Abstract
Mitochondrial population maintenance in neurons is essential for neuron function and survival. Contact sites between mitochondria and the endoplasmic reticulum (ER) are poised to regulate mitochondrial homeostasis in neurons. These contact sites can facilitate transfer of calcium and lipids between the organelles and have been shown to regulate aspects of mitochondrial dynamics. Vesicle-associated membrane protein-associated protein B (VapB) is an ER membrane protein present at a subset of ER–mitochondrial contact sites. A proline-to-serine mutation in VapB at amino acid 56 (P56S) correlates with susceptibility to amyotrophic lateral sclerosis (ALS) type 8. Given the relationship between failed mitochondrial health and neurodegenerative disease, we investigated the function of VapB in mitochondrial population maintenance. We demonstrated that transgenic expression of VapBP56S in zebrafish larvae (sex undetermined) increased mitochondrial biogenesis, causing increased mitochondrial population size in the axon terminal. Expression of wild-type VapB did not alter biogenesis but, instead, increased mitophagy in the axon terminal. Using genetic manipulations to independently increase mitochondrial biogenesis, we show that biogenesis is normally balanced by mitophagy to maintain a constant mitochondrial population size. VapBP56S transgenics fail to increase mitophagy to compensate for the increase in mitochondrial biogenesis, suggesting an impaired mitophagic response. Finally, using a synthetic ER–mitochondrial tether, we show that VapB's function in mitochondrial turnover is likely independent of ER–mitochondrial tethering by contact sites. Our findings demonstrate that VapB can control mitochondrial turnover in the axon terminal, and this function is altered by the P56S ALS-linked mutation.
Significance Statement
Mitochondrial population dysfunction is tightly tied to neurodegenerative diseases, including amyotrophic lateral sclerosis (ALS). Maintenance of the mitochondrial population in neurons requires the birth of new mitochondria and the degradation of damaged organelles. Endoplasmic reticulum–mitochondrial contact site proteins are in a position to regulate both processes in neurons. Our work demonstrates that an ALS-associated mutation in the contact site protein vesicle-associated membrane protein-associated protein B (VapB) disrupts both processes, identifying VapB as a mediator of regulated mitochondrial turnover to maintain a steady-state mitochondrial population.
Introduction
In neurons, mitochondria are dynamic and present in all compartments, including the metabolically active axon terminal. To maintain a healthy population of mitochondria in the distal axon, the neuron must balance the birth of new mitochondria through biogenesis, with the degradation and removal of damaged mitochondria through mitophagy (Misgeld and Schwarz, 2017). Balancing these processes is challenging for distal, synaptic regions of the neuron. Mitochondrial biogenesis is thought to primarily occur in the cell body, and new organelles then must be transported to the distal axon to supply synaptic energy demands. Additionally, though mitophagy initiates frequently in the axon terminal, lysosomal degradation of engulfed organelles can occur either locally in the axon terminal or after retrograde transport to the cell body (Ashrafi et al., 2014; Evans and Holzbaur, 2020a). How the mitochondrial population size is controlled through regulated mitochondrial biogenesis and mitophagy at the synapse is not well understood (Ploumi et al., 2017; Liu et al., 2023).
Proteins situated at close membrane appositions between mitochondria and the endoplasmic reticulum (ER–mitochondrial contacts) are poised to regulate both mitochondrial biogenesis and mitophagy. ER–mitochondrial contact sites have been shown to regulate mitochondrial DNA (mtDNA) replication, location, and frequency (Friedman et al., 2011; Lewis et al., 2016; Abrisch et al., 2020; Kleele et al., 2021), linking these membrane contacts to mitochondrial biogenesis. ER–mitochondrial contact sites can also promote autophagosome formation in yeast and mammalian cells (Hailey et al., 2010; Hamasaki et al., 2013; Bockler and Westermann, 2014) and have been shown to be essential for mitophagy in yeast (Bockler and Westermann, 2014). Given the relationship between ER–mitochondrial contacts and both mitochondrial biogenesis and mitophagy, proteins at sites of ER–mitochondrial contacts are uniquely situated to control both processes to regulate mitochondrial density in neurons.
Many proteins have been shown to localize to and form ER–mitochondrial contact sites (Wilson and Metzakopian, 2021). Vesicle-associated membrane protein-associated protein B (VapB), a resident ER protein, and its mitochondrial binding partner PTPIP51 are one such pair (De Vos et al., 2012). VapB-dependent ER–mitochondrial contacts have previously been linked to autophagy regulation (Gomez-Suaga et al., 2017; Zhao et al., 2018; Mao et al., 2019; Senturk et al., 2019). VapB was recently shown to stabilize mitochondria at synapses in dendrites to facilitate synaptic plasticity (Bapat et al., 2024). Mutations in VapB are associated with neurodegenerative disease risk (Hartopp et al., 2022). Specifically, a proline-to-serine mutation at amino acid 56 (P56S) in VapB is linked to a familial form of amyotrophic lateral sclerosis (ALS) type 8 (Nishimura et al., 2004). The P56S VapB mutation affects synaptic structure, function, and ATP levels (Ratnaparkhi et al., 2008; Gomez-Suaga et al., 2019; Karagas et al., 2022) and also disrupts VapB-dependent autophagic flux (Gomez-Suaga et al., 2017; Zhao et al., 2018; Mao et al., 2019; Senturk et al., 2019). Based on its location at ER–mitochondrial contact sites, effect on mitochondrial function, and role in autophagy, we hypothesized VapB regulates synaptic mitochondrial turnover (Verstreken et al., 2005; Smith et al., 2016; Misgeld and Schwarz, 2017; Rangaraju et al., 2019). Using live imaging of mitophagy in vivo, we show that VapB promotes mitophagy in the axon terminal, while expression of the ALS-linked P56S VapB mutant does not. Rather, VapBP56S increases mitochondrial biogenesis without an obligate upregulation of mitophagy. Together, our work demonstrates that the P56S VapB mutation disrupts mitochondrial population turnover at the synapse, linking neurodegenerative disease with altered mitochondrial homeostasis in this neuronal compartment.
Materials and Methods
Zebrafish husbandry
All zebrafish (Danio rerio) work was done in accordance with the University of Wisconsin Institutional Animal Care and Use Committee. Adult zebrafish were kept at 28.5°C in a 14/10 h light/dark cycle, and embryos were spawned according to the established protocol (Westerfield, 1993). AB* wild-type (WT) and nacre mutants were used for all experiments (Table 1). The nacre strain is devoid of melanosomes allowing clear visualization of the posterior lateral line (pLL) system (Lister et al., 1999). Embryos were kept in embryo media, maintained at 28.5°C, and developmentally staged using established methods (Kimmel et al., 1995). Transgenic lines used include TgBAC(neurod:egfp)nl1 (Obholzer et al., 2008) and Tg(5kbneurod:pgc1α-p2a-mRFP)uwd9 (Table 1). The Tg(5kbneurod:pgc1α-p2a-mRFP)uwd9 line was derived using Tol2-mediated transgenesis and a 5kbneurod1:pgc1α-p2a-mRFP plasmid as described (Kwan et al., 2007). The vapb mutant was created using CRISPR/Cas9 gene editing. For this, Cas9 protein (500 ng) was coinjected with ∼200 pg each of two guide RNA targeting exon 1 of vapb (Table 2). A mutant line was identified with a 69 base pair deletion, encompassing the entire exon 1 of vapb. PCR-based amplification of the region surrounding the deletion was used to genotype larvae in all studies (Table 2).
Key resources
Oligonucleotides
Western blot
Western blot analysis of VapB and single-stranded binding protein 1 (SSBP1) were done as described (Drerup et al., 2017). For HEK293T cell protein extraction, cells were lysed in protein lysis buffer (150 mM NaCl; 20 mM Tris–HCl ; 1 mM ethylenediaminetetraacetic acid (EDTA); 1% NP-40; 1% sodium deoxycholate), pH 7.5, and incubated on ice for 30 min. After incubation, lysates were centrifuged at 15,000 rpm for 30 min. The supernatant was removed, the Laemmli buffer was added, and protein was heat denatured at 98°C for 10 min prior to Western blot analysis as described below. For the VapB protein extraction for Western blot analysis, individual embryos were dissected to remove tails from the rest of the larva at 4 days post-fertilization (dpf). Proximal larval segments were suspended in a protein lysis buffer and frozen at −80°C. Tail segments were lysed in a 20 μl lysis buffer (0.2 mM EDTA; 50 mM NaOH) at 95°C for 15 min, vortexed, and neutralized with 2 μl 10 mM Tris–HCl. Individual tails were then genotyped for the vapb mutation using PCR fragment length analysis. Proximal segments of vapb mutants or heterozygous siblings were combined to generate protein extracts for Western blot analysis of VapB protein. For the WT sample, whole larval extracts generated from AB larvae were prepared similarly. Protein extracts were quantified using a Bradford assay. For the Western blot, 21 μg of protein was run on a 12% sodium dodecyl–sulfate polyacrylamide gel electrophoresis gel. After transfer to the polyvinylidene difluoride membrane, the blots were probed with rabbit anti-VapB (1:500) or anti-SSBP1 (1:1,000), washed, and incubated with secondary antibody conjugated to horseradish peroxidase. The blot was developed with a Western Sure Chemiluminescent Substrate on a C-DiGit Blot Scanner.
Cloning
Zebrafish cDNA was synthesized from 4 dpf AB embryo lysates with a reverse transcription reaction using SuperScript IV First-Strand Synthesis System, following the manufacturer's protocol. Zebrafish vapb was amplified from the cDNA with PCR. The cloned vapb was verified by Sanger sequencing. Site-directed mutagenesis was used to generate the P56S mutation by creating a C to T single-nucleotide change 165 bases from start codon in the vapb plasmid. Site-directed mutagenesis was performed using QuikChange II Site-Directed Mutagenesis Kit following manufacturer's protocol. Base change was confirmed by Sanger sequencing. Primers are included in Table 2.
In situ hybridization, immunolabeling, and hybridization chain reaction RNA fluorescent in situ hybridization
Colorimetric in situ hybridization was done according to established protocols (Thisse and Thisse, 2008; Drerup and Nechiporuk, 2013). To generate the probe for in situ hybridization, we amplified a 913 base pair region, encompassing the vapb open reading frame and 3′-untranslated region, and used it as a template for probe synthesis (Logel et al., 1992; Thisse and Thisse, 2008; Table 2). In brief, zebrafish larvae were fixed overnight in 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS) at 4°C. Fixed larvae were washed in PBS/0.1% Tween-20 and dehydrated in a methanol series. Larvae were either stored at −20°C or immediately rehydrated for use. After methanol treatment, larvae were rehydrated in a reverse methanol series, treated with 10 μg/ml proteinase K for 15 min, and postfixed with 4% PFA. Embryos were prehybridized for 4 h at 65°C and then incubated with probe (1:200) in hybridization buffer [10 mM citric acid, 0.1% Tween-20, 50 μg/ml heparin, 500 μg/ml tRNA, 5× saline sodium citrate (SSC), 50% formamide] overnight at 65°C. Larvae were kept at 65°C during washes with hybridization buffer-SSC buffer series (2:1 buffer:2× SSC; 1:2 buffer:2× SSC; 2× SSC; 0.2× SSC twice). Following 65°C washes, larvae were washed in room temperature with 2:1 and then 1:2 0.2× SSC:PBS/0.1% Tween-20. Larvae were blocked in 2% goat serum, 10% bovine serum albumin in PBS/0.1% Tween-20 for 4 h, and then goat anti-digoxigenin (1:10,000) in a block solution overnight at 4°C rocking. After washing in PBS/0.1% Tween-20, larvae were washed in a coloration buffer (100 mM Tris–HCl, 50 mM MgCl2, 100 mM NaCl, 0.1% Tween-20), pH 9.5, and stain in 0.45% nitro blue tetrazolium, 0.35% 5-bromo-4-chloro-3-indolyl phosphate in a coloration buffer in dark until color development. Larvae were washed in PBS/0.1% Tween-20 and postfix in 4% PFA in PBS, followed by methanol and reverse methanol series prior to imaging.
For hybridization chain reaction (HCR) RNA fluorescent in situ hybridization (FISH), larvae were processed as described above for fixation, dehydration in methanol and rehydration, proteinase K digest, and postdigestion fixation. At this point, larvae were incubated with a manufacturer-supplied hybridization buffer at 37°C for 1 h and then incubated in 4 nM probe in a hybridization buffer at 37°C overnight. Larvae were washed with a manufacturer-supplied wash buffer and then SSC/0.1% Tween-20. Then, larvae were incubated in a manufacturer-supplied amplification buffer at room temperature for 30 min, during which hairpins were individually heated at 95°C and then cooled to room temperature for 30 min in the dark. Amplification with hairpin chain reaction was performed with a freshly prepared solution of the two hairpins, each at 60 nM in amplification buffer, and incubation at room temperature overnight. mRNA transcript-specific probe set for each gene was designed by Molecular Instruments (Table 3). Larvae were mounted between glass slides and #1.5 coverslip with Fluoromount mounting medium prior to imaging with an Olympus FV3000 confocal microscope with a 60× (NA1.42) oil objective. Probe signal at the cell body was quantified using the ImageJ/Fiji software (Schindelin et al., 2012). Probe signal outside of the cell body was excluded. To define the cytosolic volume, we generated a masked z-stack from the GFP channel with a threshold set based on the stack histogram with the maximum entropy algorithm. After using the mask to isolate cell body probe signal, the sum intensity of probe signal was measured and then divided by cell body area measured from maximum projection of the masking z-stack. For each experimental trial, one probe signal acquisition setting was used, and groups were normalized to the control mean value of that trial.
Probe sets used for HCR RNA FISH
For SSBP1 labeling, larvae were fixed in 4% PFA/0.25% Triton X-100 and incubated in water at room temperature overnight. They were then blocked (0.1% Triton X-100, 1% dimethyl sulfoxide, 0.02% sodium azide, 0.5% bovine serum albumin, 5% goat serum) at room temperature for 2 h. Larvae were incubated in chicken anti-GFP (1:2,000) and rabbit anti-SSBP1 (1:500) in a block solution overnight (Table 1). Larvae were then washed in PBS/0.1% Triton X-100 before incubation in goat anti-chicken Alexa Fluor 488 (1:1,000) and goat anti-rabbit Alexa Fluor 568 (1:1,000) at 4°C overnight. Larvae were washed in PBS/0.1% Triton X-100 prior to imaging as described for HCR RNA FISH above. For analysis in ImageJ/Fiji (Schindelin et al., 2012), the cytosolic area was measured from standard deviation projection of the z-stack and defined by manual tracing or applied threshold. SSBP1 puncta were manually counted.
Transient transgenesis and live imaging
To express VapB and subcellular markers of interest in single pLL neurons, we used transient transgenesis. Plasmid DNA encoding the 5kbneurod promotor driving constructs of interest were derived using Gateway Technology or Gibson cloning (Kwan et al., 2007; Gibson et al., 2009; Mo and Nicolson, 2011; Table 1). Expression vectors were created to express VapB or VapBP56S tagged with eGFP, mRFP, or HaloTag or followed downstream by a p2a peptide sequence and TagBFP. The construct used for each experiment is outlined in Table 4. To induce transgene expression, 3–20 pg of plasmid DNA was microinjected into zebrafish zygotes as previously described (Drerup and Nechiporuk, 2016). Zebrafish larvae were sorted at 4 dpf to identify individuals with expression in a subset of pLL neuron cell bodies using a Zeiss AxioZoom fluorescence stereomicroscope with appropriate filters. Zebrafish larvae were treated with HaloTag ligand Janelia Fluor 635 or its analog JFX635 in embryo media at 1 μM for 1 h at room temperature or 0.1 μM overnight at 28.5°C in the dark. Fluorescence signal representing exogenous VapB or VapBP56S was recorded from each sample to ensure protein expression falls within similar ranges between the two groups and between experimental days (Table 4).
Acquisition setting ranges used for VapB and VapBP56S imaging
For live imaging, larvae expressing constructs of interest in pLL neurons were anesthetized in 0.02% tricaine and mounted individually in 1.8% low melting point agarose in embryo media and imaged with an Olympus FV3000 confocal microscope with a 40× (NA 1.25) silicone oil objective. Image acquisition setting for axon terminal area measurement and transport analysis are adjusted to saturate pixel intensity of approximately half of the labeled structure to increase signal-to-noise ratio of the z-stack.
Axon terminal images were quantified using the ImageJ/Fiji software (Schindelin et al., 2012). For area quantification, used in load analysis, split-GFP-based contact site sensor (SPLICS) analysis, and mitophagy proportion analysis, areas were measured from standard deviation projection of the axon terminal z-stack. The cytosolic area was defined by manual tracing or applied threshold. Signal outside of the axon terminal was excluded. For quantification of the organelle area, a threshold was manually applied to a projected standard deviation z-stack. The selected area was limited to the axon terminal by applying an “and” operation with the cytosolic area in the ROI manager.
Axonal transport analyses were done as described (Drerup and Nechiporuk, 2016). Briefly, 30–100 μm lengths of axon were imaged as a single z-plane at a rate of 300 ms per frame, as required by the Nyquist sampling theorem. Kymograph analyses of distance and velocity of mitochondrial transport were performed using the Metamorph software (Molecular Devices). Quantification of the frequency of mitochondrial transport was done manually, using the generated kymographs to classify anterograde, retrograde, and bidirectional movement. Stationary mitochondria were counted manually. Mitochondrial lengths were measured from the same mitochondrial transport movies. The line tool was used to measure the long axis of the mitochondria.
Mitophagosome exit from axon terminal was imaged at a single z-plane on an Olympus FV3000 confocal (40×/NA 1.25 silicone objective) at 300 ms per frame. Only axon terminals where the boundary between the axon terminal and axon lays on a single optical plane were imaged to ensure tracking of complete cargo exit. Acquisition setting for the green and red channels is adjusted to minimize but not eliminate saturated pixels to achieve consistent identification of red-only mitochondria using the minimum required laser power. Exit events were manually counted from 20 min imaging sessions.
Analysis of LC3 association with mitochondria and mitochondrial volume
Voxel colocalization of mitochondria and LC3 was performed on larvae that were fixed with 4% PFA in PBS at room temperature for 2 h. Larvae were washed in PBS/0.1% Triton X-100 prior to imaging. For axon terminal z-stacks used to assess microtubule-associated protein 1 light chain 3B (Map1LC3B/LC3) puncta association with mitochondria, acquisition setting is made to minimize but not eliminate saturated pixel to achieve consistent manual counting and volume measurement with algorithmically determined intensity threshold. Acquired z-stacks were deconvolved using the Olympus CellSens software. Constrained iterative deconvolution was performed on each channel of each z-stack with the Advanced Maximum Likelihood filter algorithm with adaptive point spread function applied for five iterations. LC3 puncta representing autophagic vesicles were manually counted and separated based on association with mitochondria. To quantify the volume of each organelle and their colocalization, we set a threshold for each channel with the maximum entropy algorithm computed from the stack histogram to generate masked z-stacks from which the mitochondrial and LC3 volumes were calculated. The overlapping volume between the two masked z-stacks was the colocalized volume. Volume measurements were performed on the 3D Manager V4.1.5 of the 3D Suite plugin (Ollion et al., 2013). To quantify the individual mitochondrion volume in the axon terminal, we identified the individual mitochondrion manually and cropped it from the z-stack, and we measured the volume as described above.
Experimental design and statistical analysis
Statistical analyses and data plots were performed with Prism 9 (GraphPad). Values of data with error bars on graphs and in text are expressed as mean ± SEM unless indicated otherwise. Each datapoint represents data collected from one larva unless otherwise stated. Each experiment is performed with a minimum of two biological and experimental replicates. Based on the variance and effect sizes measured in this study, these numbers were adequate to provide statistical power to avoid both Type 1 and Type 2 errors. Where appropriate, data were confirmed for normality using a D’Agostino–Pearson’s normality test and for equal variances using an F test to compare variances. The statistical significance between two conditions was determined by either an unpaired t test or Mann–Whitney U test as appropriate. For comparison of multiple conditions, ANOVA with Dunnett's post hoc or a Kruskal–Wallis test with Dunn's post hoc was used as appropriate. For analysis, all datasets were analyzed blind or by a secondary reviewer naive to the experimental hypothesis.
Results
Mitochondrial density in the axon terminal is regulated by VapB
We used the zebrafish larvae pLL neuron as a model to study the role of VapB in mitochondrial population maintenance in the axon terminal. In larval zebrafish, pLL neurons extend axons beneath the skin along the length of the trunk. At the end of the axon, the axon terminal forms synapses with sensory organs in the skin (Metcalfe et al., 1985; Fig. 1A). These neurons are postmitotic and participate in a functional sensorimotor neural circuit by 4 dpf (Kindt et al., 2012). Given its superficial and planar location, this neural circuit provides a unique and efficient model to study mitochondrial populations in vivo (Drerup and Nechiporuk, 2016; Mandal et al., 2018). We investigated the role of VapB in mitochondrial population turnover using pLL neurons. First, we confirmed vapb is expressed in pLL neurons using in situ hybridization. vapb is ubiquitously expressed but shows elevated expression in neurons, including those of the pLL (Fig. 1B). Then, we assessed the effect of VapB manipulation on mitochondrial density in neurons, concentrating on the synapse. To study VapB's effect on mitochondrial populations in the axon terminal, we used transient transgenesis to express exogenous VapB and the ALS-associated P56S VapB mutant (VapBP56S) mosaically in neurons. In mammalian cells, VapB localizes primarily to the ER when overexpressed, while VapBP56S aggregates and can cause changes to overall ER structure (Fasana et al., 2010; Papiani et al., 2012; Kuijpers et al., 2013; Yamanaka et al., 2020). We expressed VapB and the VapBP56S in pLL neurons and noted that both proteins localized to a large area of the axon terminal (Fig. 1C,D). We then quantified mitochondrial density in the axon terminal with expression of VapB and VapBP56S and found both mildly increased mitochondrial population size in this compartment of the neuron (F(2,40) = 8.887; p < 0.001; Fig. 1E–H).
VapB regulates mitochondrial density in the axon terminal. A, Schematic of the pLL in the zebrafish larva. Neuronal cell bodies reside in the pLL ganglion and extend axons down the trunk. Single axon in red. Axon terminal in inset. A fraction of the axon terminal is occupied by mitochondria (red in inset). B, In situ hybridization of vapb at 4 dpf. vapb is enriched in pLL neuron cell bodies. pLL ganglion boxed in B, outlined in B`. C, D, Representative images of pLL axon terminals expressing VapB (C) or VapBP56S (P56S; D) tagged with a HaloTag. E–G, Representative images of mitochondria visualized with matrix-localized TagRFP (tRFP) in axon terminal of WT (E), VapB transgenic (F), or VapBP56S transgenic (G) larvae at 4 dpf. (H) Quantification of mitochondrial load (mitochondrial area/axon terminal area) with manipulation of VapB. I–K, Images of tfam HCR RNA FISH from single pLL neuron (purple outline) coexpressing GFP (I), VapB (J), or VapBP56S (K). L, Quantification of tfam mean fluorescence intensity normalized to WT. M–O, Representative images of SSBP1 immunostaining in single pLL neurons (purple outline) coexpressing GFP (M), VapB (N), or VapBP56S (O). P, Quantification of the SSBP1 puncta number normalized to the cell body area. Q, Anti-SSBP1 Western blot of WT larval protein extract and HEK293T cell lysate. Scale bar: 200 μm in B; 50 μm in B`; 5 μm in C and E; 10 μm in I and M. Each data point represents the average calculated from an individual animal. All data represented as mean ± SEM. ANOVA with Dunnett's post hoc contrasts in H, L, and P.
The mitochondrial population size could be increased through two primary processes, increased mitochondrial biogenesis or decreased mitochondrial degradation. We assessed VapB's impact on the mitochondrial biogenesis by quantifying transcriptional readouts of mitochondrial biogenesis and measures of mtDNA replication in the cell soma. First, we quantified expression of tfam using HCR RNA FISH. TFAM is essential for mtDNA replication (Larsson et al., 1998), and its expression is upregulated during mitochondrial biogenesis (Virbasius and Scarpulla, 1994; Ekstrand et al., 2004; Nisoli et al., 2005; Picca and Lezza, 2015). Transgenic expression of VapB and VapBP56S increased tfam fluorescence intensity (F(2,117) = 9.350; p < 0.0005), suggesting increased mitochondrial biogenesis (Fig. 1I–L). Next, we directly assayed mtDNA replication by immunolabeling for SSBP1. SSBP1 binds to mtDNA during replication, and the presence of concentrated SSBP1 puncta can be used to identify replicating mtDNA nucleoids (Ruhanen et al., 2010; Rajala et al., 2014). Expression of VapBP56S increased the number of replicating mtDNA nucleoids in pLL neurons (F(2,60) = 5.894; p < 0.005), confirming increased mitochondrial biogenesis (Fig. 1M–P). Conversely, WT VapB expression had no effect on the number of replicating mtDNA nucleoids. Specificity of the anti-SSBP1 antibody was confirmed by Western blot analysis (Fig. 1Q). These data suggest that the VapBP56S mutant protein augments mitochondrial biogenesis in neurons.
Because of the inconsistent effects of VapB transgenesis on mitochondrial biogenesis markers, we next assayed additional phenotypes associated with mitochondrial biogenesis to better gauge the impact that expression of the two transgenes has on the mitochondrial population. In addition to mtDNA replication, mitochondrial biogenesis is associated with an increase in mitochondrial fission (Kleele et al., 2021). Mitochondrial fission is difficult to visualize in the axon due to high mitochondrial density in this compartment of the neuron. Instead, we inferred the rate of mitochondrial fission by measuring axonal mitochondrial length in pLL axons (Korobova et al., 2013). Analysis of the motile mitochondrial population showed no change in the average mitochondrial length with VapB or VapBP56S expression (H(3) = 0.4749; p = 0.7887, ns; Fig. 2A,B). However, stationary mitochondria were significantly shorter in VapBP56S transgenics compared with those in WT controls (H(3) = 5.752; p = 0.0564, ns; Dunn's post hoc WT vs VapB: Z = 1.641; p = 0.2014, ns; WT vs VapBP56S, Z = 2.327; p < 0.05), suggesting increased mitochondrial fission (Fig. 2A,C). Expression of VapB failed to change mitochondrial length in the stationary pool.
VapBP56S expression decreases mitochondrial length and increases mitochondrial anterograde transport. A, Average length of motile (left) and stationary (right) mitochondria with VapB and VapBP56S transgenic expression. Schematic of mitochondrial length measurement in A. The long axis of mitochondria was measured as the length. B, C, Frequency distribution histograms of mitochondrial lengths for motile (B) and stationary mitochondria (C) in pLL axons. The number of mitochondria and animals analyzed: WT, 124, 14; VapB, 110, 9; and P56S, 182, 11 for motile mitochondria (B), and WT, 109, 13; VapB, 98, 9; and P56S, 110, 11 for stationary mitochondria (C). D, Example kymographs of mitochondrial transport in WT, VapB transgenic, and VapBP56S transgenic backgrounds. E, F, Quantification of the number of mitochondria moving in the anterograde direction (E) and total number of mitochondria (F) in pLL axons. Each data point represents the average calculated from an individual animal. G, Analysis of average anterograde (Ant; left) and retrograde (Ret; right) transport velocity. The number of mitochondria and animals analyzed: WT, 161, 17; VapB, 331, 9; and P56S, 625, 13 for anterograde transport, and WT, 262, 16; VapB, 427, 8; and P56S, 471, 12 for retrograde transport. All data represented as mean ± SEM. Kruskal–Wallis test with Dunn's post hoc contrasts in A and G. ANOVA with Dunnett's post hoc contrasts in E and F.
Mitochondrial biogenesis is predicted to occur predominantly in the cell body where nuclear gene transcription can produce the RNA encoding the thousands of proteins necessary for mitochondrial structure and function (Misgeld and Schwarz, 2017). After synthesis, new organelles are transported into the axon by microtubule-dependent anterograde transport. We assessed the mitochondrial transport in pLL axons expressing VapB or VapBP56S according to established methods (Drerup and Nechiporuk, 2016). Live imaging revealed no change in frequency of anterograde mitochondrial transport, total number of axonal mitochondria, or the velocity of mitochondrial transport with VapB expression. Conversely, in VapBP56S transgenics, there is a significant increase in the number of mitochondria moving toward the axon terminal (anterograde; F(2,31) = 3.331; p < 0.05) but no significant change in the total number of mitochondria in the axon (F(2,31) = 1.251; p = 0.3004, ns) or transport velocity compared with control (anterograde, H(3) = 23.32; p < 0.0001; Dunn's post hoc WT vs VapB, Z = 1.809; p = 0.1410, ns; WT vs VapBP56S, Z = 1.730; p = 0.1671, ns; retrograde, H(3) = 4.561; p = 0.1022, ns; Dunn's post hoc WT vs VapB, Z = 1.899; p = 0.1150, ns; WT vs VapBP56S, Z = 1.956; p = 0.1009, ns; Fig. 2D–G). Together, these data indicate that exogenous VapBP56S expression increased the axon terminal mitochondrial population size through increased biogenesis and anterograde transport, while a distinct mechanism is perturbed by overexpression of wild type VapB.
VapB regulates mitophagy in the axon terminal
Decreased mitochondrial degradation could also increase the mitochondrial population size. We next asked if VapB transgenesis changed the rate of mitophagy in the axon terminal. We assayed mitophagy in the axon terminal in vivo using a previously developed genetically encoded acidity reporter localized to the mitochondrial matrix (Allen et al., 2013; Rojansky et al., 2016). This indicator is a fusion of GFP and mCherry. In acidic environments, such as that found in lysosomes, GFP fluorescence is quenched (Fig. 3A). Consequently, the ratio of the red-only mitochondrial area to the total mitochondrial area can be used as a proxy for the proportion of mitochondria undergoing lysosomal degradation. Quantification of the proportion of mCherry-only–positive mitochondria demonstrated that VapB expression increased the proportion of acidified mitochondria in axon terminals; however, expression of VapBP56S had no impact on this measure of mitophagy (F(4,81) = 6.942; p < 0.0001; Fig. 3B–D,G). To verify that protein fusion does not modify VapB function, we replicated this analysis with expression constructs in which VapB and VapBP56S were coexpressed, but not fused, with TagBFP using a p2a-cleavable peptide (p2a-TagBFP). Similar to the fusion constructs, VapB-p2a-TagBFP but not VapBP56S-p2a-TagBFP expression increased the proportion of acidified mitochondria in the axon terminal (Fig. 3E–G).
VapB expression increases mitophagy in the axon terminal. A, Schematic of mitophagy indicator mito-GFP-mCherry localization and fluorescence in the mitochondrion. In low pH, such as in a lysosome, the GFP is quenched. B–D, Representative images of axon terminals expressing the mitophagy indicator in WT (B), VapB-HaloTag transgenic (VapB-Halo; C), VapBP56S-HaloTag transgenic (P56S-Halo; D), VapB-p2a-TagBFP transgenic (VapB-2a; E), and VapBP56S-p2a-TagBFP transgenic (P56S-2a; F). Distribution of acidified (magenta) and unacidified (outline) mitochondria is illustrated (B```–F```). G, Proportion of acidified mitochondria relative to the total mitochondrial population. H, Proportion of mitochondrial load acidified (mCherry-only mitochondrial area/average mitochondrial axon terminal load derived from Fig. 1H). Scale bar, 5 μm in B``. Each data point for G represents the average calculated from an individual animal. All data represented as mean ± SEM. ANOVA with Dunnett's post hoc contrasts in G.
Our in vivo imaging suggested that VapB transgenics have increased mitochondrial density in the axon terminal and increased mitophagy. The fluorescent protein used to assess mitochondrial density, TagRFP, is not sensitive to pH, so it could detect both healthy and acidifying mitochondria. We next asked if the increased mitochondrial density could be accounted for by the increased acidified fraction. We compared the relative increase in mitochondrial density in VapB and VapBP56S expressing neurons with the fraction undergoing mitophagy. This revealed that the excess mitochondria in VapB transgenics are acidified, indicating that they are undergoing mitophagy (Fig. 3H). In contrast, the excess mitochondrial population in VapBP56S transgenics is not acidifying.
After engulfment by an autophagosome in the axon terminal, the majority of mitochondria undergoing mitophagy are thought to move via retrograde transport to the cell body for lysosomal degradation (Maday et al., 2012; Zheng et al., 2019; Evans and Holzbaur, 2020a; Cason et al., 2021). Our data suggest that engulfed mitochondria (mitophagosomes) in VapB transgenics are not leaving the axon terminal but are instead acidifying and remaining in the distal axon. To determine if these mitophagosomes were remaining in the terminal, we assessed mitophagosome exit from the axon terminal using live imaging. We expressed the mitophagy indicator with VapB or VapBP56S and assessed the number of mCherry-only (acidified) mitochondria exiting the axon terminal (Movie 1). VapB transgenics had a significant decrease in mitophagosome exit frequency from the axon terminal compared with WT and VapBP56S transgenics (H(3) = 10.54; p < 0.01; Fig. 4A,B). We further analyzed the mitophagosome transport in pLL axons (Fig. 4C). This population has either exited the axon terminal and is in the process of moving toward the cell body, or mitophagosomes are generated in the axon proper. Overexpression of VapB or VapBP56S had no effect on the overall frequency of retrograde mitophagosome transport (H(3) = 0.9186; p = 0.6317, ns; Fig. 4D). However, VapB transgenics show a significantly increased number of stationary mitophagosomes in the axon (H(3) = 11.20; p < 0.005; Fig. 4E). These data suggest that VapB expression disrupts initiation or persistence of mitophagosome transport in both the axon and axon terminal, decreasing the rate of mitophagosome clearance.
Exogenous VapB expression stalls mitophagosome transport. A, Representative image of mitophagosome exits from a WT axon terminal. Branchpoint connecting axon to axon terminal boxed and magnified in A` with axon highlighted in blue and axon terminal in yellow. The 18 s excerpt of Movie 1 axon branchpoint time lapse shows mitophagosome (yellow arrowhead) exit from the example axon terminal (A``). The arrow denotes direction of retrograde transport toward cell body, two lines border the path of mitophagosome travel. See Movie 1. B, Quantification of mitophagosome exit event frequency calculated from 20 min time-lapse imaging sessions. C, Example kymographs from axons of WT, VapB transgenic, and VapBP56S transgenic expressing the mitophagy indicator showing transport of unacidified (GFP) and all mitochondria (mCh). Schematic traces of acidified mitochondrial transport are illustrated on the right. Stationary acidified mitochondria traced in blue. D, E, Number of retrograde (D) and stationary (E) acidified mitochondria (mCherry only) in pLL axons expressing the mitophagy indicator. Scale bar: 5 μm in A, 1 μm in A` and A``. Each data point for B, D, and E represents the average calculated from an individual animal. All data represented as mean ± SEM. Kruskal–Wallis with Dunn's post hoc contrasts in B, D, and E.
Increasing mitochondrial biogenesis induces a compensatory increase in mitophagy
While expression of VapB increases mitophagy, expression of VapBP56S does not. Instead, expression of VapBP56S displays a neomorphic or gain-of-function effect on mitochondrial biogenesis which increases the mitochondrial population size in the axon terminal. This observation suggests that VapB's WT function controls the rate or progression of axonal mitophagy, while the P56S mutation impairs this function. Previous work in Caenorhabditis elegans demonstrated that rates of mitophagy and biogenesis are typically correlated. Specifically, they show that increasing mitophagy increased biogenesis (Palikaras et al., 2015). Because of the VapBP56S phenotype, we asked the converse question: To maintain a consistent mitochondrial population in the axon terminal, is mitochondrial biogenesis balanced by mitophagy? To address this, we increased mitochondrial biogenesis and measured mitophagy in pLL axon terminals. To upregulate mitochondrial biogenesis, we created a transgenic line to express human PPARG coactivator 1 α (PGC1α; master regulator of mitochondrial biogenesis; Dominy and Puigserver, 2013) in pLL neurons [Tg(-5kbneurod1:PGC1α-p2a-mRFP)uwd9]. We analyzed mitochondrial density using transgenic expression of mitochondrially localized HaloTag (targeted to the matrix using the signal sequence from Cox8a) based on methods described previously (Drerup et al., 2017). Similar to previous work in zebrafish and other systems (Wareski et al., 2009; O’Donnell et al., 2013), overexpression of PGC1α increased mitochondrial density in pLL axon terminals (t = 3.506; p < 0.005; Fig. 5A–C). To confirm that PGC1α overexpression increased mitochondrial biogenesis, we used HCR RNA FISH to quantify expression of tfam and assayed mtDNA replication using immunolabeling for SSBP1. Expression of PGC1α increased the tfam transcript level (U = 213; p < 0.0005) and the number of replicating mtDNA nucleoids in pLL neurons (t = 2.194; p < 0.05), confirming increased mitochondrial biogenesis (Fig. 5D–I). After confirming increased biogenesis with PGC1α overexpression, we asked if increased mitochondrial biogenesis induces a compensatory increase in mitophagy in axons. We found increased mitophagy in the axon terminal of PGC1α transgenic zebrafish (t = 2.261; p < 0.05; Fig. 5J–L). Together, these data show that increasing mitochondrial biogenesis and subsequent mitochondrial density in the axon terminal led to an increase in the rate of mitophagy.
PGC1α regulates mitochondrial density and mitophagy in the axon terminal. A, B, Representative images of mitochondria, visualized with HaloTag localized to the mitochondrial outer membrane via a Tomm20 signal sequence, in axon terminals of a WT (A) and PGC1α transgenic (B) at 4 dpf. C, Quantification of axon terminal mitochondrial load in PGC1α transgenic. D, E, Representative images of HCR RNA FISH labeling tfam mRNA in the pLL ganglion (purple outline) of WT (D) and PGC1α transgenic (E). F, Quantification of tfam mean fluorescence intensity normalized to WT. G, H Images of SSBP1 immunostaining in single pLL neurons (purple outline) in WT (G) and PGC1α transgenic (H). I, Quantification of SSBP1 puncta number normalized to the cell body area. J, K, pLL axon terminal of WT (J) and PGC1α transgenic (K) expressing the mitophagy indicator. Distribution of acidified (magenta) and unacidified (outline) mitochondria is illustrated (J```, K```). L, Quantification of axon terminal mitophagy in PGC1α transgenic. Scale bar, 5 μm in B and J``; 10 μm in E and H. Each data point for C, F, I, and L represents the average calculated from an individual animal. All data represented as mean ± SEM. Student's t test in C, I, and L. Mann–Whitney U test in F.
VapBP56S does not change mitochondrial association with the autophagosome membrane in the axon terminal
The lack of changes in mitophagy in response to increased biogenesis in VapBP56S transgenics suggests a deficit in the mitophagic process. We reasoned this could be due to disrupted autophagosome formation or maturation. For autophagosome formation, isolation membranes are nucleated and then expand to engulf cargo. After cargo engulfment, autophagosomes fuse with lysosomes which results in organelle acidification, protease activation, and, ultimately, cargo degradation. VapB has been implicated in autophagosome formation through promotion of isolation membrane expansion via interaction with Fip200 and WIPI2. This process is disrupted by the P56S mutation in VapB (Zhao et al., 2018). To determine if mitochondrial association with the autophagosome membrane was disrupted by expression of VapBP56S, we measured colocalization between autophagosomes and mitochondria in deconvolved confocal images of axon terminals (Fig. 6A–C). Somewhat surprisingly, axon terminals expressing VapB and VapBP56S had similar incidences of mitochondria–autophagosome colocalization to WT (normalized to mito, F(2,56) = 1.013; p = 0.3695, ns; normalized to LC3, F(2,56) = 1.249; p = 0.2948, ns; Fig. 6D,E), suggesting autophagosome membrane association with mitochondria is not perturbed. We also assessed the average autophagosome and mitochondrial size in the axon terminal in these deconvolved images and found no change with expression of VapB or VapBP56S (autophagosome, F(2,56) = 0.8033; p = 0.4530, ns; mitochondria, H(3) = 4.475; p = 0.1067, ns; Fig. 6F–J). These results suggest that VapBP56S expression does not alter mitochondrial association with the autophagosome isolation membrane necessary for mitochondrial engulfment.
Mitochondrial association with autophagosome membranes is not disrupted by VapB or VapBP56S expression. Α–C, Deconvolved images of pLL axon terminals coexpressing mitochondrially localized GFP (green) and RFP-LC3 to label autophagosomes (RFP in magenta in merge, black on the right inset) in WT (A), VapB transgenic (B), or VapBP56S transgenic (C). Inset below; yellow circles denote mitochondria-associated autophagosome; blue circles denote unassociated autophagosomes. D, Proportion of the mitochondrial volume colocalized with autophagosomes normalized to mitochondrial volume. E, Proportion of the mitochondrial volume colocalized with autophagosomes normalized to autophagosome volume. F, Average autophagosome volume in the axon terminal. G–I, Example mitochondria from deconvolved images of axon terminals of WT (G), VapB transgenic (H), or VapBP56S transgenic (I). J, Average mitochondrial volume in the axon terminal. Scale bar: 5 μm in A; 1 μm in A inset and G. Each data point for D–F represents the average calculated from an individual animal. Data points in J represent 158 mitochondria from WT, 10; VapB, 8; P56S, 8 animals per group. All data represented as mean ± SEM. ANOVA with Dunnett's post hoc contrasts in D, E, and F. Kruskal–Wallis with Dunn's post hoc contrasts in J.
The VapBP56S mutation is neomorphic or gain of function
Our work used VapB transgenesis to overexpress WT or VapBP56S mutant protein. This approach was done based on data suggesting that VapBP56S functions as a dominant negative protein (Kanekura et al., 2006; Teuling et al., 2007; Suzuki et al., 2009; De Vos et al., 2012; Papiani et al., 2012). However, additional data suggest that the P56S mutation in VapB could instead either result in haploinsufficiency (Kabashi et al., 2013) or have a toxic gain of function (Kuijpers et al., 2013). To better gauge the effect of the P56S mutation on VapB function in mitochondrial turnover, we assessed mitochondrial phenotypes in a VapB null background. We generated a vapb null mutant (vapbuwd6) using CRISPR/Cas9 mutagenesis (Fig. 7A; Shah et al., 2016a,b). In the vapb mutant, vapb mRNA is reduced (H(3) = 8.268; p < 0.05), and VapB protein is absent, confirming loss of VapB function (Fig. 7B–F). Furthermore, vapal, a close homolog of vapb, shows no evidence of compensatory upregulation, which can occur in CRISPR-induced mutant zebrafish lines (H(3) = 0.1918; p = 0.9086, ns; Fig. 7G–J; El-Brolosy et al., 2019). These data confirm loss of VapB function in the vapb mutant line. Then, we analyzed mitochondrial biogenesis and mitophagy in mutant pLL axon terminals. We found that neither markers of biogenesis (F(2,54) = 1.132; p = 0.3298, ns) nor mitophagy (F(2,21) = 0.5169; p = 0.6038, ns) were significantly altered (Fig. 7K–Q). Together, these data suggest that the P56S mutation in VapB has multiple effects on VapB protein function. While the effect on biogenesis appears neomorphic, the inability of the VapBP56S to increase mitophagy suggests it functions as a dominant negative regulator of mitophagy, as has been suggested previously for autophagy (Genevini et al., 2014; Wu et al., 2018; Mao et al., 2019; Tripathi et al., 2021).
vapb knock-out does not phenocopy VapBP56S or WT VapB transgenic expression. A, vapb exon 1 with deletion in the vapb mutant in red. Guide RNAs used to generate line are shown. This mutation eliminates the start site for protein translation (shown in yellow). B, Anti-VapB Western blot of whole larval protein extracts from WT, heterozygous siblings, and vapb mutants. C–E, Images of the pLL ganglion (outlined) of HCR RNA FISH labeling vapb mRNA in WT (C), heterozygous (D; vapb+/−), or vapb mutant (E; vapb−/−) animals. F, Quantification of vapb mRNA mean fluorescence intensity normalized to WT. G–I, Images of the pLL ganglion (outlined) of HCR RNA FISH labeling the vapb homolog vapal. WT (G), heterozygous (H; vapb+/−), and vapb mutant (I; vapb−/−) animals are shown. J, Quantification of vapal mRNA mean fluorescence intensity normalized to WT. K–M, Images of HCR RNA FISH labeling tfam mRNA in the pLL ganglion (outlined) in WT (K), heterozygous (L; vapb+/−), or vapb mutant (M; vapb−/−) animals. N, Quantification of tfam mean fluorescence intensity normalized to WT. O, P, Representative images of axon terminals expressing the mitophagy indicator in WT (O) and vapb−/− animals (P). Distribution of acidified (magenta) and unacidified (outline) mitochondria illustrated (O```–P```). Q, Proportion of acidified mitochondrial area relative to total mitochondrial population area. Kruskal–Wallis with Dunn's post hoc contrasts for F, J. ANOVA with Dunnett's post hoc contrasts in N, Q. Scale bar, 10 μm in C, G, and K, 5 μm in O``. Each data point for F, J, N, and Q represents the average calculated from an individual animal. All data represented as mean ± SEM.
ER–mitochondrial tethering alone does not coordinate mitochondrial biogenesis and mitophagy
VapB's function in mitophagy and mitochondrial biogenesis could be through its function as an ER–mitochondrial tether or a specific function for the VapB protein. Phenotypes due to disruption of some ER–mitochondrial contact site proteins can be compensated for by artificially increasing mitochondrial–ER tethering (Kornmann et al., 2009; Gomez-Suaga et al., 2017). We tested the role of ER-mitochondrial tethering in mitochondrial population turnover by increasing ER–mitochondrial tethering using a synthetic tether. The synthetic tether is a fusion of the mitochondrial targeting signal sequence from TOMM20 and the ER targeting sequence from SEC61β bridged by a HaloTag sequence for visualization (Fig. 8A). This design was based on similar tethers previously shown to increase ER–mitochondrial contacts (Csordas et al., 2006; Kornmann et al., 2009). First, we determined the effectiveness of this tether for increasing ER–mitochondrial contacts using a genetically encoded proximity sensor, SPLICS (Cieri et al., 2018). We expressed the synthetic tether in pLL neurons and saw that it localized to the axon terminal (Fig. 8B). Quantification of the fraction of SPLICS-positive mitochondrial area revealed a 40% increase in ER–mitochondrial contacts compared with control (t = 3.456; p < 0.005; Fig. 8C–E), confirming the efficacy of the tether. Next, we assessed mitochondrial turnover in our increased ER–mitochondrial contact condition. We assayed mitochondrial biogenesis and mitophagy using tfam HCR RNA FISH and expression of the mitophagy indicator. Expression of the synthetic tether did not alter measures of mitochondrial biogenesis (U = 1029; p = 0.7787, ns) or mitophagy (t = 0.6019; p = 0.5516, ns) in the axon terminal (Fig. 8F–K). These data support a model in which mitochondrial population turnover is regulated specifically by VapB and is not due to general changes in ER–mitochondrial tethering at contact sites.
Mitochondrial biogenesis and mitophagy regulation are independent of ER–mitochondrial tethering. A, Schematic of synthetic ER–mitochondrial tether. This tether is composed of the mitochondrial localization sequence from Tomm20 and the ER localization sequence from Sec61β linked with a HaloTag sequence. B, Expression of the synthetic tether (yellow in B, black in B`) in a pLL axon terminal. GFP labels the cytoplasm (blue). C, D, SPLICS signal (green on top, black on bottom) visualizing mitochondria (magenta) proximity to the ER membrane. E, Quantification of SPLICS-positive mitochondrial area in pLL axon terminals. F, G, A representative image of HCR RNA FISH labeled tfam fluorescence in single pLL neuron (purple outline) in individual pLL neurons expressing cell-fill GFP (F) or synthetic tether (G). H, Quantification of mean tfam fluorescence intensity normalized to WT levels. I, J, A representative image of axon terminal GFP and mCherry signals in pLL neuron expressing the mitophagy indicator in WT (I) or the synthetic tether transgenic (J). K, Quantification of the proportion of acidified mitochondria with tether expression. Student's t test for E and K. Mann–Whitney U test for H. Scale bar, 5 μm. Each data point for E, H, and K represents the average calculated from an individual animal. All data represented as mean ± SEM.
Mitophagosome exit from axon terminal in a 4 dpf WT larva. Related to Figure 4. Live imaging of pLL neuron axon terminal labeled with mitochondrial matrix-localized mitophagy indicator shown in two-color merged, unacidified (GFP; green in merged) and all mitochondria (mCh; magenta in merged). Retrograde transport is directed to the left end of the axon (toward cell body). The movie begins at the “0 s” frame in Figure 4A``. The movie plays 10 times real-time speed. The acquisition rate is 0.3 s/frame. Scale bar, 1 μm. [View online]
Discussion
Together, our results identify VapB as a modulator of axon terminal mitochondrial turnover. Expression of VapB or VapBP56S both increase axon terminal mitochondrial load but for different underlying reasons. In VapB transgenics, the excess mitochondria in the axon terminal are in the process of undergoing mitophagy. They are engulfed by autophagosome membrane and actively acidifying. Conversely, VapBP56S transgenics have increased mitochondrial biogenesis and no commensurate increase in mitochondrial degradation. Our data suggest that axon terminal mitochondria in VapBP56S transgenics effectively associate with autophagosomes but fail to undergo acidification and subsequent mitophagy. Together, these data support a model in which VapB normally facilitates mitochondrial degradation, and this process is disrupted by the P56S ALS-linked disease mutation. Furthermore, our data suggest a neomorphic function, either direct or indirect, for VapBP56S in mitochondrial biogenesis. When taken together with previous work on VapB, our data links VapBP56S mutation and impaired mitochondrial homeostasis in ALS.
VapB regulation of mitophagy in the axon terminal
The work presented here is the first to link VapB to mitophagy but not the first to link VapB to autophagy generally. VapB has been shown to interact with Fip200 and WIPI2 to promote autophagosome biogenesis through regulation of isolation membrane expansion (Zhao et al., 2018). Additionally, VapB has been implicated in lipid handling between the Golgi and endosomes. Disruption of this lipid handling impairs lysosomal activation and, consequently, autophagosome maturation (Mao et al., 2019). Further studies on VapB in autophagy have shown that VapB knockdown induces autophagy through upregulation of Beclin-1 (Wu et al., 2018) or by increasing the concentration of calcium transients on the cytoplasmic surface of the ER (Gomez-Suaga et al., 2017; Zheng et al., 2022). Together, though a role of VapB in autophagy is clear, the mechanism controlled by VapB to facilitate the degradation of cellular components is still disputed.
Based on our data, autophagosomal degradation of mitochondria can also be impacted by VapB manipulation. Increasing VapB protein levels does not alter mitochondrial association with autophagosomes in our model; however, mitophagy progression appears stalled by the P56S disease variant. In the axon terminal, this leads to accumulation of mitophagosomes which do not leave or effectively acidify. Conversely, VapB overexpression has the opposite effect on mitochondrial degradation in the axon terminal. VapB transgenics display an increased proportion of acidified mitochondria in the axon terminal, indicating that VapB may increase the rate of mitophagy progression, decrease mitophagosome exit frequency, or affect both processes in this neuronal compartment. Indeed, initiation of retrograde mitophagosome transport to clear them from the axon terminal is impaired in VapB transgenics, which suggests that autophagosomes could be anchored in place. Recent work has shown that VapB transgenesis can anchor mitochondria in dendrites (Bapat et al., 2024). Alternatively, VapB may accelerate mitophagosome maturation in the axon terminal leading to local mitophagy rather than transport of the mitophagosome back to the cell body for complete degradation. In Drosophila and cultured mammalian cells, VAP protein is required for proper transport of phosphatidylinositol-4-phosphate on the Golgi and endosomes, without which autophagosome maturation is impaired (Peretti et al., 2008; Mao et al., 2019). Based on these data, overexpression of VapB would be expected to accelerate acidification. In our model, this would be predicted to lead to excess acidified mitophagosomes such as those we observed in the axon terminal.
VapB and mitochondrial biogenesis in neurons
Increased mitochondrial biogenesis (through PGC1α overexpression) has been associated with improved neuronal health in neurodegeneration models (Cui et al., 2006; St-Pierre et al., 2006; Weydt et al., 2006; Zheng et al., 2010; Zhao et al., 2011; Ye et al., 2016). It is therefore surprising that the VapBP56S variant, which is associated with neurodegenerative disease, increases mitochondrial biogenesis. Curiously, mitochondrial accumulation in neurons has been observed previously in ALS (Sasaki and Iwata, 2007; Delic et al., 2018), suggesting a scenario where mitochondrial biogenesis may exacerbate the defective mitophagy in this neurodegenerative disease. The increase in biogenesis in VapBP56S transgenics could be a result of direct modulation of biogenesis mechanism by VapBP56S, possibly through regulation of ER–mitochondrial contact sites to promote mtDNA replication or mitochondrial fission. We think this is unlikely, however, because increasing mitochondrial–ER tethering does not alter measures of biogenesis. Alternatively, altered mitochondrial biogenesis may be indirectly due to feedback from mitophagy disruption. Future work will focus on the mechanism of VapBP56S-mediated mitochondrial biogenesis and the impact it has on mitochondrial population health and function.
VapB, mitochondrial maintenance, and neurodegeneration
Mitochondrial dysfunction is closely tied to the pathogenic mechanism of ALS. Oxidative stress and impairment in mitochondrial dynamics and function were found in postmortem examinations of spinal cords in ALS patients (Wiedemann et al., 2002; Sasaki and Iwata, 2007), a finding recapitulated in animal models and cell culture (Shaw et al., 1995; Kraft et al., 2007; Onesto et al., 2016; Brown and Al-Chalabi, 2017). Furthermore, as mentioned above, mitochondrial distribution is altered in sporadic ALS patients, with increased mitochondrial density in the spinal cord tissue (Sasaki and Iwata, 2007; Delic et al., 2018). These features are consistent with impaired mitochondrial turnover in the nervous system (Chan, 2006; Pickles et al., 2018; Smith et al., 2019). Mitophagy disruption has also been linked to other neurodegenerative diseases, highlighting its potential pathogenicity. The most widely associated example is Parkinson's disease which is linked to mutations in genes encoding key mitophagy proteins including PINK1, Parkin, LRRK2, and others (Wong and Holzbaur, 2015; Pickles et al., 2018; Evans and Holzbaur, 2020b).
Taken together, our data suggest that a general disruption of mitochondrial turnover by VapBP56S could contribute to reduced neuronal health in disease states. Disruptions to mitochondrial homeostasis and mitophagy are pathogenic in neurodegenerative conditions including ALS and Parkinson's disease (Schon and Przedborski, 2011; Nunnari and Suomalainen, 2012; Burte et al., 2015; Smith et al., 2019; Evans and Holzbaur, 2020b). Elucidating the involvement of VapB in mediating the complex interplay between the ER and mitochondrial maintenance will improve our current model of ALS pathogenesis to advance therapeutic strategies for such conditions.
Footnotes
Funding for this work from the National Institutes of Health–National Institute of Neurological Disorders and Stroke (R01 NS124692) and the OVCRG at University of Wisconsin–Madison to C.M.D, the Integrative Biology Post-doctoral fellowship to H.-T.C.W, and the National Science Foundation/Graduate Research Fellowship Program (DGE-2137424) to A.E.L. We thank members of the Drerup lab for their helpful comments on this work and the manuscript.
The authors declare no competing financial interests.
- Correspondence should be addressed to Catherine M. Drerup at drerup{at}wisc.edu.
