Lipid and protein imbalances in muscle of a FAR1-patient with a heterozygous de novo variant

To the Editor: Disorders of lipid metabolism can affect skeletal muscle function and homeostasis, resulting in so-called “lipid myopathies,” which represent rare multi-systemic diseases and are often caused by genetic defects. Clinically, lipid myopathies can have acute or chronic clinical presentation also including a variable age of onset. Muscular atrophy is rarely present and muscular involvement can be fluctuant or stable and may manifest as fatigue, exercise intolerance, and/or muscular weakness. Muscle biopsies may show increased lipid content based on light or electron microscopic studies and on the etiology.1 Alterations in lipid levels and composition can influence protein patterns, particularly those associated with mitochondria, wherein lipid pathologies can impact mitochondrial-related proteins and vice versa.1–3 Fatty acyl-CoA reductase 1 (FAR1) encodes for the correspondent protein FAR1. FAR1 catalyzes the reduction of saturated and unsaturated C16 or C18 fatty acyl-CoA to fatty alcohols4 and contributes to the synthesis of ether lipids/plasmalogens, a process that necessitates fatty alcohols.4–7 Heterozygous de novo variants in the corresponding gene affecting the Arg480 residue of FAR1 lead to an autosomal dominant disorder, which differs in its disease mechanism from bi-allelic variants in terms of the biochemical phenotype.7 Based on fibroblast studies, FAR1 deficiency caused by bi-allelic variants results in impaired ether lipid synthesis and plasmalogen deficiency, whereas dominant de novo variants result in elevated plasmalogen levels.7 Patients harboring dominant FAR1 variants are less severely affected compared to those with recessive variants; they present with cataracts, spastic paraparesis, combined with speech and gross motor developmental delay and truncal hypotonia. However, muscle cell effects in dominant FAR1-patients have not been studied to date. We report on a girl born as preterm infant in the 34th week of gestation (birth weight 1800 g; 14th centile) with congenital cataracts. She showed a global developmental delay. Her motor milestones were severely delayed, that is, sitting at 4 years, standing with help at 6 years, presenting generalized muscular weakness, especially in her lower extremities in combination with axial weakness (accompanied by reduced head control), walking with orthotic devices at 9 years. At the age of 3 years, bilateral cataracts were surgically removed and intraocular lenses were implanted. She required acetabuloplasty and intertrochanteric valgization on both sides due to hip subluxation at the age of 3.5 years. Macrocephaly was noticed at the age of 6 years (head circumference 3 mm > 97th percentile). The parents are not blood-related and are of German descent. Standardized intelligence testing at the age of 10 years (HAWIK-IV) revealed an IQ of 55. Clinical examination at 15 years of age showed increased muscular tone with hyperreflexia and positive Babinski reflex. Muscle strength was generally reduced (proximal more than distal, legs more affected than arms). She was wheelchair-bound, no standing or walking were possible. She had right convex lumbar scoliosis and truncal hypotonia as well as contractures of elbows and knees. Cerebral MRI, creatine kinase level, and nerve conduction velocities were normal. Spirometry at 15 years of age demonstrated moderate to severe restrictive ventilation disorder (forced vital capacity 41%). Thus, clinical presentation in our patient accords with the overall phenotype described for dominant FAR1 variants.7 Schematic representation of the diagnostic workup along with clinical findings is shown in Figure 1A. Clinical data of our FAR1-patient carrying a heterozygous de novo p.(Arg480Cys) variant including microscopic examination of quadriceps muscle. (A) Schematic representation of the clinical findings and diagnostic workup. (B) Histological studies on quadriceps muscle cryosections show minor unspecific findings including fiber size variations, atrophic, and partially angular as well as polygonal fibers (based on H&E-stain). No profound mitochondrial changes were observed based on Gomori trichrome and NADH trichrome stains. Combined COX-SDH stain showed very mild reduced COX-reactivity in some fibers (appearing as faint COX staining). Mild elevation of lipid droplets was identified based on Oil Red O stain only in single muscle fibers. Some areas of fiber-type grouping were identified based on ATPase stain (white arrows). (C) Ultrastructural analysis of skeletal muscle biopsy. C1: a skeletal muscle fiber shown regular myofibrillar architecture and nuclei. C2: Abnormal folding of sarcolemma, consistent with muscle fiber atrophy. C3: Regular Z-band structures. C4: There are subtle physiological mitochondrial aggregates; mitochondria show normal morphology. (D) Neutral lipid content (HCS LipidTox green neutral lipid stain, yellow), muscle fiber sarcolemma (anti-Spectrin antibody, magenta), mitochondria (Mitotracker, cyan), and nuclei (DAPI, white) staining revealed no remarkable differences in neutral lipid distribution patterns as well as no significant association of neutral lipids with the mitochondrial network in the FAR1-patient biopsy (upper panels) compared to a representative control shown (lower panels). (E) Combined neutral lipid (HCS LipidTox green neutral lipid stain, yellow), phospholipid (HCS LipidTox Phospholipidosis Detection Reagent, magenta), muscle fiber sarcolemma (anti-Spectrin antibody, magenta), and nuclei (DAPI, white) staining showed no remarkable differences in distribution or quantity of the lipid species between the healthy control and the FAR1-patient samples. Scale bars: D, E = 10 µm. A quadriceps muscle biopsy was collected for diagnostic purposes at the age of 6 years due to the presence of muscular hypotonia and weakness in combination with cataracts and intellectual disability hinting toward a mitochondrial or glycosylation defect-driven neuromuscular disease. Histological studies revealed mild nonspecific myopathic changes, that is fiber size variation, atrophic and partially angular and polygonal fibers accompanied by mild elevation of lipid droplets (based on Oil Red O stain) only in single myofibers. There were occasional areas of fiber-type grouping (based on ATPase stain) identified. Of note, only very mild COX reduction was observed in few muscle fibers. Gomori trichrome and NADH trichrome did not show pathological findings (Figure 1B). Ultra-structural studies did not reveal a generalized and profound increase of lipid droplets, altered cytoskeleton, or perturbations of mitochondrial architecture. There was folding of sarcolemma as a pathomorphological finding consistent with muscle fiber atrophy (Figure 1C). Molecular genetic testing (exome sequencing and consecutive Sanger sequencing in the family) was initiated at the age of 14 years and revealed a previously described heterozygous pathogenic c.1438C>T variant in FAR1 as de novo (NM_032228.6). This variant is leads to the substitution of arginine to cysteine at the moderately conserved amino acid position 480, which is located within the transmembrane domain of FAR1 (according to UniProt). Of note, no variants in genes covered by exome sequencing known to be causative for a lipid storage myopathy were identified. Prompted by the known function of FAR1, the knowledge about elevated plasmalogen levels upon the presence of this pathogenic missense variant and our combined clinical and mild histological findings, we hypothesized that this variant may also affect lipid homeostasis in muscle cells. To test this assumption, we next performed further microscopic studies. However, fluorescence-based lipid and mitochondrial staining showed no differences compared to age-matched controls (Figure 1D and E). To further elucidate potential lipid pathology in the skeletal muscle more precisely, lipidomic studies were carried out (Figure 2A1); they revealed a significant decrease of alkyl-linked phosphatidylcholine (aPC), alkyl-linked phosphatidylethanolamine (aPE), and lyso-phosphatidylethanolamine (LPE) lipid classes in the patient’s muscle (Figures 2A2 and A3). LPE is derived from PE resulting from the partial hydrolysis of PE and the removal of a fatty acid group. In addition, a statistically significant increase of sphingomyelins belonging to the group of phosphosphingolipids was identified (Figure 2A3). In line with FAR1 function and previous descriptions in fibroblasts derived from FAR1-patients harboring the dominant p.(Arg480Cys) variant,7 ether-linked phosphatidylcholine (ePC) and ether-linked phosphatidylethanolamine (ePE) were increased in patient muscle (Figure 2A4). Calculation of respective lipid ratios revealed a consistent increase in patient-derived muscle ePC/ePE, increase of 22.7%, aPC/aPE, increase of 95.7%, ePC/total PC: increase of 47.1%, ePE/total PE: increase of 53.5% (Figure 2A5 and A6). Biochemical profiling on quadriceps muscle derived from the FAR1-patient. (A1) Schematic representation of lipidomic profiling of FAR1-mutant and control muscle samples. (A2-A6) Lipid composition (in mol% of total lipids) of muscle biopsies from our FAR1-patient or controls, (n = 3 technical replicates). Shown are major lipid class species in A2, and with an expanded scale for minor lipid species in A3. A4 shows % of ether species of the total lipid species in PC or PE lipid class and the different ratios are presented in A5 for the patient and pooled control data, respectively. Lipid class ratios in the FAR1-patient and pooled controls are summarized in A6. Data represent mean values and standard deviation. Statistical significance was assessed with the two-tailed unpaired Student t-test; *P ≤ .05. (B1) Schematic representation of proteomic profiling of FAR1-mutant and control muscle samples. (B2-B5) Proteomic signature of FAR1-mutant muscle: liquid chromatography coupled to tandem mass spectrometry in a data-independent-acquisition approach (performed on n = 3 technical replicates) led to the identification of 2542 with at least two unique peptides (B2). Data analysis revealed statistically significant dysregulation of 244 proteins; 142 are increased (lilac dots) and 102 are decreased (orange dots), respectively, as depicted in the volcano plot (B3). GO-term-based in silico analyses of proteomic data unveiled affected biological processes including regular mitochondrial function and cytoskeleton (B4). Protein gene changes are shown in the heatmap on the individual protein level (B5). To further elucidate the impact of these lipid findings on proteostasis, proteomic analysis was carried out as described previously8 (Figure 2B1). This enabled robust quantification of 2542 proteins based on at least 2 unique peptides (Figure 2B2) and revealed significant dysregulation of 244 proteins (142 were increased and 102 were decreased; Figures 2B3 and B4). A gene ontology-term-based analysis of biological processes demonstrated increased proteins that have major impacts on mitochondrial functions, for example electron transport (NADH to ubiquinone), ATP synthesis-coupled proton transport, and respiratory chain complex I assembly. They also affect nitric oxide-mediated signal transduction, protein folding, actin cytoskeleton organization, immune responses, and complement activation (classical pathway) (Figure 2B4). Decreased proteins also affect mitochondrial functions such as mitochondrial-mediated protein translation, ATP synthesis-coupled proton transport, and respiratory chain complex I assembly (Figure 2B4). Moreover, an impact on DNA repair and mRNA modification in addition to protein maturation, regulation of protein stability, and regulation of macroautophagy were observed. Proteinogenic dysregulations on the individual protein level are shown in the heatmap depicted in Figure 2B5. Our overall biochemical findings in combination with the results of our microscopic studies indicate that elevation of cytoskeletal and mitochondrial proteins may act toward successful maintenance of the contractile apparatus and mitochondrial functions. The need for the activation of compensatory mechanisms toward the maintenance of mitochondrial function might result from the above-mentioned lipid imbalances. Along this line, the literature provides evidence that the balance of skeletal muscle phosphatidylcholine (PC) and phosphatidylethanolamine (PE) is a key determinant of muscle contractile function and metabolism.9 The basal aPC:aPE ratio in our patient showed a 95.7% increase and the ePC:ePE ratio a 22.7% increase compared to controls indicating an imbalance due to the presence of the dominant FAR1 variant. Altered mitochondrial metabolism and actin cytoskeleton organization are suggested by the results of our proteomic study. The increases of ePC and ePE accord with the results of previous in vitro studies on fibroblasts expressing the dominant p.(Arg480Cys) variant.7 Moreover, these data indicate that diverse cellular populations including muscle cells are affected. Hence, our biochemical observations may correlate with muscular weakness observed in FAR1-patients. PC is a main component of the lipid bilayer in the sarcolemma, which forms a hydrophobic barrier that prevents the unregulated passage of ions and other charged molecules that would otherwise lead to myofiber death and muscle wasting. Abnormal folding of the sarcolemma indicated by our electron microscopic findings correlates with muscle fiber atrophy, which is in turn reflected by our clinical observations. To sum up, given that cellular functional studies already clarified pathological enzymatic activity of FAR1 upon the presence of the dominant p.(Arg480Cys) variant,7 results of our biochemical studies are in line with an assumed gain-of-function mechanism also in muscle cells most likely contributing to muscle hypotonia in FAR1-patients. Of note, lipid imbalances are not accompanied by accumulation of lipid droplets in terms of a lipid storage myopathy. Hence, the combined data unveil a significant muscle vulnerability on a biochemical level and also prove that this does not necessarily have to be associated with a profound structural pathology. The latter aspect represents a particular challenge when diagnosing muscle involvement in complex clinical pictures such as the FAR1-related neurological phenotype. A.R. and U.S.-S. received funding from the Deutsche Gesellschaft für Muskelkranke (DGM). A.H. acknowledges the support by the “Ministerium für Kultur und Wissenschaft des Landes Nordrhein-Westfalen” and “Der Regierende Bürgermeister von Berlin, Senatskanzlei Wissenschaft und Forschung”. T.R. also acknowledges the support by the “Ministerium für Kultur und Wissenschaft des Landes Nordrhein-Westfalen” for MODS [grant no. PROFILNRW-2020-107-A]. The authors declare to not have any conflict of interest.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.