFP 16.3 + 1.5; Mfn1-myc 15.5 + 1.4; P ¼ 0.68). Consistent with a pro-fusion function of Mfn1, Mfn1 overexpression led to more fused mitochondria, compared with neurons transfected with Mito-EGFP only. Interestingly,

co-expression of Parkin-EGFP (Fig. 5A and B) or PINK1-HA (Fig. 5A and B) suppressed Mfn1-induced mitochondrial fusion, indicating that Parkin and PINK1 functionally interact with Mfn1 (Mito index: control 66.3 + 2.1%; Mfn1 82.6 + 1.7%; Mfn1 + EGFP 81.2 + 2.2%; Mfn1 + Parkin-EGFP 53.6 + 2.3; Mfn1 + PINK1 56.5 + 1.4%. Mito number: control 13.5 + 0.8; Mfn1 6.3 + 0.5; Mfn1 + EGFP 6.8 + 0.5; Mfn1 + Parkin-EGFP 24.9 + 1.2; Mfn1 + PINK1 25.3 + 1.3. Mito size: control 5.3 + 0.3; Mfn1 15.5 + 1.4; Mfn1 + EGFP 13.0 + 1.7; Mfn1 + Parkin-EGFP 2.2 + 0.1; Mfn1 + PINK1 2.4 + 0.1; n ¼ 20). Note that since Parkin was tagged with EGFP, we used Mfn1-myc staining instead of Mito-EGFP to measure mitochondrial morphology parameters in the Mfn1-myc and Parkin-EGFP co-expression experiment.

Manipulation of the PINK1/Parkin pathway affects dendritic spine formation in rat hippocampal neurons To determine the functional consequences of altering PINK1/ Parkin pathway activity, we examined dendritic spine morphogenesis in rat hippocampal neurons, as mitochondrial fission/ fusion can influence synaptic morphogenesis and synaptic plasticity in these neurons (23). An EGFP reporter with a membrane-targeting signal (gap-EGFP, Fig. 6B) was expressed in hippocampal neurons together with PINK1-wt,

3234


Figure 6. Effects of altered PINK1 and Parkin activities on dendritic spine formation and vulnerability to excitotoxicity in cultured rat hippocampal neurons. (A) Cultured hippocampal neurons were transfected with gap-EGFP and myc-tagged human PINK1 or Parkin. Immunofluorescence was performed with chicken anti-EGFP and mouse anti-myc antibodies. In the PINK1 RNAi panels, neurons were transfected with the lentiviral vector that can simultaneously express PINK1 shRNA and gap-EGFP. Immunofluorescence was performed with chicken anti-EGFP and rabbit anti-PINK1 antibodies. Arrows in PINK1 RNAi panels mark PINK1 expression in a neuron not transfected with PINK1 shRNA. (B) A diagram of the PINK1 RNAi construct that can simultaneously express PINK1 shRNA and gap-EGFP. (C) Quantification of spine numbers per 10 mm length of dendritic processes in neurons transfected with the constructs shown in (A) (∗∗ P , 0.01; ∗∗∗ P , 0.001 in Student’s t-test). (D) PINK1-RNAi neurons showed increased cell death induced by excitotoxicity when cultured in high K+ condition, whereas PINK1-wt or Parkin-wt transfected neurons showed some resistance to excitotoxicity compared with the control neurons, as measured by both trypan blue exclusion assay and annexin-V labeling assay. Excitotoxicity was blocked by glutamate receptor antagonists CNQX and AP5 (∗∗ P , 0.01; ∗∗∗ P , 0.001 in Student’s t-test).

Parkin-wt or PINK1-RNAi. The gap-EGFP reporter allows examination of neuronal processes such as dendritic spines in great detail. We found that PINK1-wt or Parkin-wt overexpression significantly decreased dendritic spine number, whereas PINK1 RNAi had the opposite effects (Fig. 6A and C; control 5.3 + 0.4; PINK1 RNAi 6.9 + 0.4; PINK1-wt 2.7 + 0.3; Parkin-wt 2.5 + 0.3; n ¼ 30). These results demonstrate that altered PINK1/Parkin activities affect synaptic morphogenesis in rat hippocampal neurons.

PINK1 deficiency increases neuronal vulnerability to excitotoxicity Excessive stimulation through imbalanced neurotransmission could lead to neuronal death found in PD and other neurological diseases. To test whether enhanced spine formation by loss of PINK1 function might confer neuronal vulnerability to excitotoxicity, neurons transfected with lentiviral constructs expressing PINK1-shRNA, PINK1-wt or Parkin-wt were


3235

Figure 7. Effects of altered PINK1 and Parkin activities on mitochondrial morphology in cultured rat midbrain dopaminergic neurons. (A) Cultured dopaminergic neurons were transfected with lentiviral vectors expressing Mito-EGFP and PINK1 shRNA, PINK1-wt or Parkin-wt constructs. Immunofluorescence was performed with a chicken anti-EGFP antibody to visualize mitochondrial morphology and a rabbit anti-TH antibody to identify transfected dopaminergic neurons. (B) Quantification of mitochondrial index in the dopaminergic neurons transfected with the constructs is shown in (A) (∗∗ P , 0.01; ∗∗∗ P , 0.001 in Student’s t-test).

subjected to cell death assays under high potassium condition. As shown in Figure 6D, high potassium induced significantly more neuronal death in PINK1-RNAi neurons compared with control neurons, as measured by both trypan blue exclusion assay and annexin-V labeling assay. Conversely, neurons transfected with PINK1-wt or Parkin-wt showed certain degree of resistance to such excitotoxicity (trypan bluepositive neurons: control 51.5 + 1.1%; PINK1-RNAi 62.7 + 0.9%; PINK1-wt 42.3 + 1.2%; Parkin-wt 41.6 + 1.0%; n ¼ 6. Annexin-V-labeled neurons: control 54.8 + 1.4%; PINK1-RNAi 65.6 + 0.7%; PINK1-wt 45.0 + 1.1%; Parkin-wt 45.2 + 0.9%; n ¼ 6). This effect was abolished when glutamate receptor antagonists (25 mM CNQX, 50 mM D-AP5) were included in the potassium solution (Fig. 6D), confirming that neuronal death was caused by released glutamate acting on postsynaptic receptors (trypan blue-positive neurons: control 9.4 + 1.0%; PINK1-RNAi 11.1 + 1.2%; PINK1-wt 10.5 + 1.0; Parkin-wt 9.7 + 0.7; n ¼ 6. Annexin-V-labeled neurons: control 10.9 + 1.0%; PINK1-RNAi 12.2 + 1.2%; PINK1-wt 12.8 + 1.5; Parkin-wt 11.3 + 1.2; n ¼ 6).

PINK1 and Parkin tip the balance of mitochondrial fission/fusion dynamics toward more fission in rat dopaminergic neurons Dopaminergic neurons in the midbrain region are particularly vulnerable in PD. To test the effect of PINK1/Parkin in these neurons, cultured rat midbrain dopaminergic neurons were transfected with lentiviral vectors expressing Mito-EGFP and PINK1 shRNA, PINK1-wt or Parkin-wt constructs. Similar to the effects seen in hippocampal neurons, lentiviral expression of PINK1-wt or Parkin-wt led to more fragmented mitochondria in these neurons, compared with neurons transfected with Mito-EGFP only, which showed more tubular mitochondria (Fig. 7A and B). Conversely, PINK1 RNAi led to more elongated and fused mitochondria (Fig. 7A and B). Similar trends in changes of mitochondrial number, size and index were also observed (Mito index: control 67.3 + 3.2%; PINK1 RNAi 76.9 + 1.5%; PINK1-wt 52.3 + 2.5%; Parkin-wt 46.3 + 1.9%. Mito number: control 14.8 + 1.3; PINK1 RNAi 6.5 + 0.4; PINK1-wt 18.2 + 1.4. Parkin-wt 20.5 + 1.3. Mito size: control 5.0 + 0.4; PINK1 RNAi 13.5 + 1.4; PINK1-wt

3236


3.0 + 0.1; Parkin-wt 2.5 + 0.1; n ¼ 19). These results demonstrate that the PINK1/Parkin pathway plays similar roles in dopaminergic neurons and hippocampal neurons. Thus, our findings of PINK1/Parkin function in hippocampal neurons are likely applicable to the disease-relevant dopaminergic neurons.

DISCUSSION Mitochondria play crucial roles in eukaryotic cells because they not only serve as cellular energy stores but also execute other cellular functions such as maintaining calcium balance and regulating apoptosis. Neurons have particular demand for mitochondrial function due to their extremely polarized morphology, with multiple dendrites and a disproportionately long axon that separate the energy-consuming nerve terminus from the cell body. In addition to the energy-demanding steps in synaptic transmission, other processes such as synaptic differentiation and plasticity that involve extensive protein synthesis, trafficking, targeting and modifications also have particular requirements for mitochondrial content and motility. It is thus not surprising that mitochondrial dysfunction is associated with a large number of neurodegenerative diseases. However, the cause-and-effect relationship between mitochondrial dysfunction and neurodegeneration remains to be established for most of these diseases. Previous in vivo studies in Drosophila have established a functional relationship between PINK1/Parkin and the mitochondrial fission/fusion pathway, implicating defective mitochondrial dynamics in PD pathogenesis. In this study, using both dopaminergic and hippocampal neurons from mammals, we have demonstrated that the PINK1/Parkin pathway plays evolutionarily conserved roles in regulating mitochondrial dynamics by tipping the fission/fusion balance toward more fission. The proper function of the PINK1/Parkin pathway requires PINK1’s kinase activity and Parkin’s E3 ubiquitin ligase activity and their functional interactions with the fission molecule Drp1 and fusion molecules OPA1 and Mfn1. Parkin has been shown to directly ubiquitinate the Mfn, inhibiting its function in mitochondrial fusion. This molecular function of Parkin also appears to be conserved (24,25). This molecular interaction between PINK1/Parkin and the Mfn could at least partially explain the genetic interactions between PINK1/ Parkin and the mitochondrial fission/fusion machinery observed in this study, as well as the photo-activation data showing inhibition of mitochondrial fusion by PINK1. Moreover, manipulation of the PINK1/Parkin pathway significantly affects synapse morphology. PD thus joins diseases such as optical atrophy and Charcot – Marie– Tooth type 2A to form a category of neurodegenerative diseases in which defective mitochondrial dynamics play a pathogenic role. Previous studies based on mammalian cell lines have produced results different from those reported in the fly models and described here. It is possible that this discrepancy could be attributed to the different cell types used. Essentially, all of the earlier mammalian studies used proliferating cells, not post-mitotic cells. The work in Drosophila has focused on post-mitotic cells (muscle, neurons and sperms), and the work reported here also used post-mitotic neurons. It has

been reported that the cell cycle critically regulates mitochondrial fission and fusion, with filamentous mitochondria present in G1 and early S phases and fragmented mitochondria in late S and M phases (26,27). This is probably due to the phosphorylation of mitochondrial fission/fusion machinery by cell-cycle-specific kinases (28). HeLa, neuroblastoma cell lines or fibroblasts are actively dividing cells. Thus, interpretation of effects on mitochondrial fission and fusion in these cells could be complicated by their cell-cycle stages. In contrast, cultured primary neurons are post-mitotic cells that lack the complications caused by the cell cycle. Interestingly, despite the role of OPA1 in promoting mitochondrial fusion in various cell types reported previously (29) and in this study, OPA1 overexpression in HeLa and COS7 cells could actually promote mitochondrial fission (30,31), suggesting that the fission/fusion molecules may regulate mitochondrial morphology in a cell-type-specific manner. Another concern over the utilization of tumor-derived cell lines is that they can derive ATP primarily from glycolysis (32), unlike other cells in which the oxidative phosphorylation (OXPHOS) system is more involved in ATP production. It is known that mitochondrial morphology could be extensively remodeled to accommodate changes in OXPHOS (32). In the case of neurons, OXPHOS is likely the sole source of ATP production since these cells have limited ability to switch to glycolysis when OXPHOS is compromised. Thus, different modes of metabolism between tumor-derived cell lines and primary neurons could also affect their mitochondrial dynamics. Recent studies, carried out primarily in tumor-derived cell lines, have suggested a role for the PINK1/Parkin pathway in mitochondrial quality control through promotion of mitophagy (33,34). PINK1 is shown to be stabilized on damaged mitochondria, which then recruits Parkin to mitochondrial outer membrane to elicit the removal of damaged mitochondria through autophagy. The pro-fission effect of the PINK1/ Parkin pathway would offer a potential mechanism of their involvement in mitochondrial maintenance via mitophagy. However, we have failed to observe a convincing effect of PINK1/Parkin on mitophagy in cultured rat hippocampal neurons (data not shown), and a recent study also showed that in mammalian neurons, unlike other cells, mitochondrial depolarization by CCCP treatment did not induce Parkin translocation to mitochondria or mitophagy (35), suggesting that additional mechanisms may be involved in regulating mitophagy in neurons. In characterizing mitochondrial morphology under different genetic manipulations, we used parameters such as mitochondrial number, size and index, and genetic interactions with known mitochondrial fission/fusion molecules, to draw conclusions about the roles of PINK1/Parkin in mitochondrial fission/fusion. Since we did not observe obvious changes in the overall distribution of mitochondria between neuronal soma and neuritis, it is unlikely that the phenotypes we observed can be explained by defects in mitochondrial trafficking, although we cannot completely rule out possible contribution by mitochondrial trafficking, since it is known that mitochondrial dynamics and trafficking are linked. Future live imaging experiments will be needed to address this point. In our experiments, we used appropriate PINK1 and Parkin controls to conclude that the effects we observed are


due to the specific activity of PINK1 and Parkin, instead of overexpression artifact. The fact that overexpression of wildtype and pathogenic forms of PINK1 had opposite effects on mitochondrial morphology argues against the overexpression artifact scenario. We did notice that the effect of Drp1 overexpression on mitochondrial index observed in our study is somewhat different from that reported in a previous study (23). This could be due to different culture conditions, the stages of the neurons or the levels of protein expression from transfection. Finally, since we observed similar effects in hippocampal neurons and dopaminergic neurons, our results are likely to be generally applicable to multiple mammalian neuronal types. In this regard, it is worth noting that in PINK1 knockout mice, enlarged mitochondria were observed in striatal neurons (36), and reduced mitochondrial fission was observed in cortical neurons under stress (37), suggesting that the findings in this study are relevant to in vivo conditions. Manipulation of PINK1/Parkin expression also affects dendritic spine numbers, supporting that the PINK1/Pakin pathway regulates synapse morphology and possibly function. Inhibition of Parkin function by RNAi or overexpression of Parkin FPD mutants in hippocampal neurons has been shown to increase excitatory glutamatergic synapses and enhanced synaptic efficacy, whereas the opposite effects have been observed in Parkin overexpression neurons (38). Our results from the manipulations of mammalian PINK1 agree with the previous studies on Parkin. It is not yet clear whether the effect of PINK1/Parkin on mitochondrial fission/ fusion is directly responsible for their effects on spine morphogenesis. In this regard, it is worth noting that although PINK1/ Parkin and Drp1 have similar effects on mitochondrial morphogenesis, their effects on spine morphogenesis are different (23), suggesting that additional pathways may also mediate the effects of PINK1/Parkin on spine morphogenesis. In addition to regulating mitochondrial fission/fusion, Parkin may also counter excitotoxicity by mono-ubiquitinating PICK1 (39). It is thus possible that neuronal Mfns may help maintain a more fused mitochondrial network to meet their high energy demands, and the PINK1/Parkin pathway could keep this process in check by negatively regulating Mfn. Loss of PINK1/Parkin function could therefore, on the one hand, cause imbalanced energy production that could lead to the accumulation of free radicals and ensuing oxidative stress, and, on the other hand, it could cause excitotoxicity due to the deregulation of PICK1 and possibly other molecules. Both effects could contribute to neuronal death as seen in PD.

MATERIALS AND METHODS Hippocampal and dopaminergic neuronal culture Primary rat hippocampal neurons were cultured as described before (40), with certain modifications. Briefly, embryonic E18 neurons from Sprague – Dawley rats were cultured on poly-L-lysine-coated cover glass in plating medium. The next day, plating medium was replaced with medium containing 1% N2 supplement (Invitrogen, Carlsbad, CA, USA). Cytosine arabinoside (1-beta-D-arabinofuranosyl-cytosine, Calbiochem, 5 × 1026 M final concentration) was added to the culture 3 days after plating to inhibit glial cell

3237

proliferation. Cultured 9-day-old neurons were transfected with 2 – 3 mg of plasmid combination per well using CalPhos Mammalian Transfection Kit (Clontech, San Jose, CA, USA). Fluorescence immunocytochemistry was performed 3 days after transfection. For dopaminergic neuronal culture, dissection of mesencephalic region containing the substantia nigra and ventral tegmental area from E18 embryos from Sprague– Dawley rats was performed as described (41). After isolation, the tissues were treated with 50 mg/ml trypsin in Hank’s balanced salt solution without calcium and magnesium for 30 min at room temperature. The dissociated cells were cultured on poly-L-lysine-coated cover glass in plating medium (DMEM medium supplemented with 10% horse serum). The next day, plating medium was replaced with medium containing 10% fetal calf serum and 1% N2 supplement (Invitrogen). Cytosine arabinoside was added to the culture 3 days after plating to inhibit glial cell proliferation. Lentiviral particle production and transfection Preparation and concentration of lentiviral particles were performed according to a previously described protocol (42). Viral transfection of hippocampal and dopaminergic neurons was carried out at 9 days after in vitro culture. The next day, the transfection medium was replaced with a mixture of 50% conditioned medium and 50% fresh medium. DNA constructions PINK1-wt and K219M and L347P mutants (16) with HA tag fused at the C-terminus were inserted into the lentiviral vector FHUGW (43). Human PINK1 cDNA and human Parkin cDNA in pcDNA3 vector were described before (6). Human Parkin-wt, Parkin-mut and Parkin-DC with myc tag fused at the C-terminus were inserted into FHUGW. Human Parkin-EGFP was constructed by inserting full-length Parkin cDNA in frame with EGFP into FHUGW. HA-tagged rhesus monkey Drp1-wt and dominant-negative Drp1-K38A in pcDNA3 were from Dr A.M. van der Bliek (18). Myc-tagged mouse Mfn1 in pcDNA3 was obtained from Dr David Chan (44). Rat OPA1 cDNA cloned in pExpress-1 was purchased from Open Biosystem and subcloned into FHUGW with a Myc-tag at its C-terminus. Lentiviral vectors expressing Mito-EGFP or Mito-DsRed were constructed by replacing EGFP in FHUGW with Mito-EGFP or Mito-DsRed between the XbaI and BsrGI sites. Lentiviral vectors expressing gap-EGFP was constructed by inserting into FHUGW a fragment containing gap-EGFP from the LZRS-CA-GAP-EGFP plasmid (45). shRNAs were targeting rat gene sequences listed here: PINK1: CCAGCGAAGCCATCTTAAGCA (GI: 157817866, 719– 739); Drp1: GCAGAAGAATGGGGTAAAT (GI: 171460913, 259– 277); OPA1: AAGTCATCAGTCTGAGCCAGGTT (GI: 148747458, 1941 – 1963). Mito-Dendra2 was obtained from Dr Xiongwei Zhu (46). The lentiviral vector simultaneously expressing PINK1 shRNA and

3238


Mito-Dendra2 was constructed by inserting inverted repeats of PINK1 shRNA target sequence and Mito-Dendra2 into FHUGW. Transfection of human embryonic kidney 293 cells and immunoprecipitation Human embryonic kidney 293 (HEK293) cells were cultured in six-well plates and transfected with 4 mg of plasmid mixed with FuGENE 6 Transfection Reagent (Roche, Palo Alto, CA, USA) according to manufacturer’s instructions. Immnoprecipitation from HEK293 cells was performed as described before (47), with 20 ml of EZview Red Anti-c-Myc Affinity Gel (Sigma-Aldrich, St Louis, MO, USA) for each sample used in this experiment. The bound proteins were eluted from the affinity gel by incubation with 2× sample buffer (125 mM Tris– HCl, pH 6.8, 20% glycerol, 10% b-mercaptoethanol and 4% SDS) at 1008C for 5 min, subjected to SDS – polyacrylamide gel electrophoresis, followed by immunoblotting with a mouse anti-Myc antibody (1:3000, Millipore, Bedford, MA, USA) or a rat anti-HA antibody (1:500, Roche). Samples were detected by chemiluminescence with the Amersham ECL Plus Western Blotting Detection System (GE Healthcare). Immunofluorescence Chicken anti-EGFP (1:3000, Abcam, Cambridge, MA, USA), mouse anti-Myc (1:1500, 4A6, Millipore), rabbit anti-Myc (1:600, Cell Signaling, Danvers, MA, USA), rabbit antiPINK1 (1:300, Abcam), rat anti-HA (1:200, Roche) and rabbit anti-TH (1:2000) (5) were used in the immunofluorescence study. Immunofluorescence analysis was done essentially as described (47). Excitotoxicity assay DIV14 hippocampal neurons were infected with lentiviral constructs expressing Mito-EGFP, PINK1RNAi-Mito-EGFP, PINK1wt or Parkin-wt. Seven days later, neurons were transferred into high K+ solution (90 mM KCl, 31.5 mM NaCl, 2 mM CaCl2, 25 mM HEPES, 1 mM glycine, 30 mM glucose) for 10 min and then incubated in conditioned medium for 2 h. Glutamate receptor antagonists were used in parallel experiments at the following concentrations: CNQX (6-cyano7-nitroquinoxaline-2,3-dione, Sigma-Aldrich; 25 mM), D-AP5 (Sigma-Aldrich; 50 mM). Neuron viability was assayed with 0.4% trypan blue or annexin-V-Alexa 568 (Invitrogen) labeling in conditioned medium via microscopy. Approximately 200 – 300 neurons were scored from each cover slip, and 6 cover slips were scored from 3 independent cultures.

Mito-EGFP or Mito-DsRed along the dendrites with either the LSM Image Browser (Version 4.2.0.121, Zeiss) or the LAS AF Lite (Version 2.0.2, build 2038, Leica) software. The measurements were consistently done on dendrites 20– 150 mm away from the soma to avoid overlapping mitochondria at the proximal dendrites. The individual mitochondrial segments were defined as units that are at least 0.1 mm apart from each other. The length of each mitochondrial segments and the length of the corresponding dendrite were exported to Microsoft Excel and the following calculations were based on these raw data: mitochondrial index (Mito index) was calculated as summed mitochondrial length within a dendrite divided by the corresponding dendrite length and presented as percentage (mean + SEM). Mitochondrial number was calculated as total numbers of mitochondrial segments per 100 mm dendrite length (mean + SEM). Mitochondrial segment size (Mito size) was calculated as the average length of each mitochondrial segment within dendrites (mean + SEM). Dendritic spine number was manually counted as the number of spines along the 10 mm length of dendritic process and presented as mean + SEM. Each experiment has been done three times and a total of 9 – 12 neurons were randomly selected. The number of dendrites selected for quantification was in the range of 15– 30 and was indicated in the text. All analyses were done in a manner blind to the conditions.

Live cell imaging and photo activation Hippocampal neurons cultured in glass bottom dishes (In Vitro Scientific, Sunnyvale, CA, USA) were transfected with Mito-Dendra2, Mito-Dendra2 + PINK1-wt or lentiviral vector expressing both Mito-Dendra2 and PINK1 shRNA. Three days after transfection, imaging of live neurons was performed with a Zeiss LSM510 Meta inverted confocal microscope equipped with the Zen 2009 software and environmental control chamber (5% CO2, 378C and humidity). Mito-Dendra2 conversion was achieved by the scanning of selected mitochondria with a 405 nm laser (30% intensity) for five iterations. Time-lapse recordings were performed every 5 min. Images were exported from Zen 2009 and analyzed with Adobe Photoshop CS4.

SUPPLEMENTARY MATERIAL Supplementary Material is available at HMG online.

ACKNOWLEDGEMENTS Image acquisition, analysis and quantification Confocal images were collected using a Zeiss inverted LSM510 confocal microscope (Carl Zeiss, Inc.) or Leica TCS SP5 confocal microscope with 63× (NA 1.4) objective at 0.5 mm/step through the z-dimension. Processed confocal images were used to make 2D projections. Mitochondrial length was measured by tracing the fluorescence of

We thank Drs David Baltimore, David Chan, Nobutaka Hattori, Chenjian Li, Robert Malenka, Yusuke Nakamura, Ami Okada, Alexander M. van der Blie, Weifeng Xu, Richard Youle and Xiongwei Zhu for reagents, and members of the Lu Laboratory for discussions. Conflict of Interest statement. None declared.


FUNDING This work is supported by grants from the McKnight Foundation (Brain Disorders Award to B.L.) and the National Institutes of Health (R01AR054926 and R01MH080378 to B.L. and DA023904 to S.G.), CIRM (RS1-00215-1 to S.G.), and a Stanford University School of Medicine Dean’s Postdoctoral Fellowship (to W.Y.).

17.

18. 19.

REFERENCES 1. Dawson, T.M., Ko, H.S. and Dawson, V.L. (2010) Genetic animal models of Parkinson’s disease. Neuron, 66, 646– 661. 2. Lu, B. (2009) Mitochondrial dynamics and neurodegeneration. Curr. Neurol. Neurosci. Rep., 9, 212 –219. 3. Clark, I.E., Dodson, M.W., Jiang, C., Cao, J.H., Huh, J.R., Seol, J.H., Yoo, S.J., Hay, B.A. and Guo, M. (2006) Drosophila pink1 is required for mitochondrial function and interacts genetically with parkin. Nature, 441, 1162– 1166. 4. Park, J., Lee, S.B., Lee, S., Kim, Y., Song, S., Kim, S., Bae, E., Kim, J., Shong, M., Kim, J.M. et al. (2006) Mitochondrial dysfunction in Drosophila PINK1 mutants is complemented by parkin. Nature, 441, 1157– 1161. 5. Yang, Y., Gehrke, S., Imai, Y., Huang, Z., Ouyang, Y., Wang, J.W., Yang, L., Beal, M.F., Vogel, H. and Lu, B. (2006) Mitochondrial pathology and muscle and dopaminergic neuron degeneration caused by inactivation of Drosophila Pink1 is rescued by Parkin. Proc. Natl Acad. Sci. USA, 103, 10793– 10798. 6. Yang, Y., Ouyang, Y., Yang, L., Beal, M.F., McQuibban, A., Vogel, H. and Lu, B. (2008) Pink1 regulates mitochondrial dynamics through interaction with the fission/fusion machinery. Proc. Natl Acad. Sci. USA, 105, 7070– 7075. 7. Deng, H., Dodson, M.W., Huang, H. and Guo, M. (2008) The Parkinson’s disease genes pink1 and parkin promote mitochondrial fission and/or inhibit fusion in Drosophila. Proc. Natl Acad. Sci. USA, 105, 14503– 14508. 8. Park, J., Lee, G. and Chung, J. (2009) The PINK1-Parkin pathway is involved in the regulation of mitochondrial remodeling process. Biochem. Biophys. Res. Commun., 378, 518–523. 9. Poole, A.C., Thomas, R.E., Andrews, L.A., McBride, H.M., Whitworth, A.J. and Pallanck, L.J. (2008) The PINK1/Parkin pathway regulates mitochondrial morphology. Proc. Natl Acad. Sci. USA, 105, 1638–1643. 10. Exner, N., Treske, B., Paquet, D., Holmstrom, K., Schiesling, C., Gispert, S., Carballo-Carbajal, I., Berg, D., Hoepken, H.H., Gasser, T. et al. (2007) Loss-of-function of human PINK1 results in mitochondrial pathology and can be rescued by parkin. J. Neurosci., 27, 12413– 12418. 11. Dagda, R.K., Cherra, S.J. III, Kulich, S.M., Tandon, A., Park, D. and Chu, C.T. (2009) Loss of PINK1 function promotes mitophagy through effects on oxidative stress and mitochondrial fission. J. Biol. Chem., 284, 13843– 13855. 12. Lutz, A.K., Exner, N., Fett, M.E., Schlehe, J.S., Kloos, K., Lammermann, K., Brunner, B., Kurz-Drexler, A., Vogel, F., Reichert, A.S. et al. (2009) Loss of parkin or PINK1 function increases Drp1-dependent mitochondrial fragmentation. J. Biol. Chem., 284, 22938– 22951. 13. Sandebring, A., Thomas, K.J., Beilina, A., van der Brug, M., Cleland, M.M., Ahmad, R., Miller, D.W., Zambrano, I., Cowburn, R.F., Behbahani, H. et al. (2009) Mitochondrial alterations in PINK1 deficient cells are influenced by calcineurin-dependent dephosphorylation of dynamin-related protein 1. PLoS One, 4, e5701. 14. Mortiboys, H., Thomas, K.J., Koopman, W.J., Klaffke, S., Abou-Sleiman, P., Olpin, S., Wood, N.W., Willems, P.H., Smeitink, J.A., Cookson, M.R. et al. (2008) Mitochondrial function and morphology are impaired in parkin-mutant fibroblasts. Ann. Neurol., 64, 555–565. 15. Hatano, Y., Li, Y., Sato, K., Asakawa, S., Yamamura, Y., Tomiyama, H., Yoshino, H., Asahina, M., Kobayashi, S., Hassin-Baer, S. et al. (2004) Novel PINK1 mutations in early-onset parkinsonism. Ann. Neurol., 56, 424–427. 16. Liu, W., Vives-Bauza, C., Acin-Perez, R., Yamamoto, A., Tan, Y., Li, Y., Magrane, J., Stavarache, M.A., Shaffer, S., Chang, S. et al. (2009) PINK1 defect causes mitochondrial dysfunction, proteasomal deficit and

20. 21. 22. 23. 24. 25.

26.

27.

28. 29. 30.

31.

32.

33. 34. 35.

36. 37.

3239

alpha-synuclein aggregation in cell culture models of Parkinson’s disease. PLoS One, 4, e4597. Schlehe, J.S., Lutz, A.K., Pilsl, A., Lammermann, K., Grgur, K., Henn, I.H., Tatzelt, J. and Winklhofer, K.F. (2008) Aberrant folding of pathogenic Parkin mutants: aggregation versus degradation. J. Biol. Chem., 283, 13771–13779. Smirnova, E., Shurland, D.L., Ryazantsev, S.N. and van der Bliek, A.M. (1998) A human dynamin-related protein controls the distribution of mitochondria. J. Cell Biol., 143, 351– 358. Delettre, C., Lenaers, G., Griffoin, J.M., Gigarel, N., Lorenzo, C., Belenguer, P., Pelloquin, L., Grosgeorge, J., Turc-Carel, C., Perret, E. et al. (2000) Nuclear gene OPA1, encoding a mitochondrial dynamin-related protein, is mutated in dominant optic atrophy. Nat. Genet., 26, 207–210. Santel, A. and Fuller, M.T. (2001) Control of mitochondrial morphology by a human mitofusin. J. Cell Sci., 114, 867 –874. Eura, Y., Ishihara, N., Yokota, S. and Mihara, K. (2003) Two mitofusin proteins, mammalian homologues of FZO, with distinct functions are both required for mitochondrial fusion. J. Biochem., 134, 333– 344. Santel, A., Frank, S., Gaume, B., Herrler, M., Youle, R.J. and Fuller, M.T. (2003) Mitofusin-1 protein is a generally expressed mediator of mitochondrial fusion in mammalian cells. J. Cell Sci., 116, 2763– 2774. Li, Z., Okamoto, K., Hayashi, Y. and Sheng, M. (2004) The importance of dendritic mitochondria in the morphogenesis and plasticity of spines and synapses. Cell, 119, 873 –887. Ziviani, E., Tao, R.N. and Whitworth, A.J. (2010) Drosophila parkin requires PINK1 for mitochondrial translocation and ubiquitinates mitofusin. Proc. Natl Acad. Sci. USA, 107, 5018–5023. Gegg, M.E., Cooper, J.M., Chau, K.Y., Rojo, M., Schapira, A.H. and Taanman, J.W. (2010) Mitofusin 1 and mitofusin 2 are ubiquitinated in a PINK1/parkin-dependent manner upon induction of mitophagy. Hum. Mol. Gen., 19, 4861–4870. Margineantu, D.H., Gregory Cox, W., Sundell, L., Sherwood, S.W., Beechem, J.M. and Capaldi, R.A. (2002) Cell cycle dependent morphology changes and associated mitochondrial DNA redistribution in mitochondria of human cell lines. Mitochondrion, 1, 425–435. Mitra, K., Wunder, C., Roysam, B., Lin, G. and Lippincott-Schwartz, J. (2009) A hyperfused mitochondrial state achieved at G1-S regulates cyclin E buildup and entry into S phase. Proc. Natl Acad. Sci. USA, 106, 11960– 11965. Taguchi, N., Ishihara, N., Jofuku, A., Oka, T. and Mihara, K. (2007) Mitotic phosphorylation of dynamin-related GTPase Drp1 participates in mitochondrial fission. J. Biol. Chem., 282, 11521– 11529. Cipolat, S., Martins de Brito, O., Dal Zilio, B. and Scorrano, L. (2004) OPA1 requires mitofusin 1 to promote mitochondrial fusion. Proc. Natl Acad. Sci. USA, 101, 15927–15932. Griparic, L., van der Wel, N.N., Orozco, I.J., Peters, P.J. and van der Bliek, A.M. (2004) Loss of the intermembrane space protein Mgm1/OPA1 induces swelling and localized constrictions along the lengths of mitochondria. J. Biol. Chem., 279, 18792–18798. Misaka, T., Miyashita, T. and Kubo, Y. (2002) Primary structure of a dynamin-related mouse mitochondrial GTPase and its distribution in brain, subcellular localization, and effect on mitochondrial morphology. J. Biol. Chem., 277, 15834– 15842. Rossignol, R., Gilkerson, R., Aggeler, R., Yamagata, K., Remington, S.J. and Capaldi, R.A. (2004) Energy substrate modulates mitochondrial structure and oxidative capacity in cancer cells. Cancer Res., 64, 985– 993. Narendra, D., Tanaka, A., Suen, D.F. and Youle, R.J. (2008) Parkin is recruited selectively to impaired mitochondria and promotes their autophagy. J. Cell Biol., 183, 795–803. Narendra, D.P., Jin, S.M., Tanaka, A., Suen, D.F., Gautier, C.A., Shen, J., Cookson, M.R. and Youle, R.J. (2010) PINK1 is selectively stabilized on impaired mitochondria to activate Parkin. PLoS Biol., 8, e1000298. Van Laar, V.S., Arnold, B., Cassady, S.J., Chu, C.T., Burton, E.A. and Berman, S.B. (2011) Bioenergetics of neurons inhibit the translocation response of Parkin following rapid mitochondrial depolarization. Hum. Mol. Genet., 20, 927–40. Gautier, C.A., Kitada, T. and Shen, J. (2008) Loss of PINK1 causes mitochondrial functional defects and increased sensitivity to oxidative stress. Proc. Natl Acad. Sci. USA, 105, 11364–11369. Gispert, S., Ricciardi, F., Kurz, A., Azizov, M., Hoepken, H.H., Becker, D., Voos, W., Leuner, K., Muller, W.E., Kudin, A.P. et al. (2009)

3240

38. 39.

40. 41. 42.


Parkinson phenotype in aged PINK1-deficient mice is accompanied by progressive mitochondrial dysfunction in absence of neurodegeneration. PLoS One, 4, e5777. Helton, T.D., Otsuka, T., Lee, M.C., Mu, Y. and Ehlers, M.D. (2008) Pruning and loss of excitatory synapses by the parkin ubiquitin ligase. Proc. Natl Acad. Sci. USA, 105, 19492–19497. Joch, M., Ase, A.R., Chen, C.X., MacDonald, P.A., Kontogiannea, M., Corera, A.T., Brice, A., Seguela, P. and Fon, E.A. (2007) Parkin-mediated monoubiquitination of the PDZ protein PICK1 regulates the activity of acid-sensing ion channels. Mol. Biol. Cell, 18, 3105– 3118. Kaech, S. and Banker, G. (2006) Culturing hippocampal neurons. Nat. Protoc., 1, 2406– 2415. Fasano, C., Thibault, D. and Trudeau, L.E. (2008) Culture of postnatal mesencephalic dopamine neurons on an astrocyte monolayer. Curr. Protoc. Neurosci., 3, 21. Salmon, P. and Trono, D. (2007) Production and titration of lentiviral vectors. Curr. Protoc. Hum. Genet., 12, 10.

43. Xu, W., Schluter, O.M., Steiner, P., Czervionke, B.L., Sabatini, B. and Malenka, R.C. (2008) Molecular dissociation of the role of PSD-95 in regulating synaptic strength and LTD. Neuron, 57, 248– 262. 44. Koshiba, T., Detmer, S.A., Kaiser, J.T., Chen, H., McCaffery, J.M. and Chan, D.C. (2004) Structural basis of mitochondrial tethering by mitofusin complexes. Science, 305, 858–862. 45. Okada, A., Lansford, R., Weimann, J.M., Fraser, S.E. and McConnell, S.K. (1999) Imaging cells in the developing nervous system with retrovirus expressing modified green fluorescent protein. Exp. Neurol., 156, 394–406. 46. Wang, X., Su, B., Siedlak, S.L., Moreira, P.I., Fujioka, H., Wang, Y., Casadesus, G. and Zhu, X. (2008) Amyloid-beta overproduction causes abnormal mitochondrial dynamics via differential modulation of mitochondrial fission/fusion proteins. Proc. Natl Acad. Sci. USA, 105, 19318– 19323. 47. Yu, W. and De Blas, A.L. (2008) Gephyrin expression and clustering affects the size of glutamatergic synaptic contacts. J. Neurochem., 104, 830 –845.