p150glued-Associated Disorders Are Caused by Activation of Intrinsic Apoptotic Pathway

Mutations in p150glued cause hereditary motor neuropathy with vocal cord paralysis (HMN7B) and Perry syndrome (PS). Here we show that both overexpression of p150glued mutants and knockdown of endogenous p150glued induce apoptosis. Overexpression of a p150glued plasmid containing either a HMN7B or PS mutation resulted in cytoplasmic p150glued-positive aggregates and was associated with cell death. Cells containing mutant p150glued aggregates underwent apoptosis that was characterized by an increase in cleaved caspase-3- or Annexin V-positive cells and was attenuated by both zVAD-fmk (a pan-caspase inhibitor) application and caspase-3 siRNA knockdown. In addition, overexpression of mutant p150glued decreased mitochondrial membrane potentials and increased levels of translocase of the mitochondrial outer membrane (Tom20) protein, indicating accumulation of damaged mitochondria. Importantly, siRNA knockdown of endogenous p150glued independently induced apoptosis via caspase-8 activation and was not associated with mitochondrial morphological changes. Simultaneous knockdown of endogenous p150glued and overexpression of mutant p150glued had additive apoptosis induction effects. These findings suggest that both p150glued gain-of-toxic-function and loss-of-physiological-function can cause apoptosis and may underlie the pathogenesis of p150glued-associated disorders.


Introduction
The dynactin subunit p150 glued is encoded by the DCTN1 gene. Mutations in this gene have been detected in patients with slowly progressive autosomal dominant distal hereditary motor neuropathy with vocal cord paralysis (HMN7B) and autosomal dominant Perry syndrome (PS), the latter of which is characterized by rapidly progressive, devastating neurodegeneration of dopaminergic neurons in the substantia nigra [1,2].
Dynactin has various molecular functions including minus-end vesicular transport, protein degradation, and cell division. p150 glued is the largest polypeptide of the dynactin complex, and it binds directly to microtubules and to cytoplasmic dynein. Disruption of the p150 glued CAP-Gly domain in neurons causes insufficient retrograde axonal transport [3,4]. Transgenic mice expressing p150 glued with a G59S mutation develop progressive degeneration of motor neurons similar to that seen in amyotrophic lateral sclerosis [5][6][7]. The mutated p150 glued polypeptide that causes PS is unable to bind to microtubules and forms intracytoplasmic aggregates. These aggregates include abnormally accumulated mitochondria [11]. Despite these findings, it is unclear whether decreased levels of endogenous p150 glued or increased levels of the mutant form dominantly contribute to the neurodegeneration seen in PS.
Here we report that knockdown of endogenous p150 glued and overexpression of p150 glued with pathogenic HMN7B or PS mutations independently induced apoptosis. However, only overexpression of mutant forms of p150 glued induced intracytoplasmic p150 glued -aggregates and accumulation of damaged mitochondria, resulting in intrinsic apoptosis induction. Importantly, mutant p150 glued overexpression with endogenous p150 glued knockdown showed additive effects on apoptosis induction, suggesting that both a gain-and loss-of-function contribute to the disease pathogenesis.

Cells overexpressing various p150 glued mutants produce cytoplasmic aggregates
To investigate the effects of overexpression of mutant p150 glued , we first generated plasmid DNAs encoding GFP-or 3xFLAGtagged wild-type (WT) and mutant p150 glued with each pathogenic mutation: G59S, which causes HMN7B, and G71A, G71E, G71R, T72P, or Q74P, which cause PS. All of these mutations are within the p150 glued CAP-Gly microtubule binding domain ( Figure 1A).
To determine if a mutation in p150 glued affected its intracellular localization, we transfected GFP-tagged WT or mutant p150 glued into HeLa cells followed by immunocytochemical analysis. HeLa cells overexpressing GFP-WT p150 glued showed complete colocalization with tubulin ( Figure 1B). By contrast, those with a pathogenic mutation were diffusely distributed in the cytoplasm and showed no apparent colocalization with tubulin ( Figure S1A). Additionally, cytoplasmic, but not nuclear, aggregates were observed in cells with high expression levels of the mutant p150 glued plasmids as early as 24 h after transfection, most frequently in the perinuclear region of the cells with G59S p150 glued ( Figure 1B, C). These findings are consistent with previous reports [8,11]. Analogous results were detected with the overexpression of 3xFLAG-tagged WT and mutant p150 glued in SH-SY5Y ( Figure S1B) and HeLa cells ( Figure S1C, D). Previous studies examining the overexpression of mutant G59S p150 glued showed decreased affinity of the mutant form of p150 glued for microtubules, indicating that mutant p150 glued dissociated from microtubules and formed aggregates [11].
To confirm the formation of cytoplasmic aggregates, we performed conventional electron microscopy (EM) analysis. High-density aggregates around the nuclei were detected in cells overexpressing G59S or G71R p150 glued ( Figure S1E). Next, using immuno-EM analysis with anti-GFP antibodies to recognize GFPtagged mutant p150 glued , we detected p150 glued localized in high density aggregates, particularly in the perinuclear region of the cells overexpressing G59S ( Figure 1D, E) or G71R (data not shown) p150 glued . Unfortunately, because of the fixation technique for immuno-EM, we could not use the same specimens to assess morphological changes in organelles, including mitochondria, in the cells that showed the aggregates.
We next sought to determine the characteristics of the aggregates by immunocytochemistry. The mutant p150 glued aggregates were partially positive for endogenous ubiquitin but not for FLAG-tagged TAR DNA-binding protein 43 (TDP-43) ( Figure S1F, G). This is consistent with previous reports showing that dynactin subunits p50 and p62 were present in less than 5% of TDP-43-expressing neurons in the globus pallidus of the autopsied brain of a PS patient [8].

Apoptotic changes occurred in cells with cytoplasmic aggregates
To elucidate the pathogenesis of the p150 glued -associated diseases, we focused on the association of cytoplasmic aggregates induced by mutant p150 glued overexpression with cell death. We tested the death rate of cells expressing the GFP-tagged p150 glued , assessed by nuclear morphological changes described in a previous report [12]. The rate of cell death was significantly increased by overexpression of G59S or G71R p150 glued both 24 and 48 h after transfection when compared with control cells (Figure 2A).
To examine the characteristics of the cell death induced by G59S or G71R p150 glued , we performed the same analysis 24 h after transfection with controls cells and cells treated with the pancaspase inhibitor carbobenzoxy-valyl-alanyl-aspartyl-[O-methyl]fluoromethylketone (z-VAD-fmk). The rate of cell death induced by overexpression of G59S or G71R p150 glued was significantly suppressed by z-VAD ( Figure 2B) [13]. In a population of cells selected based on GFP (p150 glued ) intensity, the numbers of early apoptotic cells (annexin V-positive, propidium iodide (PI)-negative) and late apoptotic or necrotic cells (annexin V-positive, PIpositive) were both increased by overexpression of G59S or G71R p150 glued ( Figure 2C, D). Likewise, z-VAD treatment significantly decreased cell death in both G59S and G71R p150 gluedoverexpressing cells.
To examine the activation of the intrinsic apoptotic pathway, we determined whether or not cells with aggregates were positive for cleaved caspase-3 using fluorescent-activated cell sorting (FACS) and immunocytochemistry analyses. The number of cells positive for both cleaved caspase-3 and GFP (p150 glued ) in cells overexpressing G59S or G71R p150 glued was markedly increased compared with control cells ( Figure 2E, F). We found that siRNA knockdown against caspase-3 blocked the increase of cell death caused by the overexpression of the mutant p150 glued ( Figure 2G, H). Next, we wanted to rule out the possibility that extrinsically induced apoptosis via caspase-8 cleavage was causing some or all of the cell death seen in these experiments. Therefore, to exclude this possibility, we examined whether siRNA knockdown of caspase-8 inhibited the apoptosis induced by overexpression of mutant p150 glued . Knockdown of caspase-8 did not inhibit apoptosis induced by overexpression of G59S p150 glued ( Figure  S2), suggesting that extrinsically induced apoptosis is not the cause of the cell death seen in these cells.
Aggregate formation caused by the G59S mutant led to cell death. Both aggregate formation and cell death are inhibited by overexpression of Hsp70, a molecular chaperone. These findings suggest that mutant p150 glued aggregates play an important role in the mechanism of cell death in HMN7B [11]. We conclude that mutant p150 glued aggregates cause apoptosis via activation of the intrinsic apoptotic pathway.
Cells with cytoplasmic aggregates have more mitochondria with reduced mitochondrial membrane potentials Levy et al. showed that mutant p150 glued aggregates are usually associated with mitochondria [11]. Therefore, we hypothesized that an accumulation of damaged mitochondria may cause apoptosis. Live-cell imaging analysis in cells overexpressing WT or mutant p150 glued detected elongated tubular mitochondria in control cells, while cells overexpressing mutant p150 glued mainly showed fragmented mitochondria in the vicinity of the nuclei ( Figure 3A). Overexpression of G59S or G71R p150 glued also increased the expression levels of Tom20, a mitochondrial outer membrane protein that is commonly used for assessing mitochondria numbers ( Figure 4B, C) [14][15][16].
To determine the health status of the accumulated mitochondria, we next analyzed cells stained with MitoTracker-Red CMXRos by FACS analysis. Only intact mitochondria with preserved respiration activities and membrane potentials absorb this dye. For an accurate assessment, we only analyzed cells that expressed a high GFP intensity, determined using a flow cytometer. Intriguingly, we detected a marked increase in the number of cells with decreased uptake of MitoTracker-Red CMXRos in cells overexpressing G59S or G71R p150 glued compared with the control cells ( Figure 3D, E). Also, overexpression of WT p150 glued decreased mitochondrial membrane potentials, which might be associated with insufficient mitochondria dynamics [11]. Based on the collected results, we conclude that mutated p150 glued causes the accumulation of damaged mitochondria, which is followed by activation of the intrinsic apoptotic pathway. p150 glued knockdown does not affect mitochondrial membrane potentials and activates apoptotic pathway via caspase-8 cleavage Next, we tested whether or not WT p150 glued siRNA knockdown affects mitochondrial functions in a manner similar to mutant p150 glued overexpression. As shown in Figure 4A-D, total levels of Tom20 and mitochondria complex I were not changed and the levels of damaged mitochondria without MitoTracker-Red CMX-Ros intake were not significantly increased by p150 glued knockdown ( Figure 4E, F). Next, we examined caspase-8 activation because of its association with the extrinsic apoptotic pathway. As shown in Figure 5A-H, the levels of total caspase-8 and caspase-3 were decreased with p150 glued knockdown, whereas the levels of their cleaved forms of caspase-8 and PARP were increased. This suggests that p150 glued knockdown activated caspase-8, leading to caspase-3 activation. Accordingly, treatment with a caspase-8 inhibitor suppressed caspase-3 activation ( Figure 5I, J), and caspase-8 siRNA knockdown also decreased apoptotic cell death ( Figure 5K, L). Taken together, these data show that the loss-of-function of endogenous p150 glued significantly activates caspase-8, inducing apoptosis.
Cells with mutant p150 glued overexpression and wildtype endogenous p150 glued knockdown showed more apoptosis Finally, to address the pathogenesis of the p150 glued -associated disorders more precisely, we performed mutant p150 glued overexpression experiments with or without siRNA knockdown against endogenous p150 glued . As shown in Figure 6, siRNA knockdown of endogenous p150 glued along with overexpression of either the G59S or G71R mutant form caused many more cells to display early apoptotic changes (GFP-positive and Annexin V-positive) compared with the cells overexpressing a p150 glued mutant and a control siRNA. Therefore, we concluded that both excess levels of mutant p150 glued and decreased levels of endogenous p150 glued contribute to the pathogenesis of p150 glued -associated disorders via the activation of apoptosis.

Discussion
In this study, we sought to determine the pathogenesis of p150 glued -associated diseases caused by p150 glued mutations. Overexpression of mutant p150 glued in HeLa and SH-SY5Y cells induced p150 glued -positive aggregate formation, accumulation of damaged mitochondria, and activation of the intrinsic apoptotic pathway. Endogenous p150 glued knockdown in the same cell lines also activated a caspase-8-dependent apoptotic pathway without apparent mitochondrial abnormalities. Importantly, cell death induced by p150 glued knockdown was markedly enhanced by simultaneous overexpression of mutant p150 glued , suggesting the disease pathogenesis may be associated with both p150 glued gainof-toxic-function and loss-of-function.
All of the HMN7B and PS associated mutations are located within the CAP-Gly microtubule domain [8], and various reports have suggested that mutant p150 glued proteins have the tendency to lose their affinity to microtubules [4,[8][9][10]. Our studies support these reports as we observed decreased colocalization of mutant p150 glued with microtubules as well as increased intracytoplasmic aggregates in our immunocytochemistry experiments. However, the in vivo binding activity changes of mutant p150 glued remain unclear. Therefore, further studies should be performed to determine precisely how mutant p150 glued proteins are detoured from their original distribution pattern and how they form aggregates.
The dynein complex plays various critical roles in mitochondrial function (such as retrograde transport and fission/fusion) [17][18][19][20][21]. In this study, we could detect mitochondrial abnormalities (loss of membrane potential and morphological abnormalities) only in cells that expressed mutant forms of p150 glued . A report by Varadi et al. found that disruption of dynein function by either p50 overexpression or microinjection of anti-dynein intermediate chain antibodies led to mitochondrial morphology and distribution changes [22]. Our data, however, showed that p150 glued siRNA knockdown did not induce mitochondrial abnormalities. Only mutant p150 glued overexpression led to these abnormalities, implying that each dynein subunit might have a specific association with mitochondrial function.
Abnormal protein accumulation has been implicated in the pathogenesis of various neurodegenerative diseases [23]. In this study, we revealed that mutant p150 glued (G59S, G71R) overexpression induced aggregate formation and caspase activation associated with mitochondrial abnormalities. This supports the findings by Levy et al. who reported that aggregate formation by the G59S mutant leads to cell death, and that both aggregate formation and the induced cell death were inhibited by overexpression of Hsp70, a molecular chaperone [11]. Likewise, other studies have shown that overexpression of a p150 glued plasmid with a truncated C-terminal as well as knockdown of endogenous p150 glued in rat hippocampal neurons induces caspase-3-positive cell death, which is consistent with our p150 glued knockdown results [24,25].
According to studies with in vivo models, neither expression of a mutant nor the DCAP-Gly domain of p150 glued affects axonal transport, but WT p150 glued is needed for initiation of retrograde transport at synaptic termini in Drosophila motor neurons and mouse dorsal root ganglion neurons [3,4]. Knock-in and transgenic mice that are heterozygous for the G59S p150 glued mutation exhibit late-onset slowly progressive muscle weakness, motor neuron death, and vesicle accumulation [5][6][7]. This is in contrast to heterozygous p150 glued knockout mice, which did not display a neurodegenerative phenotype. Taken together, this evidence suggests that the pathogenesis of p150 glued -associated diseases might be caused mainly by a gain-of-function effect [6,26].
Neurons of the hypoglossal nucleus and ventral horn in HMN7B and the substantia nigra and locus coeruleus in PS are substantially affected [1,2]. Although this means that neuronal cell lines are the most appropriate for studying the mechanisms that underlie the pathology of p150 glued -associated diseases, various studies have been performed using non-neuronal cell lines. For example, the CAP-Gly domain of p150 glued was needed for proper Golgi morphology in HeLa cells [10] and to activate cell division in Drosophila S2 cells [27]. G59S overexpression induced mitochondria-containing p150 glued aggregates and insufficient recovery of Golgi distribution following nocodazole treatment in COS7 cells. Overexpression of this mutant also increased cell death induction in MN1 mouse embryonic motor neuron cells [11]; however, this overexpression did not promote apoptosis induction by caspase-3 cleavage in COS7 cells or in rat primary motor neurons [28]. Although our data are from non-polarized and non-neuronal HeLa cells and thus might not reflect the precise physiological state of neurodegenerative disease, we believe that these results give us at least partial insight into the mechanisms underlying p150 glued -associated diseases. Further assessment with more appropriate cell lines, like neurons differentiated from induced pluripotent stem cells from the disease patients should be performed in the future.

Cell culture and transfection
HeLa and SH-SY5Y cells were maintained as previously described [29]. after treatment with or without z-VAD (100 mM) for 24 hours. Values are relative to the GFP-empty value, which is set at 1. (C, D) Transfected SH-SY5Y cells after treatment with or without z-VAD (100 mM) for 48 h were stained with Annexin V and PI, and GFP-positive cells were analyzed by flow cytometry. (E) HeLa cells transfected with GFP-tagged wild-type or mutant p150 glued were fixed and stained with an antibody against cleaved caspase-3. Bars, 10 mm. (F) Twenty-four hours after transfection, SH-SY5Y cells were fixed and stained with a cleaved caspase-3 antibody. GFP-positive cells were analyzed by flow cytometry and the mean fluorescent intensity was calculated. (G) HeLa cells were transfected with control scrambled siRNA or caspase-3 siRNA for 72 h and immunoblotting analysis was performed to monitor the knockdown efficiency of the caspase-3 siRNA. (H) Twenty-four hours after transfection with control siRNA or caspase-3 siRNA, HeLa cells were transfected with GFP-empty, GFP-tagged wild-type or mutant p150 glued . Forty-eight hours after transfection, cells were fixed and stained with DAPI, and ratios of GFP-positive cells with nuclear abnormalities were analyzed. Values are relative to the GFP-empty value, which is set at 1. The error bar indicates each standard deviation. Statistics are from three independent experiments: N.S., not significant; *, p,0.05; **, p,0.01; ***, p,0.001. doi:10.1371/journal.pone.0094645.g002 Louis, MO, USA, D2650) were added at the indicated times and concentrations.

Plasmids
The wild-type DCTN1 coding region was PCR-amplified from a cDNA plasmid kindly provided by Dr. Farrer MJ (University of British Columbia) using the following primers (Sigma): 59-TCAAGGGAATTCAATGGCACAGAGCAAGAGGCAC-39 and 59-TCAAGGGATATCAGGGAGATGAGGCGACTGT-GAA-39. The resulting fragment was inserted into the pFLAG-CMV5a vector (Sigma) using EcoRV and EcoRI. The plasmid was cut with EcoRI and KpnI, and the insert was subcloned into pAcGFP-N3 (Clontech, Mountain View, CA, USA). Mutagenesis to create the six mutated p150 glued plasmids was performed using the Quikchange Lightning site-directed mutagenesis kit (Stratagene, La Jolla, CA, USA). The pCIneo-TDP43-FLAG plasmid   The percentage of cleaved caspase-3-positive cells is shown (J). Bars, 100 mm (K,L). Twenty-four hours after transfection with control scrambled siRNA confocal microscope (Zeiss, Oberkochen, Germany) using a 636 water-immersion objective lens (NA = 1.2). Images were magnified using Zeiss LSM510 v3.2 software. Colocalization was quantified using the colocalization plugin of ImageJ 1.43 (NIH).

Quantification of aggregate formation and cleaved caspase-3 positive cells
Aggregate formation and cleaved caspase-3 positive cells were assessed using a fluorescence microscope (Axio Imager 2, Zeiss) with a 406 objective. GFP-positive or FLAG-positive cells were selected and the population of cells with aggregates was counted. Quantification was based on at least three independent experiments, each carried out in triplicate, and 100-300 cells were counted in each slide. The scorer was blinded to the identity of the slides.

Cell viability assays
GFP-positive cells were scored 24 and 48 h after transfection for abnormal cell nuclei, according to previously reported criteria [30] using a fluorescence microscope (Axio Imager 2) with a 406 objective. Analysis was performed with at least three independent experiments, each carried out in triplicate, and 100-400 cells were counted on each slide. The scorer was blinded to the identity of the slides. Detection of apoptotic cells was also determined using an annexin V/propidium iodide (PI) detection kit (Invitrogen), according to the manufacturer's protocol. Briefly, cells were harvested and washed with 16PBS 24 and 48 h after transfection. They were then incubated at room temperature with annexin V/ Alexa350 and PI for 15 min and analyzed by flow cytometry (LSRFortessa, BD Biosciences, San Jose, CA, USA).

Electron microscopy
HeLa cells were plated on Thermanox plastic coverslips (Nunc, Penfield, NY, USA) and transfected with pAcGFP-empty, wildtype, G59S, and G71R p150 glued plasmids. Twenty-four hours after transfection, one set of cells was pre-fixed in 2% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4) at 4uC, and post-fixed with 2% OsO 4 in phosphate buffer for 1 h at 4uC. After fixation, they were dehydrated in a graded series of ethanol, placed in propylene oxide, and embedded in epoxy resin (Quetol 812, Nisshin-EM, Tokyo, Japan). Ultra-thin sections (90-100 nm) were cut using an ULTRACUT-UCT (Leica, Wetzlar, Germany) with a diamond knife. Sections were stained with 2% uranyl acetate in distilled water for 15 min followed by a lead staining solution for 5 min.
For immune electron microscopic analysis, another set of cells was pre-fixed in 4% PFA and 0.1% glutaraldehyde in phosphate buffer at 4uC, and post-fixed with 1% OsO 4 and 1.5% potassium ferricyanide in phosphate buffer for 1 h at 4uC. After fixation, they were dehydrated in a graded series of ethanol and embedded in LR White resin. Ultra-thin sections were cut, and samples were incubated in 3.8% sodium periodate for 1 h at room temperature. Samples were blocked with 2% BSA in PBS for 30 min at room temperature and then immunolabeled with primary anti-GFP antibody (Living Colors A.v. Peptide Antibody, Clontech, 632377; 1:10) followed by anti-rabbit immunogold (BB International, Cardiff, UK; 1:100). Afterwards, these samples were stained with 2% uranyl acetate in distilled water for 5 min followed by a lead staining solution for 1 min. All sections were examined with a JEM-1200EX (JEOL, Peabody, MA, USA) electron microscope at 80 KV.

Statistical analysis
Densitometry analysis was performed on immunoblots from three independent experiments using ImageJ 1.43. Differences among means were analyzed using 1-or 2-way ANOVA, followed, when results showed significant differences, by pair-wise comparisons between means using Tukey's Honestly Significant Difference Test. When only two groups were compared, the Student's t test was used. In all analyses, the null hypothesis was rejected at the 0.05 level. SYSTAT 13 software (Hulinks, Tokyo, Japan) was used for statistical calculations. Figure S1 Overexpression of mutant p150 glued disrupts p150 glued distribution and causes aggregate formation. (A) HeLa cells transfected with GFP-tagged wild-type or mutant p150 glued were fixed after 24 h and analyzed using confocal microscopy. Bars, 10 mm. (B) SH-SY5Y cells transfected with GFP-tagged wildtype or mutant (G59S or G71R) p150 glued were fixed and stained with an antibody against a-tubulin (red) after 24 h and analyzed using confocal microscopy. Bars, 10 mm. (C) HeLa cells transfected with 3xFLAG-tagged wild-type or mutant p150 glued were fixed and co-stained with antibodies against FLAG (green) and atubulin (red) after 24 h. Bars, 10 mm.