Aberrant mitochondrial morphology and function associated with impaired mitophagy and DNM1L-MAPK/ERK signaling are found in aged mutant Parkinsonian LRRK2R1441G mice

Abstract

Mitochondrial dysfunction causes energy deficiency and nigrostriatal neurodegeneration which is integral to the pathogenesis of Parkinson disease (PD). Clearance of defective mitochondria involves fission and ubiquitin-dependent degradation via mitophagy to maintain energy homeostasis. We hypothesize that LRRK2 (leucine-rich repeat kinase 2) mutation disrupts mitochondrial turnover causing accumulation of defective mitochondria in aging brain. We found more ubiquitinated mitochondria with aberrant morphology associated with impaired function in aged (but not young) LRRK2^R1441G knockin mutant mouse striatum compared to wild-type (WT) controls. LRRK2^R1441G mutant mouse embryonic fibroblasts (MEFs) exhibited reduced MAP1LC3/LC3 activation indicating impaired macroautophagy/autophagy. Mutant MEFs under FCCP-induced (mitochondrial uncoupler) stress showed increased LC3-aggregates demonstrating impaired mitophagy. Using a novel flow cytometry assay to quantify mitophagic rates in MEFs expressing photoactivatable mito-PAmCherry, we found significantly slower mitochondria clearance in mutant cells. Specific LRRK2 kinase inhibition using GNE-7915 did not alleviate impaired mitochondrial clearance suggesting a lack of direct relationship to increased kinase activity alone. DNM1L/Drp1 knockdown in MEFs slowed mitochondrial clearance indicating that DNM1L is a prerequisite for mitophagy. DNM1L knockdown in slowing mitochondrial clearance was less pronounced in mutant MEFs, indicating preexisting impaired DNM1L activation. DNM1L knockdown disrupted mitochondrial network which was more evident in mutant MEFs. DNM1L-Ser616 and MAPK/ERK phosphorylation which mediate mitochondrial fission and downstream mitophagic processes was apparent in WT using FCCP-induced stress but not mutant MEFs, despite similar total MAPK/ERK and DNM1L levels. In conclusion, aberrant mitochondria morphology and dysfunction associated with impaired mitophagy and DNM1L-MAPK/ERK signaling are found in mutant LRRK2 MEFs and mouse brain.Abbreviations: ATP: adenosine triphosphate; BAX: BCL2-associated X protein; CDK1: cyclin-dependent kinase 1; CDK5: cyclin-dependent kinase 5; CQ: chloroquine; CSF: cerebrospinal fluid; DNM1L/DRP1: dynamin 1-like; ELISA: enzyme-linked immunosorbent assay; FACS: fluorescence-activated cell sorting; FCCP: carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone; GAPDH: glyceraldehyde-3-phosphate dehydrogenase; LAMP2A: lysosomal-associated membrane protein 2A; LRRK2: leucine-rich repeat kinase 2; MAP1LC3/LC3: microtubule-associated protein 1 light chain 3; MAPK1/ERK2: mitogen-activated protein kinase 1; MEF: mouse embryonic fibroblast; MFN1: mitofusin 1; MMP: mitochondrial membrane potential; PAmCherry: photoactivatable-mCherry; PD: Parkinson disease; PINK1: PTEN induced putative kinase 1; PRKN/PARKIN: parkin RBR E3 ubiquitin protein ligase; RAB10: RAB10, member RAS oncogene family; RAF: v-raf-leukemia oncogene; SNCA: synuclein, alpha; TEM: transmission electron microscopy; VDAC: voltage-dependent anion channel; WT: wild type; SQSTM1/p62: sequestosome 1.

Keywords: Aging; Dnm1l/DRP1; SQSTM1/p62; knockin mice; macroautophagy; mitochondria dysfunction; mitochondrial fission; mitophagy; parkinson disease; ubiquitination.

Figure 1.

Accumulation of ubiquitinated mitochondria in aged LRRK2^R1441G mutant mouse striata compared to levels in age-matched wild-type (WT) controls. (a) Representative flow cytometry outputs of isolated mitochondria from 6- and 24-month-old WT and LRRK2^R1441G mutant mouse striatum co-stained with MitoTracker Red (Texas Red channel) and anti-ubiquitin antibody (FITC channel). A total of ~65 µg of freshly isolated mitochondria were obtained from each 6- and 24-month-old WT and LRRK2 mutant mouse whole striatum, and were resuspended in 125 µl suspension buffer for staining. (b) Relative amount of ubiquitinated mitochondria (Q2) was significantly higher in aged mutant mice (N = 3). (c) Representative immunoblot showing levels of ubiquitinated protein in isolated striatal mitochondrial lysates in 24 month-old WT and mutant mice (WT: wild-type; RG: LRRK2^R1441G mutant); COX4 levels indicate loading of mitochondria; Densitometry analysis showing higher level of ubiquitination in striatal mitochondria of aged mutant mice compared with age-matched WT controls. Data represents mean ± standard error of mean (SEM) from three independent experiments (N = 5). Statistical significance between groups was analyzed by unpaired Student’s t-test. * represents statistical significance at p < 0.05. The “N” represents the number of mice used for measurements. (d) Western blot of mitochondrial lysates to demonstrate the purity of mitochondrial isolates. Mitochondrial lysates showed highly enriched mitochondrial marker protein, VDAC, whereas only trace amounts of lysosomal LAMP2A and ACTB protein were detected in the isolates. (e) MitoTracker Red staining efficiency in isolated mitochondria from aged WT and mutant mice. Total mitochondria were stained against a mitochondrial specific marker protein, VDAC (voltage-dependent anion channel), and subsequently stained with MitoTracker Red. The percentage of double-stained mitochondria (Q2) between WT and mutant mice were similar

Figure 2.

Small-sized mitochondria and large, “dumb bell”-shaped mitochondria are more abundant in aged LRRK2^R1441G mutant mouse striatum compared to numbers in age-matched wild-type mice. (a) Representative TEM photomicrographs showing cross-section of dorsal striatum of aged 24-month old WT and LRRK2 mutant mice; Scale bar: 2 µm. (b) Cross-sectional area of each individual mitochondrion was calculated using image analysis software. Mitochondria of different sizes were ranked to demonstrate their distribution. (c) There were more small-sized (≤0.1 μm²) mitochondria in LRRK2^R1441G mutant mice compared with age-matched WT mice (N = 3). Total mitochondria were counted in 5 random micrographs from each of 3 pairs of WT and mutant mice. (Total mitochondria counted: WT: 557; mutant: 615) (d) Representative EM photomicrographs showing more “dumb bell”-shaped mitochondria in aged LRRK2 mutant mouse striatum than in age-matched WT mice. “Dumb-bell” shaped mitochondria are magnified and indicated by yellow arrows (scale bar: 2 µm) and magnified (scale bar: 1 µm). (e) There was no significant difference in mitochondria density between WT and mutant mice. (f) However, the ratio of “dumb-bell” shaped to total mitochondria was higher in mutant than WT mice (N = 3). Total mitochondria were counted in 20 random micrographs from each of 3 pairs of WT and mutant mice. (Total mitochondria counted: WT: 2330; mutant: 2362). Statistical significance between groups was analyzed using Student’s unpaired t-test, * and ** represent p < 0.05 and p < 0.01, respectively

Figure 3.

Mitochondria from aged LRRK2^R1441G mutant mouse brain showed disordered cristae and lower ATP production. (a) Representative EM photomicrographs showing four cross-sections of mitochondria (M) in striatal presynaptic terminals of aged (18-month old) WT and LRRK2 mutant mice. In WT mice the cristae are in an ordered formation, whereas in the LRRK2 mutant mice the cristae are in a disordered state; Scale bar: 0.4 µm. (b) Typical oxygraph produced using freshly isolated mitochondria were incubated in a microcell of Clarke-type oxygen electrode, maintained at 37°C with glutamate/malate (complex I substrate) and ADP (0.3 mM). ATP extraction buffer was added into the microcell to stop mitochondrial activity after 2 min of incubation. ATP level was determined by luciferase bioluminescence assay to determine the amount of newly synthesized ATP in mitochondria. (c) Isolated mitochondria from aged LRRK2 mutant mice produced less ATP compared with mitochondria from the age-matched WT controls. Basal ATP level of striatal mitochondria isolated from 18-month old WT and mutant mice was similar (WT: 2.668 ± 0.224 nmol/mg mitochondrial protein; LRRK2^R1441G: 3.352 ± 0.443 nmol/mg mito protein). (d) Representative western blots of mitochondrial marker protein, VDAC, demonstrates equal loading of mitochondria in microcell. Data represents mean ± standard error of mean (SEM) from five independent experiments (N = 5). The “N” represents the number of mice used for measurements. Each experiment was performed in duplicate. Statistical significance between groups was analyzed by Student’s unpaired t-test. *P < 0.05. M: mitochondria

Figure 4.

LRRK2^R1441G mutant mouse embryonic fibroblasts (MEFs) were more susceptible to FCCP-induced ATP depletion than WT control cells. (a) FCCP is an oxidative phosphorylation uncoupler which depolarizes and causes damage to mitochondria. A treatment time point of FCCP (10 µM) before cell death occurred was determined in WT and LRRK2 mutant MEFs using the ToxiLight™ Nondestructive Cytotoxicity BioAssay Kit. No obvious cell death was observed under FCCP (10 μM) toxicity within 6 h of treatment in both WT and mutant MEFs (N = 4). Each experiment was performed in triplicate. Statistical significance between groups was analyzed by one-way ANOVA. **P < 0.01. (b) Intracellular ATP level was significantly lower in mutant MEFs compared to WT at 0.5, 2 and 6 h after FCCP (10 μM) treatment. Data represents mean ± standard error of mean (SEM) from five independent experiments (N = 5). Statistical significance among groups was analyzed by one-way ANOVA. ^##P < 0.01 represents statistical significance between WT and mutant MEFs at the same time point of FCCP treatment

Figure 5.

Levels of SQSTM1 recruitment to mitochondria was significantly higher in LRRK2^R1441G mutant MEFs. (a) Depolarization of mitochondrial membrane potential (MMP) by FCCP treatment was determined by TMRM staining. FCCP (10 µM) treatment caused mitochondrial depolarization as indicated by decrease of TMRM fluorescence over 6 h as measured by flow cytometry. Mitochondria from mutant MEFs were more depolarized compared to those from WT cells (unpaired t-test; p < 0.01; N = 4). Also, mutant mitochondria were significantly more depolarized compared to mitochondria of WT cells under the same FCCP treatment (unpaired t-test; p < 0.01 for 15 & 30 min; N = 4). (b) Total mitochondria were isolated from WT and mutant MEFs after treatment with FCCP to determine the levels of SQSTM1 recruitment to mitochondria. Isolated mitochondria were immuno-stained against SQSTM1 (FITC-channel) and counter-stained with MitoTracker Red (Texas-Red channel). Double-stained particles as illustrated in Q2 represent SQSTM1 conjugated mitochondria. (c) LRRK2 mutant MEFs under normal condition showed significantly higher level of mitochondrial SQSTM1 recruitment than WT cells under normal condition (p < 0.01; N = 3). However, FCCP exposure for 2 h caused 54% increase (i.e. 18.4% → 30.4%; p < 0.05) in SQSTM1 recruitment for WT MEFs, whereas only 12% increase (i.e. 25.6% → 26.6%; not statistically significant) in mutant cells. Data represents mean ± standard error of mean (SEM). Statistical significance between groups was analyzed by unpaired Student’s t-test. *P < 0.05, **P < 0.01

Figure 6.

Cytosolic LC3-II:(I+II) ratio and SQSTM1 level were lower in LRRK2^R1441G mutant MEFs than in WT under FCCP-induced toxicity. (a) Level of autophagic LC3 activation was compared between WT and mutant MEFs after FCCP treatment. MEFs were refreshed with new medium for 2 h to minimize basal autophagy before FCCP treatment. Chloroquine (CQ) was added to prevent degradation of autophagosomes and allowed quantitation of activated LC3-II; (b) Two representative sets of western blots showing changes of autophagy markers, LC3-II:I and SQSTM1 in WT and mutant MEFs under FCCP toxicity. (c) Both ratios of LC3-I+II:ACTB and (d) LC3-II:ACTB in mutant MEFs were significantly lower than WT cells (all p < 0.05; N = 4). (e) Similarly, ratios of LC3-II (active):LC3-I (inactive) and (f) LC3-II:(LC3-I+II) (total LC3) were also significantly lower in mutant MEFs compared to WT cells after FCCP treatment for 2 h (all p < 0.05; N = 4). (g) SQSTM1, a reporter protein of autophagy activity, was significantly lower in mutant MEFs. Data represent mean ± standard error of mean (SEM) from four independent experiments (N = 4). Statistical significance between groups was analyzed by one-way ANOVA, *P < 0.05; **p < 0.01; or unpaired Student’s t-test, #p < 0.05; ##p < 0.01 compared to WT cells

Figure 7.

Abnormal accumulation of GFP-LC3 aggregates in LRRK2^R1441G mutant MEFs under mitochondrial stress indicating impaired mitophagy. (a) Mitochondrial-specific mito-PAmCherry (red) and GFP-LC3 (green) were co-expressed in WT and LRRK2 mutant MEFs. Cells were treated with FCCP (10 μM) for 6 h to induce mitochondrial stress and recruitment of mitochondria to LC3 with and without chloroquine (autophagy inhibitor; CQ). Magnification: 1 × 630. (b–d) Magnified images of WT and mutant MEFs after FCCP treatment. Mitochondria were visualized by photoactivated mito-PAmCherry (Red) after photoactivation. GFP-LC3 aggregation was demonstrated by formation of green puncta, which indicated formation of autophagosomes after LC3 activation. LRRK2 mutant MEFs showed abnormal accumulation of green puncta after mitochondrial stress by FCCP without degradation. This contrasted with what occurred in WT cells undergoing the same treatment, where significantly fewer green puncta were accumulated in the cells. Data represent mean ± standard error of mean (SEM). Statistical significance between groups was analyzed by unpaired Student’s t-test, **P < 0.01

Figure 8.

The relative rate of mitochondrial clearance was slower in LRRK2^R1441G mutant than in WT MEFs. (a) To address the effects of LRRK2^R1441G mutation on rate of mitochondrial clearance, WT and mutant MEFs were transduced by lentivirus to stably express mitochondria-specific photoactivatable PAmCherry (mito-PAmCherry), which emitted red fluorescence after photoactivation. The initial (level) and the subsequent decline in the levels of red fluorescence from mito-PAmCherry in each treatment group at 6 and 24 h after FCCP exposure were measured by flow cytometry; (b) Western blotting of mitochondrial COX4 showed that levels of basal amount of mitochondria in WT and mutant MEFs were similar before mitochondria clearance assay (t = 0); (c) The rate of photoactivated mito-PAmCherry clearance indicating mitochondria clearance in mutant MEFs was significantly slower than in WT MEFs under both normal (p < 0.01; N ≥ 7) and 2 μM FCCP-treated conditions (p < 0.05; N ≥ 7). Statistical significance between groups was analyzed by three-way ANOVA and Tukey’s multiple comparison (Table 1); Data represents mean ± standard error of mean (SEM) from seven independent experiments (N = 7). Statistical significance between groups was analyzed by unpaired Student’s t-test. *P < 0.05, **P < 0.01. (d) No significant difference in the cell proliferation rates of WT and mutant MEFs during the time course of the assay. Thus, the observed difference in the rate of mitochondrial clearance was not due to difference in rate of cell proliferation between the two cell lines (N ≥ 4). Data represents mean ± standard error of mean (SEM) from seven independent experiments (N ≥ 4). Statistical significance between groups was analyzed by unpaired Student’s t-test, **P < 0.01

Figure 9.

Knockdown DNM1L expression perturbed mitochondrial network resulting in mitochondrial aggregation and reduced degradation in LRRK2^R1441G mutant MEFs under FCCP toxicity. (a) WT and LRRK2 mutant MEFs were transfected with either DNM1L siRNA(S) or scrambled negative siRNA for 72 h to knockdown DNM1L expression. Cells were then treated with FCCP (2 µM) for 6 h to study their corresponding changes in mitochondrial morphology. The rates of photoactivated mito-PAmCherry clearance indicating mitochondrial clearance were assayed by flow cytometry over 24 h. (b and c) Transfection of Dnm1l siRNA(s) significantly reduced endogenous DNM1L (total and p-Ser616) protein expression in both WT and mutant MEFs to similar extent; (d) DNM1L knockdown did not affect mitochondrial content in both WT and mutant MEFs (N = 3); (e) Cartoon representation of different mitochondrial morphologies under confocal microscopy of WT and LRRK2 mutant MEFs. (f) Photomicrographs in red demonstrate changes in mitochondria morphologies in WT and mutant MEFs after DNM1L knockdown with/without FCCP treatment (Magnification: 1 × 630); Photomicrographs in white are the enlarged images of the designated square boxes in the red photomicrographs. Morphological differences in mitochondria between WT and LRRK2 mutant cells were observed under normal culture conditions. Mitochondria in WT MEFs existed as an interconnected tubular network, whereas mitochondria in mutant MEFs appeared as swollen tubes. DNM1L knockdown in WT MEFs caused the interconnected tubular mitochondrial network to become individual swollen tubes. Similar morphology of swollen tubes was also observed in mutant MEFs after DNM1L knockdown. Incomplete mitochondrial fragmentation (short separate tubes & swollen tubes) in DNM1L-knockdown WT MEFs was observed when these cells were treated with FCCP. An even stronger effect of DNM1L knockdown was observed in LRRK2 mutant MEFs as shown by accumulation of large rounded aggregates of mitochondria in these cells under FCCP toxicity. (g) Diagrammatic illustration of experimental protocol of mitochondrial clearance assay with/without FCCP treatment after DNM1L knockdown. FCCP treatment resulted in decreased rate of mitochondrial clearance in both WT and mutant cells compared to their corresponding untreated control groups in 24 h (all p < 0.01; N ≥ 3). Knockdown of DNM1L expression in WT and mutant cells with FCCP treatment resulted in even lower rate of mitochondrial clearance. Data represents mean ± standard error of mean (SEM) from at least three independent experiments (N ≥ 3). Statistical significance between groups with/without DNM1L knockdown was analyzed by three-way ANOVA and Tukey’s multiple comparison test (Table 2), * and ** represent statistical significance at p < 0.05 and p < 0.01, respectively

Figure 10.

FCCP toxicity-induced phosphorylation of DNM1L (p-Ser616) and MAPK1/ERK2-MAPK3/ERK1 in WT MEFs but not in LRRK2^R1441G mutant MEFs. (a) Diagrammatic representation of experimental protocol to induce DNM1L:MAPK/ERK phosphorylation by FCCP (10 μM) toxicity; (b) Representative immunoblots of two independent sets of MEFs treatments with FCCP for 0, 0.5 and 1 h, showing changes in level of phosphorylated DNM1L (p-Ser616) and phosphorylated MAPK1/ERK2-MAPK3/ERK1 (p-ERK); (C-F) FCCP exposure for 1 h increased phosphorylation of DNM1L (Ser616) and MAPK1/ERK2-MAPK3/ERK1 (p-MAPK/ERK) in WT, but such phosphorylation in mutant MEFs was not observed. Data represent mean ± standard error of mean (SEM) from four independent experiments (N = 4). Statistical significance between groups was analyzed by one-way ANOVA, *P < 0.05 represents statistical significance between the level at designated treatment time and its corresponding level at t = 0. ^#p < 0.05 represents statistical significance between WT and mutant MEFs at designated time point