Vmp1, Vps13D and Marf/Mfn2 function in a conserved pathway to regulate mitochondria and ER contact in development and disease

SUMMARY

Mutations in Vps13D cause defects in autophagy, clearance of mitochondria and human movement disorders. Here we discover that Vps13D functions in a pathway downstream of Vmp1 and upstream of Marf/Mfn2. Like vps13d, vmp1 mutant cells exhibit defects in autophagy, mitochondrial size and clearance. Through the relationship between vmp1 and vps13d, we reveal a novel role for Vps13D in the regulation of mitochondria and endoplasmic reticulum (ER) contact. Significantly, the function of Vps13D in mitochondria and ER contact is conserved between fly and human cells, including fibroblasts derived from patients suffering from VPS13D mutation-associated neurological symptoms. vps13d mutants have increased levels of Marf/MFN2, a regulator of mitochondrial fusion. Importantly, loss of marf/MFN2 suppresses vps13d mutant phenotypes, including mitochondria and ER contact. These findings indicate that Vps13d functions at a regulatory point between mitochondria and ER contact, mitochondrial fusion and autophagy, and help to explain how Vps13D contributes to disease.

Graphical Abstract

eTOC blurb

The clearance of mitochondria by mitophagy is important for cell health. Shen et al. identify Vps13D as functioning in a pathway with Vmp1 and Marf/Mfn2 to regulate mitophagy. Loss of Vps13D in flies and cells derived from patients with movement disorders also impacts mitochondria and ER contact, and these cellular defects depend on Marf/MFN2.

INTRODUCTION

Autophagy, the lysosome-dependent clearance of intracellular contents, plays important roles in organism development and health. The failure to remove mitochondria by autophagy, or mitophagy, results in defects in cellular homeostasis and health, and contributes to multiple diseases.¹ For example, mutations in genes responsible for mitophagy manifest as inheritable forms of Parkinson’s disease and Alzheimer’s disease.² As a result, it is becoming clear that understanding the mechanisms that regulate mitophagy under different cellular contexts is crucial to our understanding of biology and health.

The mechanisms underlying mitophagy in animals have been defined through studies of derived cell lines. These elegant studies of PINK1- and Parkin-dependent mitophagy have significantly advanced our understanding of this important process. Importantly, these studies have focused on mitophagy in response to mitochondrial damage, where cells are treated with mitochondrial depolarizing agents that are not typically seen in a physiological setting.³ Mitophagy in animals can occur in response to other stimuli, suggesting mitophagy in cells and tissues under physiological conditions do not always utilize the same regulatory pathways.^4,5 During Drosophila development, the larval intestine undergoes an autophagy driven remodeling process where cells reduce in size and mitochondria are cleared by mitophagy.^{6, 7} Mitophagy is driven by the steroid hormone ecdysone, and occurs without the need to induce mitochondrial damage through incubation with uncoupling reagents.^{8, 9} This system allowed us to identify vps13d and other genes as regulators of autophagy under physiological conditions.^6,7 Importantly, vps13d is an essential and conserved gene that regulates mitochondrial clearance, mitochondrial morphology, and has been implicated in human movement disorders.^{7, 10, 11}

Vacuolar protein sorting 13(vps13) was discovered in yeast, and animals possess four evolutionarily conserved Vps13 family members Vps13A-D.^12,13 Yeast VPS13, as well as mammalian Vps13A and C, have been implicated in the regulation of inter-organelle contact and lipid transport.^14,15 However, these studies fail to address if loss-of-function mutants of these human paralogs repress or enhance membrane contacts. Furthermore, no study has linked VPS13D specifically to regulation of membrane contacts.

Members of the Vps13 family possess unique functional requirements. In contrast to VPS13A-C, VPS13D is one of the most vital genes for survival in human cell lines,^16,17 and is essential for Drosophila development.⁷ Vps13D is the only Vps13 family member that contains a ubiquitin binding domain, which is required for proper mitochondrial clearance. Furthermore, loss of this ubiquitin binding domain results in enlarged mitochondria. This phenotype is unique in that most autophagy genes required for mitochondrial clearance do not affect mitochondrial size. For example, loss of the autophagy receptor Ref2p/P62 does not result in enlarged mitochondria.⁷ Vps13D is also the only Vps13 family member in flies that is required for autophagy.⁷ Significantly, mutations in VPS13D have been associated with multiple diseases, including altered septic shock mortality and a unique group of familial neurological movement disorders involving ataxia, chorea and dystonia.^10,11,18-20 Despite these findings, it remains unclear how VPS13D regulates autophagy, mitochondrial morphology and contributes to human diseases.

Here we investigate the relationship between vps13d and genes that regulate autophagy and mitochondrial morphology. We discover that Vps13D acts downstream of Vmp1/EPG-3, a regulator of autophagy and mitochondria and endoplasmic reticulum (ER) contact. Like Vps13D, loss of Vmp1 disrupts autophagy and mitochondrial morphology. Through this relationship, we identify a novel role for Vps13D as a regulator of mitochondria and ER contact in Drosophila and human cell lines, including fibroblasts derived from patients symptomatic for VPS13D associated neurodegenerative disease. Importantly, we also discover that Vps13D interacts with the regulator of mitochondrial fusion Marf (Mitofusins in mammals), and that loss of either marf in flies or MFN2 in patient-derived cell lines suppresses vps13d mutant phenotypes. Our findings establish Vmp1, Vps13D and Marf/MFN2 as important factors in a pathway that regulates inter-organelle contacts in autophagy and mitochondrial morphology, and help explain the role of vps13d in disease.

RESULTS

Vmp1 regulates autophagy, mitophagy and mitochondrial morphology

The essentiality (Figure S1A) and unique role of Vps13D in autophagy among Vps13 family members prompted us to consider if other factors implicated in both autophagy and inter-organelle contact may possess phenotypes that are similar to vps13d. Vmp1 (EPG-3 in C. elegans) is a conserved regulator of autophagy in worms and mammals, and also influences inter-organelle contacts.^21,22 To test if Drosophila Vmp1 (also known as Tango5) has a similar function to Vps13D, we analyzed the function of vmp1 in larval intestine cells where vps13d functions in autophagy, cell size reduction, mitochondrial clearance and maintenance of mitochondrial size. Like vps13d mutant cells,⁷ cells with reduced Vmp1 function that express vmp1 RNAi and green fluorescent protein (GFP) did not accumulate mCherryAtg8a autophagy reporter puncta and were unable to reduce in size, unlike neighboring GFP-negative control cells (Figure 1A and B). This was the same phenotype seen in vps13d RNAi-expressing and loss of function intestine cells.⁷ Similar results were obtained using a distinct RNAi targeting a different vmp1 sequence (Figure S1B and C).

Figure 1. Vmp1 is required for autophagy in Drosophila intestines.

(A) vmp1 RNAi intestine cells (green) 2 hours after pupariation exhibit decreased mCherryAtg8a reporter puncta formation (red) compared to neighboring control cells (non-green). (B) Quantification of mCherryAtg8a puncta in vmp1 RNAi cells (n=8) compared to control cells (n=16). (C) vmp1(Δ) mutant cells (non-red, white dotted line) possess increased Ref2p/p62 puncta (green) compared to neighboring control cells (red) in intestines 2 hours after pupariation. (D) Quantification of Ref2p puncta in vmpl(Δ) mutant (n=9) and control cells (n=8) 2 hours after pupariation. (E) vmp1(Δ) loss-of-function mutant cells (non-red, white dotted line) possess elevated mitochondrial ATP5a puncta (green) compared to neighboring control cells (red) in intestines 2 hours after pupariation. (F) Quantification of ATP5a puncta in vmp1(Δ) mutant (n=6) and control cells (n=16) 2 hours after pupariation. Scales bars in (A), (C) and (E) represent 40 μm. Error bars in (B), (D) and (F) are SEM. Thresholding in (A), (C), and (E) were based on maximizing the quality of signals without over-saturation. Representative of 3 or more independent biological experiments. See also Figure S1.

We next used CRISPR/CAS9 gene editing to create a loss-of-function vmp1 mutant Drosophila named vmp1(Δ) (Figure S1D). Homozygous vmp1(Δ) mutant animals die during development with a small number of animals surviving until the 3^rd instar larval stage. Importantly, an X chromosome duplication containing the vmp1 open reading frame complemented the vmp1(Δ) lethal phenotype (Figure S1E).

We analyzed vmp1(Δ) mutant cells for phenotypes that are similar to homozygous vps13d mutant intestine cells. Consistent with vmp1 RNAi knockdown, intestines with homozygous vmp1(Δ) mutant cells lacking red fluorescent protein (RFP) accumulated the autophagic cargo receptor Ref2p (p62 in mammals) compared to neighboring control cells that possess RFP (Figure 1C and D), indicating that autophagy is impaired. Similar to homozygous vmp1(Δ) mutant cells, Ref2p accumulated in vps13d (MiMic) mutant cells (Figure S1F and G).

Mitochondria are cleared by autophagy during intestine development.⁶ Therefore, we investigated if Vmp1, like Vps13D, is required for clearance of mitochondria in the intestine. Significantly, homozygous vmp1(Δ) mutant intestine cells lacking RFP were unable to clear mitochondria compared to neighboring control cells that express RFP based on persistence of the mitochondrial protein ATP5a (Figure 1E and F). Combined, these data indicate that Vmp1 has similar functions to Vps13D, including the regulation of autophagy and clearance of mitochondria.

We next investigated if the presence of mitochondria in homozygous vmp1(Δ) mutant intestine cells was due to a defect in mitophagy. We used the mito-QC system which utilizes a mitochondrial protein tagged with GFP and RFP to detect when mitochondria are delivered to autolysosomes.^4,5 Control intestines that expressed control luc RNAi throughout the intestine cleared most mitochondria by 2 hours after pupariation as shown by the presence of RFP-positive and GFP-negative puncta (Figure 2A and B). By contrast, intestines that expressed vps13d RNAi or expressed either of two distinct vmp1 RNAi constructs retained mitochondria that were both RFP- and GFP-positive 2 hours after pupariation (Figure 2A and B). In addition, transmission electron microscopy (TEM) analyses revealed enlarged mitochondria in vmp1 RNAi-expressing intestine cells compared to control intestine cells at 2 hours after pupariation (Figure 2C and D). Larger mitochondria were also observed by TEM analyses of intestine cells expressing a different vmp1 RNAi (Figure S2). These data indicate that Vmp1 and Vps13D have similar functions in regulating autophagy, mitophagy and mitochondrial morphology in Drosophila intestines.

Figure 2. Vmp1 is required for mitophagy and normal mitochondrial morphology in Drosophila intestines.

(A) Mito-QC expression was driven by intestine-specific NP1-GAL4 in different genotypes and analyzed in all intestine cells 2 hours after pupariation. Control luciferase (luc) RNAi-expressing cells possessed mostly red puncta (reflecting mitochondria in autolysomes, mitolysosomes), while intestine cells expressing RNAi against either vps13d or 2 distinct vmp1 RNAi expressing constructs (#46667 and #100745) all exhibited yellow puncta, reflecting mitochondria that failed to get cleared by mitophagy. (B) Quantification of the percentage of mitolysosomes to total mitochondria puncta in luc (n=10) RNAi-, vps13d (n=10) RNAi-, vmp1 (#46667) (n=10) RNAi-, and vmp1 (#100745) (n=8) RNAi-expressing cells 2 hours after pupariation. (C) TEM images of cells from intestines expressing either control luciferase (luc) RNAi or vmp1 RNAi (#100745) 2 hours after pupariation. Enlarged regions are outlined by a black box. (D) Quantification of the size of mitochondria in either control luc (n=53) RNAi- or vmp1 (n=51) RNAi-expressing intestine cells 2 hours after pupariation. Scales bars in (A) represent 40 μm. Scale bars in (C) represent 2.0 μm. Error bars in (B) and (D) are SEM. Thresholding in (A) was based on maximizing the quality of signal without over-saturation. Representative of 3 or more independent biological experiments. See also Figure S2.

Vps13D regulates mitochondria and endoplasmic reticulum contact

Vmp1 is a repressor of membrane contact, and the failure to disassemble mitochondria and endoplasmic reticulum (ER) contact alters mitochondrial morphology in vmp1 mutant mammalian and C. elegans cells.²² We investigated if Vmp1 influences mitochondria and ER contact in Drosophila through TEM analyses of intestines 2 hours after pupariation. Intestines with decreased Vmp1 function that express vmp1 RNAi possessed increased contact between mitochondria and ER compared to luciferase (luc) RNAi control cells (Figure 3A and B). The majority of these contact sites were between mitochondria and rough ER. Similar results were obtained with a different vmp1 RNAi line (Figure S3). These results indicate that Vmp1 regulates mitochondria and ER contact in Drosophila.

Figure 3. Vmp1 and Vps13D regulate mitochondria and ER contact.

(A) TEM images of cells from intestines expressing either control luciferase (luc) RNAi or vmp1 RNAi (#100745) 2 hours after pupariation. Enlarged regions are outlined by a black box. Mitochondria (M) and ER (arrows) are indicated. (B) Quantification of the perimeter of mitochondria in contact with ER in either control luc (n=100) RNAi- or vmp1 (n=78) RNAi-expressing intestine cells 2 hours after pupariation. Mitochondria and ER contact sites are defined as being separated by 0.03 μm or less and having a contact length of at least 0.02 μm (16). (C) TEM images of cells from either control +/vps13d (ΔUBA), vps13d (ΔUBA)/vps13d (ΔUBA), or vps13d (ΔUBA)/Df intestines 2 hours after pupariation. (D) Quantification of mitochondria and ER contact in either control +/vps13d (ΔUBA) (n=50), vps13d (ΔUBA)/vps13d (ΔUBA) (n=50), or vps13d (ΔUBA)/Df (n=50) intestines 2 hours after pupariation. (E) TEM images of either wild-type control, VPS13D (ΔUBA), or VPS13D KO (exon 3 deletion) HeLa cells. (F) Quantification of mitochondria and ER contact in either control (n=96), VPS13D (ΔUBA) (n=116), or VPS13D KO (exon 3 deletion) (n=100) HeLa cells. In (A), (C), and (E), arrows represent regions of contact between mitochondria (M) and ER. Scale bars in top panels represent 0.5 μm and bottom panels represent 0.03 μm. Error bars in (B), (D), and (F) are SEM. Representative of 3 or more independent biological experiments. See also Figure S3.

Given the role of Vmp1 in mitochondria and ER contact, as well as the similarities between vmp1 and vps13d mutant cell phenotypes, we investigated if vps13d functions in mitochondria and ER contact by TEM analyses. Intestine cells of either homozygous vps13d (ΔUBA), a mutant lacking the ubiquitin binding domain, or vps13d(ΔUBA)/chromosome deficiency (Df) for the vps13d genomic region had significantly increased mitochondria and ER contact compared to heterozygous vps13d (ΔUBA)/wild type control cells 2 hours after pupariation (Figure 3C and D). Like vmp1 depleted cells, the majority of the ER in contact with mitochondria in Vps13d mutant cells were rough ER.

The function of Vps13D in the regulation of mitochondria and ER contact in Drosophila prompted us to consider if this phenotype is conserved in humans. Therefore, we analyzed HeLa cells that either lack the ubiquitin binding domain, VPS13D(ΔUBA), or are thought to be a strong loss-of-function mutant, VPS13D(KO). Significantly, we found that mitochondria and ER contact were increased in both VPS13D mutant human HeLa cell lines (Figure 3E and F). These results indicate that both vmp1 and vps13d regulate mitochondria and ER contact in Drosophila and human cells.

Mutations in VPS13D have been associated with familial neurological movement disorders, including ataxia, dystonia, and chorea.^{10, 11} Given the conserved function of VPS13D in inter-organelle contact between fly and human HeLa cells, we investigated if patient-derived cells with VPS13D mutations have altered mitochondria and ER contact by TEM. Remarkably, mitochondria in fibroblasts that were derived from the symptomatic VPS13D mutant (G1190D/Q1106*) patient had increased mitochondria and ER contact compared to the mitochondria in fibroblasts derived from a relative (G1190D/+) and unrelated control (Figure 4A and B).¹⁰ In addition, we analyzed mitochondria and ER contact in a second set of fibroblasts derived from an unrelated family with symptoms associated with the VPS13D mutations.⁸ Mitochondria from the symptomatic VPS13D mutant patient from this family (A4210V/Y1803*) also exhibited increased mitochondria and ER contact compared to mitochondria in fibroblasts derived from both an asymptomatic relative (A4210V/+) and a separate unrelated control (Figure 4C and D). A poor correlation exists between the perimeter of mitochondria in Vps13d mutant intestine cells (R² = 0.1456), HeLa cells (R² = 0.00554), and patient fibroblasts (R² = 0.00037) (Figure S4), suggesting that the increase in mitochondria and ER contact in Vps13d mutants was not simply due to increased mitochondrial perimeter. Therefore, VPS13D regulates mitochondria and ER contact, this function is conserved from flies to humans, and this phenotype likely contributes to cell health and neurological disease.

Figure 4. Fibroblasts derived from patients with neurological symptoms associated with VPS13D mutations have increased mitochondria and ER contact.

(A) TEM images of fibroblast cells derived from a family with mutations in VPS13D (Family 1). Cells were derived from either an unrelated donor without mutations in VPS13D (+/+), a relative carrying the G1190D allele for VPS13D (G1190D/+), or a patient with neurological symptoms carrying the G1190D and Q1106* mutations in VPS13D (G1190D/Q1106*). Enlarged regions are outlined by a black box, mitochondria (M) and ER (arrows) are indicated. (B) Quantification of mitochondria and ER contact in VPS13D (+/+) (n=54), (G1190D/+) (n=50), and (G1190D/Q1106*) (n=50) fibroblasts derived from Family 1. (C) TEM images of fibroblast cells derived from a family with mutations in VPS13D (Family 2). Cells were derived from either an unrelated donor without mutations in VPS13D (+/+), a relative carrying the A4210V allele for VPS13D (A4210V/+), or the patient with neurological symptoms carrying the A4210V and Y1803* mutations in VPS13D (A4210V and Y1803*). (D) Quantification of mitochondria and ER contact in VPS13D (+/+) (n=50), (A4210V/+) (n=56), and (A4210V/Y1803*) (n=50) fibroblasts derived from Family 2. In (A) and (C), scale bars in top panels are 0.5 μm and in bottom panels are 0.03 μm. Error bars in (B) and (D) are SEM. Representative of 3 or more independent biological experiments. See also Figure S4.

Vps13D functions downstream of Vmp1 to regulate mitochondrial morphology and mitophagy

The similarities between vmp1 and vps13d mutant phenotypes suggests that these genes may be in the same genetic pathway regulating mitophagy and mitochondrial size. We analyzed Vps13D protein localization in control and homozygous vmp1(Δ) mutant intestine cells 2 hours after pupariation, and found that Vps13D protein puncta were significantly decreased in vmp1 mutant cells compared to neighboring control cells (Figure 5A and B). These results indicate that Vps13D puncta are dependent on Vmp1. We next investigated if Vps13d influences Vmp1. Antibodies do not exist to detect Vmp1 in Drosophila. Therefore, CRISPR/CAS9 was used to tag Vmp1 with GFP on the N terminus (GFP-Vmp1) (Figure S5A). These flies are viable, fertile and complemented the lethal phenotype associated with the vmp1(Δ) mutant. GFP-Vmp1 co-localizes with the ER markers Sec61β and SERCA in intestine cells (Figure S5B). In addition, vmp1 RNAi expression in the entire GFP-Vmp1 larval intestine resulted in retention of mitochondria and increased cell size compared to controls. vmp1 RNAi throughout the intestine also caused an almost complete ablation of GFP signal, verifying that the GFP puncta were indeed GFP-Vmp1 (Figure S5C). Interestingly, vps13d(MiMic) loss-of-function mutant cells did not possess altered GFP-Vmp1 localization (Figure 5C and D), indicating that Vmp1 localization is not dependent on vps13d function. Combined, these results suggest that Vps13d functions downstream of Vmp1.

Figure 5. Vps13d and Vmp1 function in a pathway to regulate mitophagy and mitochondrial morphology.

(A) vmp1 (Δ) loss-of-function mutant cells (non-red, white dotted line) possess fewer Vps13D puncta (green) compared to neighboring control cells (red) in intestines 2 hours after pupariation. (B) Quantification of Vps13D puncta in vmp1(Δ) mutant (n=6) and control (n=16) intestine cells 2 hours after pupariation. (C) vps13d (MiMic) mutant cells (lacking nuclear RFP, white dotted line) do not have altered GFP-Vmp1 in larval intestines 2 hours after pupariation. Antibody against GFP was used to enhance GFP-Vmp1 signal. (D) Quantification of GFP-Vmp1 puncta in vps13d (MiMic) mutant (n=8) and control (n=14) intestine cells 2 hours after pupariation. (E) vmp1(Δ) and vps13d (ΔUBA) double mutant cells (non-red, white dotted line) exhibit similar levels of mitochondrial ATP5a protein compared to neighboring control vmp1(Δ)/+ and vps13d (AUBA) single mutant cells (red) 2 hours after pupariation. (F) Quantification of ATP5a puncta in vmp1(Δ) and vps13d (ΔUBA) double mutant (n=8) and vmp1(Δ)/+ and vps13d (ΔUBA) single mutant (n=14) control intestine cells 2 hours after pupariation. (G) Mito-QC expression was driven by intestine-specific NP1-GAL4 in different genotypes and analyzed in intestine cells 2 hours after pupariation. Control vps13d (ΔUBA)/+ cells possessed mostly red puncta (reflecting mitochondria in autolysomes, mitolysosomes), while vps13d (ΔUBA/ΔUBA) homozygous mutant, vps13d (ΔUBA/ΔUBA) mutant expressing flp, and vps13d (ΔUBA/ΔUBA) mutant with vmp1 RNAi-expressing intestine cells all exhibited large yellow puncta (reflecting mitochondria that fail to be cleared by mitophagy). (H) Quantification of the percentage of mitolysosomes to total mitochondria puncta in vps13d (ΔUBA)/+ (n=10), vps13d (ΔUBA/ΔUBA) (n=10), vps13d (ΔUBA/ΔUBA), UAS-flp (n=11), and vps13d (ΔUBA/ΔUBA), vmp1 RNAi (100745) (n=10), cells 2 hours after pupariation. (I) TEM images of cells from either control vps13d (ΔUBA)/MiMic expressing rfp RNAi or vps13d (ΔUBA)/MiMic expressing vmp1 RNAi intestines 2 hours after pupariation. Insets in (A) and (C) represent enlarged regions, with dashed lines inside representing cell borders. Enlarged regions in (I) are outlined by a black box, mitochondria (M) and ER (arrows) are indicated. (J) Quantification of either mitochondrial size or mitochondria and ER contact in either control vps13d (ΔUBA)/MiMic expressing rfp RNAi (n=55) or vps13d (ΔUBA)/MiMic expressing vmp1 RNAi (#100745) (n=62) intestine cells 2 hours after pupariation. Scale bars in (A), (C), (E), and (G) are 40 μm. Scale bars in top panel of (I) represent 0.5 μm while scale bars in bottom panels represent 0.03μm. Error bars in (B), (D), and (F), (H), and (J) are SEM. Thresholding in (A), (C), (E) and (G) were based on maximizing the quality of signals without over-saturation. Representative of 3 or more independent biological experiments. See also Figure S5.

We investigated the relationship between Vmp1 and Vps13d in the clearance of mitochondria. We compared mitochondrial clearance in vmp1(Δ) and vps13d(ΔUBA) double mutant intestine cells with vmp1(Δ)/+ and vps13d(ΔUBA) single mutant control cells 2 hours after pupariation. Double mutant cells had similar amounts of mitochondrial ATP5a protein compared to neighboring control cells (Figure 5E and F), suggesting that these genes function in the same pathway to clear mitochondria. Consistent with these findings, loss of vps13d function failed to enhance the mitochondrial clearance phenotype caused by expression of vmp1 RNAi throughout the intestine (Figure S5D).

We next used mito-QC to investigate if Vmp1 and Vps13D function in a shared mitophagy pathway. Control intestines that were heterozygous for the vps13d (ΔUBA) mutation cleared most mitochondria by 2 hours after pupariation as shown by the presence of RFP-positive and GFP-negative puncta. By contrast, intestines that were homozygous for the vps13d (ΔUBA) mutation retained mitochondria that were both RFP- and GFP-positive 2 hours after pupariation (Figure 5G and H). Combined knockdown of vmp1 by RNAi in a homozygous vps13d (ΔUBA) mutant background failed to enhance the vps13d mutant mito-QC phenotype (Figure 5G and H), further indicating that vmp1 and vps13d function in the same mitophagy pathway.

To investigate if Vps13D and Vmp1 function in the same pathway to regulate mitochondria and ER contact, we analyzed vps13d (ΔUBA)/DF expressing either vmp1 or control rfp RNAi by TEM. Importantly, the combined reduction of both vmp1 and vps13d function failed to enhance either the increased mitochondrial size or mitochondria and ER contact phenotypes compared to the loss of vps13d alone (Figure 5I and J). These findings indicate that Vps13D and Vmp1 function in the same pathway to regulate mitophagy and mitochondria and ER contact, and that Vps13D functions downstream of Vmp1.

marf/MFN2 suppresses vps13d and vmp1 mutant phenotypes

The role of Vps13D in regulating mitochondria and ER contact sites prompted us to consider the relationship between Vps13D and known regulators of inter-organelle contact. Mfn2 (Marf in flies) has been implicated in mitochondria and ER contact,^23-25 and knockdown of marf suppressed the vps13d knockdown mitochondria size phenotype.⁷ Therefore, we investigated how loss of Vps13D influences Marf in Drosophila. Antibodies against Marf do not work for immunohistochemistry, but can be used for immunoblotting (Figure S6A). Therefore, we analyzed Marf levels in intestines isolated from control and vps13d mutant animals 2 hours after pupariation. Intestines from vps13d (ΔUBA)/DF trans-heterozygous mutants have increased levels of Marf and ATP5a compared to vps13d (ΔUBA)/+ and Df/+ controls 2 hours after pupariation (Figure 6A and B), suggesting that Vps13D influences Marf levels. Similarly, patient fibroblasts incubated with mitochondrial uncoupler to induce mitophagy were unable to degrade ATP5a and MFN2 as effectively as heterozygous control fibroblasts even though MFN1 degradation was unaffected (Figure S6B and C).^8,9

Figure 6. Vps13d functions upstream of Marf to mediate mitochondrial clearance.

(A) Western blot of lysates from vps13d(ΔUBA)/+, Df/+, and vps13d(ΔUBA)/Df intestines 2 hours after pupariation that was probed with antibodies against Marf, ATP5a and Actin. (B) Quantification of relative levels of Marf and ATP5a in vps13d(ΔUBA)/+, Df/+, and vps13d(ΔUBA)/Df intestines 2 hours after pupariation compared to Actin. (C) Intestine cells that overexpress marf using the Act-GAL4 (green) were stained with antibodies against ATP5a (purple) and compared to neighboring control cells (non-green). (D) Quantification of levels of ATP5a puncta in marf overexpressing intestine cells 2 hours after pupariation compared to control cells. (E) Intestines 2 hours after pupariation containing marf(B) loss-of-function mutant cells (non-RFP) were stained with antibody against ATP5a (purple). (F) Quantification of levels of ATP5a puncta in marf(B) loss-of-function mutant cells (n=6) 2 hours after pupariation compared to control cells (n=14). (G) Intestines 2 hours after pupariation containing marf(B) loss-of-function mutant cells (non-RFP) were stained with antibody against Vps13D (purple). (H) Quantification of levels of Vps13D puncta in marf(B) loss-of-function mutant cells (n=6) 2 hours after pupariation compared to control cells (n=13). Scale bars in (C), (E), and (G) represents 40 μm. Error bars in (B), (D), (F), and (H) are SEM. Thresholding in (C), (E), and (G) were based on maximizing the quality of signals without over-saturation. Representative of 3 or more independent biological experiments. See also Figure S6.

We next investigated if Marf interacts with Vps13D in Drosophila. The large size of Vps13D made it difficult to conduct biochemical experiments using exogenous expression in animal and cell culture models. Thus, we used CRISPR to tag the endogenous Drosophila vps13d gene with 3xflag on the C terminus of the open reading frame (Figure S6D). Unlike vps13d mutants, these flies are viable, fertile, and do not have altered mitochondrial morphology in intestine cells at 2 hours after pupariation (Figure S6E). Co-staining of intestine cells with anti-FLAG and anti-Vps13D at 2 hours after pupariation revealed co-localization (Figure S6F). Furthermore, vps13d-3xFLAG pupal lysates had a unique band at the approximate weight of Vps13d-3xFLAG that was absent in the untagged w1118 pupae (Figure S6G). The 3xFLAG epitope was used to immuno-precipitate Vps13D and potential interacting proteins. We used a Marf-specific antibody to reveal the presence of a band corresponding to endogenous Marf in the vps13d-3xflag eluate that was absent in the w1118 negative control eluate (Figure S6H and I). We did not obtain significant enrichment of endogenous COXIV, an inner mitochondrial membrane protein serving as a control for non-specific mitochondrial protein interactions, in the vps13d-3xFLAG eluate. These results suggest a specific physical interaction between Vps13d and Marf.

We next sought to characterize the role that Marf may play in mitochondrial clearance in intestines 2 hour after pupariation. Interestingly, over-expression of Marf inhibited mitochondrial clearance (Figure 6C and D), a phenotype that is similar to vps13d loss-of-function mutants. Similar results were obtained by expression of Marf in all intestine enterocyte cells 2 hours after pupariation (Figure S6J). Unlike vps13d loss-of-function mutants, marf(B) loss-of-function mutant cells did not possess a defect in mitochondrial clearance (Figure 6E and F). In addition, marf(B) mutant cells did not have altered Vps13d puncta (Figure 6G and H), suggesting that Vps13D functions upstream of Marf in the regulation mitochondrial clearance and morphology.

MFN2 is an established mitochondria and ER tether that regulates mitochondrial dynamics and mitophagy.^23-25 Given the physical and genetic relationship between Vps13d, Vmp1, and Marf, we hypothesized that Vmp1 and Vps13D may regulate mitochondria morphology and mitochondria and ER contact sites upstream of Marf. Knockdown of marf suppressed the enlarged mitochondrial phenotypes seen in vps13d (ΔUBA/MiMic) mutants and vmp1 knockdown intestine cells (Figure 7A and B). Knockdown of marf also suppressed the Mito-QC and Ref2p accumulation phenotype in vps13d mutant intestine cells (Figure S7A and B). Consistent with findings in other cell lines,²² knockdown of VMP1 in heterozygous control fibroblasts increased the number of round mitochondria, similar to the VPS13D mutant patient-derived fibroblasts. VMP1 knockdown in patient-derived fibroblasts did not significantly increase the ratio of round mitochondria to tubular mitochondria, suggesting that like in Drosophila intestines, VMP1 and VPS13D are functionally linked in a pathway in human fibroblasts. Significantly, MFN2 knockdown in patient-derived fibroblasts (Figure S7C) also suppressed the abnormal mitochondrial phenotype in VPS13D mutant patient-derived fibroblasts (Figure 7C and D). Interestingly, MFN1 knockdown in patient-derived fibroblasts (Figure S7C) did not suppress this VPS13D associated phenotype (Figure 7C and D). These findings suggest that the mechanistic relationship between VPS13D, VMP1, and Marf/MFN2 are conserved from Drosophila to humans, and that this relationship likely contributes to disease pathology.

Figure 7. Reduction of Marf/Mfn2 function suppresses Vps13D and Vmp1 phenotypes.

(A) vps13d (ΔUBA/MiMic) and vmp1 RNAi-expressing intestines 2 hours after pupariation were stained with antibody against ATP5a (purple) with control rfp RNAi or marf RNAi expression. (B) Quantification of ATP5a puncta size in vps13d (ΔUBA/MiMic) and vmp1 RNAi-expressing intestines 2 hours after pupariation with control rfp RNAi (n=13 for vps13d, n=11 for vmp1) or marf RNAi (n=12 for vps13d and vmp1) expression. (C) Fibroblasts from a patient (mutant) with transheterozygous VPS13D mutations (G1190D/Q1106*) were stained with TOMM20 antibody (green) and compared to heterozygous control (G1190D/+) fibroblasts (control). Control fibroblasts were transfected with negative control mock and VMP1 RNAi and mutant fibroblasts were transfected with mock, MFN2 and VMP1 RNAi. (D) Quantification of mitochondria morphology in control fibroblasts transfected with mock RNAi (n=11) and VMP1 RNAi (n=10) compared to mutant fibroblasts transfected with mock RNAi (n=11), VMP1 RNAi (n=15), and MFN2 RNAi (n=14). (E) Representative TEM images of cells from vps13d (ΔUBA)/(MiMic) intestine cells expressing either rfp (control) or marf RNAi (left panels) 2 hours after pupariation, and VPS13D (A4210V/Y1803*) patient fibroblasts treated with either negative control mock or MFN2 RNAi (right panels). Enlarged regions are outlined by a black box, mitochondria (M) and ER (arrows) are indicated. (F) Quantification of mitochondrial size in vps13d (MiMic)/+ (n=62) intestine cell expressing rfp RNAi, vps13d (MiMic)/+ (n=82) intestine cells expressing marf RNAi, vps13d (ΔUBA)/(MiMic) (n=84) intestine cell expressing rfp RNAi, and vps13d (ΔUBA)/(MiMc) (n=72) intestine cells expressing marf RNAi 2 hours after pupariation. (G) Quantification mitochondria and ER contact in vps13d (MiMic)/+ (n=62) intestine cells expressing rfp RNAi, vps13d (MiMic)/+ (n=82) intestine cells expressing marf RNAi, vps13d (ΔUBA)/(MiMic) (n=84) intestine cells expressing rfp RNAi, and vps13d (ΔUBA)/(MiMic) (n=72) intestine cells expressing marf RNAi 2 hours after pupariation. (H) Quantification of mitochondria and ER contact in VPS13D (A4210/+) heterozygous control fibroblasts treated with mock (n=50) and MFN2 (n=51) RNAi compared to VPS13D (A4210V/Y1803*) mutant fibroblasts treated with mock (n=50) and MFN2 (n=50) RNAi. Scale bar in top panels of A) and C) represents 40 μm, bottom panels represent 10μm. Scale bars in the upper panels of (E) represent 0.5 μm while scale bars in bottom panels represent 0.03μm. Error bars in (B), (D), (F), (G), and (H) are SEM. Thresholding in (A) and (C) were based on maximizing the quality of signals without over-saturation. Representative of 3 or more independent biological experiments. See also Figure S7.

We next investigated if decreased marf/MFN2 function can suppress the vps13d mutant intestine cell mitochondria and ER contact phenotype. Consistent with our previous TEM analyses of mitochondria in vps13d RNAi-expressing intestine cells,⁷ reduction of marf function by RNAi suppresses the enlarged mitochondrial phenotype in vps13d (ΔUBA)/MiMic mutants (Figure 7E and F). Significantly, expression of marf RNAi also suppressed the increased mitochondria and ER contact phenotype in vps13d mutant intestine cells (Figure 7E and G). Importantly, MFN2 knockdown in VPS13D mutant fibroblasts also suppressed the mitochondria and ER contact phenotype (Figure 7E and H). Therefore, these data indicate that Vps13d mechanistically regulates mitochondria and ER contact sites through Marf/MFN2 in Drosophila and human fibroblasts.

DISCUSSION

Mitophagy plays an important role in organism health and disease. vps13d is not only a regulator of mitophagy, but is also required for proper mitochondrial morphology.⁷ However, the pathway and mechanism by which vps13d regulates mitochondrial size and clearance, and subsequently why vps13d is crucial for cell health and its associated diseases, remains elusive. Here, we establish vmp1 as a regulator of vps13d mediated mitophagy and regulation of mitochondria morphology. In doing so, we identify a new function of Vps13D as a regulator of mitochondria and ER contact in multiple systems, including fibroblasts derived from patients with vps13d-associated neurological symptoms.

The regulation of mitochondria and ER contact sites plays important roles in autophagy and mitochondrial morphology. Autophagosomes form at mitochondria and ER contact sites,²⁶ making these sites crucial for autophagy. In yeast, it has been shown that mitochondria and ER contact sites are required for mitophagy.²⁷ Furthermore, mitochondria and ER contact regulate mitochondrial dynamics in cultured human cells,²⁸ and altered mitochondrial dynamics influences vps13d mutant intestine cell mitochondrial size and clearance.⁷ The inability to disassemble mitochondria and ER contact sites, such as is required in Pink1/Parkin mitophagy,²⁵ can explain the mitophagy deficiency and altered mitochondrial morphology phenotypes that we observe in diverse vps13d mutant models.

Here we have identified the first direct link between Vmp1 and the Vps13D family. Vmp1 is an ER localized transmembrane protein that regulates autophagy through disassembly of inter-organelle membrane contacts, where loss of Vmp1 leads to failure in the disassembly of these contact sites.²² Vps13d interacts with proteins that localize to mitochondria and ER contact sites, including Marf.^23,29,30 Consistent with our data in Drosophila, Vps13D was recently reported to interact with Mfn2 in 293 human embryonic kidney cells.²⁹ Marf has been linked to mitophagy by being degraded by the PINK1/PARKIN pathway.^{25, 31,} It may thus be possible that Vps13dfacilitates Marf degradation to decrease mitochondria and ER contact sites, enabling better access for Pink1/Parkin substrates during mitophagy.²⁵ Recent studies have shown that other proteins known to regulate mitophagy and mitochondrial dynamics also localize to mitochondria and ER contact sites.^29,30 Along with Marf, these proteins provide a possible mechanism by which Vps13d regulates mitophagy and mitochondrial morphology.

Our data indicate that Vps13d functions as a link between Vmp1 and regulators of mitochondrial dynamics, such as Marf. In mammalian cell lines, failure to disassemble mitochondria and ER contact sites is believed to contribute to the abnormal spherical mitochondria in vmp1 mutant cells.²² Similarly, VPS13D mutant HeLa cells display a spherical mitochondrial morphology.⁷ Vmp1 is an integral ER membrane protein and localization is unlikely to be affected by Vps13D function,²¹ and this is consistent with our findings (Figure 5C and D). Since Vps13d acts downstream of Vmp1, it is thus likely that Vps13d helps mediate disassembly of membranes at mitochondria and ER contact sites via a mechanism that is similar to Vmp1. Thus, Vps13d may function at a nodal point that links mitochondria and ER contact to influence both mitochondrial dynamics and mitophagy, processes that are mostly studied in isolation. Curiously, Vmp1 also regulates membrane contacts between the ER and other organelles, including endosomes and lipid droplets. The influence of Vmp1 on these organelles does not appear to impact autophagy, suggesting that Vmp1 also has autophagy-independent functions.²² Furthermore, unlike Vps13D, Vmp1 appears to be required for starvation induced autophagy.^7,22 Thus, while Vps13D appears to function downstream of Vmp1 in some cellular contexts, it is likely that Vmp1-dependent and Vps13D-independent processes exist. Investigation of the mechanism by which Vmp1 affects Vps13D puncta localization, including if Vmp1 regulates Vps13D protein levels or stability, may clarify this relationship between Vps13D and Vmp1. Furthermore, while we have focused on the role of Vps13D and Vmp1 in mitophagy, there remains the possibility that these proteins are also responsible for the clearance of other cargos.

Vps13 family members are known to act as lipid transfer proteins.^14,32 Autophagosomes form at mitochondria and ER contact sites, and lipids from both organelles are thought to contribute to the autophagosome membrane.^26,33 As a lipid transporter, Vps13D could facilitate the transfer of important lipids to the forming autophagosome membrane. Given the importance of lipid transfer in organelle signaling pathways,³⁴ Vps13D could also play a role in membrane disassembly by allowing the formation of the autophagosome between the mitochondria and ER.

vps13d loss-of-function mutant phenotypes are different than those of other VPS13 family members. Knockdown of VPS13A in cell lines results in decreased mitochondria and ER contact, these cells do not have spherical mitochondria, which is in direct contrast to phenotypes in VMP1 and VPS13D mutant cells.^22,35 Furthermore, VPS13C does not localize to mitochondria and ER contact sites.¹⁴ These findings suggest VMP1 functions in a pathway with VPS13D, but not other VPS13 family members. In Drosophila, only loss of vps13d, and neither vps13a nor vps13b, results in the same autophagy deficiency phenotype as loss of vmp1.⁷ In addition to being the only VPS13 family member that is essential for survival, VPS13D is the only VPS13 protein with a ubiquitin binding domain, which is crucial for regulating mitochondria morphology and clearance. Even the single yeast VPS13 does not have this ubiquitin binding domain,^7,13 hinting that the divergence of VPS13D from other VPS13 family members may have been the result of an evolutionary need for regulating more complex functions in higher life forms. These observations further illustrate the unique role that VPS13D plays compared to other VPS13 family members.

Mitochondria and ER contact sites are responsible for a wide range of cellular processes. In addition to autophagy, these sites are crucial for mitochondrial biogenesis, stress response, lipid exchange, calcium signaling, intracellular trafficking and immune responses.³⁶ As many cells depend on these processes, it is possible that mitochondria and ER contact disruption in vps13d mutants is why vps13d is such an essential gene (Figure S1A).⁷ Furthermore, Vps13D could regulate membrane contact sites between other organelles, potentially increasing the number of vital roles of Vps13D. For example, over-expression of a peroxisome-localizing variant of Miro increased recruitment of Vps13D to peroxisomes in COS7 cells.³⁷ These findings suggest that in addition to mitophagy, there are still uncharacterized and context-dependent functions of Vps13D.

Dysfunction of mitochondria and ER contacts is believed to contribute to neurodegenerative conditions through impairment of these processes.^28,38-40 Individuals with VPS13D mutations possess a range of symptoms, from asymptomatic to early childhood onset neurological disabilities,^{10, 11} raising the possibility that the severity of disease may be proportional to the extent of altered mitochondria and ER contact. Knockdown of vps13d in Drosophila neurons results in enlarged mitochondria, consistent with Vps13D playing a role in neurological health in Drosophila.⁴¹ There are no known treatments for patients experiencing symptoms from VPS13D mutation-driven diseases. However, the relationship between vps13d and marf/mfn2 raises the possibility that therapies designed to reduce MARF/MFN2 activity could alleviate such symptoms. The link between vps13d and vmp1 also raises the novel possibility that VMP1 may play a role in neurological disease. This link not only includes the diseases VPS13D is involved in, but unique diseases that MFN2 is involved in, such as Charcot-Marie-Tooth disease.⁴² Future studies will determine if the membrane contact function of vps13d and relationship to either vmp1 or marf/MFN2 contributes to other associated diseases.

STAR METHODS

RESOURCE AVAILABILITY

Lead Contact

Please contact the Lead Contact, Dr. Eric Baehrecke, for additional information and requests for resources and reagents (eric.baehrecke@umassmed.edu).

Materials Availability

Please contact the Lead Contact to request all strains and protocols pertaining to this study.

Data and Code Availability

Please contact the Lead Contact to obtain any data or related images generated in this study.

EXPERIMENTAL MODELS AND SUBJECT DETAILS

Drosophila strains

Drosophila melanogaster strains used in this study are listed in the Key Resources Table. Drosophila strains used for each experiment are listed in Table S1. w¹¹¹⁸ were used as controls. All flies were raised on standard cornmeal/molasses/agar media at 25°C.

KEY RESOURCES TABLE

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Vps13D	Eric Baehrecke	[7]
ref(2)p	Gabor Juhasz	[46]
ATP synthase complex V	Abcam	Ab14748
GFP	Abcam	ab13970
SERCA	Mani Ramaswami	[47]
Actin	Proteintech	60008-1-Ig
TOMM20	Abcam	ab78547
Marf	Alexander Whitworth	[31]
COX IV	Abcam	ab14744
FLAG	Abcam	Ab1162
FLAG M2	Millipore Sigma	F1804
MFN2	Abnova	H00009927-M03
MFN1	Proteintech	13798-1-AP
anti-mouse AlexaFluor 647	Invitrogen	A-21235
anti-rabbit Alexafluor 546	Invitrogen	A-11035
anti-chicken AlexaFluor 488	Invitrogen	A-11039
anti-mouse Alexa Fluor 488	Invitrogen	A-11029
Oregon Green 488 goat anti-rabbit	Molecular Probes	O-6381
Bacterial and Virus Strains
E. coli/One Shot TOP10	Invitrogen	C404010
Chemicals, Peptides, and Recombinant Proteins
Fetal bovine serum	Gemini Bio Products	100-106
DMEM	Life Technologies	31053-028
Penicillin-streptomycin	GIBCO	15140122
PBS	GIBCO	70011
25% Glutaraldehyde	Electron Microscopy Sciences	16220
Sodium cacodylate	Electron Microscopy Sciences	11650
osmium tetroxide	SPI	2601
propylene oxide	SPI	75-56-9
SPI-pon/Araldite 6005 epoxy embedding kit	SPI	02635-AB
Paraformaldehyde	Electron Microscopy Sciences	15710
Dimethyl sulfoxide	Sigma	D2438
Triton x-100	Sigma	T8787
Vectashield	Vector Laboratories	H-1200
Tris-HCl	Sigma	T3253
NaCl	Fisher	BP358-212
Goat serum	Sigma	G9023
EDTA	Quality Biological	351-027-101
3xFLAG peptide	Sigma	F4799-4MG
Carbonyl cyanide 4-(trifluoromethoxy)phenylhydrazone (FCCP)	Sigma	C2920
Antimycin A from Streptomyces sp.	Sigma	A8674
Oligomycin A	Sigma	75351
Critical Commercial Assays
Deposited Data
Project Achiles	Broad Institute	[48][49]
Experimental Models: Cell Lines
HeLa	ATCC	CCL-2
HeLa VPS13D ゔUBA deletion (ΔUBA)	Eric Baehrecke	[7]
HeLa VPS13D exon 3 deletion (KO #45)	Eric Baehrecke	[7]
Family 1 Patient Fibroblasts (G1190D/Q1106*): UM1.2	Margit Burmeister	[10]
Family 1 Heterozygote Fibroblasts (Q1106*/+): UM1.17	Margit Burmeister	[10]
Family 1 Unrelated Fibroblasts (+/+): UMCtrl1	Margit Burmeister	[10]
Family 2 Patient Fibroblasts (A4210V/Y1803*): LUB1.1	Katja Lohmann	[10]
Family 2 Heterozygote Fibroblasts (Y1803*/+): LUB1.3	Katja Lohmann	[10]
Family 2 Unrelated Fibroblasts (+/+): LUBCtrl1	Katja Lohmann	[10]
Experimental Models: Organisms/Strains
Drosophila melanogaster/w1118	Bloomington Drosophila stock center	5905
Drosophila melanogaster/y1 w1118 hsFlp; If/CyO; Act>CD2>GAL4, UAS–nlsGFP/TM6B	Eric Baehrecke	N/A
Drosophila melanogaster/y1 w1118 hsFlp; pmCherry–Atg8a; Act>CD2>GAL4, UAS–nlsGFP/TM6B	Eric Baehrecke	N/A
Drosophila melanogaster/w1118; P{GD15932}v46667	VDRC	46667
Drosophila melanogaster/w1118; P{KK108366}VIE-260B	VDRC	100745
Drosophila melanogaster/w1118; P{KK105681}VIE-260B	VDRC	105261
Drosophila melanogaster/w1118; P{KK105681}VIE-260B; P{GD15932}v46667	Eric Baehrecke	N/A
Drosophila melanogaster/ y1 sc* v1 sev21; P{TRiP.HMS01784}attP2	Bloomington Drosophila stock center	38320
Drosophila melanogaster/ y1 v1; P{TRiP.JF01355}attP2	Bloomington Drosophila stock center	31603
Drosophila melanogaster/ y1 sc* v1 sev21; P{TRiP.HMS05713}attP40	Bloomington Drosophila stock center	67852
Drosophila melanogaster/ y1 sc* v1 sev21; P{TRiP.HMS05713}attP40; Mi{MIC}Vps13DMI11101/TM6B	Eric Baehrecke	N/A
Drosophila melanogaster/ y1 w*; Mi{MIC}Vps13DMI11101/TM6B	Bloomington Drosophila stock center	56282
Drosophila melanogaster/ w1118; Df(3L)BSC631/TM6C, cu1 Sb1	Bloomington Drosophila stock center	25722
Drosophila melanogaster/ w1118; Df(3L)BSC631/TM6B	Eric Baehrecke	N/A
Drosophila melanogaster/w1118; Dp(1;3)RC025, PBac{RC025}VK00033/TM6C, Sb1	Bloomington Drosophila stock center	38486
Drosophila melanogaster/ w1118; P{20XUAS-tdTomato-Sec61β}attP2	Bloomington Drosophila stock center	64747
Drosophila melanogaster/ y1 w*; P{Act5C-GAL4}25FO1/CyO, y+	Bloomington Drosophila stock center	4414
Drosophila melanogaster/ y1 w* MarfB P{neoFRT}19A/FM7c, P{GAL4-Kr.C}DC1, P{UAS-GFP.S65T}DC5, sn+	Bloomington Drosophila stock center	67154
Drosophila melanogaster/y1 w*; P{UAS-Marf.HA.S}3/T(2;3)TSTL, CyO: TM6B, Tb1	Bloomington Drosophila stock center	67157
Drosophila melanogaster/w*; P{GawB}Myo31DFNP0001, P{UAS-mito-HA-GFP.AP}2/CyO	Eric Baehrecke	N/A
Drosophila melanogaster/w*; P{w[+mW.hs]=FRT(w[hs])}2A	Bloomington Drosophila stock center	1997
Drosophila melanogaster/ w1118; P{Ubi-GFP.nls}3L1 P{Ubi-GFP.nls}3L2 P{FRT(whs)}2A	Bloomington Drosophila stock center	5825
Drosophila melanogaster/ P{hsFLP}12, w1118; P{Ubi-GFP.nls}3L1 P{Ubi-GFP.nls}3L2 P{FRT(whs)}2A	Eric Baehrecke	N/A
Drosophila melanogaster/ P{Ubi-mRFP.nls}1, w1118, P{neoFRT}19A, P{hsFLP}12	Eric Baehrecke	N/A
Drosophila melanogaster/w*; P{w[+mC]=Ubi-GFP.D}61EF, P{w[+mW.hs]=FRT(w[hs])}2A	Bloomington Drosophila stock center	1626
Drosophila melanogaster/ w1118;; P{w[+mW.hs]=FRT(w[hs])}2A, Mi{MIC}Vps13DMI11101/TM6B	Eric Baehrecke	[7]
Drosophila melanogaster/w*; P{w[+mC]=UAS-FLP.D}JD2	Bloomington Drosophila stock center	4540
Drosophila melanogaster/w*; P{GawB}Myo31DFNP0001 / CyO, P{UAS-lacZ.UW14}UW14	Drosophila Genetic Resource Center	112001
Drosophila melanogaster/w*; P{GawB}Myo31DFNP0001; vps13d (ΔUBA)/TM6B	Eric Baehrecke	N/A
Drosophila melanogaster/ P{UAS-mito-QC}attP16	Alexander J. Whitworth	[5]
Drosophila melanogaster/ P{UAS-mito-QC}attP16, P{GawB}Myo31DFNP0001	Eric Baehrecke	N/A
Drosophila melanogaster/ P{UAS-mito-QC}attP16, P{GawB}Myo31DFNP0001; vps13d (ΔUBA)/TM6B-GFP	Eric Baehrecke	N/A
Drosophila melanogaster/ w1118; P{UAS-mito-HA-GFP.AP}2/CyO	Bloomington Drosophila stock center	8442
Drosophila melanogaster/ vps13d (ΔUBA)/TM6B	Eric Baehrecke	[7]
Drosophila melanogaster/w*; P{w[+mC]=UAS-FLP.D}JD2; vps13d (ΔUBA)/TM6B	Eric Baehrecke	N/A
Drosophila melanogaster/w*; P{KK108366}VIE-260B; vps13d (ΔUBA)/TM6B	Eric Baehrecke	N/A
Drosophila melanogaster/w*; P{KK108366}VIE-260B; Mi{MIC}Vps13DMI11101/TM6B)/TM6B	Eric Baehrecke	N/A
Drosophila melanogaster/w*; P{KK108366}VIE-260B; {w[+mW.hs]=FRT(w[hs])}2A, Mi{MIC}Vps13DMI11101/TM6B	Eric Baehrecke	N/A
Drosophila melanogaster/ P{Ubi-mRFP.nls}1, w1118, P{neoFRT}19A, P{hsFLP}12;; vps13d (ΔUBA)/TM6B	Eric Baehrecke	N/A
Drosophila melanogaster/ w1118, vmp1(Δ)	This paper	N/A
Drosophila melanogaster/ w1118, P{neoFRT}19A, vmp1(Δ)	This paper	N/A
Drosophila melanogaster/ w1118, P{neoFRT}19A, vmp1(Δ);; vps13d (ΔUBA)/TM6B	This paper	N/A
Drosophila melanogaster/ w1118, gfp-vmp1	This paper	N/A
Drosophila melanogaster/ w1118, gfp-vmp1;; P{w[+mW.hs]=FRT(w[hs])}2A, Mi{MIC}Vps13DMI11101/TM6B	This paper	N/A
Drosophila melanogaster/ w1118, gfp-vmp1; P{GawB}Myo31DFNP0001	This paper	N/A
Drosophila melanogaster/ P{hsFLP}12;; P{Ubi-mRFP.nls}3L, P{w[+mW.hs]=FRT(w[hs])}2A	Eric Baehrecke	N/A
Drosophila melanogaster/ P{hsFLP}12; P{GawB}Myo31DFNP0001; P{Ubi-mRFP.nls}3L, P{w[+mW.hs]=FRT(w[hs])}2A	Eric Baehrecke	N/A
Drosophila melanogaster/w1118; P{KK105681}VIE-260B; Mi{MIC}Vps13DMI11101/TM6B	Eric Baehrecke	N/A
Drosophila melanogaster/w1118;; vps13d-3xFLAG	This paper	N/A
Oligonucleotides
Primer to design sgRNA1: TGTTGTTGTGACGATTGCTC	Integrated DNA Technologies	N/A
Primer to design sgRNA2: TTACGGGACTAGAAAATCAG	Integrated DNA Technologies	N/A
Primer to design sgRNA3: TGCTGTGACATTTAAGCGGT	Integrated DNA Technologies	N/A
Primer to design sgRNA4: CGAATGCTGTGACATTTAAG	Integrated DNA Technologies	N/A
Single-stranded donor DNA to design vmp1(Δ): CAAAAACCGTGAAAAACACAGCCGCTTGCAAGCCAACCGCTTAAATGTCACAGCATTCGAAAAAGGAAACCACCAGCAGCAGCGGTAGCCATGCGGCCCAAAAATCAGTGGATCAGTCTTCTGTTGATCTTTTTCCCCAACCCTTGGTTTTCGTTTAAATTATTTTGTATATTGCGGCTTTCGCTCTTAGTCAAATGATG	Integrated DNA Technologies	N/A
Primer to screen and verify 5’ of vmp1(Δ) and gfp-vmp1: CTTTTCGAATCGCCGGCATTTACATCAC	Integrated DNA Technologies	N/A
Primer to screen and verify 3’ of vmp1(Δ) and gfp-vmp1: CATCATTTGACTAAGAGCGAAAGCCGC	Integrated DNA Technologies	N/A
Primer to verify 3’ N terminus vmp1 of gfp-vmp1: CTGATGCTGTTGCTATTGCCGTTTCC	Integrated DNA Technologies	N/A
siRNA NC negative (mock) control: UUCUCCGAACGUGUCACGUTT	Hong Zhang	[22]
siRNA Human VMP1: 5’- GGAAUGGACCUCAAAAUUATT-3’	Hong Zhang	[22]
siRNA SMARTpool for Human MFN2: GACUAUAAGCUGCGAAUUA CAUGAGGCCUUUCUCCUUA GCAACUCUAUCGUCACAGU GGUGGACGAUUACCAGAUG	Dharmacon	N/A
siRNA SMARTpool for Human MFN1: CGAUGAAGUAAACGCCUUA CAUGAUAGGAGGAAACGAA CAGAAUAUAUGGAAGACGU GGAAGUUCUUAGUGCUAGA	Dharmacon	N/A
Recombinant DNA
Plasmid: U6droBsagRNA	Drosophila Genomics Resource Center	1341
Plasmid: pCR™2.1-TOPO® vector	Thermo Fisher Scientific	K450002
Software and Algorithms
ImageJ	NIH	https://imagej.nih.gov/ij/
Zen Black	Zeiss	N/A
Prism	Graphpad Software, Inc.	https://www.graphpad.com/scientific-software/prism/
Biorender	Biorender	https://biorender.com/
Other

Human cell lines

All cells used in this study are listed in Key Resources Table. Cells were cultured at 37°C in 5% CO₂ in DMEM supplemented with 5% FBS and Penicillin/Streptomycin, as previously described.⁷ VPS13D mutant HeLa cells were designed as previously described.⁷ Fibroblasts from patients with VPS13D mutations were obtained and cultured as described.¹⁰ Cells were transfected with the Lipofectamine™ RNAiMAX Transfection Reagent (Invitrogen) following manufacturer protocol and processed 2 days later.

METHOD DETAILS

vmp1(Δ), gfp-vmp1, and vps13d-3xFLAG fly construction

vmp1 loss-of-function, vmp1(Δ) and N terminal GFP-tagged (gfp-vmp1) vmp1 strains were edited using CRISPR/Cas9.⁴³ All germline injections were done by the UMass Medical School CRISPR Core. Oligonucleotide donor used to design gfp-vmp1 is listed in Table S2. All other primers and oligonucleotides used in this study are listed in the Key Resources Table. For vmp1(Δ), the following sgRNA targeting sequences were used (5’ to 3’): sgRNA1: TGTTGTTGTGACGATTGCTC, sgRNA2: TTACGGGACTAGAAAATCAG. A 200bp ultramer donor with 100bp regions flanking the site of the deletion designed by IDT (San Diego, California) was used to facilitate the deletion, resulting in a single female fly with the deletion that was validated by DNA sequence. Dp(1;3)RC025 from the Bloomington Duplication Project was used to verify that lethality was due to loss of the vmp1. For gfp-vmp1, the following sgRNA targeting sequences were used (5’ to 3’): sgRNA3: TGCTGTGACATTTAAGCGGT, sgRNA4: CGAATGCTGTGACATTTAAG. A 2kb gblock with 1kb regions flanking the site of insertion and the GFP open reading frame was designed by IDT (San Diego, California) and used to tag the N terminal of vmp1 with gfp. A single female fly containing the insertion was collected, and validated by DNA sequencing. For vps13d-3xflag, the following sgRNA targeting sequence was used (5’ to 3’): sgRNA5:TTTATAAAATGCAATAGGT. A 2kb region flanking the C terminal of genomic vps13d was amplified by PCR and site-directed mutagenesis was used to insert the 3xflag sequence in frame immediately before the stop codon. This fragment was inserted into a TOPO vector via TOPO cloning and sequenced to ensure no additional mutations were present and was used to tag the C terminal of vps13d with 3xflag. A single female fly containing the insertion was collected and validated by DNA sequencing.

Induction of mosaic RNAi/mutant cell clones and whole intestine RNAi expression

Mosaic GFP positive RNAi- and transgene-expressing cell clones and fluorescentnegative cell clones were induced as described.⁷ To induce mosaic vmp1(Δ) and vps13d(MiMic) loss-of-function mutant cell clones, we used the hsflp, FRT19A, mRFP and hsflp;;FRT2A, Ubi-nlsGFP flies and crossed them with vmp1(Δ) FRT19A/ FM7i-pAct-GFP and vps13d(MiMic) FRT2A/TM6B flies, respectively. Mosaic clonal analyses were only used as indicated, otherwise entire mutant or organ specific RNAi animals were used. 8-hour eggs lays were heat shocked for 90 minutes at 37°C. Myo31DFNP0001 (NP1-GAL4) was used to drive expression of transgenes, including RNAi and the MitoQC reporter, throughout the entire larval intestine.⁴⁴

Dissection and immuno-labeling of Drosophila larval intestines and Cell Lines

White prepupae were collected and allowed to develop on wet filter paper for 2 hours prior to dissection. Intestines were immuno-stained as previously described with modifications.⁷ Intestines were removed in cold PBS before being placed in 4% paraformaldehyde solution for fixation at 4°C overnight. Intestines were washed twice with PBS and then twice with 0.1% PBSTx before blocking in 5% normal goat serum for 90 minutes and incubation with primary antibody in 0.1% PBSTx overnight. Intestines were then stained with secondary antibody for 3 hours before nuclei staining and mounting. All antibodies used in this study are listed in Key Resources Table. The following primary antibodies were used: rabbit anti-ref(2)p (1:1000, from Gabor Juhasz), mouse anti-ATP synthase complex V (1:1000, Abcam #ab14748), anti-GFP (1:1000, Abcam # ab13970), rabbit anti-TOMM20 (1:200, Abcam # ab78547), rabbit anti-SERCA (1:1000, from Mani Ramaswami), rabbit anti-FLAG (1:1000, Abcam #Ab1162), and anti-Vps13D (1:50).⁷ The following secondary antibodies were used: anti-mouse AlexaFluor 647 (Invitrogen #A-21235), anti-rabbit Alexafluor 546 (Invitrogen # A-11035) and anti-chicken AlexaFluor 488 (#A-11039). Nuclei were stained with Hoescht (Invitrogen) and samples were mounted with Vectashield (Vector Lab). Intestines expressing mCherryAtg8a puncta driven by the Atg8a promoter were fixed overnight at 4°C in 4% paraformaldehyde before being imaged the next day.⁴⁵ Cell lines were fixed in 4% paraformaldehyde for 30 minutes at room temperature, permeabilized with digitonin for 15 minutes, and blocked in 5% normal goat serum in PBS before overnight staining with primary antibody overnight at 4°C. Cells were then washed and stained with secondary antibody before mounting in Vectashield with DAPI. Images were acquired using a Zeiss LSM 700 Axioobserver confocal microscope with Plan apochromate 63x/1.40 NA oil immersion objective. Signal in images were thresholded to only include distinct puncta and gaussian blur with sigma 0.8 was applied to reduce background noise. Puncta were quantified using Fiji (ImageJ) and only the brightest 2.5% of puncta were included in quantification.

Transmission electron microscopy

Transmission electron microscopy (TEM) was conducted as previously described with modifications.⁷ Intestines were dissected in PBS (GIBCO) 2 hours after pupariation and fixed in a solution of 2.5% glutaraldehyde and 2% paraformaldehyde in 0.1M sodium cacodylate buffer, pH 7.4 (Electron Microscopy Sciences) for 1 hour at room temperature followed by overnight fixation at 4°C in fresh fix. Intestines were washed in 0.1M sodium cacodylate buffer, pH 7.4, post-fixed in 1% osmium tetroxide in distilled water for 1 hour at room temperature and washed in distilled water. Preparations were stained en bloc in 1% aqueous uranyl acetate for 1 hour at 4°C in the dark, washed in distilled water, dehydrated through a graded ethanol series, treated with propylene oxide and infiltrated in SPI-pon/Araldite for embedding. We cut ultrathin sections on a Leica UC7 microtome. Sections were stained with uranyl acetate and lead citrate and examined on a Phillips CM10 TEM. Images were taken down the length of the anterior region of the midgut to ensure an unbiased approach. For each genotype, at least 3 intestines were embedded and sectioned for analyses and quantification. We reviewed all images and selected representative images for analyses.

For cell culture, plated cells were prefixed in 50% media: 50% fix, 2.5% glutaraldehyde and 2% paraformaldehyde in 0.1M sodium cacodylate buffer, pH 7.4 (Electron Microscopy Sciences) for 5 minutes followed by fixation in full fix for 1 hour at room temperature. Cells were then washed with 0.1M cacodylate buffer, pH 7.4, post-fixed in 1% osmium tetroxide in distilled water for 1 hour at room temperature and washed in distilled water. Preparations were stained en bloc in 1% aqueous uranyl acetate over night at 4°C in the dark and then washed in distilled water. The cells were then scraped and pelleted. Cell pellets were embedded in agarose, dehydrated through a graded ethanol series, treated with propylene oxide and infiltrated in SPI-pon/Araldite for embedding. We cut ultrathin sections on a Leica UC7 microtome. Sections were stained with uranyl acetate and lead citrate and examined on a Phillips CM10 TEM. For each cell line, at least (3) 10 cm² dishes at 60-80% confluency were embedded independently of each other and sectioned in an unbiased manner for analyses and quantification.

Western Blot and Immunoprecipitation

Tissue was lysed in 1X Laemli Sample Buffer diluted in RIPA lysis buffer (10mM Tris-Cl PH 8.0, 1mM EDTA PH 8.0, 0.5mM EGTA, 2.4mM Sodium Deoxycholate 140mM Sodium Chloride) at a ratio of 10μL lysis buffer per intestine and 30μL per whole pupa. Samples were crushed in solution using a plastic pestle for 30 seconds before being boiled at 99°C for 6 minutes. Samples were run on 7.5% polyacrylamide gel, transferred onto 0.45μm PVDF membranes (Millipore Sigma), and probed with antibodies using standard protocols. Primary antibodies used were mouse anti-FLAG (1:1000, Millipore Sigma), rabbit anti-Marf (1:1000, from Alexander Whitworth) mouse anti-CoxIV (1:1000, Abcam), mouse anti-Actin (1:1000, Proteintech), mouse anti-MFN2 (1:1000, Abnova #H00009927-M03), anti-MFN1 (1:1000, Proteintech #13798-1-AP) mouse anti-ATP synthase complex V (1:1000, Abcam).

For immunoprecipitations, 2-hour-old pupae were lysed in RIPA lysis buffer supplemented with 1mM NEM, 1mM PMSF and Halt Protease Inhibitor Cocktail (Thermo Fisher) at a ratio of 16 pupae per 250μL lysis buffer. Pupae were crushed with a plastic pestle for 30 seconds and incubated on ice for 30 minutes before being centrifuged at 4°C at 13,000rpm for 10 minutes. Supernatant was filtered through 0.45μm Cellulose Acetate filters (Millipore Sigma). 30μL of filtered supernatant was diluted in 10μL of 4x Laemli Sample Buffer (Biorad), boiled for 6 minutes at 99°C and used as input. 200μL of filtered supernatant (approximately 1mg protein) was used for immunoprecipitation. 40μL of anti-FLAG M2 magnetic bead slurry (Millipore Sigma) warmed to room temperature was washed twice with RIPA buffer before incubation with filtered supernatant for 2 hours at 4°C on a rotator. Following incubation, supernatant was discarded and beads were washed 4 times with 1mL 0.1% PBST. Beads were eluted with 20μL 1X Laemli Sample Buffer diluted in RIPA lysis buffer and boiled for 6 minutes at 99°C. 20μL of input and eluate was run on 7.5% polyacrylamide gel for Western Blot analysis.

QUANTIFICATION AND STATISTICAL ANALYSES

Quantification of mitochondria morphology in fibroblasts were blindly characterized as tubular or spherical. All other quantifications, including cell size and puncta in immunofluorescence images were conducted using ImageJ, as described previously.⁷ Sample sizes (n) in immunofluorescent experiment quantifications represent number of cells.

TEM analyses of mitochondria area was manually calculated by individually analyzing mitochondria using ImageJ. For determination of mitochondria and ER contact sites, mitochondria and ER were individually identified on TEM sections and distance between mitochondria and ER was manually measured using ImageJ. Regions of ER that were within 30nm of mitochondria and had a contact length of at least 20nm were identified as mitochondria and ER contact sites as previously defined.²² The length of these sites of contact were manually measured for each mitochondria and compared with the entire mitochondria perimeter to determine the percentage in contact. Regions for analyses were randomly selected from embedded blocks for each sample. For each analysis, at least 50 mitochondria from at least 20 sections and 3 independent experiments were used for quantifications. Sample sizes (n) for TEM quantifications represent number of mitochondria.

All experiments are representative of 3 independent replicates. No statistical methods were used to predetermine sample sizes. Preliminary experiments were conducted to achieve similar sample sizes as previous published studies using our model systems. Animals were not excluded for statistical analyses. Researchers were not blinded. Unless otherwise stated, p values were calculated using the two-tailed student t-test without Welch’s correlation for experiments where a single comparison was made. For experiments with multiple comparisons, p values were calculated using one-way ANOVA with Tukey post-hoc analysis. p values greater than 0.05 were considered non-significant (n.s.). All bars are means and error bars are SEM unless otherwise.

Highlights.

• Vps13D and Vmp1 are linked as regulators of autophagy and mitochondrial morphology

• Vps13D, like Vmp1, regulates mitochondria and ER contact sites

• Vps13D regulates mitophagy and mitochondrial morphology downstream of Vmp1

• Vps13D mitochondria and ER contact phenotypes depend on Marf/MFN2