(C) PLOS One [1]. This unaltered content originally appeared in journals.plosone.org.
Licensed under Creative Commons Attribution (CC BY) license.
url:https://journals.plos.org/plosone/s/licenses-and-copyright

------------



Mitochondrial fission, integrity and completion of mitophagy require separable functions of Vps13D in Drosophila neurons

['Ryan Insolera', 'Molecular', 'Cellular', 'Developmental Biology Department', 'University Of Michigan', 'Ann Arbor', 'Michigan', 'United States Of America', 'Péter Lőrincz', 'Department Of Anatomy']

Date: 2021-10

A healthy population of mitochondria, maintained by proper fission, fusion, and degradation, is critical for the long-term survival and function of neurons. Here, our discovery of mitophagy intermediates in fission-impaired Drosophila neurons brings new perspective into the relationship between mitochondrial fission and mitophagy. Neurons lacking either the ataxia disease gene Vps13D or the dynamin related protein Drp1 contain enlarged mitochondria that are engaged with autophagy machinery and also lack matrix components. Reporter assays combined with genetic studies imply that mitophagy both initiates and is completed in Drp1 impaired neurons, but fails to complete in Vps13D impaired neurons, which accumulate compromised mitochondria within stalled mito-phagophores. Our findings imply that in fission-defective neurons, mitophagy becomes induced, and that the lipid channel containing protein Vps13D has separable functions in mitochondrial fission and phagophore elongation.

Neurons are reliant on the maintenance of mitochondrial function to fuel their high-energy demands, and mitochondrial dysfunction is a common pathology associated with neurodegenerative disease. The Vps13 family of proteins are hypothesized to mediate lipid transport between organelles and have been linked to multiple neurological diseases. In Drosophila neurons, we delineate separable functions for the essential family member Vps13D in (1) mitochondrial fission, and (2) mitochondrial degradation via mitophagy, which becomes induced in response to fission defects in neurons. Prior studies in Drosophila intestinal cells have suggested that Vps13D is required for developmental clearance of mitochondria downstream of its role in mitochondrial fission. In contrast, we find that in neurons, rather than blocking mitophagy, impairments in fission lead to an induction in mitophagy. Vps13D depleted neurons initiate but fail to complete mitophagy, leading to the accumulation of stalled, toxic mitophagy intermediates with compromised integrity. These studies establish a new experimental framework for studying mitophagy in neurons in vivo.

Funding: This work is funded by the following grants: NIH grants R21NS107781 and R01NS069844 to C.A.C.; NIH grant K99NS111000 to R.I.; National Research Development and Innovation Office grants KKP129797, GINOP-2.3.2-15-2016-00032, and NKFIH-871-3/2020 to G.J.; and Magyar Tudományos Akadémia grant PPD-222/2018 (Hungarian Academy of Sciences) to P.L. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Copyright: © 2021 Insolera et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

However, in neurons we noticed additional phenotypes associated with loss of Vps13D. Here we report the accumulation of objects that appear to be stalled intermediates which have initiated but fail to complete mitophagy. These intermediates lack matrix components, hence evade detection by traditional means of mitochondrially targeted reporters. This led us to question whether these objects also appear in other genetic conditions. Indeed, we found similar objects in neurons mutant or depleted for Drp1. Through our study of mitophagy in the fission-deficient conditions of Drp1 or Vps13D depletion, and found that: (1) neurons deficient in their capacity for mitochondrial fission can still undergo functional mitophagy; (2) neurons require Vps13D for a separable function in phagophore elongation, in addition to mitochondrial fission; and (3) mitophagy within neurons is robustly induced following impairments in mitochondrial fission. These studies establish a new experimental framework for studying the induction and completion of mitophagy in neurons in vivo.

In 2018 our collaborators and others identified VPS13D as a cause of familial ataxia [ 30 ], developmental movement disorders [ 30 , 31 ] and spastic paraplegia [ 32 ]. We found that Vps13D depleted neurons accumulate severely enlarged mitochondria which fail to be trafficked to distal axons [ 30 ]. Indeed, loss of Vps13D function in many different cell types leads to severely enlarged mitochondria, including Drosophila intestinal cells, HeLa cells [ 33 , 34 ], and cultured fibroblasts from human ataxia patients containing point mutations in the VPS13D gene [ 30 ]. These findings suggest a conserved role for Vps13D in mitochondrial dynamics.

Here we have used Drosophila melanogaster to investigate the relationship of mitochondrial fission with mitophagy within an in vivo nervous system. Our starting point was to understand the neuronal function of the neurodegenerative disease-associated protein Vps13D (Vacuolar Protein Sorting protein 13D). The Vps13 protein family (Vps13A-D) has been characterized as phospholipid transporters that specifically localize to inter-organelle contact sites [ 27 , 28 ]. While a specific cellular function has not yet been established for each of the Vps13 proteins, it is clear that they are all critical for neuronal health, as mutations in all family members are associated with neurological disorders in humans [ 29 – 31 ].

Another potential discrepancy between mitophagy in cultured cells compared to neurons in vivo is the understood role of mitochondrial fission. Multiple studies [ 20 – 22 ], with some exceptions [ 23 ], have indicated that mitochondrial fission is required for the induction of mitophagy in cultured cells subjected to toxins. Conditional knockout of the essential fission protein dynamin-related protein 1 (Drp1) in Purkinje cells of the mouse cerebellum results in the accumulation of autophagy components (ubiquitin, p62, and microtubule-associated proteins 1A/1B light chain 3B (LC3)) on mitochondria [ 24 , 25 ]. These observations suggest that fission via Drp1 is not required for the initiation of mitophagy in neurons, however an understanding of Drp1’s role requires a better understanding of what these autophagy-targeted mitochondria, termed halted mitophagy intermediates [ 26 ], represent. One possibility is that Drp1 loss leads to induction of mitophagy, which allows detection of the intermediates because of their frequency. An opposite possibility, which was inferred by these reports, is that Drp1 loss leads to a block in the progression of mitophagy, causing the accumulation of stalled intermediates in the cerebellum. Further in vivo studies are needed, with the acknowledgment that mechanisms that support the survival of neurons over the long life time of an animal may diverge considerably from mechanisms sufficient in cultured cells.

The development of specialized acid-sensitive fluorescent reporters have opened opportunities to monitor mitophagy in vivo [ 13 – 16 ], and have thus far been tested in conditions shown to alter in vitro toxin-induced mitophagy, with mixed results. Different reporters, mitoQC and mitoKeima, targeted to the outer mitochondrial membrane (OMM) or matrix, respectively [ 15 , 17 , 18 ], yielded different interpretations of in vivo phenotypes. While some of these differences can be attributed to the use of different reporters and model organisms [ 19 ], the range of differences in existing studies emphasizes the remaining large gap in our understanding of mitophagy mechanisms in neurons.

Research in the past decade has uncovered numerous cellular components of mitophagy machinery, primarily through studies that follow toxin-induced damage to the entire mitochondrial population [ 7 – 10 ]. However, neurons strongly require mitochondria, so are unlikely to undergo widespread clearance of mitochondria or survive such harsh insults [ 11 , 12 ]. Instead, neurons are expected to selectively degrade only damaged mitochondria, however molecular tools to study this type of mitophagy in neurons have been limited.

Neurons are among the most sensitive cell types to mitochondrial perturbation, and are heavily reliant upon a proper balance of mitochondrial biogenesis and degradation, as well as fission and fusion dynamics [ 1 ]. Mutations that disrupt mitochondrial fission and fusion machinery lead to severe neurological dysfunction in humans and animal models [ 2 – 4 ]. Likewise, quality control mechanisms, including the autophagic clearance of damaged mitochondria, known as mitophagy [ 5 ], play important protective roles in neurons; impaired mitophagy is thought to contribute to pathology of multiple neurodegenerative diseases [ 6 ].

Results

Similar mitophagy intermediates lacking matrix accumulate in neurons disrupted for Drp1 Mitophagy intermediates lacking matrix were not previously observed In Drosophila intestinal cells depleted for Vps13D [33]. We then wondered whether these intermediates reflect a neuron-specific consequence of impaired mitochondrial fission. We first utilized Gal4-driven motoneuron expression (D42-Gal4) of Drp1 RNAi [39], simultaneous with expression of mitoGFP in motoneurons. Knockdown of Drp1 resulted in extreme enlargement of mitochondria, to a greater degree than loss of Vps13D (S5 Fig), and also led to the presence of Ref(2)p+/ATP5A+/mitoGFP- mitophagy intermediates (Fig 4A and 4B). Further, these mitophagy intermediates are associated with a phagophore, as shown by the colocalization of ATP5A+/PolyUb+ mitochondria with Atg8A/B (Fig 4C) 84.2% of the time. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. Neurons depleted for Drp1 contain mitophagy intermediates that lack matrix proteins, similar to Vps13D depleted neurons. A) Representative images of motoneurons in the larval VNC which co-express the mitochondrial marker mitoGFP (cyan) and RNAi targeting luciferase (control) versus Drp1 (BL# 67160), via the D42-Gal4 driver. Tissue was stained for Ref(2)p (red) and ATP5A (yellow). Dashed box outlines a single Gal4-expressing neuronal cell body that is shown in high magnification via the inset in the bottom right corner. Arrowheads highlight examples of mitophagy intermediates (Ref(2)p+/ATP5A+/mitoGFP-). Scale bars = 10μm, 2μm. B) Distribution of intensities for mitoGFP (top) and ATP5A (bottom) in conventional mitochondria (ATP5A+/mitoGFP+ and Ref(2)p-), compared to the Ref(2)p+ objects. Based on the criteria of intensity greater than 2.5% of conventional mitochondria (shading), 3.4% of Ref(2)p+ objects contained mitoGFP, while 89.7% contained ATP5A. n = 528 mitochondria and 29 PolyUb+ from 8 larval VNCs. C) Representative image from a Drp1-RNAi expressing neuron which contains two enlarged ATP5A+ mitochondria (yellow). One, highlighted by the arrowhead, co-localizes with PolyUb (cyan), and endogenous Atg8A/B (red). 84.2% of PolyUb+/ATP5A+ objects co-localize with Atg8A/B (n = 38 PolyUb+ mitochondria from 15 larval VNCs). Scale bar = 2μm. D) Representative image of motoneurons in the larval VNC of drp12/drp1KG mutant larvae, with UAS-mitoGFP (cyan) expressed via the D42-Gal4 driver. Dashed box outlines a single Gal4-expressing neuronal cell body that is shown in high magnification in the inset in the bottom right corner. Arrowheads highlight examples of mitophagy intermediates that are positive for Ref(2)p (red) and ATP5A (yellow), but lack mitoGFP (cyan). Scale bars = 10μm, 2μm. E) Graph showing the distribution of intensities for mitoGFP (top) and ATP5A (bottom) in conventional mitochondria (ATP5A+/mitoGFP+ and Ref(2)p-), compared to the Ref(2)p+ objects (n = 648 mitochondria and 29 Ref2(p)+ objects from 7 larval VNCs). Based on the criteria of intensity greater than 2.5% of conventional mitochondria (shading), 4.5% of Ref(2)p+ objects contain mitoGFP and 68.2% of Ref(2)p+ objects contain ATP5A. F) Representative image of a mitophagy intermediate (arrowhead) in the VNC of a drp12/drp1KG mutant larva, with UAS-mitoGFP (cyan) expressed via the D42-Gal4 driver. The highlighted intermediate in this motoneuron stains for PolyUb (red) and endogenous Atg8A/B (yellow), but lacks the matrix marker mitoGFP. 74.4% of PolyUb+/mitoGFP- objects co-localized with Atg8A/B (n = 43 mitophagy intermediates lacking matrix from 8 drp1 mutant VNCs). Scale bar = 2μm. https://doi.org/10.1371/journal.pgen.1009731.g004 To verify that the mitophagy intermediates are linked to loss of Drp1 function, we examined independent genetic methods to impair Drp1. Based on the morphological enlargement of mitochondria, knockdown of Drp1 via RNAi is strong, as the phenotype most closely compared to the loss-of-function drp1KG/Df mutant condition (Df = deficiency) [39,40] (S5 Fig). Regardless of the allelic combination, we observed the presence of Ref(2)p+/ATP5A+/mitoGFP- mitophagy intermediates in all drp1 mutants (Figs 4D, 4E and S5), with a frequency that correlated with the severity of the mitochondrial enlargement phenotype. Consistent with other control conditions (Control RNAi and Vps13D heterozygous animals), these objects were not observed in drp1 heterozygous animals. Similar to Vps13D RNAi, and Drp1 RNAi conditions, the PolyUb+/mitoGFP- intermediates in drp1 mutants engage with a phagophore, as they stain positive for Atg8A/B (Fig 4F) 74.4% of the time. Overall, these results demonstrate that neurons with either disrupted Vps13D or Drp1 contain mitophagy intermediates that uniquely lack mitochondrial matrix proteins, and are engaged with a phagophore.

Completion of mitophagy differs between the fission-deficient conditions of Vps13D and Drp1 loss The presence of mitophagy intermediates could reflect a blockage in mitophagy completion, which was suggested to occur in the cerebellum of Drp1 knockout mice [24]. To estimate successful trafficking to the lysosome, we turned to fluorescent reporters that contain dual acid-labile (GFP) and acid-stable (mCherry) tags. We chose not to test the mitophagy reporter mitoKiema [18] because mitoGFP, which utilizes the same mitochondrial targeting sequence to localize the acid-sensitive fluorescent protein Keima to the matrix, was largely absent from stalled mitophagy intermediates in Vps13D and Drp1 depleted neurons (Figs 1B and 4A). In addition, for unknown reasons, the OMM targeted mitoQC reporter [13,15,41] did not consistently localize to mitophagy intermediates in our experimental conditions (S6 Fig); we did not further optimize the use of this reporter. We instead turned to a previously characterized mCherry-GFP-Atg8A reporter for autophagy [42–44]. This reporter showed co-localization with the Ref(2)p+ mitochondria in Vps13D and Drp1 depleted neurons (S7 Fig), consistent with the endogenous Atg8A/B localization (Fig 3A). Dual GFP/mCherry-containing phagophores were only detectable in a small subpopulation of Vps13D depleted neurons in live VNC preparations (S7 Fig), likely due to lower fluorescence intensity in active phagophores compared to accumulated signal in lysosomes [45]. However, bright mCherry-only signal, indicating the accumulation of reporter delivered to the acidic lysosome compartment, were detectable in the majority of neurons in all conditions in live preparations. The mCherry-only signal was increased in larval motoneurons overexpressing Atg1, which is expected to induce autophagy [45] (S8 Fig). Conversely, knockdown of the essential autophagy component Atg5 resulted in less overall mCherry-only signal in neurons containing mCherry-only puncta (38.5% compared to Control RNAi), with a high proportion of neurons lacking mCherry-only signal altogether (47.8%, 34 of 71 neurons) (S8 Fig). These results indicate that the total levels of mCherry-only signal in larval motoneurons expressing the mCherry-GFP-Atg8A reporter can represent an estimation of delivery of autophagic substrate to the lysosome. An experimental condition that has been rigorously demonstrated to induce mitophagy in the Drosophila larval nervous system has not been previously established. Based on our observations of mitophagy intermediates in the fission-deficient conditions we tested above, we expected to observe an increased delivery of reporter to lysosomes in both conditions due to elevated mitophagy. However, instead we observed strikingly different results between Vps13D and Drp1 depleted neurons. Drp1 depletion led to a strong induction of mCherry-only mCherry-GFP-Atg8A puncta (magenta in merged image), however this induction was not observed in Vps13D depleted neurons (Fig 5A and 5B). Because both Vps13D and Drp1 impaired conditions revealed mitophagy intermediates associated with a phagophore, we interpret that mitophagy was induced in both conditions, however these conditions may differ in their ability to complete clearance via the lysosome. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 5. Reporter data imply that Vps13D, but not Drp1, is required for mitophagy completion. A) Representative images of live larval motoneurons expressing UAS-mCherry-GFP-Atg8A, expressed via the D42-Gal4, simultaneous with expression of UAS-RNAi lines targeting luciferase (control) (BL# 31603), Vps13D (BL# 38320), or Drp1 (BL# 67160). White dashed lines indicate the outlines of individual cell bodies. Scale bar = 10μm. B) Quantification of the sum pixel intensity of the mCherry-only signal per neuronal cell body (normalized to Control RNAi). Each point represents a single neuronal cell body, bars represent the mean ± SEM (Control RNAi n = 69 cell bodies; Vps13D RNAi n = 59 cell bodies; and Drp1 RNAi n = 82 cell bodies from 6 larval VNCs each). **** represents p value <0.0001. n.s. (not significant) represents p value >0.5 (p = 0.09). C) Representative images of live larval motoneurons expressing UAS-mCherry-GFP-Atg8A, expressed via the D42-Gal4 driver, in the indicated genetic backgrounds. White dashed lines indicate the outlines of individual cell bodies. Scale bar = 10μm. D) Quantification of the sum pixel intensity of the mCherry-only signal per neuronal cell body (normalized to WT). Each point represents a single neuronal cell body, bars represent the mean ± SEM (WT n = 84 cell bodies; drp12/+ n = 93 cell bodies; and drp12/drp1KG n = 96 cell bodies from 7 larval VNCs each). *** represents p value <0.001. **** represents p value <0.0001. n.s. (not significant) represents p value >0.5 (p = 0.09). https://doi.org/10.1371/journal.pgen.1009731.g005 A potential confound to this interpretation was an unexplained observation that the mCherry-GFP-Atg8A reporter was detected at significantly higher levels in neurons expressing Drp1 RNAi. We therefore repeated the experiment in drp1 mutants, which did not show elevated expression levels of the reporter. We found that drp1 mutant neurons also showed an increased number of mCherry-only puncta compared to WT and heterozygous animals (Fig 5C and 5D), and this increase was suppressed in the condition of Atg5 knockdown (S8 Fig). Therefore the mCherry-only puncta reflect an increase in autophagic/mitophagic flux in drp1 mutant conditions (and not simply an increase in reporter expression levels). The observation that Vps13D depleted neurons still show the presence of mCherry-GFP-Atg8A puncta that fluoresces only in the mCherry channel suggests that at least some degree of macroautophagy still occurs in this condition. This is consistent with the observation that starvation-induced autophagy can still occur in Vps13D mutant fat body cells [33]. However, in contrast to Drp1-impaired conditions, we failed to see an increase in mCherry-only puncta, even though the presence of intermediates indicates that mitophagy is initiated in both Vps13D and Drp1-impaired conditions. We infer that Vps13D function may not be essential for macroautophagy but is specifically required for completion of mitophagy, a form of selective autophagy.

[END]

[1] Url: https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1009731

(C) Plos One. "Accelerating the publication of peer-reviewed science."
Licensed under Creative Commons Attribution (CC BY 4.0)
URL: https://creativecommons.org/licenses/by/4.0/


via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.