mTORC1 is required for expression of LRPPRC and cytochrome-c oxidase but not HIF-1α in Leigh syndrome French Canadian type patient fibroblasts

Abstract

Leigh syndrome French Canadian type (LSFC) is a mitochondrial disease caused by mutations in the leucine-rich pentatricopeptide repeat-containing (LRPPRC) gene leading to a reduction of cytochrome-c oxidase (COX) expression reaching 50% in skin fibroblasts. We have shown that under basal conditions, LSFC and control cells display similar ATP levels. We hypothesized that this occurs through upregulation of mechanistic target of rapamycin (mTOR)-mediated metabolic reprogramming. Our results showed that compared with controls, LSFC cells exhibited an upregulation of the mTOR complex 1 (mTORC1)/p70 ribosomal S6 kinase pathway and higher levels of hypoxia-inducible factor 1α (HIF-1α) and its downstream target pyruvate dehydrogenase kinase 1 (PDHK1), a regulator of mitochondrial pyruvate dehydrogenase 1 (PDH1). Consistent with these signaling alterations, LSFC cells displayed a 40–61% increase in [U-¹³C₆]glucose contribution to pyruvate, lactate, and alanine formation, as well as higher levels of the phosphorylated and inactive form of PDH1-α. Interestingly, inhibition of mTOR with rapamycin did not alter HIF-1α or PDHK1 protein levels in LSFC fibroblasts. However, this treatment increased PDH1-α phosphorylation in control and LSFC cells and reduced ATP levels in control cells. Rapamycin also decreased LRPPRC expression by 41 and 11% in LSFC and control cells, respectively, and selectively reduced COX subunit IV expression in LSFC fibroblasts. Taken together, our data demonstrate the importance of mTORC1, independent of the HIF-1α/PDHK1 axis, in maintaining LRPPRC and COX expression in LSFC cells.

INTRODUCTION

Mitochondrial diseases, also known as oxidative phosphorylation (OXPHOS) disorders, are the most frequent inherited metabolic diseases globally with a prevalence rate of 1:5,000 (15, 25, 27). This includes Leigh syndrome French Canadian type (LSFC), a recessive neurodegenerative infantile mitochondrial disease with an incidence of 1:2,000 in the Saguenay-Lac-Saint-Jean region of Quebec, Canada (22). LSFC is caused by mutations in the nuclear gene leucine-rich pentatricopeptide repeat-containing (LRPPRC; 21). Most Quebec LSFC patients are homozygous for an A354V substitution that causes a decrease in the expression of the LRPPRC protein (21, 41). Recently, novel disease-causing mutations in this gene have also been reported outside the French Canadian population (26).

LRPPRC is a multifunctional protein that plays an important role in mRNA metabolism including posttranscriptional processes and translation (30, 32). Its deficiency produces a tissue-specific mitochondrial defect that, in fibroblasts, is characterized by an isolated respiratory chain complex IV or cytochrome-c oxidase (COX) defect (32, 33, 41). LSFC fibroblasts present several mitochondrial functional abnormalities including reduced mitochondrial membrane potential, fragmentation of the mitochondrial network, and impaired OXPHOS capacity (2). However, despite these abnormalities, LSFC fibroblasts have preserved ATP levels in basal conditions, suggesting the activation of a compensatory mechanism (2). For instance, upregulation of glycolysis and the concomitant production of lactate could potentially overcome the decrease in mitochondrial respiration.

Among the numerous signaling pathways that regulate glycolysis, mechanistic target of rapamycin (mTOR) is a key regulator of energy metabolism. mTOR exists as two functional complexes distinguished by their composition: mTOR complexes 1 (mTORC1) and 2 (mTORC2; 19). The best characterized complex, mTORC1, is sensitive to rapamycin and is regulated positively by Akt/protein kinase B (PKB) and negatively by AMP-activated protein kinase (AMPK) in response to multiple signals as well as the cellular AMP-to-ATP ratio (17, 37). In immune and cancer cells, mTORC1 regulates aerobic glycolysis through increased expression of several proteins including the transcription factor hypoxia-inducible factor 1α (HIF-1α) and its target pyruvate dehydrogenase kinase 1 (PDHK1; 4, 11, 16, 34). PDHK1 phosphorylates and inhibits pyruvate dehydrogenase 1 (PDH1), the gatekeeper of glucose oxidative metabolism, leading to a metabolic switch away from mitochondria and toward glycolysis (reviewed in Refs. 8, 34, 39).

On the basis of these findings, we hypothesize that activation of the mTORC1 pathway and its downstream targets HIF-1α/PDHK1 could be a compensatory mechanism in LSFC fibroblasts, enabling them to maintain their ATP levels. Our data demonstrate that contrary to our hypothesis, inhibition of mTORC1 does not alter the HIF-1α/PDHK1 signaling pathway or ATP levels in LSFC fibroblasts. However, our study has uncovered a potential role for mTORC1 in LRPPRC and COX expression in LSFC fibroblasts.

METHODS

Cell culture.

Skin fibroblasts from one LSFC female patient (8 yr old) homozygous for the common A354V mutation and one paired female control (5 yr old) were provided by the LSFC Consortium Biobank (Université du Québec à Chicoutimi, Chicoutimi, QC, Canada). The protocol was approved by the Research Ethics Committee of the Centre Intégré Universitaire de Santé et de Services Sociaux du Saguenay-Lac-Saint-Jean and the Montreal Heart Institute. Unless stated, all chemicals were obtained from Sigma-Aldrich (St. Louis, MO). Primary fibroblasts were grown (up to passage 18) at 37°C and 5% CO₂ in high-glucose (4.5 g/l) Dulbecco’s modified Eagle’s medium (DMEM; Mediatech, Manassas, VA) supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin, 10 mg/ml streptomycin (Invitrogen, Carlsbad, CA), and 1% MEM vitamin and nonessential amino acid solutions from Mediatech. For the labeling experiments, which require a larger number of cells, we used fibroblasts that were immortalized by transduction with a retroviral vector expressing the HPV-16 E7 gene and the catalytic component of human telomerase as described (42).

Cell treatment.

To examine mTORC1 function, cells were treated for 16 h with 100 nM rapamycin or vehicle (DMSO) in supplemented DMEM. On the day of the study, cells were cultured in a serum-free nonsupplemented DMEM with inhibitor or vehicle, as specified, for an additional 4 h.

Gel electrophoresis and immunoblotting.

Cells were washed twice with ice-cold PBS. They were then lysed in buffer containing 25 mM Tris·HCl pH 7.4, 150 mM NaCl, 1 mM sodium orthovanadate, 20 mM sodium fluoride, 10 mM sodium pyrophosphate, 2 mM EGTA, 2 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 0.1% sodium dodecyl sulfate (SDS), 1.5% dodecyl maltoside, and protease inhibitors. Lysates were kept on ice for 30 min and then centrifuged at 20,000 g for 30 min at 4°C to remove insoluble material. An equal amount of protein was resolved on 8 or 12% SDS-PAGE gels, as appropriate, and transferred to nitrocellulose membranes (Bio-Rad, Hercules, CA) for 1 h at 4°C. Membranes were blocked for 1 h with 5% (wt/vol) nonfat dry milk in TBST (25 mM Tris·HCl pH 7.6, 150 mM NaCl, and 0.05% Tween 20) and then incubated overnight at 4°C with the following primary antibodies: AMPK, phospho-AMPK (Thr172), acetyl-CoA carboxylase (ACC), phospho-ACC (Ser79), mTOR, phospho-mTOR (Ser2448), Akt, phospho-Akt (Ser473), p70 ribosomal S6 kinase (p70S6K), phospho-p70S6K (Thr389), eukaryotic translation initiation factor 4E-binding protein-1 (4E-BP1), phospho-4E-BP1 (Thr37/46), HIF-1α, PDHK1, COX subunit IV (COX IV), and NADH dehydrogenase (ubiquinone) iron-sulfur protein-2, mitochondrial (NDUFS2), from Cell Signaling Technology (Danvers, MA); β-actin from Santa Cruz Biotechnology (Dallas, TX); PDH1-α, phospho-PDH1-α (Ser293), and an OXPHOS cocktail from Abcam (Toronto, ON, Canada); and a polyclonal rabbit anti-LRPPRC antibody raised against a 22-amino acid peptide having the sequence CEPPESFEFYAQQLRKLRENSS (antibody 295-313; Zymed Laboratories, San Francisco, CA). Membranes were then incubated at room temperature for 1.5 h with the appropriate horseradish peroxidase-conjugated secondary antibody from Cell Signaling Technology. Immunoreactive bands were then visualized using chemiluminescence and quantified using ImageJ software (http://imagej.net/Downloads). All protein levels were expressed in arbitrary units and normalized to β-actin, which was used as loading control. For treatment conditions, results were presented as fold change over basal conditions.

ATP levels.

ATP levels were measured with the ATPlite luminescence ATP detection assay system from PerkinElmer (Waltham, MA), according to the manufacturer’s instructions. Fibroblasts were seeded in opaque 96-well plates at a density of 10,000 cells per well. Determinations were performed in triplicate. Luminescence was quantified using a Synergy 2 Alpha Microplate Reader (BioTek Instruments, Winooski, VT).

Cell treatments and [U-¹³C₆]glucose labeling.

The relative contribution of exogenous glucose to the formation of intracellular pyruvate, lactate, and alanine was assessed using modifications of previously published methods (18, 28). For these studies, confluent immortalized fibroblasts were cultured in supplemented DMEM with the following modifications: 5 mM glucose, 0.2 mM pyruvate, and 0.5 mM glutamine. On the day of the study, cells were washed with PBS and then incubated in a serum-free nonsupplemented DMEM containing 5 mM [U-¹³C₆]glucose, 0.2 mM pyruvate, and 0.5 mM glutamine for 4 h. Cells were washed again with PBS and recovered in a 70% methanol buffer. Hydroxylamine was added to the supernatants to a final concentration of 25 mM. Labeled metabolites (pyruvate, lactate, and alanine) were extracted with 100% methanol, sonicated for 1 min, incubated at 70°C for 15 min, and then centrifuged at 14,000 rpm at 4°C for 15 min. The resulting supernatants were dried, taken up in 25 μl pyridine, incubated at 45°C for 90 min, derivatized in 100 μl N-methyl-N-tert-butyldimethylsilyltrifluoroacetamide (MtBSTFA), and then incubated at 90°C for 4 h. Metabolites were then measured by gas chromatography-mass spectrometry (GC-MS). Details for the measurement of ¹³C enrichment of lactate, pyruvate, and alanine by GC-MS, as well as the equations used to calculate the relative contribution of exogenous glucose to the formation of these metabolites, have been previously described (28, 29).

Statistical analysis.

Statistical analysis was performed using GraphPad Prism 5 software. Significant differences were determined using a paired two-tailed Student’s t-test. Data were log transformed (base 10) for analysis when necessary. The results are expressed as means ± SE, and significance was accepted at P ≤ 0.05.

RESULTS

Increased contribution of exogenous glucose to the formation of pyruvate, lactate, and alanine in LSFC fibroblasts.

As previously reported (32), compared with controls, LSFC fibroblasts displayed a reduction of LRPPRC and COX IV expression by 72 and 50%, respectively (Fig. 1, A and B). Despite mitochondrial dysfunction and COX deficiency, control and LSFC fibroblasts had similar ATP levels under basal conditions (2). We hypothesized that this was the result of increased glycolytic flux. As a first step to address this question, we evaluated the relative contribution of exogenous glucose to pyruvate, lactate, and alanine using [U-¹³C₆]glucose (Fig. 1C) as described in the methods section. In control fibroblasts, 32% of labeled pyruvate derives from [U-¹³C₆]glucose, while the remainder comes from unlabeled substrates, most likely extracellular pyruvate (Fig. 1D). Furthermore, the formation of ¹³C-labeled pyruvate, lactate, and alanine from [U-¹³C₆]glucose was increased by 41%, 61%, and 52%, respectively, in LSFC cells compared with control cells (P ≤ 0.05).

Fig. 1.

Higher contribution of exogenous glucose to pyruvate formation in Leigh syndrome French Canadian type (LSFC) fibroblasts. A and B: immunoblot and densitometric analysis of leucine-rich pentatricopeptide repeat-containing (LRPPRC, n = 12; A) and cytochrome-c oxidase subunit IV (COX IV, n = 9; B) protein levels in primary fibroblasts from LSFC and control donors. C and D: immortalized control and LSFC fibroblasts were incubated with 5 mM [U-¹³C₆]glucose, 0.2 mM pyruvate, and 0.5 mM glutamine for 4 h in serum-free DMEM, and the relative contribution of glucose to the formation of pyruvate, lactate, and alanine was analyzed by GC-MS (n = 7). Results represent means ± SE. Difference between control and LSFC cells was assessed with a paired Student’s t-test. *P ≤ 0.05; ***P ≤ 0.001.

Upregulation of the Akt/mTORC1 signaling pathway in LSFC fibroblasts.

We next examined the AMPK and Akt/mTORC1 signaling pathways, two master regulators of cellular energy homeostasis (4, 7, 38). As shown in Fig. 2A, we did not detect any changes in the phosphorylation levels of AMPKα or its downstream target ACC in LSFC fibroblasts, which is consistent with their unchanged ATP levels under basal conditions. Next, we determined the activity of the Akt/mTOR signaling pathway and found an increase of total (28%; P ≤ 0.01) and phospho-Akt (138%; P ≤ 0.01) resulting in an unchanged ratio. We did not detect any changes in total or phospho-mTOR (Fig. 2B). However, mTOR regulation is complex, and it has been suggested that activation of downstream targets should be used as proxy for its activation. As shown in Fig. 2C, we observed enhanced phosphorylation of p70S6K and 4E-BP1 by 59% (P ≤ 0.05) and 29% (P ≤ 0.001), respectively (Fig. 2C), as well as a 36% increase in total 4E-BP1 (P ≤ 0.001). This resulted in a 75% increase in the ratio of phosphorylated to total protein for p70S6K (P ≤ 0.05) but an unchanged ratio for 4E-BP1 in LFSC fibroblasts compared with controls.

Fig. 2.

Upregulation of the Akt/mechanistic target of rapamycin (mTOR) complex 1 (mTORC1) signaling pathway in Leigh syndrome French Canadian type (LSFC) fibroblasts. A: AMPK activity of primary fibroblasts from control and LSFC donors was assessed by immunoblot and densitometric analysis of total and phosphorylated levels of AMPK (n = 7) and its downstream target acetyl-CoA carboxylase (ACC; n = 6). B and C: activity of the mTORC1 signaling pathway in control and LSFC primary fibroblasts was assessed by immunoblot and densitometric analysis of total and phosphorylated levels of Akt (n = 10), mTOR (n = 13), and their downstream targets p70 ribosomal S6 kinase (p70S6K, n = 9) and eukaryotic translation initiation factor 4E-binding protein-1 (4E-BP1, n = 12). Results represent means ± SE. Difference between control and LSFC cells was assessed with a paired Student’s t-test. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001.

PDH1 is inhibited in LSFC fibroblasts.

In cancer cells, mTORC1-mediated metabolic remodeling has been shown to occur, in part, through activation of the HIF-1α/PDHK1 axis (8, 11). Thus, we examined whether this pathway was upregulated in LSFC cells. As illustrated in Fig. 3A, the expression of HIF-1α and PDHK1 was increased by 200% (P ≤ 0.05) and 155% (P ≤ 0.05), respectively, in LSFC fibroblasts. Consistent with this result, the phosphorylated and inhibited form of PDH1-α (phospho-PDH1-α) was increased by 63% (P ≤ 0.05) in LSFC fibroblasts compared with controls, whereas total PDH1-α protein levels remained unchanged (Fig. 3B). This leads to a 72% increase in the ratio of phosphorylated to total PDH1-α (P ≤ 0.05) in LFSC fibroblasts compared with controls.

Fig. 3.

Pyruvate dehydrogenase 1 (PDH1) is inhibited in Leigh syndrome French Canadian type (LSFC) fibroblasts. Immunoblot and densitometric analysis of total and/or phosphorylated levels of hypoxia-inducible factor 1α (HIF-1α, n = 5) and pyruvate dehydrogenase kinase 1 (PDHK1, n = 5; A) and PDH1-α (n = 8; B) in primary fibroblasts from control and LSFC donors. Results represent means ± SE. Difference between control and LSFC cells was assessed with a paired Student’s t-test. *P ≤ 0.05.

Suppression of mTORC1 activity does not alter HIF-1α and PDHK1 levels.

To further our understanding of mTORC1’s role, cells were treated with 100 nM rapamycin for 20 h. As anticipated, rapamycin reduced the ratio of phosphorylated to total mTOR by 58 and 49% (P ≤ 0.05) in LSFC and control fibroblasts, respectively (Fig. 4A). This resulted in a 98% and 92% reduction (P ≤ 0.001) in the ratio of phosphorylated to total p70S6K in control and LSFC fibroblasts (Fig. 4A), respectively. It has been shown that inhibition of p70S6K prevents its negative feedback on phosphatidylinositol 3-kinase leading to mTORC2 activation and phosphorylation of its downstream target Akt (31, 47). Consistent with this, rapamycin treatment was associated with a 1.8- and 2.1-fold increase in the ratio of phosphorylated to total Akt (P ≤ 0.01) in control and LSFC fibroblasts (Fig. 4A), respectively. However, contrary to our hypothesis, rapamycin did not affect HIF-1α and PDHK1 levels either in LSFC or in control fibroblasts (Fig. 4B). On the other hand, PDH1-α phosphorylation increased by 1.42- (P ≤ 0.05) and 1.51-fold (P ≤ 0.01) in control and LSFC fibroblasts, respectively, in response to rapamycin (Fig. 4C). This occurs in the absence of any change in total PDH1-α protein levels leading to a 1.6-fold increase in the ratio of phosphorylated to total PDH1-α in control and LSFC fibroblasts (Fig. 4C). In light of these results, we determined whether mTOR inhibition alters ATP levels. As illustrated in Fig. 4D, rapamycin reduced ATP levels by 15% (P ≤ 0.05) in control cells but did not provoke any significant changes in LSFC cells. Thus, in rapamycin-treated cells, ATP levels were higher in LSFC fibroblasts compared with controls (P ≤ 0.05).

Fig. 4.

Suppression of mechanistic target of rapamycin (mTOR) complex 1 (mTORC1) activity does not alter hypoxia-inducible factor 1α (HIF-1α) and pyruvate dehydrogenase kinase 1 (PDHK1) levels. Primary fibroblasts from control and Leigh syndrome French Canadian type (LSFC) donors were treated for 16 h with 100 nM rapamycin or vehicle (DMSO) in supplemented DMEM. On the day of the study, cells were cultured in a serum-free nonsupplemented DMEM, with inhibitor or vehicle as indicated, for an additional 4 h. A: suppression of mTORC1 activity was confirmed by immunoblot and densitometric analysis of phosphorylated levels of mTOR (n = 3), p70 ribosomal S6 kinase (p70S6K, n = 3), and Akt (n = 3). B and C: immunoblot and densitometric analysis of total levels of HIF-1α (n = 7) and PDHK1 (n = 8; B) and phosphorylated pyruvate dehydrogenase 1α (PDH1-α, n = 7; C) in response to rapamycin treatment. D: primary fibroblasts were cultured with inhibitor or vehicle, as indicated, in opaque 96-well plates at a density of 10,000 cells per well in triplicate. Cellular ATP levels were measured with the ATPlite luminescence ATP detection assay system (n = 4). Results represent means ± SE. Difference between baseline and rapamycin treatment was assessed with a paired Student’s t-test. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001.

Suppression of mTORC1 activity decreases LRPPRC and COX IV expression in LSFC fibroblasts.

Since mTORC1 also regulates mitochondrial protein expression (23, 24), we investigated the effect of rapamycin on the expression of OXPHOS complexes. Whereas no effect on OXPHOS complexes I, II, III, and V was observed in control or LSFC fibroblasts (Fig. 5A), there was a selective reduction of COX IV expression by 32% (P ≤ 0.01) in LSFC fibroblasts (Fig. 5B). Given the link between COX and LRPPRC, we examined the effect of mTORC1 inhibition on the latter. Rapamycin decreased LRPPRC expression by 11% (P ≤ 0.05) in control fibroblasts but by 41% (P ≤ 0.01) in LSFC fibroblasts (Fig. 5B). This supports a potential role for mTORC1 in the regulation of LRPPRC and COX expression.

Fig. 5.

Suppression of mechanistic target of rapamycin complex 1 (mTORC1) activity decreases leucine-rich pentatricopeptide repeat-containing (LRPPRC) and cytochrome-c oxidase (COX) expression in Leigh syndrome French Canadian type (LSFC) fibroblasts. Primary fibroblasts from control and LSFC donors were treated for 16 h with 100 nM rapamycin or vehicle (DMSO) in supplemented DMEM. On the day of the study, cells were cultured in serum-free nonsupplemented DMEM, with inhibitor or vehicle as indicated for an additional 4 h. A: immunoblot and densitometric analysis of mitochondrial complex levels [NADH dehydrogenase (ubiquinone) iron-sulfur protein-2, mitochondrial (NDUFS2, n = 5), succinate dehydrogenase (ubiquinone) iron-sulfur subunit, mitochondrial (SDHB, n = 3), cytochrome–b-c1 complex subunit 2, mitochondrial (UQCRC2, n = 4), and ATP synthase subunit-α, mitochondrial (ATP5A, n = 5)] in response to rapamycin treatment. B: immunoblot and densitometric analysis of LRPPRC (n = 8) and COX subunit IV (COX IV, n = 11) content in response to rapamycin treatment. Results represent means ± SE. CI–CV, oxidative phosphorylation complexes I–V. Difference between baseline and rapamycin treatment was assessed with a paired Student’s t-test. *P ≤ 0.05; **P ≤ 0.01.

DISCUSSION

LSFC is a metabolic disease characterized by decreased COX activity and mitochondrial dysfunction, in the absence of any changes in ATP levels under basal conditions (2). This suggests that activation of compensatory pathways can maintain ATP production. Our aim was to examine the remodeling of signaling pathways in LSFC fibroblasts. Our data demonstrate that LSFC fibroblasts present a series of adaptations including increased contribution of glucose to lactate formation, upregulation of the mTORC1/p70S6K/HIF-1α/PDHK1 signaling pathway, and increased PDH1-α phosphorylation. However, contrary to expectation, inhibition of mTORC1 with rapamycin was associated with increased phospho-PDH1-α in the absence of any changes in HIF-1α/PDHK1 levels. Importantly, rapamycin treatment further decreased LRPPRC and COX IV expression in LSFC fibroblasts. To our knowledge, this is the first time that LSFC metabolic signaling alterations have been reported and that mTORC1 regulation of LRPPRC levels has been revealed.

Our results demonstrate that the relative contribution of exogenous glucose to intracellular pyruvate, lactate, and alanine formation is enhanced in LSFC fibroblasts. This is consistent with the clinical phenotype of LSFC patients, which shows elevated plasma levels of these metabolites (40). A redirection of intracellular pyruvate toward the formation of cytosolic lactate and alanine, to the detriment of its mitochondrial metabolism, could be due to reduced pyruvate entry into the mitochondria due to PDH1 inactivation as well as reduced COX IV levels. Another possibility is enhanced glycolysis. Interestingly, an increase in basal extracellular acidification rate has been reported in fibroblasts of patients with novel LRPPRC mutations (not the French Canadian mutations; 26). Increased glycolysis in other models of mitochondrial diseases is also supported by indirect measurements such as higher levels of tissue glycolytic intermediates or use of inhibitors (6, 14, 26). In cancer cells, glycolysis can be upregulated as much as 200-fold allowing for rapid cell proliferation (reviewed in Ref. 7). A metabolic shift including enhanced glycolysis would enable greater cytosolic ATP production and could possibly explain the maintenance of ATP levels in LSFC fibroblasts.

Two major signaling pathways regulate glucose homeostasis, namely, AMPK and mTORC1. Whereas we did not observe any changes in AMPK activity, we found that LSFC fibroblasts display an upregulation of the mTORC1/p70S6K signaling cascade. mTORC1 activation has also been reported in Ndufs4 knockout mice, a model of Leigh syndrome (14), as well as in cell lines from patients with primary respiratory chain diseases such as mitochondrial encephalopathy, lactic acidosis, and stroke-like episodes (MELAS), Leigh syndrome, and mitochondrial myopathy (13, 46). Thus, mTORC1 activation may represent an adaptive response to electron transport chain defects.

Our data show that LSFC cells overexpressed HIF-1α, an important regulator of both glycolysis and glucose oxidation. Consistent with results obtained in other cells, upregulation of HIF-1α was found to be associated with increased PDHK1 expression and phosphorylation (i.e., inactivation) of PDH1 (16, 34). Increased HIF-1α expression may be a compensatory mechanism to palliate the mitochondrial deficit of LSFC cells. Indeed, HIF-1α stabilization improved survival of zebrafish embryos exposed to the complex III inhibitor antimycin (12). This study also showed that Ndufs4 knockout mice appeared healthier after chronic hypoxia (12). In cancer cells, HIF-1α impairs fatty acid oxidation through decreased medium- and long-chain acyl-CoA dehydrogenase levels (10). Interestingly, LSFC patients display elevated plasma levels of medium- and long-chain acyl-carnitines. The latter may be due to an electron chain defect, which reduces the mitochondrial capacity to metabolize acetyl-CoA produced by β-oxidation leading to increased acyl-carnitine levels (40). A reduction of both glucose and fatty acid oxidation (3, 16, 36) in LSFC cells may be protective in view of the reduced mitochondrial oxidative capacity (2, 26). However, although these mechanisms may compensate for mitochondrial dysfunction under basal conditions, we have previously shown that LSFC fibroblasts are more susceptible to cell death when exposed to high levels of palmitate (2) suggesting that these adaptive mechanisms may not be sufficient in cells exposed to nutrient overload.

Although the mechanism underlying HIF-1α expression in LSFC cells is unknown at present, numerous mechanisms have been shown to contribute to increased HIF-1α levels including lactate production, activation of the mTORC1 pathway, or reduction of succinate dehydrogenase activity (16a, 35). Furthermore, HIF-1α expression can be regulated at multiple levels including transcription, translation, or degradation (reviewed in Ref 16a). It is unlikely that the increase in HIF-1α levels is due to increased transcription since we did not observe any change in HIF-1α mRNA levels in LSFC fibroblasts compared with controls (data not shown). Numerous studies have highlighted the role of mTORC1 in HIF-1α activation (9, 11, 43). Surprisingly, inhibition of mTORC1 did not affect the expression of HIF-1α or its downstream target PDHK1 in LSFC cells, suggesting that upregulation of HIF-1α is independent of this pathway. Potentially, HIF-1α upregulation may result from inhibition of its degradation or increased translation. It should be noted that HIF-1α protein can be stabilized by end products of glycolysis, such as lactate and pyruvate, both of which are increased in LSFC fibroblasts (20). Further studies should explore the mechanisms underlying HIF-1α expression in LSFC cells.

Our results demonstrated that rapamycin increased PDH1 phosphorylation in both LSFC and control cells. A similar response to mTOR inhibition has also been observed in head and neck squamous cell carcinoma cells (3); however, this effect was attributed to Akt inhibition. In contrast, we observed an increase in phospho-Akt in rapamycin-treated cells, which probably occurs through activation of its upstream kinase, mTORC2 (31, 47). Since we did not observe any change in PDHK1 expression, rapamycin treatment may lead to overexpression of another PDHK isoenzyme (45) or, alternatively, modulate the expression or activity of phosphatases (reviewed in Ref. 1).

An important finding of our study is that mTORC1 inhibition resulted in a reduction of COX IV expression in LSFC cells. Morita et al. demonstrated that mTORC1 regulates the translation of a subset of mitochondrial proteins including complexes I and V (23). Since COX IV expression is dependent on LRPPRC, we hypothesized that mTORC1 may modulate COX IV expression through changed LRPPRC levels. Consistent with our hypothesis, rapamycin reduced LRPPRC levels in both control and LSFC cells. To our knowledge, this is the first report of LRPPRC regulation by mTORC1. Data on the regulation of LRPPRC expression are scarce. Cooper et al. have shown reduced LRPPRC levels in PGC-1α and PGC-1β double-deficient cells, suggesting that LRPPRC is downstream of these proteins (5). Future studies should investigate the underlying mechanism leading to the reduction of LRPPRC and COX IV in response to rapamycin. Interestingly, rapamycin treatment also resulted in reduction of ATP levels in control but not in LSFC cells. Furthermore, despite reduced COX IV expression, ATP levels of LSFC fibroblasts remained elevated compared with controls in rapamycin-treated cells. This strengthens the hypothesis that LSFC fibroblasts depend on extramitochondrial metabolism to maintain ATP levels.

Although patient fibroblasts are commonly used as a research model when studying mitochondrial diseases, the consequences of reduced LRPPRC expression are tissue specific (33). Thus, although our data cannot be extrapolated to tissues such as brain and liver, which have different energetic metabolism, they provide a better understanding of the adaptive molecular mechanisms for energy homeostasis that should be considered in future studies of mitochondrial diseases as well as important new information related to the potential role of mTORC1 in regulating LRPPRC expression.

In conclusion, we demonstrated that LSFC fibroblasts display molecular adaptation, including an mTORC1-independent upregulation of the HIF-1α/PDHK1 axis. However, mTORC1 upregulation appears to be required for LRPPRC and COX IV expression in LSFC fibroblasts. Further studies should examine in detail the effect of mTOR inhibition on LRPPRC and mitochondrial function in brain and liver, the two primary tissues affected by the disease.

GRANTS

This work was supported by a Canadian Institutes of Health Research Emerging Team Grant on LSFC (CPG 102168) and by the Fondation du Grand défi Pierre Lavoie. Y. Mukaneza was supported by a studentship from the Université de Montréal and Corporation de Recherche et d’Action sur les Maladies Héréditaires.

DISCLAIMERS

The funders had no role in study design, data collection and analysis, or preparation of the manuscript.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Y.M., J.T., C.L., C.D.R., and L. Coderre conceived and designed research; Y.M., A.C., M.-E.R., and S.D. performed experiments; Y.M., A.C., M.-E.R., S.D., and L. Coderre analyzed data; Y.M., C.D.R., and L. Coderre interpreted results of experiments; Y.M. prepared figures; Y.M. drafted manuscript; Y.M., M.R., C.L., C.D.R., and L. Coderre edited and revised manuscript; Y.M., A.C., M.-E.R., J.T., S.D., M.R., LSFC Consortium, C.L., C.D.R., and L. Coderre approved final version of manuscript.