Skin fibroblast metabolomic profiling reveals that lipid dysfunction predicts the severity of Friedreich’s ataxia

Friedreich’s ataxia (FRDA) is an autosomal recessive neurodegenerative disorder caused by a triplet guanine-adenine-adenine (GAA) repeat expansion in intron 1 of the FXN gene, which leads to decreased levels of the frataxin protein. Frataxin is involved in the formation of iron-sulfur (Fe-S) cluster prosthetic groups for various metabolic enzymes. To provide a better understanding of the metabolic status of patients with FRDA, here we used patient-derived fibroblast cells as a surrogate tissue for metabolic and lipidomic profiling by liquid chromatography-high resolution mass spectrometry. We found elevated HMG-CoA and β-hydroxybutyrate-CoA levels, implying dysregulated fatty acid oxidation, which was further demonstrated by elevated acyl-carnitine levels. Lipidomic profiling identified dysregulated levels of several lipid classes in FRDA fibroblast cells when compared with non-FRDA fibroblast cells. For example, levels of several ceramides were significantly increased in FRDA fibroblast cells; these results positively correlated with the GAA repeat length and negatively correlated with the frataxin protein levels. Furthermore, stable isotope tracing experiments indicated increased ceramide synthesis, especially for long-chain fatty acid-ceramides, in FRDA fibroblast cells compared with ceramide synthesis in healthy control fibroblast cells. In addition, PUFA-containing triglycerides and phosphatidylglycerols were enriched in FRDA fibroblast cells and negatively correlated with frataxin levels, suggesting lipid remodeling as a result of FXN deficiency. Altogether, we demonstrate patient-derived fibroblast cells exhibited dysregulated metabolic capabilities, and their lipid dysfunction predicted the severity of FRDA, making them a useful surrogate to study the metabolic status in FRDA.Supplementary key words: frataxin, ceramides, fatty acids oxidation, triglycerides, phospholipids, lipidomics, lipid remodeling, neurodegenerative disorders, triplet repeat expansion, stable isotope tracing

Friedreich’s ataxia (FRDA) is an autosomal recessive neurodegenerative disorder with an incidence of 1 in 29,000 (1). Currently it has no approved treatment (1). The main clinical features in FRDA include gait and limb ataxia, dysarthria, sensory loss, and cardiomyopathy (2). Heart failure from cardiomyopathy is the primary cause of death in the majority of patients with FRDA (3). FRDA is caused by a triplet guanine-adenine-adenine (GAA) repeat expansion in intron 1 of the FXN gene that leads to gene silencing and decreased levels of the mitochondrial protein frataxin (4). The number of GAA repeats inversely correlates with frataxin protein level and age of disease onset, both of which determine disease severity (5, 6). The tissues most affected are the heart, dorsal root ganglia, posterior columns of the spinal cord, dentate nucleus, and corticospinal tracts. The exact mechanism by which frataxin deficiency leads to neuro- and cardiodegeneration is not completely understood.One function of frataxin is in the formation of the iron-sulfur (Fe-S) cluster prosthetic groups that are critical for enzymes in the Krebs cycle (aconitase), oxidative phosphorylation (electron transport chain components of complexes I–III), and fatty acid breakdown (β-oxidation) (7, 8). Frataxin localization in the mitochondria (9) further suggests that mitochondrial dysfunction plays a role in FRDA. Decreased conversion of labeled glucose to acetyl-CoA in platelets from patients with FRDA (10) is consistent with studies that show diminished pyruvate oxidation in FRDA (10). Increased incorporation of labeled palmitate into HMG-CoA, an important intermediate in ketogenesis and sterol synthesis, in patients with FRDA suggests increased fatty acid metabolism through β-oxidation (11). Increased β-oxidation produces FADH₂ and NADH that can be utilized to maintain the electrochemical gradient across the inner mitochondrial membrane needed for ATP synthesis. Therefore, increased lipid metabolism observed in FRDA could be important to maintain cellular homeostasis during mitochondrial dysfunction.A recent study found reactive oxygen species-independent accumulation of iron in the nervous system of an FRDA fly model with a mutant frataxin homolog, associated with enhanced sphingolipid synthesis (12). Sphingolipids are linked to increased inflammation (13) and activate 3-phosphoinositide dependent protein kinase-1 (Pdk1) and myocyte enhancer factor-2 (Mef2) to trigger neurodegeneration (12). The findings in the fly model were replicated in a frataxin knockdown mouse model suggesting that the mechanism is evolutionarily conserved (14). PDK1 activity and sphingolipid levels were also elevated in heart tissues of patients with FRDA compared with healthy controls suggesting that a similar pathway may be activated in humans with FRDA (14).Ceramides are central intermediates in sphingolipid metabolism and have been implicated in several cellular processes including apoptosis (15). Dysregulated ceramides have been the focus of study in a variety of cardiac diseases. High ceramide ratios of Cer 16:0 and 18:0 to Cer 24:0 in plasma are strongly associated with increased risk for major adverse cardiac events (16). Furthermore, increased ceramide levels have been associated with diabetic cardiomyopathy (17) and increased de novo ceramide synthesis has been linked to advanced heart failure (18). The observation of elevated ceramides in FRDA heart tissue raises the question of whether sphingolipids will be dysregulated in other affected and nonaffected tissues.Ideally, metabolic and lipidomic abnormalities should be studied in the most affected tissues, but frataxin deficiency is present in all tissues to different extents (19). Since it is difficult to sample human cardiac tissue from living individuals, peripheral tissues, such as fibroblasts, can be used as models to study metabolic profiles of FRDA. Fibroblasts in culture have the additional advantage of not being influenced by diet or environment, thus providing a stable system for comparing metabolic flux between patients and controls. Recently, RNA sequencing and gene ontology analysis was used to identify differentially expressed genes between FRDA and healthy control fibroblasts and indicated that fibroblasts are an accessible system to study dysregulated pathways in FRDA (20). In the present study, we used highly sensitive and specific liquid chromatography-high resolution mass spectrometry (LC-HRMS) assays to perform metabolomic and lipidomic profiles in fibroblast cells from patients with FRDA with different disease severities. This study complements the RNA sequencing data and gives new insights into the disease mechanism.