Mitochondria catabolize nutrients by generating sequentially-ordered organic acid intermediates that are oxidized through the tricarboxylic acid cycle. Pathogenic accumulation of metabolic organic acids manifests as devastating organic acidemias/acidurias and other severe diseases, but the underlying mechanisms are largely unknown. Using unbiased C. elegans genetic screening, we here reveal that mutations in the phosphoenolpyruvate carboxykinases PCK-1 and PCK-2 cause buildup of oxaloacetate, a key tricarboxylic acid cycle intermediate, leading to severe mitochondrial damage. Depletion of mitochondrial GOT-2.1 or GOT-2.2, which catalyze oxaloacetate conversion to aspartate, also causes oxaloacetate accumulation and defective mitochondria with disrupted cristae. We demonstrate that oxaloacetate binds the MICOS complex subunit CHCH-3/MIC19 and inhibits its function of promoting IMMT-1/MIC60-dependent membrane shaping and remodeling. In mammalian cells, aberrant OAA buildup similarly causes mitochondrial impairment through MIC19 and MIC60. These findings not only provide important mechanistic insights into mitochondrial damage in the context of defective oxaloacetate metabolism, but also suggest therapeutic strategies for oxaloacetate-related mitochondriopathies.

Loss of pck-2 causes mitochondrial abnormalities

By mutagenizing C. elegans hermaphrodites expressing MITO-GFP and screening mutants with abnormal mitochondrial morphology, we identified two mutants, yq270 and yq271, together with many other mutants that displayed varying abnormal mitochondrial morphologies compared to the predominantly tubular mitochondria in the N2 wild-type animals [7,8,35]. yq270 and yq271 mutants were characterized by the appearance of enlarged and spherical mitochondria in hypodermal cells in a development-dependent manner (S1A Fig). The mitochondrial defect was confirmed by co-labeling mitochondria with MITO-GFP and mCherry-tagged TOMM-20 (TOMM-20::mCh), an outer mitochondrial membrane protein (Fig 1A). Spherical mitochondria also existed in the body-wall muscle cells and intestinal cells in these mutants (Fig 1B). yq270 and yq271 animals showed significant reduction in ATP levels and body-bending rates (Fig 1C and 1D). In addition, both mutants exhibited a lower percentage of hatched embryos, reduced life span and shorter body lengths (Fig 1E–1G). Together these results suggest that the yq270 an yq271 mutations cause mitochondrial abnormalities, leading to defective development and growth.

Genetic mapping and sequencing revealed that yq270 and yq271 both cause mutations in the pck-2 gene on linkage group I (LGI). yq270 (g.4060 G > A) results in a premature stop codon at W212 of the PCK-2 protein, while yq271 (g.4375 C > T) results in a S317F missense mutation (S1B and S1C Fig). C. elegans PCK-2 and its homolog PCK-1 are homologous to the mammalian PEPCKs PCK1 and PCK2 (S1C Fig), which catalyze the conversion of OAA to PEP in gluconeogenesis [24]. Animals carrying a previously identified pck-2 deletion allele, pck-2(ok2586), similarly exhibited spherical mitochondria, impaired mitochondrial functions, and defective development and growth (Fig 1A–1G). We created a rescue construct consisting of transgenic PCK-2 tagged with mCherry, with expression driven by the pck-2 promoter (P_pck-2pck-2::mCh). The fusion protein was cytoplasmic, and it successfully rescued the spherical mitochondria in both yq270 and yq271 mutants to wild-type levels of mitochondrial tubules (Fig 1H). Collectively, these results suggest that loss of pck-2 function causes mitochondrial abnormalities.

Humans have two PEPCK homologs, the cytosolic PCK1 (hPCK1) and mitochondrial PCK2 (hPCK2) (S1C, S1D, and S2 Figs). We tested their rescue ability via transgenic expression of hPCK1 (P_semo-1hPCK1::gfp) or hPCK2 with deletion of its N-terminal mitochondrion-targeting sequence (P_semo-1hPCK2(∆MTS)::gfp). Both proteins were distributed in the cytoplasm, and they mostly rescued the abnormal mitochondrial morphology in C. elegans pck-2(ok2586) mutants (Fig 1I). However, expression of the full-length hPCK2 (P_semo-1hPCK2::gfp), which localized to mitochondria, failed to rescue the defective mitochondria in C. elegans pck-2 mutants (S1E Fig). These results suggest that PCK activity is required in the cytoplasm to maintain mitochondrial morphology, and this function is evolutionarily conserved.

C. elegans PCK-1 and PCK-2 act together to affect mitochondrial homeostasis

In addition to PCK2, C. elegans possesses an additional PEPCK, PCK-1 (S1B, S1C, and S2 Figs). To investigate whether PCK-1 is also required for normal mitochondrial morphology, we analyzed pck-1(ok2098) deletion mutants (S1B Fig) and double mutants of pck-1(ok2098) with pck-2(ok2586). pck-1(ok2098) animals exhibited spherical mitochondria in the hypodermis and body-wall muscles, and pck-2;pck-1 double mutants displayed abnormal mitochondria, similar to single mutants of either pck gene (Fig 2A). Using transmission electron microscopy (TEM), we confirmed that both pck-1(ok2098) and pck-2(ok2586) single and double mutants had spherical mitochondria with disrupted crista organization and significantly reduced crista numbers (Fig 2B). Moreover, these pck single and double mutants exhibited similar reductions in oxygen consumption rate (OCR), ATP production and body bending (Fig 2C–2E), as well as shortened body lengths and life span (Fig 2F and 2G). These results suggest that pck-1 function is required for normal mitochondrial morphology, and that pck-1 acts in the same genetic pathway with pck-2. Using in vitro GST pull-down assays, we found that recombinant His₆-PCK-2 directly interacted with GST-PCK-1 (Fig 2H). Transgenic PCK-1::GFP driven by the pck-1 promoter (P_pck-1pck-1::gfp), which localized to the cytoplasm, successfully rescued the spherical mitochondria in pck-1(ok2098) mutants but not pck-2(ok2586) mutants (Fig 2I). In comparison, transgenic PCK-2::GFP (P_pck-2pck-2::gfp) fully rescued the spherical mitochondria in both pck-1 and pck-2 deletion mutants (Fig 2I). Taken together, these results suggest that PCK-1 and PCK-2 function together, and PCK-2 acts downstream of PCK-1 in maintaining normal mitochondrial structure and functions.

Loss of pck-1 and/or pck-2 leads to OAA buildup that impairs mitochondria

As PEPCKs, PCK-1 and PCK-2 act in the cytoplasm to convert OAA to PEP (Fig 3A and 3B). We thus analyzed OAA levels in single mutants and double mutants of pck-1 and pck-2. Indeed, OAA levels were significantly increased to similar levels in these single and double mutants (Fig 3C). Further analysis indicated that OAA was increased in both cytosol and mitochondria (Fig 3D). Thus, loss of pck-1 and/or pck-2 probably blocks the conversion of OAA to PEP, leading to OAA accumulation in both cytosol and mitochondria. To determine whether OAA accumulation accounts for the abnormal mitochondria in pck mutants, we generated double mutants of pck-1 or pck-2 with got-1.2(yq163), a mutation causing a premature stop at W312 in the GOT-1.2 transaminase that converts Asp to OAA (S3A and S3B Fig). Unlike the spherical mitochondria in pck mutants, single mutants of got-1.2(yq163) and double mutants with pck-1 or pck-2 displayed very well-connected tubular mitochondria, which were confirmed by TEM analysis (Fig 3E and 3F). Importantly, these mutants had significantly lower levels of cytosolic and mitochondrial OAA than pck-1 and pck-2 mutants (Fig 3G), and their ATP levels were significantly higher than in pck-1 and pck-2 mutants (Fig 3H). Collectively, these findings suggest that the aberrantly accumulated OAA is responsible for mitochondrial abnormalities in pck mutants.

Mitochondrial accumulation of OAA causes mitochondrial impairment

While PCK-2 and PCK-1 localize and function in the cytoplasm (Fig 2I), their deficiencies lead to OAA elevation in mitochondria (Fig 3C and 3D). To determine whether the elevated mitochondrial OAA directly causes mitochondrial abnormalities, we investigated mutants of got-2.1 and got-2.2, which encode mitochondrial glutamic-oxaloacetic transaminases that convert OAA to Asp in mitochondria [15] (S3C–S3F Fig). As expected, got-2.1(gk109644) and got-2.2(yq341) mutants had significantly higher mitochondrial OAA levels than the wild type (Fig 4A). Consistent with the elevated mitochondrial OAA, got-2.1(gk109644) and got-2.2(yq341) animals exhibited spherical mitochondria with disrupted crista organization and reduced crista numbers, as in pck-1 and pck-2 mutants (Fig 4B and 4C). In addition, mitochondrial functions were impaired in these mutant animals, manifesting as significantly reduced ATP levels, body-bending rates, OCRs and lifespans (Fig 4D–4G).

Mitochondrial OAA is a pivotal metabolic intermediate primarily produced by malate dehydrogenase 2 (MDH-2) in the TCA cycle (Fig 3A). We sought to reduce OAA levels with RNAi knockdown of mdh-2, the gene that encodes mitochondrial MDH. mdh-2 RNAi significantly reduced mitochondrial OAA levels in got-2.1(gk109644) animals and got-2.2(yq341) animals (Fig 4H). Consistent with this, mdh-2 RNAi-treated got-2.1(gk109644) mutants and got-2.2(yq341) mutants displayed well-connected tubular mitochondria (Fig 4I). These findings suggest that aberrantly accumulated OAA in mitochondria causes mitochondrial impairment. To corroborate this conclusion, we fed N2 animals OAA and found that it caused mitochondrial fragmentation in a concentration-dependent manner (Fig 4J). In contrast, similar supplementation of Asp or Mal did not induce an obvious change in mitochondrial morphology (Fig 4J). By TEM analysis, we found that OAA treatment disrupted mitochondrial cristae and reduced crista numbers, as observed in pck-1/-2 and got-2.1/-2.2 mutants (Fig 4K). Importantly, both cytosolic and mitochondrial OAA levels were significantly elevated following exogenous OAA supplementation (Fig 4L). Taken together, these results suggest that excessive accumulation of OAA in mitochondria causes mitochondrial impairment and dysfunction.

Reinforced expression of IMMT-1 or CHCH-3 ameliorates mitochondrial defects in C. elegans mutants with OAA accumulation

We hypothesized that aberrantly accumulated OAA in mitochondria probably binds to and interferes with proteins that are required for cristae organization and mitochondrial functions. These include EAT-3/Opa1, FZO-1/Mfn1/2, the MICOS complex, respiratory chain complexes, and the ATP synthase complex [3,4,36]. To investigate this possibility, we reinforced the expression of individual proteins in pck-1/-2 mutants (S4 Fig). Our results revealed that reinforced expression of IMMT-1/MIC60, which is capable of membrane binding and shaping [37,38], and CHCH-3/MIC19, which promotes IMMT-1/MIC60 activities [39,40], mostly restored the spherical mitochondria to tubular morphology in pck-1/-2 single and double mutants, and in got-2.1 and got-2.2 mutants (Figs 5A and S5A). TEM analysis further confirmed that the organization of cristae was restored to normal (Figs 5B and S5B). Overexpression of other components involved in crista organization did not have similar effects (S4 Fig). In addition, while exogenously supplied OAA induced mitochondrial fragmentation in the wild type, it did not cause a marked change of mitochondrial morphology in animals overexpressing either IMMT-1 or CHCH-3 (Fig 5C). Thus, reinforced expression of IMMT-1 and CHCH-3 antagonizes OAA-induced mitochondrial abnormalities. These results collectively suggest that OAA probably impairs mitochondria through IMMT-1 and/or CHCH-3.

OAA binds to CHCH-3 and inhibits its function of promoting IMMT-1-dependent membrane shaping

To determine if OAA acts through IMMT-1 and/or CHCH-3, we first examined their expression levels in pck mutants that accumulate OAA. Neither mRNA nor protein levels of IMMT-1, CHCH-3, or several other MICOS components (F54A3.5(MIC10), MOMA-1(MIC27), W04C9.2(MIC13)) were changed in single or double pck mutants (S6A and S6B Fig). This suggests that OAA does not affect mitochondria by altering the expression of these genes. We then examined whether OAA binds with IMMT-1 and CHCH-3. In microscale thermoelectrophoresis (MST) assays, OAA bound well to purified His₆-CHCH-3 (K_d = 2.28 ± 0.67 μM); however, no obvious OAA binding with His₆-IMMT-1 was detected (Fig 6A). In comparison, as the substrate of PCK-1 and PCK-2, OAA bound to these enzymes with much lower binding affinity (K_{d PCK-1} = 331.97 ± 46.11 μM) and PCK-2 (K_{d PCK-2} = 496.37 ± 75.19 μM) (S6C Fig). Further analysis with the AutoDock Vina 1.2.0 algorithm [41,42] suggests that OAA probably binds to CHCH-3 at the C-terminal CHCH domain by forming hydrogen bonds with Asn129, Ala153, and Glu160 residues (Fig 6B). Consistent with this prediction, CHCH-3(N129A) and CHCH-3(E160A) mutants did not bind with OAA in MST assays (Fig 6C). Moreover, overexpression of CHCH-3(N129A) and CHCH-3(E160A), unlike the wild-type CHCH-3, failed to restore the spherical mitochondria to tubular morphology in pck-2;pck-1 double mutants (Fig 6D). Thus, the interaction of OAA with CHCH-3 is specific.

In the fungus Chaetomium thermophilum, Mic19, the homolog of C. elegans CHCH-3, interacts with Mic60 and promotes Mic60-dependent membrane shaping [39,40]. To test this interaction in C. elegans, we knocked in GFP at the C-terminus of IMMT-1, then performed co-immunoprecipitation assays. We found that endogenous CHCH-3 was indeed precipitated with endogenous-level IMMT-1::GFP from cell lysates of both wild type (N2) and pck mutants (S6D Fig). In in vitro GST pull-down assays, increasing the OAA concentration did not markedly affect the interaction of CHCH-3 with IMMT-1 (S6E Fig). Together these results suggest that the aberrantly accumulated OAA in pck-1 and pck-2 mutants probably does not affect the interaction of CHCH-3 with IMMT-1. We next investigated whether OAA interferes with IMMT-1-dependent membrane shaping using liposome assays. We prepared liposomes and incubated them with His₆-IMMT-1 or His₆-CHCH-3. His₆-IMMT-1 induced obvious tubulation of liposomes, while His₆-CHCH-3 did not have the same effect (Fig 6E). Notably, when both His₆-CHCH-3 and His₆-IMMT-1 were together incubated with liposomes, they generated membrane tubules with branches or button-like structures (Fig 6E). These results suggest that CHCH-3 acts through IMMT-1 to promote membrane shaping, which is consistent with previous reports that fungal Mic19 promotes Mic60-dependent membrane shaping [39,40]. Importantly, incubation of CHCH-3(N129A) and CHCH-3(E160A) with His₆-IMMT-1 did not markedly promote the formation of membrane branches and buttons (Fig 6E), suggesting that the OAA binding sites in CHCH-3 are important for its function to promote IMMT-1/Mic60-dependent membrane shaping. In the presence of OAA, the branching of membrane tubules by the coordinated action of His₆-CHCH-3 with His₆-IMMT-1 was inhibited in a concentration-dependent manner, while Asp and Mal did not have marked inhibitory effects (Fig 6F). These results suggest that OAA binds to CHCH-3 and prevents it from promoting IMMT-1-dependent membrane shaping. Taking these results together with the findings that reinforced expression of either CHCH-3 or IMMT-1 ameliorated the mitochondrial defects in animals accumulating OAA (Figs 5A and S5A), we reasoned that immt-1 probably acts downstream of chch-3 in the maintenance of mitochondrial homeostasis. To test this possibility, we overexpressed IMMT-1::GFP in chch-3(tm26333) mutants, and untagged CHCH-3 in immt-1(yq160) mutants. Overexpressing IMMT-1::GFP mostly restored the mitochondria in chch-3 mutants to wild-type tubular morphology, while overexpressing chch-3 failed to ameliorate the mitochondrial fragmentation in immt-1 mutants (Fig 6G). Altogether, these findings provide genetic evidence that CHCH-3 acts through IMMT-1, and suggest a mechanism in which excessively accumulated OAA disrupts mitochondrial homeostasis by inhibiting CHCH-3 function.

OAA accumulation impairs mitochondria in mammalian cells

We finally investigated whether OAA accumulation similarly causes mitochondrial impairment in mammalian cells. 3-MPA (3-mercaptopicolinic acid) is an inhibitor of human PCK1 and PCK2 [43,44]. In U2OS and HeLa cells, 3-MPA treatment induced obvious mitochondrial fragmentation in a concentration-dependent manner (Figs 7A and S7). Consistent with this, 3-MPA led to a significant elevation of total and mitochondrial OAA levels in U2OS cells (Fig 7B). These results suggest that inhibition of hPCK1 and hPCK2 causes mitochondrial accumulation of OAA, which induces mitochondrial fragmentation as in C. elegans. To consolidate this point, we treated cells with OAA. As in C. elegans, exogenously supplied OAA led to mitochondrial fragmentation in U2OS and HeLa cells (Figs 7C and S7), and the total cellular and mitochondrial OAA levels were significantly increased (Fig 7D). TEM analysis further revealed that both 3-MPA and OAA similarly disrupted mitochondrial crista structures and reduced the numbers of cristae and crista junctions (Fig 7E). In contrast, adding Asp and Mal to U2OS and HeLa cells did not induce mitochondrial fragmentation (Figs 7C and S7). To investigate whether OAA acts through MIC19 and MIC60 to impair cristae in mammalian cells, we overexpressed MIC60-EGFP or MIC19-Flag in U2OS cells and treated them with 3-MPA or OAA. Cells overexpressing MIC60-EGFP or MIC19-Flag were resistant to mitochondrial fragmentation induced by 3-MPA or OAA (Fig 7F and 7G). Taken together, these results suggest that OAA accumulation damages mitochondria in mammalian cells in a similar way to that in C. elegans.

C. elegans strains and genetics

C. elegans cultures and genetic analysis were performed with standard procedures. The Bristol N2 strain was used as the wild type. pck-2(yq270), pck-2(yq271), and got-1.2(yq163) mutants were obtained by ethyl methanesulfonate (EMS) mutagenesis. The got-2.2(yq341) mutants were generated by CRISPR/Cas9 system in this laboratory. Integrated arrays, deletion mutants and mutants obtained by EMS mutagenesis were outcrossed with the N2 strain at least four times. All strains used in this study are listed in Supporting information (S1 Table).

Expression vectors

The expression vectors used in the study were constructed using standard protocols and are listed in Supporting information (S2 Table).

Ethyl methanesulfonate (EMS) mutagenesis and gene cloning

Synchronized L4-stage yqIs157 animals expressing MITO-GFP were treated in 50 mM EMS solution for 4 h at 25 °C and then transferred to NGM plates at 20 °C to recover and produce progeny. F2 animals at 24–48 h post-L4 stage were examined for mitochondrial morphology. yq163, yq270 and yq271 mutants were obtained from a screen of 12,000 haploid genomes. yq270 and yq271 mutants showed spherical mitochondria and yq163 mutants displayed better mitochondrial connection. Mapping of mutations was performed using SNP (single-nucleotide polymorphism) mapping and mutation sites were identified with whole-genome sequencing.

Generation of deletion mutants

To generate got-2.2(yq341) deletion mutants, a single-guide RNA (sgRNA) targeting sequence (5′-TTTTCTCTACCGCCGTGCG-3′) in the first exon of the got-2.2 gene was cloned into the Cas9-expressing pDD162 vector. A repair template (5′-TCAACAATGAGCGTTTCCAAGAAGCTTTTCTCTACCGCCGGCTAGCGAAAGTCGTGGTGGTCGCATGTTGAGATGGGACCACCAGA-3′) containing the mutation of interest was designed to remove the cleavage site. An Nhe I restriction site was introduced into the repair template. dpy-10 was used as a positive control marker. The resulting pDD162 construct (20 ng/μl) containing the sgRNA sequence and repair templates (2 μmol/l) were co-injected with the dyp-10 sgRNA construct (20 ng/µl) into gonads of young adult animals. Roller F1 worms were singled into new NGM plates, and the F2 animals were examined by PCR amplification (forward primer: 5′-ATCATGCGTTGTACGTACTC-3′; reverse primer: 5′-CAACAATTCCGGCGTACTCC-3′) and restriction enzyme digestion (Nhe I). All mutations were confirmed by sequencing.

ATP measurements

Three hundred adult animals (48 h post-L4) were homogenized in 200 μl of ATP Lite lysis buffer (Vigorous) and subjected to five freeze-thaw cycles in liquid nitrogen. The lysates were boiled for 5 min and then centrifuged at 15,000g for 5 min at 4 °C to collect the supernatants. Twenty microliters of supernatant was applied to the luciferin-luciferase ATP assay. Luminescence was recorded on a Clarity microplate luminometer (BioTek) between 300 and 600 nm.

Analysis of body-bending rates

Synchronized adult worms (48 h post-L4) were transferred from bacterial lawns to a fresh, unseeded agar plate to remove external bacteria. Individual worms were then transferred to a new unseeded plate containing 1 ml of M9 buffer. Head swings per minute, defined as the return of the head to the original position, were manually counted under a Motic SMZ-168 microscope.

Measurement of OCR

OCRs were measured at 25 °C using a Seahorse XFe24 analyzer (S7801B). Briefly, 300 adult worms (48 h post-L4) were transferred in quadruplicate to a 24-well microplate containing 500 µl of M9 buffer. Basal respiration was measured for a total of 22.5 min with 4.5-min intervals, with each interval consisting of 2 min of mixing, 0.5 min of waiting and 2 min of measuring). Maximal respiration was induced with 20 mM FCCP (carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone) and detected for a total of 40.5 min with 4.5-min intervals, with each interval consisting of 2 min of mixing, 0.5 min of waiting and 2 min of measuring.

Lifespan assays

For each group, 120 synchronized L4 animals were maintained on 4 fresh NGM plates (30 animals per plate) and transferred to a new plate every day until the end of reproduction. Animals that failed to respond to mechanical stimulation were defined as dead. Viable worms were scored every day.

Body length measurements

Synchronized L4 animals were cultured on fresh NGM plates at 20 °C for 48 h. Worms were then anesthetized with 2.5 mM levamisole in M9 buffer and mounted on 3% agar pads. Images were acquired using an Axio Imager M2 microscope (Carl Zeiss). Body lengths were quantified from head to tail tip using ZEN software (v.2.3).

Microscopy and image analysis

C. elegans animals were anesthetized in 2.5 mM levamisole in M9 buffer and immobilized on 3% agar pads. Differential interference contrast (DIC) and fluorescence images were captured at 20 °C with an inverted Zeiss LSM 880 confocal microscope equipped with an Alpha Plan-Apochromat 100×/1.46 oil objective. Images were processed using ZEN software (v.2.3).

TEM analysis

Adult C. elegans animals were frozen under high pressure by using a Leica EM ICE system. Samples were subjected to freeze-substitution in a Leica EM AFS2 unit with anhydrous acetone containing 1% osmium tetroxide and 0.1% uranyl acetate. The parameters were: −90 °C for 72 h, −60 °C for 15 h, −30 °C for 15 h, and 0 °C for 5 h. The samples were washed with cold anhydrous acetone, and infiltrated with Embed-812 resin in a graded series (resin:acetone 1:3 for 3 h, 1:1 for 5 h, 3:1 overnight), followed by pure resin for 4 h. The samples were then incubated 60 °C for 48 h for polymerization. Ultrathin sections (70 nm) were cut on a Leica EM UC7 microtome, followed by staining with 2% uranyl acetate for 10 min and lead citrate for 5 min. Images were acquired with a HT7800 (Hitachi) at 80 kV.

For TEM analysis of U2OS cells, U2OS cell monolayers were first fixed with 2.5% glutaraldehyde in 0.1 M PBS at 4 °C for 2 h, and then post-fixed with 1% OsO₄ at 4 °C for 1 h, followed by ethanol dehydration and embedding in EMbed-812. Ultrathin sections (70 nm) were prepared using an UC6 ultramicrotome equipped with a 45° diamond knife (Diatome). The grids were stained with aqueous uranyl acetate and Reynolds lead citrate. Imaging was performed as above.

Recombinant proteins and GST pull-down

Individual cDNAs were cloned into pET-21a (+) or pGEX-4T-1 and expressed in E. coli Rosetta (DE3) strain. His₆-tagged or GST-fused recombinant proteins were purified with Ni-Sepharose beads (GE Health Care) or Glutathione Sepharose 4B beads (GE Health care), respectively. The lysis buffer, which is typically used for membrane protein purification, contained 50 mM HEPES, pH 7.4, 150 mM NaCl, 10% glycerol, and 1% n-dodecyl-β-D-maltoside (DDM), and 1 mM phenylmethylsulfonyl fluoride (PMSF). The elution buffer for purification of His-tagged proteins contained 50 mM HEPES, pH 7.4, 150 mM NaCl, 10% glycerol, 0.03% DDM and 250 mM imidazole. The eluted protein was transferred to an Amicon Ultra ultrafiltration tube (Merck Millipore) and centrifuged to remove imidazole and DDM by buffer exchange against 50 mM HEPES (pH 7.4) and 150 mM NaCl.

For GST pull-down, His₆-tagged proteins were incubated with GST-fusion proteins immobilized on glutathione Sepharose 4B beads at 4 °C overnight. The beads were washed extensively and bound proteins were resolved by SDS-PAGE and detected with western blotting.

Western blotting and immunoprecipitation (IP)

For western blotting, adult worms (48 h post-L4 stage) grown on 20 NGM plates (10 cm in diameter) were used to prepare cytosolic and mitochondrial fractions. Cytosolic or mitochondrial samples were separated on 12.5% SDS-PAGE and analyzed by western blotting. For immunoprecipitation, worms expressing GFP-tagged proteins were lysed in worm lysis buffer (25 mM Tris-HCl, pH 7.5, 50 mM NaCl, 0.1% NP-40, 1 mM PMSF (phenylmethylsulfonyl fluoride) and 1% glycerol, Complete Protease Inhibitor Cocktail (Roche) on ice. The crude lysates were collected by centrifugation at 12,000g for 10 min at 4 °C. The supernatants were incubated with GFP-Nanoab-Agarose Beads (Lablead) overnight at 4 °C, and washed three times with wash buffer (25 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.1% NP-40, and 1 mM PMSF). Bound proteins were resolved by SDS-PAGE and detected by western blotting. Antibodies and reagents used in this study are listed in Supporting information (S3 Table).

RNA interference (RNAi)

Five L4 animals were grown on NGM plates supplemented with 1 mM IPTG (isopropyl β-D-thiogalactoside) and seeded with E. coli OP50 expressing double-stranded RNA of genes of interest. The F1 progeny at the L4 stage were transferred to fresh RNAi plates and cultured for 48 h. The animals were then examined for mitochondrial morphology under confocal fluorescence microscope.

Measurement of OAA levels

Cytosolic and mitochondrial fractions were prepared from C. elegans animals expressing TOMM-20::RFP::3× HA and mammalian cells by following previously described methods [6,52,53]. Briefly, pellets (approximately 500 μl) of worms synchronized at 48 h post-L4 were homogenized in mitochondrial buffer (10 mM MOPS, pH 7.4, 70 mM sucrose, 1 mM EGTA, 210 mM sorbitol) on ice using a Dounce homogenizer. The crude lysates were fractionated by differential centrifugation (450g, 550g, and 12,000g, 10 min each) at 4 °C to pellet the mitochondria. The supernatants were used as cytosolic fractions. The pelleted mitochondrial fractions were resuspended in PBS (Phosphate Buffered Saline) and incubated with HA-conjugated magnetic beads (Thermo Scientific) at 4 °C for 60 min. The mitochondria-capture beads were collected by using DynaMag-2 (Thermo Scientific). Mitochondria were lysed with PBS containing 0.05% Triton X-100 and used for further analysis. To prepare mitochondria from cultured cells, cell pellets (100 μl) were broken in homogenization buffer (3.5 mM Tris-HCl, pH 7.8, 2.5 mM NaCl, 0.5 mM MgCl₂, 1 mM PMSF) on ice using a Dounce homogenizer. The lysates were centrifuged sequentially at 1,200g for 3 min and 15,000g for 5 min at 4 °C to pellet the mitochondria. Mitochondrial fractions were examined by western blotting using antibodies against organelle-specific proteins. OAA levels in the mitochondrial and cytosolic fractions were measured by using the Oxaloacetate Assay Kit (Sigma) following the procedures provided by the supplier.

Immunostaining

One hundred worms synchronized at 48 h post L4 were fixed with 2% PFA for 10 min at room temperature (RT), and further incubated with pre-cold methanol for 5 min at −20 ℃. The worms were collected and washed twice in Tris-Triton buffer (100 mM Tris, PH7.4, 1% Triton X-100, 1 mM EDTA PH8.0). The worms were resuspended in TTB buffer (100 mM Tris PH7.4,1% Triton X-100, 1 mM EDTA PH8.0, 1% β-mercaptoethanol) for 1 h at 37 ℃, and collected. Following wash once in Boric acid buffer (25 mM H₃BO₃, 0.01% Triton X-100), worms were sequentially incubated in Boric acid buffer containing 25 mM DTT for 15 min at 37 ℃ and Boric buffer containing 0.3% H₂O₂ for 15 min at RT. The samples were washed with Boric buffer again and incubated in PBST (PBS, 2% Tween-20) for 30 min at RT, and then incubated with the primary antibody in the same buffer overnight at 4 °C. After extensive wash, the samples were incubated with Alexa Fluor 488- or Fluor 568-conjugated secondary antibodies in PBST for 2–4 h at RT. Worms were then imaged with confocal microscopy.

For immunostaining of cultured mammalian cells, cells grown on confocal glass dishes (Cellvis) were fixed with 4% PFA for 20 min and then permeabilized with 0.3% Triton X-100 for 10 min at RT. After extensive wash with PBS, cells were first incubated in PBS containing 5% BSA and 0.1% Triton X-100 at RT for 1 h, and then incubated with primary antibody in the same buffer overnight at 4 °C. Cells were extensively washed and incubated with Alexa Fluor 488- or Fluor 568-conjugated secondary antibodies in PBST for 1 h at RT, and washed extensively. Imaging was carried out with confocal microscopy.

Treatment with OAA, Asp, Malate or 3-MPA

C. elegans embryos were cultured on NGM plates supplemented with OAA (0.1, 1.0, or 5.0 mM), aspartate (5.0 mM), or malate (5.0 mM) and allowed to develop to adults (48 h post-L4). The mitochondrial morphology was examined using confocal microscopy. For treatment of mammalian cells, U2OS and HeLa cells were grown to approximately 40% density and then cultured in complete medium supplemented with 3-MPA (0.25, 0.50, or 1.00 mM), OAA (1.0 or 5.0 mM), aspartate (5.0 mM), or malate (5.0 mM). Twenty-four hours later, the cells were stained with Mito-CMXRos for 20 min and imaged by confocal microscopy to assess mitochondrial morphology.

Real-time quantitative PCR (qPCR)

Total RNA was isolated from C. elegans at 48 h post-L4 stage using TRIzol reagent (Life Technologies). cDNA was synthesized using oligo-(dT18) primers and M-MLV reverse transcriptase (Promega). qPCR was performed on a CFX-96 real-time system (Bio-Rad) with SYBR Premix Ex Taq (Medchem Express). The ΔΔCT method was used for quantification with act-1 as the reference gene. Oligos used for qPCR are listed in Supporting information (S4 Table).

Microscale thermophoresis (MST) assays

MST assays were carried out on a Monolith NT.115 system (Nanotemper Technologies). Briefly, 200 nM of target protein (IMMT-1 or CHCH-3) was incubated with varying concentrations of OAA in 20 µl and loaded into NT.115 standard-coated capillaries (Nanotemper Technologies). MST measurements were conducted in triplicate at 25 °C with 40% excitation power and high MST power. Data were processed with NanoTemper software.

Virtual docking of OAA on CHCH-3

The protein structure file of CHCH-3 was retrieved from the AlphaFold Database (https://alphafold.ebi.ac.uk/) and preprocessed using AutoDockTools 1.5.6 [54]. Molecular docking of OAA to CHCH-3 was performed using AutoDock Vina [41,42]. The resulting ligand-protein poses were analyzed in PyMOL (https://pymol.org/2/) with the most favorable binding pose selected and the hydrogen-bond interactions annotated.

Liposome assays

Type I lipids (Folch Fraction I, Sigma) were dissolved in chloroform and dried under a nitrogen stream, and further placed under high vacuum (≥3 h) to remove residual solvent. To generate the liposomes, the lipid film was hydrated in the buffer containing 50 mM HEPES, pH 7.4, 150 mM NaCl at 60 °C for 1 h. Liposomes at 1 mg/ml were incubated with 5 µM of target protein (IMMT-1 and CHCH-3) and varying concentrations of metabolites (OAA, aspartate, or malate) for 30 min at 20 °C. Five to seven microliters of liposome suspension was loaded on a grid coated with formvar and incubated for 1 min. After removal of excess liposome solution, the grid was stained with 2% uranyl acetate and 0.45% methyl cellulose solution for 5 min and imaged with an HT7800 (Hitachi) transmission electron microscope operating at 80 kV.

Statistics and reproducibility

Statistical analyses were performed using GraphPad Software (version 8.2.1, Dotmatics) and IBM SPSS statistics (version 21.0). Data are presented as mean ± s.e.m (standard error of the mean). Comparisons between two groups were made by using the two-tailed unpaired Student t test. Comparisons of multiple groups were assessed by using one-way or two-way analysis of variance (ANOVA) followed by Dunnett’s, Sidak’s, or Tukey’s post-hoc test. *** P < 0.001; ** P < 0.01; * P < 0.05; ns, not significant (P > 0.05).

S1 Fig. Mutations in pck-2 cause abnormal mitochondrial morphology in C. elegans.

(A) Images (left) and quantification (right) of mitochondria in the hypodermis of N2 (wild type), yq270, yq271 and ok2586 animals. Adult animals were examined at the indicated ages after the 4th larval (L4) stage. A total of 120 animals were examined in n = 4 repeats for each genotype. Bars, 5 µm. (B) Schematic depiction of C. elegans pck-2 and pck-1 genes. Filled boxes and thin lines represent exons and introns, respectively. The locations of the yq270 and yq271 point mutations are indicated with red asterisks. ok2586 and ok2098 deletions are indicated with red horizontal lines. (C) Diagram of C. elegans (C.e.) PCK-2, C.e. PCK-1, human (Homo sapiens, H.s.) PCK1 and H.s PCK2 proteins. Mutation sites and deletion regions are indicated in C.e. PCK-2 and C.e. PCK-1. The mitochondrion-targeting sequence (MTS) of PCK2 is indicated in the red circle. Yellow and green boxes indicate PEPCK_N and PEPCK_GTP domains, respectively. (D) Representative images of mitochondria in HeLa cells with transient expression of human PCK1-EGFP and human PCK2-EGFP. Mitochondria were stained with Mito-CMXRos. Bars, 10 µm. (E) Full-length hPCK2 fails to ameliorate the abnormal mitochondria in C. elegans pck-2(ok2586) mutants. hPCK2::GFP driven by the semo-1 promoter (P_semo-1hPCK2::gfp) was expressed in hypodermis of C. elegans pck-2(ok2586) mutants expressing MDH-2::mKATE. Left, representative images of mitochondria; right, quantification. Bars, 5 µm. A total of 120 animals were examined for each group in n = 4 independent repeats. For quantifications, data points represent (mean ± s.e.m.). P values were determined using one-way ANOVA in (A) or the two-tailed unpaired Student t test in (E). ***P < 0.001; **P < 0.01; *P < 0.05; ns, not significant (P > 0.05). The numerical data for this figure can be found in S1 Data (Fig S1A and S1E).

https://doi.org/10.1371/journal.pbio.3003649.s001

(TIF)

S2 Fig. Amino acid sequence alignment of human (H.s.

) PCK1 and PCK2 with C. elegans PCK-1 and PCK-2. The mitochondrion-targeting sequence (MTS) of H.s. PCK2 is indicated in the violet box. The PEPCK_N and PEPCK_GTP domains are indicated in green and blue boxes, respectively. The residues affected by the yq270 and yq271 point mutations in C.e. PCK-2 are labeled in red.

https://doi.org/10.1371/journal.pbio.3003649.s002

(TIF)

S3 Fig. Characterization of glutamate-oxaloacetate transaminases.

(A) Diagram of the C. elegans got-1.2 gene (top) and the encoded protein (bottom). In the gene, filled boxes and thin lines represent exons and introns, respectively. The location of the yq163 point mutation is indicated with the red asterisk. In the protein, the yellow box indicates the aminotransferase domain. (B) Expression pattern of GOT-1.2::mCh driven by the got-1.2 promoter (P_got-1.2got-1.2::mcherry) in adult N2 hypodermal and intestinal cells. DIC (differential interference contrast) and fluorescence images are shown for each tissue. Bars, 5 µm. (C) Diagram of the C. elegans got-2.1 gene (top) and the encoded protein (bottom). The gk109644 point mutation is indicated with the red asterisk. The yellow box indicates the aminotransferase domain. (D) Expression pattern of GOT-2.1::GFP under the control of the got-2.1 promoter (P_got-2.1got-2.1::gfp) in adult N2 intestine cells. Mitochondria were labeled with MDH-2::mKATE. Bars, 5 µm. (E) Diagram of the C. elegans got-2.2 gene (top) and the encoded protein (bottom). The yq341 point mutation is indicated with the red asterisk. The yellow box indicates the aminotransferase domain. (F) Expression pattern of GOT-2.2::GFP driven by the got-2.2 promoter (P_got-2.2got-2.2::gfp) in adult N2 hypodermal and muscle cells. Mitochondria were labeled with MDH-2::mKATE. Bars, 5 µm.

https://doi.org/10.1371/journal.pbio.3003649.s003

(TIF)

S4 Fig. The effects of overexpression of key mitochondrial proteins on mitochondrial defects in pck-1 and pck-2 mutants.

(A) Schematic depiction of key regulators of mitochondrial fusion and cristae formation and maintenance. (B) Representative images of mitochondria in the hypodermis in pck-1(ok2098) and pck-2(ok2586) mutants overexpressing FZO-1::GFP (P_fzo-1fzo-1::gfp) or EAT-3::GFP (P_eat-3eat-3::gfp). Bars, 5 µm. (C) Representative images of mitochondria (labeled with MITO-GFP) in the hypodermis in pck-1(ok2098) and pck-2(ok2586) mutants overexpressing F54A3.5::mKATE (P_F54A3.5F54A3.5::mKate), MOMA-1::mKATE (P_moma-1moma-1::mKate) or W04C9.2::mKATE (P_col-19W04C9.2::mKate). Bars, 5 µm. (D) Representative images of mitochondria (labeled with MDH-2::mKATE) in the hypodermis in pck-1(ok2098) and pck-2(ok2586) mutants overexpressing NUO-3::GFP (P_nuo-3nuo-3:gfp), NUO-4::GFP (P_nuo-4nuo-4:: gfp), SDHA-2::GFP (P_sdha-2sdha-2:: gfp) or ATP-2::GFP (P_atp-2atp-2:: gfp). Bars, 5 µm.

https://doi.org/10.1371/journal.pbio.3003649.s004

(TIF)

S5 Fig. Reinforced expression of IMMT-1 and CHCH-3 restores mitochondrial structure in pck-1 and pck-2 single mutants.

(A) Representative images (left) and quantification (right) of mitochondria (labeled with MDH-2::mKATE) in the hypodermis of N2, pck-1(ok2098) single mutants, and pck-2(ok2586) single mutants without or with overexpression of IMMT-1::GFP (P_immt-1immt-1::gfp) or untagged CHCH-3 (P_chch-3chch-3). CHCH-3 was visualized by immunostaining with CHCH-3 antibody. Bars, 5 µm. A total of 120 animals were examined for each genotype in n = 4 repeats. (B) TEM images (left) and quantification (right) of mitochondria in the hypodermis of N2, pck-1(ok2098) single mutants, and pck-2(ok2586) single mutants without or with overexpression of IMMT-1::GFP (P_immt-1immt-1::gfp) or CHCH-3 (P_chch-3chch-3). Bars, 1 µm. A total of 400 mitochondria from n = 4 animals were examined for tubular mitochondria each genotype (top right). Thirty mitochondria were scored for crista numbers in each group (bottom right). For quantifications, data points represent (mean ± s.e.m.). P values were determined using one-way ANOVA in (A and B). ***P < 0.001; **P < 0.01; *P < 0.05; ns, not significant (P > 0.05). The numerical data for this figure can be found in S1 Data (Fig S5A and S5B).

https://doi.org/10.1371/journal.pbio.3003649.s005

(TIF)

S6 Fig. Analysis of the effects of OAA on expression of MICOS subunits and the interaction of IMMT-1 with CHCH-3.

(A) Relative levels of mRNAs encoding the indicated MICOS subunits. n = 3 independent experiments. (B) Western blotting of IMMT-1 and CHCH-3 in purified mitochondria from animals of the indicated genotypes. IMMT-1, CHCH-3, HSP-60 (mitochondrial marker) and α-tubulin (cytosolic marker) were detected using the corresponding antibodies. (C) Binding curves, determined by microscale thermoelectrophoresis (MST), of OAA with His₆-PCK-1 and His₆-PCK-2. Coomassie blue staining of purified His₆-PCK-1 and His₆-PCK-2 is shown on the left. n = 3 independent experiments. (D) Co-IP of endogenous CHCH-3 with IMMT-1::GFP (knock-in) from animals with the indicated genotypes. IPs were performed with α-GFP antibody and precipitated proteins were detected with IMMT-1 and CHCH-3 antibodies. (E) His₆-CHCH-3 is pulled down by GST-IMMT-1 in the presence of increasing concentrations of OAA. His₆-CHCH-3 was detected with His₆ antibody. Coomassie blue staining of purified GST and GST-IMMT-1 is shown on the left. For quantifications, data points represent (mean ± s.e.m.). P values were determined using one-way ANOVA. The numerical data for this figure can be found in S1 Data (Fig S6A and S6C).

https://doi.org/10.1371/journal.pbio.3003649.s006

(TIF)

S7 Fig. Inhibition of PCK or addition of exogenous OAA leads to abnormal mitochondria in HeLa cells.

Representative images (left) and quantification (right) of abnormal mitochondria in HeLa cells treated with 3-MPA (inhibitor of PCK1 and PCK2), OAA, Asp, and Mal at the indicated concentrations. Mitochondria were labeled with Mito-CMXRos. Boxed regions are magnified (2.5×) in the bottom left. Bars, 10 µm. A total of ≥240 cells were scored in each treatment in 4 repeats. For quantifications, data points represent (mean ± s.e.m.). P values were determined using one-way ANOVA. ***P < 0.001; **P < 0.01; *P < 0.05; ns, not significant (P > 0.05). The numerical data for this figure can be found in S1 Data (Fig S7).

https://doi.org/10.1371/journal.pbio.3003649.s007

(TIF)

S1 Data. Numeric data underlying all main figures and supporting Information (Figs 1C–1I, 2A–2G, 2I, 3C–3H, 4A–4L, 5A–5C, 6A, 6C–6G, 7A–7F, S1A, S1E, S5A, S5B, S6A, S6C, and S7).

This file contains individual replicate values, replicate numbers, and complete statistical results used to generate the plots.

https://doi.org/10.1371/journal.pbio.3003649.s013

(XLSX)