Loss of VPS13C Function in Autosomal-Recessive Parkinsonism Causes Mitochondrial Dysfunction and Increases PINK1/Parkin-Dependent Mitophagy

Abstract

Autosomal-recessive early-onset parkinsonism is clinically and genetically heterogeneous. The genetic causes of approximately 50% of autosomal-recessive early-onset forms of Parkinson disease (PD) remain to be elucidated. Homozygozity mapping and exome sequencing in 62 isolated individuals with early-onset parkinsonism and confirmed consanguinity followed by data mining in the exomes of 1,348 PD-affected individuals identified, in three isolated subjects, homozygous or compound heterozygous truncating mutations in vacuolar protein sorting 13C (VPS13C). VPS13C mutations are associated with a distinct form of early-onset parkinsonism characterized by rapid and severe disease progression and early cognitive decline; the pathological features were striking and reminiscent of diffuse Lewy body disease. In cell models, VPS13C partly localized to the outer membrane of mitochondria. Silencing of VPS13C was associated with lower mitochondrial membrane potential, mitochondrial fragmentation, increased respiration rates, exacerbated PINK1/Parkin-dependent mitophagy, and transcriptional upregulation of PARK2 in response to mitochondrial damage. This work suggests that loss of function of VPS13C is a cause of autosomal-recessive early-onset parkinsonism with a distinctive phenotype of rapid and severe progression.

Figure 1

Identification of VPS13C Mutations (A) Pedigrees of families with VPS13C mutations. Black symbols represent individuals with PD, open symbols, those unaffected. Arrows point to probands who underwent whole-exome sequencing. Abbreviations are as follows: AE, age at examination; AD, age at death; AO, age at onset. (B) Schematic representation of VPS13C and its variations. VPS13C spans 208 kb and contains 86 exons encoding a 3,753-amino acid protein with a chorein domain at its N terminus, a DUF1162 domain of unknown function, and a putative autophagy-related domain. The five variations found in the three probands are indicated. Numbers above the gene identify the exons containing VPS13C variations. Alternative splicing a and b represent skipping of exons 6+7 and of exon 82, respectively. Transcripts 1A, GenBank: NM_017684.4: splicing a + b; 2A, GenBank: NM_020821.2: splicing b; 1B, GenBank: NM_018080.3: ends at exon 82; 2B, GenBank: NM_001018088.2: splicing a and ends at exon 82.

Figure 2

Neuropathology in the Proband of Family B with c.[806_807insCAGA];[9568G>T] VPS13C Mutations Shows Abundant α-Synucleinopathy (A–C) Macroscopic appearance of the left hemisphere (fixed): lateral view (A); medial view (B); coronal section at the level of the cerebral peduncle (C). (D and E) Lewy bodies in pigmented neurons in the substantia nigra (D, arrow, hematoxylin-eosin [HE] staining) and the parietal neocortex (E, arrowhead, HE staining). (F) Representative image of α-synuclein immunoreactivity in the frontal cortex showing abundant Lewy bodies and neurites. (G) Tau-immunoreactive neurofibrillary tangles in the primary motor cortex. Scale bars for microscopic images represent 50 μm.

Figure 3

A Pool of VPS13C Is Located on the Outer Mitochondrial Membrane (A) Sucrose gradient fractionation illustrating the subcellular distribution of endogenous VPS13C in HEK293T. Note the enrichment of the protein in fractions 1–3 and 8–10. Soluble endoplasmic reticulum (ER, BiP) and mitochondrial (PMPCB) markers in fractions 1 and 2 reflect organelle damage during fractionation. (B) Western blot showing VPS13C immunoreactivity in mitochondria purified by Percoll gradient centrifugation from HEK293T cells (pM fraction). Note the enrichment in VPS13C and the mitochondrial markers TOMM70 and PMPCB in the pM fraction compared to the mitochondrion-enriched fraction (M). Abbreviation is as follows: T, total lysate. (C) Limited trypsin treatment of mitochondrion-enriched fractions (M) from HEK293T or COS-7 cells caused loss of VPS13C and the mitochondrial surface marker TOMM70; the outer mitochondrial membrane channel TOMM40 and the matrix marker PMPCB are preserved.

Figure 4

VPS13C Silencing Impacts Mitochondrial Morphology, Transmembrane Potential, and Respiration (A) Representative immunofluorescence staining illustrating mitochondrial perinuclear redistribution and fragmentation in COS-7 cells silenced for VPS13C (siVPS13C, 30 nM) compared to cells treated with control siRNA (siControl, 30 nM): green, mitochondrial matrix marker PMPCB; red, α-Tubulin. VPS13C silencing reduced VPS13C mRNA levels to no more than 25% of the control condition (see Figure S6). Scale bars represent 10 μm. Quantification of aspect ratio and form factor (see Subjects and Methods, Koopman et al., and Buhlman et al.²²) shows reduced mitochondrial network complexity in siVPS13C-treated cells (means ± SEM; ^∗∗p < 0.01; ^∗∗∗p < 0.001, of n = 88 or 86 cells scored per condition). (B) Analysis of the relative TMRM fluorescence of mitochondria in COS-7 cells transfected as in (A), illustrating the decrease in ΔΨmt in cells depleted for VPS13C. n = 40 cells per condition from one experiment representative of three carried out. ^∗∗∗p < 0.001. (C) Oxygen consumption rates in intact COS-7 cells transfected with siControl or siVPS13C. The top panel shows the oxygen flux corrected for instrumental background from one representative experiment. The graph in the bottom panel displays the respiration rates. Absence of VPS13C is associated with increased maximal respiration (= maximal uncoupled respiration under CCCP − non-mitochondrial respiration in the presence of the mitochondrial complex I and III inhibitors, rotenone, and antimycin A) and reserve capacity (= maximal uncoupled respiration − basal respiration before the addition of the complex V inhibitor oligomycin). Means ± SEM; ^∗p < 0.05, of six independent experiments.

Figure 5

Loss of VPS13C Function Enhances Mitochondrial Accumulation of PINK1, Recruitment of Parkin, and PARK2 Upregulation in Response to CCCP (A and B) Western blot (A) and corresponding VPS13C protein levels (B) (normalized to α-Tubulin or PMPCB) in cytosolic (C), mitochondrion-enriched (M), and total (T) cell fractions from HEK293T cells treated or not with CCCP (10 μM, 3 hr). VPS13C levels decreased significantly in mitochondria after CCCP treatment, but tended to increase in cytosol (means ± SEM; ^∗p < 0.05, of six independent fractionation experiments). (C–E) Western blot (C) and corresponding normalized protein levels (D, E) in cytosolic and mitochondrion-enriched fractions from HEK293T transfected with 30 nM of control siRNA (siControl) or siRNA targeting VPS13C (siVPS13C). (D) CCCP treatment resulted in accumulation of PINK1 (endogenous) in mitochondrion-enriched fractions (M) after treatment with siControl (−) and, more significantly, with siVPS13C (+). (E) Accumulation of Parkin (endogenous) on depolarized mitochondria was also strongly enhanced in cells treated with siVPS13C. In addition, Parkin levels were upregulated in the cytosolic (C) fractions, particularly in untreated cells (means ± SEM; ^∗p < 0.05 of four independent fractionation experiments). (F) Quantitative real-time RT-PCR showing relative mRNA levels, normalized to α-actin (ACTB), in HEK293T cells treated with control siRNA (siControl), or siRNA targeting VPS13C (siVPS13C) or PINK1 (siPINK1), under basal conditions or after CCCP treatment. Note the more than 30% decrease in VPS13C mRNA levels after PINK1 silencing under basal conditions, but not after CCCP treatment (left). Note also that VPS13C and PINK1 silencing enhance the upregulation of PARK2 mRNA at 48 hr of CCCP treatment (right); means ± SEM of three to nine replicates per condition from two independent experiments (^∗∗p < 0.01; ^∗∗∗p < 0.001 compared to siControl within each condition of CCCP treatment; ###p < 0.001 between the indicated conditions of CCCP treatments).

Figure 6

Loss of VPS13C Function Exacerbates PINK1/Parkin-Dependent Mitophagy (A) Immunofluorescence staining of a representative experiment illustrating PINK1/Parkin-dependent mitophagy in COS-7 cells overproducing Parkin and silenced for VPS13C or PINK1 (20 nM siRNA) after CCCP treatment (10 μM for 48 hr): red, Parkin; green, mitochondrial matrix marker PMPCB. Open arrows indicate loss of mitochondrial networks; white arrows show preserved networks. Scale bars represent 10 μm. (B) Quantification of mitophagy in the conditions described in (A), expressed as the proportion of COS-7 cells without PMPCB (black bars) or TOMM20 (gray bars) staining; the siVPS13C treatment increased and siPINK1 decreased the proportion. In the absence of exogenous Parkin (−Parkin; cells overproducing the control protein EGFP) or CCCP (not shown), all the cells harbored normal mitochondrial PMPCB staining, whether or not VPS13C was silenced (means ± SEM; ^∗p < 0.05, ^∗∗p < 0.01 of 3 independent experiments; 100 cells scored per condition). (C) Proportion of COS-7 cells without PMPCB staining after transfection with half-doses (10 nM) of each siRNA and 48 hr of CCCP treatment. The mitophagy-promoting effect of VPS13C depletion was abolished by concomitant silencing of PINK1 (means ± SEM; ^∗p < 0.05 of 3 independent experiments; 100 cells scored per condition).