Comprehensive Genetic Characterization of Mitochondrial Ca²⁺ Uniporter Components Reveals Their Different Physiological Requirements In Vivo

Summary

Mitochondrial Ca²⁺ uptake is an important mediator of metabolism and cell death. Identification of components of the highly conserved mitochondrial Ca²⁺ uniporter has opened it up to genetic analysis in model organisms. Here, we report a comprehensive genetic characterization of all known uniporter components conserved in Drosophila. While loss of pore-forming MCU or EMRE abolishes fast mitochondrial Ca²⁺ uptake, this results in only mild phenotypes when young, despite shortened lifespans. In contrast, loss of the MICU1 gatekeeper is developmentally lethal, consistent with unregulated Ca²⁺ uptake. Mutants for the neuronally restricted regulator MICU3 are viable with mild neurological impairment. Genetic interaction analyses reveal that MICU1 and MICU3 are not functionally interchangeable. More surprisingly, loss of MCU or EMRE does not suppress MICU1 mutant lethality, suggesting that this results from uniporter-independent functions. Our data reveal the interplay among components of the mitochondrial Ca²⁺ uniporter and shed light on their physiological requirements in vivo.

Graphical Abstract

Highlights

• •MCU or EMRE loss blocks fast mitochondrial calcium uptake but are relatively benign
• •MCU knockout flies are short lived compared to EMRE or MICU3 mutants
• •MICU1 mutants are developmentally lethal, and MCU or EMRE knockout fails to rescue
• •MICU1 and MICU3 are not functionally interchangeable

Tufi et al. generate a genetic toolkit for all conserved components of the Drosophila mitochondrial calcium uniporter. Under basal conditions, MCU and EMRE loss is tolerated and MICU3 mutants are mildly impaired. MICU1 mutants are lethal, but this is not suppressed by MCU or EMRE loss, suggesting an unidentified uniporter-independent role.

Introduction

The uptake of Ca²⁺ into mitochondria has long been established as a key regulator of an array of cellular homeostatic processes as diverse as bioenergetics and cell death (Granatiero et al., 2017, Mammucari et al., 2016). A series of seminal discoveries has elucidated the identity of the components that make up the mitochondrial Ca²⁺ uniporter complex. The mammalian uniporter is composed of MCU (mitochondrial calcium uniporter) as the main pore-forming protein (De Stefani et al., 2011, Baughman et al., 2011); its paralog MCUb (Raffaello et al., 2013); a small structural component, EMRE (essential MCU regulator) (Sancak et al., 2013); and the regulatory subunits MICU1–MICU3 (mitochondrial calcium uptake 1–3) (Perocchi et al., 2010, Plovanich et al., 2013, Patron et al., 2019). Reconstitution studies in yeast, which lacks a mitochondrial Ca²⁺ uniporter, have demonstrated that heterologous co-expression of MCU and EMRE is necessary and sufficient to confer uniporter activity (Kovács-Bogdán et al., 2014). The family of EF-hand-containing proteins (MICU1, MICU2, and MICU3) has been shown to exhibit a gatekeeper function for the uniporter, inhibiting Ca²⁺ uptake at low cytoplasmic concentrations (Mallilankaraman et al., 2012, Kamer et al., 2017, Patron et al., 2019). These components are generally highly conserved across eukaryotes, including most metazoans and plants, but not in many fungi and protozoans, reflecting their ancient and fundamental role (Bick et al., 2012).

Although the composition and function of the uniporter have been well characterized in vitro and in cell culture models, the physiological role of the uniporter is beginning to emerge with in vivo characterization of knockout mutants (Liu et al., 2017). Current data present a complex picture. Initial studies of MCU knockout mice described a viable strain with a modest phenotype in a mixed genetic background (Pan et al., 2013), although subsequent studies using an inbred background reported MCU loss to be lethal or semi-viable (Murphy et al., 2014) and tissue-specific conditional knockout revealed an important role in cardiac homeostasis (Luongo et al., 2015). Similarly, loss of MICU1 in mice has a complex phenotype, varying from fully penetrant perinatal lethality (Antony et al., 2016) to incomplete lethality with a range of neuromuscular defects that unexpectedly improve over time in surviving animals (Liu et al., 2016).

One explanation for the reported phenotypic variability is that perturbing mitochondrial Ca²⁺ uptake can be influenced by additional factors, the most obvious being genetic background. Hence, there is a need for greater investigation into the physiological consequences of genetic manipulation of the uniporter components in a genetically powerful model system. Here we report a comprehensive genetic analysis of the uniporter complex components that are conserved in Drosophila. This includes loss-of-function mutants for MCU, EMRE, MICU1, and MICU3 (Drosophila lack MCUb and MICU2) and corresponding inducible transgenic expression lines. Despite lacking fast Ca²⁺ uptake, MCU and EMRE mutants present a surprising lack of organismal phenotypes, although both mutants are short lived, with a more pronounced effect when MCU is lost. In contrast, loss of MICU1 causes developmental lethality, whereas mutants for MICU3 are viable with modest phenotypes. Performing genetic interaction studies with these strains, we confirm the gatekeeper function of MICU1 is conserved in flies and reveal that MICU1 and MICU3 are not functionally interchangeable. More surprisingly, we find that loss of MCU or EMRE does not suppress MICU1 mutant lethality, suggesting that the lethality results from MCU-independent functions. The generation of these genetic tools in Drosophila will facilitate further investigation of the functional roles of the uniporter components in vivo.

Results

To generate null mutants for MCU, we used a P element mobilization technique exploiting a transposon at the 5′ end, P{EPgy2}MCU^EY08610 (Figure 1A). We isolated a single imprecise excision: a deletion of 1,557 bp removing the 5′ end of MCU that includes the first three exons containing the start codon and mitochondrial targeting sequence common to most isoforms. We refer to this mutation as MCU¹ (Figure 1A). Precise excision revertants were also recovered (see STAR Methods). The MCU¹ deletion can be detected by genomic PCR, and the breakpoints were verified by Sanger sequencing (Figure S1A). The neighboring genes, sulfateless (sfl) and javelin (jv), remained intact and showed unaltered levels of expression (Figures S1B and S1C). Immunoblot analysis of crude mitochondrial extracts from homozygous MCU¹ mutant homogenates using an antibody raised against the C terminus of Drosophila MCU confirmed the absence of MCU protein (Figure 1B).

Figure 1.

The MCU¹ Mutant Abolishes Fast Mitochondrial Ca²⁺ Uptake and Shortens Lifespan without Affecting Organismal Phenotypes despite Respiratory Defects

(A) Overview of the MCU (CG18769) 5′ gene region (from FlyBase), including the neighboring sfl (sulfateless). The P{EPgy2}EY08610 transposable element used to generate MCU¹ is displayed, along with the location of the MCU¹ breakpoints.

(B) Western blot analysis of MCU¹. Immunoblots were probed with the indicated antibodies. The asterisk denotes a non-specific band. Mitochondrial ATP5A is used as a loading control.

(C) Representative traces of Ca²⁺ uptake in mitochondria isolated from adult flies of the indicated genotypes after addition of 45 μM CaCl₂. Extramitochondrial Ca²⁺ was measured by calcium green-5N fluorescence. Ca²⁺ was released from mitochondria by addition of 1 μM FCCP. Addition of the MCU inhibitor ruthenium red (RuR; 2 μM) blocks mitochondrial Ca²⁺ uptake, which is mirrored by MCU¹. Mitochondrial Ca²⁺ uptake is restored by transgenic re-expression of MCU driven by da-GAL4.

(D) Relative uptake kinetics of traces shown in (C) were determined through linear fits of the initial phase of Ca²⁺ uptake and normalized to the wild-type control (mean ± SEM, n = 3).

(E) Lifespan curves of MCU¹ male flies compared with control and transgenic rescue (MCU¹ + MCU) driven by da-GAL4. Statistical analysis: Mantel-Cox log-rank test (n ≥ 80).

(F) Climbing assay of control (a precise excision revertant, MCU^rv) and MCU¹ flies 2 and 20 days post-eclosion. Significance measured by Kruskal-Wallis test with Dunn’s post hoc correction for multiple comparisons (mean ± 95% confidence interval (CI); n > 70; ^∗p < 0.05; ns, non-significant).

(G) Relative ATP levels from control and MCU¹ flies. Statistical analysis: unpaired t test (mean ± SD; n = 2–3; ^∗∗∗∗p < 0.0001; ns, non-significant).

(H) Oxygen consumption rate (OCR) of control and MCU¹ flies at 3 and 20 days post-eclosion. Statistical analysis: unpaired t test (mean ± SEM; n = 3; ^∗∗p < 0.01, ^∗∗∗p < 0.001).

The control genotype is w¹¹¹⁸ unless otherwise stated. See also Figures S1 and S2.

Mitochondria from human or mouse cells lacking MCU fail to perform fast Ca²⁺ uptake (Baughman et al., 2011, De Stefani et al., 2011, Pan et al., 2013). To verify that MCU¹ represents a functional null mutant, mitochondria were isolated from homozygous MCU¹ adult flies and assayed for Ca²⁺ uptake. Similar to mammalian cells, the addition of Ca²⁺ to purified, energized mitochondria from wild-type Drosophila yields a rapid spike of extra-mitochondrial calcium green-5N fluorescence followed by a progressive decline in fluorescence as Ca²⁺ is buffered by mitochondria (Figures 1C and 1D). Ca²⁺ is released again upon depolarization by the uncoupling agent carbonyl cyanide-4-(trifluoromethoxy)phenylhydrazone (FCCP), as reflected by the concomitant rise in calcium green-5N fluorescence. As expected, rapid Ca²⁺ uptake is blocked by the addition of the MCU inhibitor ruthenium red (RuR). This effect is fully replicated in MCU¹ mutant mitochondria, reflecting a complete loss of fast Ca²⁺ uptake. The lack of Ca²⁺ uptake was not due to loss of membrane potential, because this was equivalent across the samples (Figures S1D and S1E). Moreover, Ca²⁺ uptake is restored upon transgenic expression of MCU (Figures 1C and 1D; Figures S1D and S1E). Altogether, these data show that MCU¹ is a null mutant incapable of fast Ca²⁺ uptake.

Despite this deficiency, MCU¹ mutants are homozygously viable and develop to adult stage in expected Mendelian proportions (Figure S1F). However, MCU¹ mutants are significantly shorter lived than controls (∼34% reduction of median lifespan), a phenotype that is fully rescued by ubiquitous expression of transgenic MCU (Figure 1E). Despite this attenuated longevity, MCU¹ mutants do not display an appreciable decline in vitality, as assessed by analyzing their motor ability using a negative geotaxis (climbing) assay (Figure 1F).

We next sought to determine the effect of MCU loss on mitochondrial metabolic function. Young MCU¹ mutants showed a modest but significant decrease in basal ATP levels compared to controls (Figure 1G), which ameliorated with age. In contrast, measurement of oxygen consumption rate (OCR) revealed a marked reduction in complex I- or complex II-linked respiration in young and older flies (Figure 1H). Assessing the impact on mitochondrial cell biology, we found no difference in mitochondrial morphology in flight muscle (Figure S2A) or mitochondrial axonal transport (Figures S2B and S2C). Altogether, these results indicate that loss of MCU affects mitochondrial respiratory capacity, but this is surprisingly well tolerated at the organismal level.

To mutate EMRE, we used a CRISPR/Cas9-based approach (Port et al., 2014) with two simultaneously expressed transgenic guide RNAs (gRNAs). We isolated several insertion/deletion (indel) events resulting in frameshift mutations that led to premature stop codons. Three such mutations are shown in Figure 2A. As with MCU¹, analysis of these EMRE mutants revealed that they all exhibit no fast mitochondrial Ca²⁺ uptake (Figures 2C and 2D), in normally energized mitochondria (Figures S3B and S3C), indicating that the three mutants are functionally equivalent. We focused on one mutant, EMRE¹, whose mutation abolishes a BcnI restriction site (Figure S3A) and shows a substantial reduction in the level of the mRNA transcript (Figure 2B), for further characterization.

Figure 2.

EMRE Mutants Exhibit No Fast Mitochondrial Ca²⁺ Uptake and Are Short Lived but Have Mild Phenotypes

(A) Sequence alignments of wild-type and EMRE mutants, with predicted protein sequences and positions of gRNA recognition sites (colored text). the box denotes the BcnI cleavage site.

(B) Relative expression of EMRE transcript for control and EMRE¹ mutants (mean ± SD; n = 3).

(C) Representative traces of Ca²⁺ uptake in mitochondria isolated from adult flies of the indicated genotypes after addition of 45 μM CaCl₂. Extramitochondrial Ca²⁺ was measured by calcium green-5N fluorescence. Ca²⁺ was released from mitochondria by addition of 1 μM FCCP. EMRE mutations prevent mitochondrial Ca²⁺ uptake equivalent to the inhibitor ruthenium red (RuR; 2 μM).

(D) Relative uptake kinetics were determined through linear fits of Ca²⁺ uptake traces and normalized to controls (mean ± SEM; n = 3).

(E) Lifespan curves of EMRE¹ male flies compared with control. Statistical analysis: Mantel-Cox log-rank test (n ≥ 91).

(F) Relative ATP levels from control and EMRE¹ flies. Statistical analysis: unpaired t test (mean ± SD; n = 3; ns, non-significant).

(G) Oxygen consumption rate (OCR) of control and EMRE¹ flies at 3 and 20 days post-eclosion. Statistical analysis: unpaired t test (mean ± SEM; n = 3; ^∗p < 0.05).

The control genotype is w¹¹¹⁸ in all cases. See also Figure S3.

Similar to the MCU¹ flies, EMRE¹ mutants are viable, eclose at expected Mendelian ratios (Figure S3D), and display a significantly shortened lifespan (23% reduction of median lifespan compared to control) (Figure 2E). The climbing ability of EMRE¹ mutants was similar to that of heterozygous controls at 2 days old, although a modest difference becomes apparent by 20 days (Figure S3E). This is mirrored in the basal ATP level of EMRE¹ mutants, being only marginally reduced at 20 days (Figure 2F). However, in contrast to the strong reduction in respiration seen in MCU¹ mutants, the complex I- or complex II-linked respiration of EMRE¹ flies is either non-significant or only modestly affected compared to controls (Figure 2G).

To target MICU1, we again used P element mobilization, using P{SUPor-P}MICU1^KG04119, and isolated a large deletion spanning ∼11 kb, removing half of MICU1 and extending some 9 kb upstream of MICU1 (Figure 3A). This region is relatively gene sparse and devoid of additional predicted protein-coding genes. Expression analysis of homozygous MICU1³² larvae yielded no detectable transcript, establishing it as a null allele (Figure 3B). In contrast to MCU and EMRE mutants, homozygous MICU1³² mutants are larval lethal, with a few animals reaching the third instar stage. Supporting this, ubiquitous expression of two independent RNAi transgenes also caused developmental lethality.

Figure 3.

MICU1³² Mutants Are Lethal, Have Reduced ATP and Mitochondrial Transport, and Are Not Rescued by MCU¹ or EMRE¹

(A) Overview of the MICU1 (CG4495) gene region (from FlyBase). The P{SUPor-P}MICU1^KG04119 transposable element used to generate MICU1³² is displayed, as well as the deleted region in MICU1³².

(B) Relative expression of MICU1 transcript for control and MICU1³² larvae (mean ± SD; n = 3).

(C) Larval crawling of control and MICU1³² larvae, expressed as the number of peristaltic waves per minute. Statistical analysis: unpaired t test (mean ± 95% CI; n = 10; ^∗∗p = 0.0034).

(D) Relative ATP levels from control and MICU1³² larvae. Statistical analysis: unpaired t test (mean ± SD; n = 4; ^∗∗p = 0.0011).

(E) Quantification of mitochondrial density in the images shown in (F) (zoomed images). Statistical analysis: one-way ANOVA (mean ± SD; n = 4; ^∗p < 0.05; ns, non-significant).

(F) Confocal microscopy analysis of epidermal cells in control and MICU1³² larvae immunostained with the mitochondrial marker anti-ATP5A. Boxed areas are enlarged to the right. Scale bars: 10 μm (left), 4 μm (right).

(G) Representative kymographs of mitochondrial axonal transport in control and MICU1³² larvae. Scale bars: 10 μm (horizontal), 50 s (vertical). Genotypes—control: M12-GAL4, UAS-mito-HA-GFP/+. MICU1³²: MICU1³²/MICU1³²; M12-GAL4, UAS-mito-HA-GFP/+.

(H) Quantification of mitochondrial transport shown in (G). Statistical analysis: one-way ANOVA (mean ± 95% CI; n = 6 (control) and 11 (mutant); ^∗∗p < 0.01, ^∗∗∗∗p < 0.0001; ns, non-significant).

(I) Table of viability of MICU1³² rescue by transgenic expression of MICU1 or MICU3 isoforms or loss of MCU or EMRE. Transgenic expression was induced using ubiquitous drivers: arm-GAL4 for MICU1 and da-GAL4 for MICU3.

(J) Climbing assay of control flies (arm-GAL4/+) and MICU1³² mutants with ubiquitous (arm-GAL4) driven transgenic re-expression of HA-tagged MICU1-A and MICU1-B isoforms. Statistical analysis: Kruskal-Wallis test with Dunn’s post hoc correction for multiple comparisons (mean ± 95% CI; n ≥ 40).

The control genotype is w¹¹¹⁸ unless otherwise stated.

Because MICU1³² mutants do not reach adulthood, we analyzed larval locomotion (crawling) as a measure of organismal vitality, revealing a behavioral deficit (Figure 3C). We further found that these mutants had significantly lower ATP levels compared to controls (Figure 3D), indicative of a substantial mitochondrial impairment. This prompted us to investigate other indicators of mitochondrial homeostasis. Visualizing mitochondria in larval epidermal cells, the morphology looked comparable to controls, although the mitochondria were more diffuse in the MICU1³² mutants occupying a smaller area of the cell (Figures 3E and 3F). Furthermore, mitochondrial axonal transport was significantly reduced in MICU1³² larvae (Figures 3G and 3H). Collectively, these data reveal the presence of multiple mitochondrial defects that together might be responsible for lethality of the MICU1³² mutants.

MICU1³² lethality was fully rescued upon ubiquitous expression of hemagglutinin (HA)-tagged MICU1 by either the A or the B isoform (Figure 3I). In addition, organismal vitality, as measured by climbing ability, was restored (Figure 3J), demonstrating that these phenotypes are specifically due to loss of MICU1.

Drosophila do not have an ortholog of the MICU1 paralog MICU2, but MICU3 is conserved, encoded by CG4662 (Figure 4A). Similar to mammals (Patron et al., 2019), Drosophila MICU3 also appears to be mainly expressed in neuronal tissue (see FlyBase and Graveley et al., 2011). Little is known about the function of MICU3, and no in vivo studies have been reported. To assess its role in vivo, we used CRISPR/Cas9 to induce indel mutations. One of these mutations, MICU3²⁷, a single-base deletion (Figure 4B; Figure S4A) that abolishes a MboII restriction site (Figure S4B), leads to a frameshift and early truncation. This mutation also substantially de-stabilizes the MICU3 transcript (Figure 4C). In contrast to loss of MICU1, homozygous MICU3²⁷ mutants are fully viable (Figure S4C), though lifespan was modestly (7% reduction in median lifespan) but significantly reduced (Figure 4D). In addition, these mutants exhibited a significant climbing defect in young and older flies (Figure S4D). These results indicate a function for MICU3 in proper maintenance of neuronal function, consistent with its neuronally restricted expression. However, analyzing mitochondrial respiration from heads of MICU3²⁷ mutants, we observed no significant differences in young and older flies compared to control (Figure 4E).

Figure 4.

MICU3 Loss of Function Is Largely Benign

(A) Overview of MICU3 (CG4662) gene region (from FlyBase), including positions of the gRNA recognition site.

(B) Sequence alignment of wild-type and MICU3²⁷, with relative predicted protein sequences. The gRNA recognition site is highlighted in red. The box denotes the MboII cleavage site.

(C) Relative expression of MICU3 transcript for control and MICU3²⁷ flies (mean ± SD; n = 3).

(D) Lifespan curves of MICU3²⁷ male flies compared with control. Statistical analysis: Mantel-Cox log-rank test (n ≥ 84).

(E) Oxygen consumption rate (OCR) of control and MICU3²⁷ flies at 3 and 20 days post-eclosion. Statistical analysis: unpaired t test (mean ± SEM; n = 4).

The control genotype is w¹¹¹⁸ in all cases. See also Figure S4.

To investigate the functional relationships among the various uniporter components, we undertook several genetic interaction studies. First, because Drosophila MICU1 and MICU3 share a fair degree of homology (∼49% similarity and ∼31% identity between MICU1-B and MICU3-C), we reasoned that they may share some functional overlap. To address this, we asked whether normally neuronally restricted MICU3 could functionally substitute for MICU1. Thus, we ectopically expressed MICU3 ubiquitously in MICU1³² mutants. Here we chose to express isoforms A and C, because these cover all predicted coding regions (Figure 4A). However, neither MICU3 isoform was able to restore viability of MICU1³² mutants or shift the lethal phase (Figure 3I), indicating that MICU3 is not functionally equivalent to MICU1 in vivo.

MICU1 has been shown to provide a gatekeeper function for the uniporter channel, with loss of MICU1 causing unregulated mitochondrial Ca²⁺ uptake. It has also been shown that in mice lacking MICU1, genetic reduction of EMRE substantially ameliorates the MICU1 phenotypes (Liu et al., 2016). Thus, we reasoned that the lethality of the MICU1³² mutants is caused by unregulated Ca²⁺ entry, which should be prevented by loss of the MCU channel. To test this, we combined homozygous MICU1³² and MCU¹ mutants and, to our surprise, found that this did not suppress the lethality or noticeably shift the lethal phase (Figure 3I). We corroborated this finding by combining MICU1³² mutants with EMRE¹ mutants, with the same result (Figure 3I). Moreover, the aberrant mitochondrial distribution in MICU1³² epidermal cells was not restored by loss of MCU (Figures 3E and 3F).

Overexpression paradigms disrupting uniporter stoichiometry have previously been used to study the functional relationships of uniporter components (Choi et al., 2017). Using a classic eye morphology assay as a readout of the impact of genetic interactions on cell and tissue viability, we first found that overexpression of any of the uniporter components alone in the eye, using a GMR-GAL4 driver, had no effect on eye or ommatidial morphology (Figure S5A), with overexpression confirmed by immunoblotting (Figures S5B–S5E). This indicates that overabundance of any uniporter component, including MCU, is insufficient to grossly disrupt mitochondrial Ca²⁺ homeostasis as expected. However, the co-expression of MCU and EMRE caused a dramatic disruption of eye morphology with a general loss of retinal pigment and ommatidia, resulting in a glazed appearance with occasional black, necrotic patches (Figure 5A). This effect is consistent with the cooperative actions of MCU and EMRE to create the channel, and in line with reconstitution experiments in yeast showing that expression of mammalian MCU and EMRE are necessary and sufficient to elicit Ca²⁺ uniporter activity. The gross disruption of eye integrity also demonstrates the catastrophic effects of unregulated mitochondrial Ca²⁺ entry.

Figure 5.

Genetic Interactions of Overexpression Uniporter Components

(A) Flies bearing the eye-specific GMR-GAL4 driver and UAS-MCU transgene (GMR > MCU) crossed to transgenes of the indicated uniporter components or to a control (+) w¹¹¹⁸ line.

(B) Flies bearing a combination of the eye-specific GMR-GAL4 driver and both UAS-MCU and UAS-EMRE transgenes (GMR > MCU+EMRE) crossed to transgenes of the indicated components or to a control (GFP) transgene.

Below the micrograph images are schematic cartoons of the proposed status of the uniporter based on the composition of the overexpressed components. Scale bar: 100 μm. See also Figure S5.

In contrast, co-expression of MCU with either MICU1 or MICU3 was not so detrimental, although in all cases, it caused a mild disruption of the ommatidial arrangement, resulting in a mild roughened appearance (Figure 5A). This system allowed us to test possible functional differences among the isoforms of MICU1 and MICU3. In general, the co-expression of all isoforms with MCU caused similar effects, though the expression of MICU3-C gave a slightly stronger phenotype that may reflect its greater level of expression compared to MICU3-A (Figure S5E). The enhanced phenotype of MCU:MICU3 co-expression is consistent with a report that MICU3 enhances MCU-mediated Ca²⁺ uptake (Patron et al., 2019).

We next reasoned that if the dramatic eye phenotype caused by MCU and EMRE co-expression was due to extra unregulated channels and excessive Ca²⁺ uptake, this may be ameliorated by co-expression of the MICU1 gatekeeper. Co-expression of either MICU1-A or MICU1-B with MCU and EMRE prevented the MCU:EMRE phenotype (Figure 5B). Co-expression of a GFP control confirmed that this was not due to titration of multiple upstream activating sequence (UAS) lines. Although both isoforms provided excellent rescue, isoform B appeared slightly more effective than isoform A. These observations are consistent with MICU1 providing the main gatekeeper function for the uniporter.

We also used this assay to test the functionality of MICU3 in this context. In contrast to MICU1 expression, co-expression of either MICU3-A or MICU3-C with MCU and EMRE provided no suppressing effect for the MCU:EMRE phenotype (Figure 5B). These results are consistent with MICU3 not being a major uniporter gatekeeper in vivo.

Discussion

The capacity of mitochondria to take up Ca²⁺ has important implications for cellular homeostasis as it regulates fundamental processes from metabolism to cell death. In this context, the mitochondrial Ca²⁺ uniporter plays a crucial function, driving the rapid entry of Ca²⁺ into mitochondria. To better understand the physiological role of mitochondrial Ca²⁺ uptake, we have used the genetically powerful model Drosophila to manipulate the conserved set of dedicated channel and regulatory proteins that form the uniporter complex.

MCU mutants are viable and fertile with no gross morphological or behavioral defects, which was initially surprising given the historical importance of mitochondrial Ca²⁺. Still, this corroborates another report of fly MCU mutants (Choi et al., 2017) and is consistent with studies in mice and worms in which deletion of MCU orthologs is essentially benign at the organismal level under basal conditions (Pan et al., 2013, Xu and Chisholm, 2014). However, fly MCU mutants are significantly shorter lived than controls. This situation is mirrored by EMRE mutants, albeit with a smaller impact on lifespan. The reason for the shortened lifespans is unknown but may reflect the effects of a chronic bioenergetic deficit evident from the OCR measurements. Accordingly, MCU mutants show a greater respiration defect compared to EMRE mutants, consistent with their respective impacts on lifespan. The respiratory impairment could be due to the previously reported increase in oxidative stress that occurs in MCU mutants (Choi et al., 2017), which has yet to be assessed in EMRE mutants. Alternatively, the short lifespan may be due to a myriad of potential metabolic imbalances, such as disruption of NADH/NAD+ levels. Chronic adaptations may also occur through transcriptional responses. Further studies analyzing the metabolic and transcriptional changes occurring in these flies will shed light on this fundamental question.

Nevertheless, the EMRE mutants are relatively benign at the organismal level, which corroborates the surprising viability of MCU mutants. Considering this, it is striking that flies, like mice and worms, consistently show an ability to compensate for the lack of fast mitochondrial Ca²⁺ uptake, suggesting the induction of some adaptive mechanism, as discussed by others (Murphy et al., 2014, Harrington and Murphy, 2015). While alternative routes of mitochondrial Ca²⁺ entry must exist, because matrix Ca²⁺ is not abolished in MCU knockout (KO) mice (Pan et al., 2013), proposed mechanisms are speculative, and it is unclear whether they constitute a compensatory adaptation for fast Ca²⁺ uptake or simply allow gradual, slow accumulation (Harrington and Murphy, 2015). However, rapid mitochondrial Ca²⁺ uptake mediated by MCU is thought to constitute a specific metabolic regulatory mechanism, e.g., to increase ATP production, under certain conditions, such as strenuous exercise or pathological conditions (Denton, 2009), which is partly evident in the MCU KO mice (Pan et al., 2013) or heart-specific conditional KO (Luongo et al., 2015). Such important physiological roles would not necessarily be apparent under basal conditions in flies. MCU has also been proposed to promote wound healing (Xu and Chisholm, 2014); however, our preliminary studies did not find evidence supporting this. The current study presents a summary of the requirements of uniporter components under basal conditions, and further work will be needed to evaluate the role of the uniporter in the full range of physiological conditions.

In seeking to understand the importance of the regulatory components of the uniporter, we also developed loss-of-function models for MICU1 and MICU3. In contrast to MCU and EMRE mutants, loss of MICU1 results in larval lethality, which is associated with alterations in mitochondrial distribution and motility, and a reduced level of total ATP. In line with its role as the principle gatekeeper of the uniporter, coupled with excess mitochondrial Ca²⁺ triggering cell death, we reasoned that the lethality was due to Ca²⁺ accumulation in the mitochondrial matrix through unregulated MCU-EMRE channels. Supporting this, we observed that dual overexpression of MCU and EMRE in the eye leads to substantial loss of retinal tissue; concomitant overexpression of MICU1 is sufficient to prevent this phenotype, consistent with MICU1 re-establishing appropriately regulated uniporter channels.

However, one observation that was most surprising to us was the inability of MCU or EMRE mutants to rescue the MICU1 mutant lethality. This result is particularly puzzling, because it has been shown that mice lacking MICU1, which present multiple pathogenic phenotypes, are substantially rescued by genetic reduction of EMRE levels (Liu et al., 2016). While the reason for the lack of rescue in flies is unclear, we postulate that this suggests the function of MICU1 is not limited to uniporter-dependent Ca²⁺ uptake. We do not know whether the lethality of MICU1 mutants is specifically due to excessive mitochondrial Ca²⁺ levels; however, it appears to be independent of fast mitochondrial Ca²⁺ uptake, because this is eliminated in MCU and EMRE mutants. As noted earlier, other routes of Ca²⁺ uptake into mitochondria exist, but the mechanisms that regulate them are uncertain. It is possible that aberrant manganese uptake, as reported to occur in cell models, may contribute to the MICU1 mutant lethality (Kamer et al., 2018, Wettmarshausen et al., 2018). However, this mechanism would presumably be expected to be mitigated by loss of MCU. Nevertheless, these Drosophila models are ideally suited for unbiased genetic screening to uncover such fundamental regulatory mechanisms.

In contrast to MICU1, loss of MICU3 was well tolerated overall at the organismal level. Functional analysis of MICU3 is extremely limited, but the neuronally restricted expression led us to anticipate that these mutants might have more neurological-specific phenotypes, which was at least partly borne out. Whereas longevity of these mutants was only minimally affected, they exhibited a notable locomotor deficit even in young flies. We initially hypothesized that MICU3 may be able to act redundantly with MICU1, but attempts to transgenically rescue MICU1 mutants by ectopic MICU3 expression were unsuccessful. This result is consistent with a report showing that MICU3 binds to MICU1 but apparently enhances mitochondrial Ca²⁺ uptake (Patron et al., 2019).

In summary, we present a comprehensive analysis of the conserved components of the mitochondrial Ca²⁺ importer and its regulators. While loss of the various components results in dramatically different organismal phenotypes, ranging from the most severe deficit exemplified by the MICU1 mutants to the mild consequences of mutating MICU3, such diverse phenotypes mirror the situation reported in humans so far. The first described patients with MICU1 mutations exhibit a severe, complex neurological condition accompanied by muscular dystrophy and congenital myopathy, clearly associated with mitochondrial dysfunction (Logan et al., 2014), whereas a later study reported MICU1 patients with a relatively mild fatigue syndrome (Lewis-Smith et al., 2016). One explanation for the reported phenotypic variability is that the consequence of perturbing mitochondrial Ca²⁺ uptake can be influenced by additional factors, the most obvious being genetic background. The genetic tools described here open up the possibility for a thorough analysis of the uniporter function in a powerful genetic model organism, which will advance our understanding of the role of mitochondrial Ca²⁺ in health and disease.

STAR★Methods

Key Resources Table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Mouse monoclonal anti-ATP5A	Abcam	RRID:AB_301447; Abcam: ab14748
Rabbit polyclonal anti-HA	Abcam	RRID:AB_307019; Abcam: ab9110
Mouse monoclonal anti-α-Tubulin, clone DM1A	Sigma	RRID:AB_477593; Sigma: T9026
Rabbit polyclonal anti-Porin	Millipore	Millipore PC548; RRID:AB_2257155
Mouse monoclonal anti-V5	Thermo Fisher Scientific	Thermo: R960-25; RRID:AB_2556564
Mouse monoclonal anti-Myc tag, clone 9B11	Cell Signaling	Cell Signaling: 2276; RRID:AB_331783
Goat polyclonal anti-Mouse IgG - H&L, HRP Conjugated	Abcam	Abcam: ab6789, RRID:AB_955439
Goat polyclonal anti-Rabbit IgG (H+L) Cross-Adsorbed Secondary Antibody, HRP	Thermo Fisher Scientific	Invitrogen G21234; RRID:AB_2536530
Goat polyclonal anti-Mouse IgG (H+L) Cross-Adsorbed Secondary Antibody, Alexa Fluor 488	Abcam	Thermo: A-11001; RRID:AB_2534069
Rabbit polyclonal anti-MCU (targeting RTQENTPPTLTE EKAERKY)	Pepceuticals	N/A
Chemicals, peptides, and recombinant proteins
NotI FastDigest	Thermo Fisher Scientific	Thermo: FD0593
XhoI FastDigest	Thermo Fisher Scientific	Thermo: FD0694
XbaI FastDigest	Thermo Fisher Scientific	Thermo: FD0684
EcoRI FastDigest	Thermo Fisher Scientific	Thermo: FD0274
Bovine Serum Albumin	Sigma	Sigma: A7030
Mannitol	Sigma	Sigma: M9546
Sucrose	Sigma	Sigma: S9378
HEPES	Sigma	Sigma: H3375
EGTA	Sigma	Sigma: E0396
Taurine	Sigma	Sigma: T0625
Triton X-100	Sigma	Sigma: T8787
Tween 20	Thermo Fisher Scientific	Thermo: BP337500
cOmplete, Mini, EDTA-free Protease Inhibitor Cocktail	Sigma	Roche: 4693159001
2-Mercaptoethanol	Sigma	Sigma: M6250
MgCl2	Sigma	Sigma: M8266
ECL-Prime	Sigma	Sigma: GERPN2232
4% Formaldehyde	Thermo Fisher Scientific	Thermo: 28908
Prolong Diamond Antifade Mounting Medium	Thermo Fisher Scientific	Thermo: P36965
Sylguard	Sigma	Sigma: 761028
Malate	Sigma	Sigma: M 1000
Glutamate	Sigma	Sigma: G 1626
Succinate	Sigma	Sigma: S 2378
ADP	Sigma	Sigma: A 5285
Rotenone	Sigma	Sigma: R 8875
Critical commercial assays
CellTiter-Glo Luminescent Cell Viability Assay	Promega	Promega: G7570
RNeasy RNA purification kit	QIAGEN	QIAGEN: 74106
ProtoScript® II first strand cDNA Synthesis Kit	New England BioLabs	NEB: E6560S
DC Protein Assay Kit	Bio-Rad	N/A
iQ SYBR® Green Supermix	Bio-Rad	170-8880
Turbo DNase Free	Ambion	AM1907
Experimental Models: organisms/strains
D. melanogaster: w[1118]	BDSC (RRID:SCR_006457)	RRID:BDSC_6326
D. melanogaster: y[1] w[67c23]; P{w[+mC] y[+mDint2] = EPgy2}MCU[EY01803]	BDSC	RRID:BDSC_16357
D. melanogaster: y[1] w[67c23]; P{y[+mDint2] w[BR.E.BR] = SUPor-P}MICU1[KG04119]	BDSC	RRID:BDSC_13588
D. melanogaster: y[1] M{w[+mC] = nos-Cas9.P}ZH-2A w[∗]	BDSC	RRID: BDSC_54591
D. melanogaster: w[∗]; Kr[If-1]/CyO; P{w[+mW.hs] = GAL4-da.G32}UH1	BDSC	RRID:BDSC_55850
D. melanogaster: w[∗]; P{w[+mW.hs] = GAL4-arm.S}11	BDSC	RRID:BDSC_1561
D. melanogaster: w[∗]; P{w[+mC] = GAL4-ninaE.GMR}12	BDSC	RRID:BDSC_1104
D. melanogaster: y[1] w[∗]; P{w[+mC] = CCAP-GAL4.P}16/CyO	BDSC	RRID:BDSC_25685
D. melanogaster: w[∗]; P{w[+mW.hs] = GawB}tey[5053A]/TM6B, Tb[+]	BDSC	RRID:BDSC_2702
D. melanogaster: w[1118]; P{w[+mC] = UAS-mito-HA-GFP.AP}2/CyO	BDSC	RRID:BDSC_8442
D. melanogaster: w[1118]; P{w[+mC] = UAS-mito-HA-GFP.AP}3, e[1]	BDSC	RRID:BDSC_8443
D. melanogaster: y[∗] w[∗];P{w[+mC] = UAS-tdTomato.mito}1	BDSC	RRID:DGGR_117015
D. melanogaster: M{UAS-MICU1.ORF.3xHA}ZH-86Fb	FlyORF	FlyORF Line ID: F000962
Oligonucleotides
MCU1 Forward Primer: GCAACTTCAGCATATGACC	This paper	N/A
MCU1 Reverse Primer: GGAATTGGGATGCCATAGC	This paper	N/A
EMRE1 Forward Primer: GCGCTTTTCAACACTACTAC	This paper	N/A
EMRE1 Reverse Primer: GGTATGACGGCACAGAAGATG	This paper	N/A
MICU327 Forward Primer: CTCGATCTCTGATCCCGCA	This paper	N/A
MICU327 Reverse Primer: TCGTGCAGAAAACAACTACATTT	This paper	N/A
qPCR Primers	This paper (See Table S1)	N/A
Recombinant DNA
pCFD4	Addgene	Addgene: 49411
pUAST-attB	Drosophila Genomics Resource Centre	DGRC: 1419
LD26402	Drosophila Genomics Resource Centre	DGRC: 2351
IP17639	Drosophila Genomics Resource Centre	DGRC: 1605128
LD23951	Drosophila Genomics Resource Centre	DGRC: 7119
RH09265	Drosophila Genomics Resource Centre	DGRC: 1153329
Software and algorithms
Datlab	Oroboros	N/A
Prism v7	GraphPad	RRID:SCR_002798
Fiji	Fiji.sc	RRID:SCR_002285
Micro-Manager	Micro-manager.org	RRID:SCR_016865
Other
Microinjection Services	BestGene Inc., or Department of Genetics, University of Cambridge	N/A

Contact for Reagent and Resource Sharing

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Alexander J. Whitworth (a.whitworth@mrc-mbu.cam.ac.uk).

Experimental Model and Subject Details

Flies were raised under standard conditions at 25°C on food consisting of agar, cornmeal, molasses, propionic acid and yeast in a 12h:12h light:dark cycle. The following strains were obtained from the Bloomington Drosophila Stock Center (RRID:SCR_006457): w¹¹¹⁸ (RRID:BDSC_6326), MCU^EY08610 (RRID:BDSC_16357), MICU1^KG04119 (RRID:BDSC_13588), da-GAL4 (RRID:BDSC_55850), arm-GAL4 (RRID:BDSC_1561), GMR-GAL4 (RRID:BDSC_1104), CCAP-GAL4 (RRID:BDSC_25685), M12-GAL4 (RRID:BDSC_2702), UAS-mito-HA-GFP (RRID:BDSC_8442, RRID:BDSC_8443), UAS-mito.tdTomato (RRID:DGGR_117015). UAS-MICU1-A-HA (F000962) was obtained from the FlyORF collection (Bischof et al., 2013). Experimental procedures and fly husbandry were performed in accordance with local Biological Services committee approval.

Method Details

Drosophila mutagenesis

Mobilisation of the MCU^EY08610 and MICU1^KG04119 transposable elements was used to generate MCU (CG18769) and MICU1 (CG4495) mutants, termed MCU¹ and MICU1³². Several precise excisions of the P-elements (revertants) were also recovered. For EMRE (CG17680) and MICU3 (CG4662), a CRISPR-based strategy was employed. Here, guide RNAs (gRNAs) were evaluated using an online tool (http://targetfinder.flycrispr.neuro.brown.edu/) (Gratz et al., 2014). Where possible, gRNAs toward the 5′ end of the protein coding sequence without predicted off-targets were selected. gRNAs were cloned into the pCFD4 vector (Addgene 49411), and the resulting constructs were verified by sequencing before being sent for transgenesis by phiC31 site-directed integration into attP40 and attP2 sites (BestGene Inc., or Department of Genetics, University of Cambridge). These gRNA-expressing flies were crossed to y¹ M{nos-cas9, w⁺}ZH-2A w^∗ (RRID: BDSC_54591) for mutagenesis. Mutant lines were screened via sequencing of PCR products spanning the targeted gene region. The mapped breakpoints for each mutant are as follows: MCU¹, 3L:6550718-6552274; EMRE¹, 2R:18147830-18147821, 2R:18147812-18147804; EMRE², 2R: 18147814-18147795; EMRE³, 2R:18147805-18147805; MICU1³², 2L:7173086-7184065; MICU3²⁷, 3R:19850278-19850278. All genomic coordinates are according to FlyBase/BDGP Release 6 (Gramates et al., 2017). All the mutants lines used in this study were backcrossed to an isogenic w¹¹¹⁸ strain (RRID:BDSC_6326), for 4-6 generations before use.

Generation of transgenic lines

The following transgenes were generated by cloning into the pUAST-attB vector (BestGene Inc.).

UAS-MCU

MCU was amplified from cDNA clone LD26402 (equivalent to isoforms A, B, C and D) and inserted between the NotI and XhoI sites. UAS-MICU1-B-HA: MICU1-B was amplified from cDNA IP17639 to include a single 3′ HA tag, and was inserted between the NotI and XbaI sites. UAS-EMRE-myc: EMRE was amplified from genomic DNA with primers encoding a single 3′ Myc tag, and was inserted between the EcoRI and XhoI sites. UAS-MICU3-A-V5, UAS-MICU3-C-V5: MICU3-A and MICU3-C were amplified from cDNA clones lD23951 and RH09265 respectively, including a single 3′ V5 tag, and were inserted between the NotI and XbaI sites.

All cDNA clones were obtained from Drosophila Genomics Resource Center (Bloomington, Indiana). Constructs were verified by sequencing before being sent for transgenesis by phiC31 site-directed integration into attP40 and attP2 sites (BestGene Inc., or Department of Genetics, University of Cambridge). For all integration events, multiple independent lines were initially isolated, verified by PCR and assessed for consistent effects before selecting a single line of each integration site for further study.

Mitochondrial membrane potential and calcium flux

Mitochondria were prepared from ∼50 whole adult flies by differential centrifugation. Samples were homogenized with a Dounce glass potter and a loose-fitting glass pestle in a mannitol-sucrose buffer (225 mM mannitol, 75 mM sucrose, 5 mM HEPES, 0.1 mM EGTA, pH 7.4) supplemented with 2% BSA. Samples were centrifuged at 1,500 × g at 4°C for 6 min. The supernatant was filtered through a fine mesh, and centrifuged at 7,000 × g at 4°C for 6 min. The resulting pellet was resuspended in mannitol-sucrose buffer without BSA before being centrifuged at 7,000 x g under the same conditions as above and resuspended in a small volume (∼50 μL) of mannitol-sucrose buffer. Protein concentration was measured using the Biuret test.

Mitochondrial membrane potential of isolated mitochondria was measured based on the fluorescence quenching of Rhodamine123 (Rh123; Molecular Probes) and mitochondrial Ca²⁺ fluxes were measured by Calcium Green 5N (Molecular Probes) fluorescence at 25°C (von Stockum et al., 2011) using a Fluoroskan Ascent FL (Thermo Electron) plate reader (excitation and emission wavelengths of 485 and 538 nm, respectively with a 10 nm bandpass filter) at a mitochondrial concentration of 1 mg/mL. The incubation medium contained 250 mM sucrose, 10 mM MOPS-Tris, 5 mM/2.5 mM glutamate/malate-Tris, 5 mM Pi-Tris, 10 μM EGTA, and 0.4 μM Rhodamine123, or 0.5 μM Calcium Green 5N, pH 7.4. Addition of Ruthenium Red (RuR, 2 μM) was made directly into the well containing the assay medium before mitochondria were added. Further additions were made as indicated in the figure legends.

Locomotor and lifespan assays

Climbing (negative geotaxis assay) was assessed as previously described with minor modifications (Greene et al., 2003). Briefly, for climbing, 20-25 males were placed into the first chamber of a ‘Benzer’ counter-current apparatus, tapped to the bottom, and given 10 s to climb a 10 cm distance. This procedure was repeated five times, and the number of flies remaining in each chamber was counted. The weighted performance of several group of flies for each genotype was normalized to the maximum possible score and expressed as Climbing Index.

For larval locomotion (crawling), mid-third instar larvae were picked from the surface of the food using a brush and cleaned using a moist Kimwipe. Excessive handling was minimized at all stages. Larvae were assayed individually by placing onto plates containing freshly set 1% agarose, where they acclimatised for 30 s, after which peristaltic waves were counted for one minute. Larvae of different genotypes were placed onto different plates, and no more than 6 larvae were assayed for any individual plate.

For lifespan experiments, flies were grown under identical conditions at low-density. Groups of approximately 20-25 males of each genotype were collected under very light anesthesia, placed into separate vials with food and maintained at 25°C. Flies were transferred into vials containing fresh food every 2-3 days, and the number of dead flies was recorded. Percent survival was calculated at the end of the experiment after correcting for any loss resulting from handling.

Mitochondrial protein enrichment

Crude mitochondrial extracts were obtained from approximately 100 flies per sample. After 5 min on ice, flies were homogenized in a glass tissue grinder containing 2 mL of cold mitochondrial isotonic buffer (225 mM mannitol, 75 mM sucrose, 5 mM HEPES, 0.5 mM EGTA, 2 mM taurine, pH 7.25) for 30-60 s until uniform. Subsequently, the homogenates were centrifuged at 500 × g at 4°C for 5 min. The resulting supernatants were passed through a 100 μm nylon sieve (Cell Strainer REF 352360, BD Falcon, USA), centrifuged at 11,000 × g for 10 min at 4°C, and pellets were stored at −80°C until use. The final mitochondrial pellets were subsequently used for lysis in 50 μL of RIPA buffer (50 mM Tris-HCl, pH 8.0; 150 mM NaCl; 1 mM EDTA, 0.5% SDS, 1% (vol/vol) Triton X-100) with cOmplete mini EDTA-free protease inhibitors (Roche) for 10 min on ice. After carrying out three freeze-thaw cycles with dry ice and a 37°C water bath, the lysates were centrifuged at 20,000 × g for 5 min and the supernatants taken for SDS-PAGE.

Immunoblotting

For MCU and EMRE expression analysis, mitochondrial proteins were isolated from whole adult flies according to the method described in the previous section. For MICU1 and MICU3 overexpression analysis flies were homogenized in a PBS-based lysis buffer with lithium dodecyl sulfate containing β-Mercaptoethanol and supplemented with cOmplete mini EDTA-free protease inhibitors (Roche). Equivalent amounts of proteins were resolved by SDS-PAGE and transferred onto nitrocellulose membrane using a semi-dry Transblot apparatus (BioRad) according to the manufacturer’s instructions. The membranes were blocked in TBST (0.15 M NaCl and 10 mM Tris-HCl; pH 7.5, 0.1% Tween 20) containing 5% (w/v) dried non-fat milk (blocking solution) for 1 h at room temperature and probed with the indicated primary antibody before being incubated with the appropriate HRP-conjugated secondary antibody. Antibody complexes were visualized by an ECL-Prime enhanced chemiluminescence detection kit (Amersham) using a ChemiDoc XRS+ molecular imager (BioRad).

Antibodies

For immunoblot experiments, the following antibodies were used: mouse anti-ATP5A (Abcam ab14748; RRID:AB_301447; 1:20000), rabbit anti-HA (Abcam ab9110; RRID:AB_307019; 1:1000), mouse anti-α-Tubulin (Sigma T9026, clone DM1A; RRID:AB_477593; 1:1500), rabbit anti-Porin (Millipore PC548; RRID:AB_2257155; 1:5000), mouse anti-V5 (Thermo Fisher Scientific R960-25; RRID:AB_2556564; 1:2000), mouse anti-Myc tag (Cell Signaling, clone 9B11; RRID:AB_331783; 1:800). Horseradish peroxidase-conjugated secondary antibodies: anti-mouse (Abcam ab6789-1; RRID:AB_955439; 1:5000-1:40000), anti-rabbit (Invitrogen G21234; RRID:AB_2536530; 1:3000 to 1:5000). Anti-MCU antiserum was raised in rabbits against a KLH-conjugated C-terminal peptide, RTQENTPPTLTEEKAERKY (Pepceuticals, 1:1000). For immunohistochemistry, tissues were incubated with mouse anti-ATP5A (Abcam ab14748; RRID:AB_301447; 1:500) and secondary antibody anti-mouse AF488 (Invitrogen: A11001; RRID:AB_2534069; 1:200).

Microscopy

Indirect flight muscle was dissected and fixed in 4% formaldehyde (Agar scientific; R1926) in PBS for 30 minutes, washed twice with PBS, and mounted on slides in Prolong Diamond Antifade mounting medium (Thermo Fisher Scientific; RRID:SCR_015961). Larval epidermal cells were prepared as previously described (Lee et al., 2018). Larvae were dissected in PBS and fixed in 4% formaldehyde, for 30 min, permeabilized in 0.3% Triton X-100 for 30 min, and blocked with 0.3% Triton X-100 plus 1% bovine serum albumin in PBS for 1 h at room temperature. Tissues were incubated with anti-ATP5A antibody diluted in 0.3% Triton X-100 plus 1% bovine serum albumin in PBS overnight at 4°C, rinsed three times 10 min with 0.3% Triton X-100 in PBS, and incubated with the appropriate fluorescent secondary antibodies for 2 h at room temperature. The tissues were washed twice in PBS and mounted on slides using Prolong Diamond Antifade mounting medium (Thermo Fisher Scientific). Fluorescence imaging was conducted with a Zeiss LSM 880 confocal microscope/Nikon Plan-Apochromat 63x/1.4 NA oil immersion objective. For adult eyes, images were acquired using a Leica DFC490 camera mounted on a Leica MZ6 stereomicroscope set at maximum zoom.

Quantifying Mitochondrial Density

Images of larval epidermal cells we acquired as above and processed in Fiji software (Schindelin et al., 2012). Briefly, images encompassing several cells were acquired using a 63x/1.4 NA oil immersion objective. Several 15 μm² regions, avoiding especially sparse or dense areas, were selected. Exposure was adjusted using the Threshold function, selecting Otsu type, B&W and Auto. Images were made binary, and Analyze Particles function was applied (size: 0.1-infinity).

Axonal Transport

Analysis of axonal transport was performed on wandering third instar larvae as previously described (Baldwin et al., 2016). Larvae were pinned at each end dorsal side up to a Sylgard (Sigma 761028) slide and cut along the dorsal midline using micro-dissection scissors. Larvae were covered in dissection solution (128 mM NaCl, 1 mM EGTA, 4 mM MgCl₂, 2 mM KCl, 5 mM HEPES and 36 mM sucrose, adjusted to pH 7 using NaOH), the sides were pinned back and the internal organs removed. Movies were taken using a Nikon E800 microscope with a 60x water immersion lens (NA 1.0 Nikon Fluor WD 2.0) and an LED light source driven by Micro-Manager 1.4.22 Freeware (Edelstein et al., 2014). A CMOS camera (01-OPTIMOS-F-M-16-C) was used to record 100 frames at a rate of 1 frame per 5 s for CCAP-GAL4 samples or 1 frame per 2.5 s for M12-GAL4 samples. Movies were converted into kymographs using Fiji (Schindelin et al., 2012) and quantified manually.

Respirometry analysis

Respiration was monitored at 30°C using an Oxygraph-2k high-resolution respirometer (OROBOROS Instruments) using a chamber volume set to 2 mL. Calibration with air-saturated medium was performed daily. Data acquisition and analysis were carried out using Datlab software (OROBOROS Instruments). Five flies per genotype (equal weight) were homogenized in respiration buffer (120 mM sucrose, 50 mM KCl, 20 mM Tris-HCl, 4 mM KH₂PO₄, 2 mM MgCl₂, and 1 mM EGTA, 1 g/l fatty acid-free BSA, pH 7.2). For MICU3²⁷ experiments, 5 heads per genotype were used. For coupled (state 3) assays, complex I-linked respiration was measured at saturating concentrations of malate (2 mM), glutamate (10 mM) and adenosine diphosphate (ADP, 2.5 mM). Complex II-linked respiration was assayed in respiration buffer supplemented with 0.15 μM rotenone, 10 mM succinate and 2.5 mM ADP.

ATP levels

The ATP assay was performed as described previously (Pogson et al., 2014). Briefly, five male flies of the indicated age or 10 larvae for each genotype were homogenized in 100 μL 6 M guanidine-Tris/EDTA extraction buffer and subjected to rapid freezing in liquid nitrogen. Homogenates were diluted 1/100 with the extraction buffer and mixed with the luminescent solution (CellTiter-Glo Luminescent Cell Viability Assay, Promega). Luminescence was measured with a SpectraMax Gemini XPS luminometer (Molecular Devices). The average luminescent signal from technical triplicates was expressed relative to protein levels, quantified using the DC Protein Assay kit (Bio-Rad). Data from 2-4 independent experiments were averaged and the luminescence expressed as a percentage of the control.

RNA extraction, cDNA synthesis and qRT-PCR

Isolation of total RNA was performed using the RNeasy RNA purification kit (QIAGEN); cDNA was synthesized from total RNA using ProtoScript® II first strand cDNA Synthesis Kit (New England BioLabs, E6560S) according to manufacturer’s instructions. Total RNA concentration was ascertained spectrophotometrically, and equivalent amounts of total RNA underwent reverse transcription for each sample. Quantitative real-time PCR (qRT-PCR) was performed on a CFX96 Touch Real-Time PCR Detection System. Gene-specific primers were designed to have oligos spanning an intron whenever possible. Primer sequences (summarized in Table S1) were as follows: EMRE forward 5′-ACATGTCCAGCGTGTACTTTC-3′ and reverse 5′-GGTATGACGGCACAGAAGATG-3′; MICU1 forward 5′-GTGGCCATGGTCAATCTTTC-3′ and reverse 5′-TTGTTGCTGAGTTGGTTGTCA-3′; MICU3 forward 5-GATCCACAAACCAAGCGAAT-3′ and reverse 5′-CCTCTTCCGGCTCTTGCT-3′; sulfateless forward 5′-AAGCTGTCGATTTGAGTAGCAA-3′ and reverse 5′-GACTGTCCACTCGCAATCAG-3′; javelin forward 5′-GCGGATTTTTCCGTGAATC-3′ and reverse 5′-TCTGGCTCTGGGTGTCATC-3′; and RpL32 forward 5′-GCCGCTTCAAGGGACAGTATCTG-3′ and reverse 5′-AAACGCGGTTCTGCATGAG-3′. Carryover DNA was removed with Turbo DNase free (Ambion, Cat. No. AM1907) according to the manufacturer’s protocol. The relative transcript levels of each target gene were normalized against RpL32 mRNA levels; quantification was performed using the comparative C_T method (Schmittgen and Livak, 2008).

Quantification and Statistical Analysis

Data are reported as mean ± SD, SEM or 95% confidence interval (CI) as indicated in figure legends. For climbing analysis, Kruskal-Wallis non-parametric test with Dunn’s post hoc correction for multiple comparisons was used. For lifespan experiments, significance levels were determined by log-rank tests and reported in the figure legends. Mitochondrial transport was analyzed by one-way ANOVA, and larval crawling, ATP, and Oroboros measurements analyzed by two-tailed unpaired t test. Unless specifically indicated, no significant difference was found between a sample and any other sample in the analysis. Where n is indicated in the figure legends, these refer to biological replicates or number of animals tested. No data points were excluded from the analyses. Analyses were performed using GraphPad Prism 7 software (RRID:SCR_002798).

Published: April 30, 2019