Fly strains

A Drosophila strain on the Oregon-R-P2 background was used as the wild type. MICOS-RNAis were expressed in the indirect flight muscle by Actin88F-GAL4 (Bloomington 38459). The UAS-RNAi used in the study were P{TRiP.HMS05459}attp40 (for CG5903/MIC26-MIC27; Bloomington 66933), P{TRiP.HMJ30307}attp40 (for Milton/MIC60; Bloomington 63994), and P{TRiP.GLC01383}attp2 (for QIL1/MIC13; Bloomington 44634). The genotypes of the flies are listed in Table 1.

TFAM-APEX2 knock-in flies were generated by CRISPR/Cas9-mediated genome editing and homology-dependent repair using a guide RNA(s) and a dsDNA plasmid donor. The PBac system was used to facilitate genetic screening (Well Genetics). The construct design was detailed as follows.

Guide RNA Primers: Sense oligo 5′- CTTCGCCAAAGCCCCGCAAGACGC;

Antisense oligo 5′- AAACCGCGTCTTGCGGGGCTTTGGC

PAM mutation: GCCAAAGCCCCGCAAGACGC[TGG] CTG→CTC/ L→L

Upstream Homology Arm: 1083 bp, −1086 to −4 nt relative to stop codon of TFAM

Forward oligo 5′- TGCCAATCCCCAGATTACCAC;

Reverse oligo 5′-TATATCTTTGGAGGCGAGCGT

Downstream Homology Arm: 1026 bp, +1 to +1028 nt relative to stop codon of TFAM

Forward oligo 5′-TTGTAGCTGCTCGGCCCGC;

Reverse oligo 5′-AAATGATGCAGAAGTGGCT

TFAM-APEX2 on a MICOS-RNAi background was expressed in the indirect flight muscle by Actin88F-GAL4 (Bloomington 38459).

Thin-section TEM for morphological observation

The experiments were carried out as previously described with some modifications (Jiang et al., 2017a,b, 2020). Flies were anesthetized on ice and embedded in 4% low melting agarose in 0.1 M phosphate buffer. Embedded flies were then sectioned at 100 μm with a vibrating blade microtome (Leica VT1200S) and fixed in 2% glutaraldehyde in buffer containing 0.1 M sodium cacodylate with 2 mM CaCl₂, pH 7 for 60 min followed by washing (2 min in the buffer, five times). The sections were post-fixed in 2% osmium tetroxide for 30 min followed by washing (2 min in the buffer, five times), after which samples were incubated in 2% uranyl acetate overnight. After dehydration in ascending percentages of ethanol, the specimens were infiltrated and embedded in Spurr's resin and polymerized at 65°C for 16 h. The specimen blocks were trimmed and sectioned using an ultramicrotome. The sections were stained with 2% uranyl acetate for 10 min, Reynold's lead citrate for 4 min, and subjected to TEM inspection.

The mitochondria size distribution was determined using Amira-Avizo 3D visualization and analysis software (Thermo Fisher Scientific), where individual mitochondria were defined manually and the areas were output and analyzed (Fig. S2a–d; Fig. 1F). The analysis showed in Fig. 1f included 113, 193, 255, and 171 of mitochondria of the control, CG5903/MIC26-MIC27-, Mitofilin/MIC60-, and QIL1/MIC13-knockdown in the analysis, respectively.

Immunofluorescence staining

The experiments were carried out as previously described with some modifications (Macchi et al., 2013). Fly thoraxes were dissected into halves in fixation buffer containing 4% paraformaldehyde and 1% Triton X-100 in PBS and fixed for 20 min at room temperature (RT) without shaking. The specimens were washed with 0.1% Triton X-100 in PBS for 20 min at RT three times. After blocking with 5% normal goat serum (Jackson ImmunoResearch) in 0.1% Triton X-100 in PBS for 2 h at RT, the specimens were stained with primary mouse anti-ATP5A (1:1000, Abcam 14748) and/or primary mouse anti-GABARAP+GABARAPL1+GABARAPL2 antibody (1:500, Abcam ab109364) in blocking buffer overnight at 4°C. After washing with 0.1% Triton X-100 in PBS for 20 min at RT three times, the specimens were stained with secondary antibody (1: 1000 dilutions, anti-mouse IgG Alexa-488, Jackson ImmunoResearch, or Alexa Fluor 647 Phalloidin, Invitrogen A22287) in blocking buffer overnight at 4°C. After washing with 0.1% Triton X-100 in PBS for 20 min at RT three times, the specimens were mounted on the glass slides for imaging. Volumes of 84.2×84.2×5 μm³ were imaged by confocal using LMS880, Zeiss.

JC-1 staining

The experiments were carried out as previously described with some modifications (Macchi et al., 2013). Fly thoraxes were dissected into halves in Schneider's medium containing 1% cyclodextrin and stained with 8 μM JC-1 (Thermo Fisher Scientific) for 30 min. After washing with Schneider's medium for 5 min two times, the specimens were mounted in Schneider's medium and imaged by confocal microscopy (LMS880, Zeiss). Triplicates of volumes of 84.2×84.2×5 μm³ were analyzed. The red and green fluorescent signals from JC1 were measured, and the ratios were calculated using Imaris image analysis software (Bitplane).

LysoTracker staining

The experiments were carried out as previously described with some modifications (Macchi et al., 2013). Fly thoraxes were dissected into halves in Schneider's medium containing 1% cyclodextrin and stained with 1 μM LysoTracker Red DND-99 (Thermo Fisher Scientific) for 5 min. After washing with Schneider's medium for 5 min two times, the specimens were mounted in Schneider's medium and imaged by confocal microscopy (LMS880, Zeiss). Triplicates of volumes of 84.2×84.2×5 μm³ were analyzed. The red fluorescent signals from LysoTracker were measured, and the corresponding volumes were calculated using Imaris image analysis software (Bitplane).

TUNEL staining

The experiments were carried out as previously described with some modifications (Macchi et al., 2013). Fly thoraxes were dissected into halves in fixation buffer containing 4% paraformaldehyde and 1% Triton X-100 in PBS and fixed for 20 min at RT without shaking. The specimens were washed with 0.1% Triton X-100 in PBS for 20 min at RT three times. After blocking with 5% normal goat serum (Jackson ImmunoResearch) in 0.1% Triton X-100 in PBS for 2 h at RT, the specimens were washed and stained with in situ cell death detection kit (Roche) reagents at 37°C for 1 h. The positive control specimens were first incubated with DNase I (2000 U/ml) for 10 min at RT. The negative control specimens were incubated without enzyme terminal transferase. Specimens were washed and stained with primary mouse anti-dsDNA (1:1000, Abcam 27156) in blocking buffer overnight at 4°C. After washing with 0.1% Triton X-100 in PBS for 20 min at RT three times, the specimens were s stained with secondary anti-mouse IgG Alexa-594 (1:500, Jackson ImmunoResearch) or Alexa Fluor 647 Phalloidin (1:1000, Invitrogen A22287) in blocking buffer overnight at 4°C. After washing with 0.1% Triton X-100 in PBS for 20 min at RT three times, the specimens were mounted on glass slides for confocal imaging (LMS880, Zeiss). Volumes of 84.2×84.2×5 μm³ were analyzed.

Apex2 staining electron microscopy (EM) of fly tissue

The protocol was performed as previously described (Hung et al., 2016) with slight modifications (Jiang et al., 2020). Vibratome sections of the fly tissues were fixed in 2% glutaraldehyde in 0.1 M sodium cacodylate with 2 mM CaCl₂, pH 7. Residual glutaraldehyde was washed off with buffer (2 min, five times) and quenched with 20 mM glycine followed by another wash (2 min, five times). The specimens were subsequently stained with SIGMA FAST^™ DAB (3,3′-Diaminobenzidine tetrahydrochloride) with Metal Enhancer Tablets (Sigma-Aldrich) for 20 min, washed in buffers (10 min, five times) and stained with 1% osmium tetroxide for 30 min. After washing with ddH2O (10 min, three times), the specimens were stained with 1% uranyl acetate overnight. The specimens were further dehydrated and embedded in resin for thin-section and TEM observation.

The TFAM-Apex2 staining signals were analyzed using Amira-Avizo 3D visualization and analysis software (Thermo Fisher Scientific). In short, the EM images were subject to threshold adjustment to select positive Apex2 staining signals. The positive signals were processed using the despeckle and closing functions of Amira-Avizo software. The areas of positive Apex2 staining signals were exported in excel format for analysis. The statistics showed in Fig. 7E included 44, 204, 167, and 94 mitochondria of the control, CG5903/MIC26-MIC27-, Mitofilin/MIC60-, and QIL1/MIC13-knockdown, respectively.

Apex2 staining EM of cell culture

S2 cells were seeded in six-well culture plates at 1×10⁶ cells/ml and grown for another day to 2–4×10⁶ cells/ml. The cells were transfected with individual vectors pMT-V5-HisB-dMIC60-Apex2-Flag, pMT-V5-HisB-dQIL1-Apex2-Flag, and pMT-V5-HisB-dCG5903/MIC26-MIC27-Apex2-Flag, using a calcium phosphate transfection kit (Invitrogen), and the protein expression was induced by CuSO₄. The cells were harvested 1-day post-induction, fixed with 2% glutaraldehyde, and subjected to the APEX2 staining procedure as described above.

Subcellular and sub-mitochondrial localization of positive Apex2 staining was analyzed by MetaMorph microscopy automation and image analysis software (Molecular Devices). In brief, the EM image was inverted and adjusted to a consistent threshold to isolate positive Apex2 signals of the target mitochondrion. The intensity in the cristae junction was determined by manual selection of the structure with a circle of 20 pixels in diameter. The intensity in the cristae was also determined based on the manual selection of the structures. The intensity in the IBM was defined by subtracting the intensity in the cristae and cristae junctions from the total intensity. The distribution ratios were calculated relative to total intensity. The analysis showed in Fig. 8E included 11, 17, and 15 mitochondria of CG5903/MIC26-MIC27-, Mitofilin/MIC60-, and QIL1/MIC13-knockdown in the analysis, respectively.

Quantitative PCR

Fly DNA was extracted by homogenizing about 50 flies in 200 μl of buffer containing 10 mM Tris (pH 8.0), 1 mM EDTA, 25 mM NaCl. Proteinase K (0.2 mg∕ml) was added to the lysates and incubated at 45°C for 30 min followed by inactivation at 95°C for 5 min. The supernatant was collected after centrifugation. Mitochondrial DNA content was analyzed by qPCR using LightCycler^® 480 SYBR Green I Master and LightCycler^® 480 instrument (Roche). Specific primers for mtDNA (COIII Forward: 5^′-CACGAGAAGGAAC ATACC-3^′; Reverse: 5^′-GCGGGTGATAAACTTC TG-3^′) and nuclear DNA (RpL32 Forward: 5^′-GCCGCTTCAAGGGACAGTATCTG −3^′; Reverse: 5^′-AAACGCGGTTCT GCATGAG −3^′) were used. The relative mtDNA COIII copy number was normalized to the nuclear RpL32 copy number. Three independent runs were performed; each run included triplicates with a standard deviation of Cp value less than 0.5.

For RT-qPCR, fly RNA was extracted using RNeasy Mini Kit (Qiagen) and reverse-transcribed using a RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific). Specific primers for CG5903/MIC26-MIC27 (Forward: 5^′-CGGTCTGGCTG GTTTCATCT; Reverse: 5^′-GGCACTACGGGAACATCCTC-3^′), MIC60 (Forward: 5^′-GATAAGCTGCTGCGCTTGCAGCTCAAAAAG-3^′; Reverse: 5^′-GCCACCTTGG CAATGGCATTGATCTCGTTC-3^′), QIL1 (Forward: 5^′-TTTCATCCACATGCTG CCCT-3^′; Reverse: 5^′-GCGAGCGGATCGAGGAATAA-3^′), and TFAM (Forward: 5^′-AACAAAGTCAGGCCCCTAGC-3^′; Reverse: 5^′-CTCGACGGTGGTAATCTG GG-3^′) were used. The relative copies of mtDNA COIII transcripts or transcripts from individual genes were normalized to the nuclear RpL32 transcript copy number. Three independent runs were performed; each run included triplicates with a standard deviation of Cp value less than 0.5.

Western blot analysis

The fly thoraxes were homogenized in RIPA buffer containing protease inhibitors (cOmplete^TM, Roche) using a Dounce tissue grinder. Cellular debris was removed by centrifugation at 14,000 ×g for 20 min, 4°C. The supernatants were collected and the protein concentrations were determined by Pierce protein assay (Pierce 660 nm Protein Assay Reagent, Thermo Fisher Scientific). Proteins were loaded at 20 μg/well for SDS-PAGE and western blot analysis.

Mouse anti-Flag M2 (1 μg/ml, Sigma-Aldrich F3165), rabbit anti-alpha tubulin (10000x, Abcam ab18251), anti-mouse IgG-HRP (2000×, Invitrogen 62-6520), and anti-rabbit IgG-HRP (5000x, Abcam ab97051) were used in the study. For quantification of band intensities, ratios of the densitometric signal of individual proteins to that of alpha-tubulin were calculated. The ratios were then normalized to the control samples.

Climbing assay

The flies were transferred to new culture tubes one day before the analysis. On the day of analysis, flies were transferred to a 100 ml graduated cylinder and knocked down to the bottom of the cylinder when starting video-typing them climbing up to the 100 ml marker of a cylinder (about 18 cm in height). Numbers of flies climbing up to the target line every 10 s in 120 s were calculated and plotted. Numbers of 56, 69, 79, and 62 of the control, CG5903/MIC26-MIC27-, Mitofilin/MIC60-, and QIL1/MIC13-knockdown flies were used in the triplicate analysis, respectively.